UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Isolation, structure determination, and biosynthetic studies of secondary metabolites from Dorid Nudibranchs Graziani, Edmund Idris 1996

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

Item Metadata

Download

Media
831-ubc_1996-147509.pdf [ 9.35MB ]
Metadata
JSON: 831-1.0059575.json
JSON-LD: 831-1.0059575-ld.json
RDF/XML (Pretty): 831-1.0059575-rdf.xml
RDF/JSON: 831-1.0059575-rdf.json
Turtle: 831-1.0059575-turtle.txt
N-Triples: 831-1.0059575-rdf-ntriples.txt
Original Record: 831-1.0059575-source.json
Full Text
831-1.0059575-fulltext.txt
Citation
831-1.0059575.ris

Full Text

ISOLATION, STRUCTURE DETERMINATION, AND BIOSYNTHETIC STUDIES OF SECONDARY METABOLITES FROM DORID NUDLBRANCHS  by  EDMUND IDRIS GRAZIANI  B.Sc, Trinity College, University of Toronto, 1991  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 July 1996 © Edmund I. Graziani 1996  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.  Department The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT  Investigations of the skin extracts from a number of dorid nudibranchs have led to the isolation of two novel compounds, lovenone (2) and limaciamine (8). from the North Sea dorid  Adalaria  loveni,  Lovenone (2), isolated  represents the first triterpenoid isolated from a  nudibranch, and is only the second triterpenoid ever isolated from a marine mollusc. The structure of lovenone (2) was solved using a number of two-dimensional NMR techniques. Similarly, the isolation of limaciamine (8) from the North Sea dorid Limacia  represents the only  clavigera,  naturally occurring analogue reported to date of triophamine (9). Triophamine (9) was originally isolated from the British Columbia dorids, Triopha  catalinae  and Polycera  (2)  tricolor.  (9)  The isolation and structure determination of these novel compounds led directly to an investigation into the biosynthesis of secondary metabolites by dorid nudibranchs. By taking advantage of the unique biology of dorid nudibranchs, a protocol has been developed whereby multiple injections over time of [1,2- C2] acetate has afforded irrefutable proof for de 13  novo  biosynthesis in a number of dorid nudibranchs. Terpenoic acid glycerides have been isolated from skin extracts of a number of dorid nudibranchs collected worldwide. Herein is reported the first unambiguous proof for the de synthesis of terpenoic acid glycerides 28 and 29, isolated from montereyensis,  Archidoris  odhneri  novo  and A.  respectively, by NMR analysis of C - C coupling arising from incorporation of  intact doubly labeled acetate.  13  13  Stable isotope incorporation studies with [1,2-^C2] acetate have been used to investigate the biosynthesis of the sesquiterpenoids nanaimoal (38), acanthodoral (39), and isoacanthodoral (40) by the dorid nudibranch Acanthodoris sesquiterpenoids are synthesized de  novo  The results have shown that: i) the  nanaimoensis.  by A.  nanaimoensis  and ii) that a previously proposed  biogenetic pathway to the isoacanthodoral skeleton was not tenable and required modification.  The use of stable isotope methodology has been extended to probe polyketide biosynthesis by the dorid nudibranch, Triopha  catalinae.  Incorporation of [ 1,2-* ^C2]acetate into triophamine  (9) has clearly shown the biogenesis of (9) from two units of butyrate and one unit of acetate. This work represents the first experimental evidence for  de novo  polyketide biosynthesis by a  dorid nudibranch; moreover, the use of doubly-labeled [1,2-13C2] acetate has provided clear evidence in support of one pathway where a number of biosynthetic pathways were possible.  iv TABLE OF CONTENTS Abstract  Page ii  Table of Contents List of Tables List of Figures  iv vi vii  List of Schemes  "  x  List of Abbreviations Acknowledgments Dedication  xi xiii xiv  I. General Introduction  1  LA. I.B. I.C. I. D.  Introduction to Marine Natural Products Introduction to Dorid Nudibranchs and Their Secondary Metabolism Research Summary Endnotes: Chapter I: General Introduction  1 3 6 6  II. Isolation and Structure Determination of Lovenone, A Cytotoxic Degraded Triterpenoid from the North Sea Dorid Nudibranch, Adalaria  8  loveni  II. A. II.B. II.C. U.D. II. E. H.F.  Taxonomy Collection and Isolation Structure Determination Biological Activity Origin and Proposed Biogenesis Endnotes: Chapter II: A. loveni  9 10 10 37 37 40  III. Isolation and Structure Determination of Limaciamine, A Diacylguanidine from the North Sea Dorid Nudibranch, Limacia  clavigera  m . A . Taxonomy III. B. Collection and Isolation ITJ.C. Structure Determination m . D . Endnotes: Chapter ITJ: Limacia  41  clavigera  42 43 52 56  v  IV. Biosynthetic Studies of Isoprenoid Secondary Metabolites from Dorid Nudibranchs Using Stable Isotopes IV.I.A. Introduction IV.I.B. Biosynthetic Studies with Marine Invertebrates: Practical Considerations IV.I.C. Introduction to Isoprenoid Biosynthesis rV.I.D. Previous Studies into the Biosynthesis of Terpenoids by Marine Invertebrates IV.II. Biosynthesis of Terpenoic Acid Glycerides by the Dorid Nudibranchs Archidoris  odhneri  and A. montereyensis  IV.II.A. Preamble IV.II.B. Preliminary Results Using Liposomes IV. II.C. Successful Incorporation Studies  57 57 59 63 67 75  76 78 82  IV.III. Stable Isotope Incorporation Studies on Sesquiterpenoids from the Dorid Nudibranch Acanthodoris nanaimoensis  99  IV. IV. Endnotes: Chapter IV: Biosynthesis of Isoprenoids  130  V. Stable Isotope Investigations on the Biosynthesis of Triophamine by  Triopha  V. A. V.B. V.C. V.D.  catalinae  Introduction to Polyketide Biosynthesis Precedents for a Polyketide Origin of Triophamine Feeding Experiments with Triopha catalinae Endnotes: Chapter V: Biosynthesis of Triophamine  VI. Concluding Remarks VLB. Endnotes: Chapter VI: Concluding Remarks VII. Experimental VII.B. Endnotes: Chapter VU: Experimental  134  135 137 149 160 162 171 172 181  VIII: Appendix A: Nuclear Magnetic Resonance Techniques  182  IX. Appendix B: Isolation of Known Compounds from New Sources  190  vi  LIST OF TABLES Page  Table 1:  H, C , COSY, HMBC, and nOe NMR Data for lovenone (2)  19  Table 2:  1  H , C , COSY, and HMBC Data for limaciamine (8)  53  Table 3:  Specific Incorporation Data for the [1,2- C2] acetate Feeding  l  13  13  13  Experiments with Archidoris odhneri and A. montereyensis Table 4: Table 5: Table 6:  NMR Data for nanaimool (41) and isoacanthodorol (43)  95 121  NMR Incorporation Data for Labeled nanaimool (41) and isoacanthodorol (43) Specific Incorporation Data for triophamine (9)  127 159  vii  LIST OF FIGURES Page  Figure 1:  Anatomy of a Typical Dorid Nudibranch  5  Figure 2:  Color Plate of Adalaria  8  Figure 3:  1  Figure 4:  13  C Spectrum of lovenone (2) [ 125 MHz, C&>6\  13  Figure 5:  APT Spectrum of lovenone (2) [125 MHz, C^Dd  14  Figure 6:  HMQC Spectrum of lovenone (2) [500 MHz, QDg]  15  Figure 7:  Expanded Upfield Region of HMQC Spectrum of lovenone (2)  loveni  H NMR Spectrum of lovenone (2) [500 MHz, C ^ ]  12  [500 MHz, C6D ]  16  Figure 8:  Electron Impact Mass Spectrum of lovenone (2)  17  Figure 9:  Fourier Transform Infrared (FT-IR) Spectrum of lovenone (2)  18  Figure 10:  Fragment Describing Ring C of lovenone (2)  21  Figure 11:  COSY Spectrum of lovenone (2) [500 MHz, C6D6]  22  Figure 12:  Expanded Upfield Region of COSY Spectrum of lovenone (2) [500 MHz, C6D ] Expanded Methyl Region of HMBC Spectrum of lovenone (2) [500 MHz, C6D ] Fragment Describing Ring B of lovenone (2) Expanded Region of HMBC Spectrum of lovenone (2) [500 MHz, C6D6] Showing Correlations from the H at 5 3.67 Expanded Region of HMBC Spectrum of lovenone (2) [500 MHz, C6D6] Showing Correlations from the *H at 8 2.05  6  6  Figure 13:  6  Figure 14: Figure 15:  l  Figure 16: Figure 17:  23 24 25 26 27  Expanded Downfield Region of HMBC Spectrum of lovenone (2) [500 MHz, C D ]  28  Figure 18:  Fragment Describing Ring D of lovenone (2)  29  Figure 19:  Downfield Region of H Spectrum of lovenone (2) Recorded in d6-DMSO [500 MHz] Downfield Region of H Spectrum of lovenone (2) Recorded in d6-DMSO + D 0 [500 MHz] Expanded Region of COSY Spectrum of lovenone (2) in d6-DMSO 500 MHz]  6  Figure 20:  6  ]  2  Figure 21:  30  !  Figure 22:  Fragment Describing Side-Chain of lovenone (2)  Figure 23:  Expanded Region (*H 8 1.5 to 1.8) of HMBC Spectrum  30 31 33  of lovenone (2) [500 MHz, C ^ ]  34  Figure 24:  Proposed Conformation of lovenone (2)  35  Figure 25:  Selected Difference nOe Spectra of lovenone (2)  36  Figure 26:  Color Plate of Limacia  41  Figure 27:  !  H Spectrum of limaciamine (8) [500 MHz, CDCI3]  44  Figure 28:  FT-IR Spectrum of limaciamine (8) [NaCl, thin film, neat]  45  clavigera  Vlll  Figure 29:  Electron Impact Mass Spectrum of limaciamine (8)  46  Figure 30: Figure 31: Figure 32: Figure 33:  13  C Spectrum of limaciamine (8) [125 MHz, CDCI3] HMQC Spectrum of limaciamine (8) [500 MHz, CDCI3] COSY Spectrum of limaciamine (8) [500 MHz, CDCI3] Selected Region of HMBC Spectrum of limaciamine (8)  47 48 49  [500 MHz,  Figure 34:  Methyl Region of HMBC Spectrum of limaciamine (8) [500 MHz,  Figure 35: Figure 36: Figure 37: Figure 38: Figure 39: Figure 40: Figure 41: Figure 42: Figure 43: Figure 44:  50  CDCI3]  51  CDCI3]  Color Plates of Archidoris odhneri (top) and A. montereyensis (bottom) H Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] C Spectrum of farnesic acid glyceride (28) [125 MHz, CDCI3] HMQC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] Expanded Upfield Region of HMQC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] Expanded Upfield Region of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] Expanded Region (Upfield H , Downfield C) of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] Expanded Region (Downfield H, Upfield C) of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDCI3] Normalized and Truncated C Resonances for Labeled (left) and Unlabeled (right) Samples of farnesic acid glyceride diacetate (35) [125MHz,CDCl ] H Spectrum of diterpenoic acid glyceride (29) [400 MHz, CDCI3]  1  13  1  Figure 46:  86 87 88  13  1  89  13  90  13  3  Figure 45:  75 83 84 85  1  94 96  Normalized and Truncated C Resonances for Labeled (left) and Unlabeled (right) Samples of diterpenoic acid glyceride (29) [125MHz,CDCl ] Color Plate of Acanthodoris nanaimoensis H Spectrum of nanaimool [500 MHz, CDCI3] 13  97 99  3  Figure 47: Figure 48: Figure 49: Figure 50: Figure 51: Figure 52: Figure 53:  (41) C Spectrum of nanaimool ( 41) [ 125 MHz, Upfield Region of APT Spectrum of nanaimool ( 41) [125 MHz, HMQC Spectrum of nanaimool ( 41) [500 MHz, COSY Spectrum of nanaimool ( 41) [500 MHz,  13  CDCI3]  CDCI3]  CDCI3]  104 105 106  Expansion of Upfield Region of HMBC Spectrum of nanaimool (41) 107  CDCI3]  Expansion of Downfield C Region of HMBC Spectrum of 13  nanaimool Figure 55:  103  CDCI3]  [500 MHz,  Figure 54:  102  l  (41) [500 MHz,  108  CDCI3]  Expansion of HI 2 Region of HMBC Spectrum of nanaimool (41) [500 MHz,  Figure 56:  1  Figure 57:  13  109  CDCI3]  (43) [500 MHz, C Spectrum of isoacanthodorol ( 43) [125 MHz,  H Spectrum of isoacanthodorol  CDCI3] CDCI3]  112 113  IX  Figure 58:  HMQC Spectrum of isoacanthodorol (43) [500 MHz, CDC1 ]  Figure 59:  Expansion of Upfield Region of HMQC Spectrum of isoacanthodorol (43)  3  [500 MHz,  115  CDCI3]  Figure 60:  COSY Spectrum of isoacanthodorol (43) [500 MHz, CDCI3]  Figure 61:  Expansion of Upfield Region of HMBC Spectrum of isoacanthodorol (43)  Figure 62:  Expansion of Downfield *H Region of HMBC Spectrum of isoacanthodorol (43) [500 MHz, CDCI3] Expansion of Downfield C Region of HMBC Spectrum of isoacanthodorol (43) [500 MHz, CDCI3] Normalized and Truncated C Resonances for Labeled (left) and Unlabeled (right) Samples of nanaimool ( 4 1 ) [125 MHz, CDCI3] Normalized and Truncated C Resonances for Labeled (left) and Unlabeled (right) Samples of isoacanthodorol ( 4 1 ) [ 125 MHz, CDCI3] Selected C Resonances (Truncated and Normalized) for Labeled (left) and Unlabeled (right) Samples of a) nanaimool (41) and •b) isoacanthodorol (43) [ 125 MHz, CDCI3]  [500 MHz,  Figure 63: Figure 64: Figure 65: Figure 66: Figure 67: Figure 68:  114  116 117  CDCI3]  118  13  119  13  122  13  124  13  Color Plate of Triopha  126  catalinae  134  'H Spectrum of triophamine (9) [500 MHz, CDCI3]  150  C Spectrum of triophamine (9) [ 125 MHz, CDCI3]  151  Figure 69:  13  Figure 70:  COSY Spectrum of triophamine (9) [400 MHz, CDCI3]  152  Figure 71:  HMQC Spectrum of triophamine (9) [500 MHz, CDC13]  153  Figure 72:  Expansion of Upfield Region of HMBC Spectrum of triophamine (9) [500 MHz,  Figure 73:  Expansion of H5 Region of HMBC Spectrum of triophamine (9) [500 MHz,  Figure 74:  154  CDCI3]  155  CDCI3]  Expanded Downfield C Region of HMBC Spectrum of triophamine (9) 13  [500 MHz,  156  CDCI3]  Figure 75:  Truncated Normalized C Resonances of Labeled (left) and Unlabeled (right) Samples of triophamine (9) [125 MHz, CDCI3]  158  Figure A l :  Pulse sequence for the acquisition of a H spectrum  183  Figure A2:  Pulse sequence for the acquisition of the nOe difference spectrum  184  Figure A3:  Pulse sequence for the acquisition of a C spectrum  185  Figure A4:  Pulse sequence for the APT spectrum  185  Figure A5:  Pulse sequence for a typical 2D NMR experiment  186  Figure A6:  Pulse sequence for the COSY experiment  187  Figure A7:  Pulse sequence for the HMQC experiment  188  Figure A8:  Pulse sequence for the HMBC experiment  189  Figure B1:  Color Plates of Anisodoris fontaini Thecacera darwinii (below)  190  13  l  13  (above) and  X  LIST OF SCHEMES Page  Scheme 1: Scheme 2: Scheme 3: Scheme 4:  Proposed Biogenesis of lovenone (2) Proposed EDMS Fragmentation Pattern for limaciamine (8) Biosynthesis of JPP/DMAPP via Acetate and Mevalonic Acid Biosynthesis of Higher Terpenoids from JPP/DMAPP  Scheme 5:  Incorporation of Acetate into farnesic acid glyceride (28) and diterpenoic acid glyceride (29) Biogenesis of Acanthodorane, Isoacanthodorane, and Nanaimoane Sesquiterpenoid Carbon Skeletons Proposed Biogenesis of the Isoacanthodorane Skeleton Formation of malonyl Co A from acetyl CoA The Principal Reactions of Polyketide Biosynthesis S-Adenosylmethionine (SAM) Methylation Biogenetic Routes to the denticulatins (48 and 49) Proposed Biogenesis of Acyl Moiety of Triophamine (9) Biosynthesis of Triophamine from Acetate Derived Butyrate Alternate Biosynthesis of Triophamine via SAM Alkylations Phylogenetic Tree for Selected Opisthobranch and Pulmonate Genera with Known or Suspected Biosynthetic Capabilities  Scheme 6: Scheme 6b: Scheme 7: Scheme 8: Scheme 9: Scheme 10: Scheme 11: Scheme 12: Scheme 13: Scheme 14:  39 55 64 66 91 101/123 128 135 136 137 140 145 147 148 170  XI  LIST OF ABBREVIATIONS [CX]D  -specific rotation at wavelength of sodium D line  Ac  -acetyl  AC2O  -acetic anhydride  APT  -attached proton test  ax  -axial  B.B. Dec.  -broad band decoupling  BIRD  -bilinear rotation decoupling  bm  -broad multiplet  br  -broad  bs  -broad singlet  °C  -degrees Celcius  C6D6 cm CoA COSY 8  -benzene-d6 -chloroform-d -wavenumbers -coenzyme A -correlation spectroscopy -chemical shift in parts per million  d  -doublet  Da  -Daltons  dd  -doublet of doublets  AM  -difference in mass  DMAPP  -dimethylallyl pyrophosphate  DMSO  -dimethyl sulfoxide  dt  -doublet of triplets  EC50  -effective concentration resulting in 50% response  EIHRMS  -electron impact high resolution mass spectrometry  EIMS  -electron impact mass spectrometry  eq  -equatorial  Et  -ethyl  EtOAc  -ethyl acetate  FTD  -free induction decay  FTIR  -Fourier transform Infrared  HETCOR  -heteronuclear correlation  HMBC  -heteronuclear multiple bond multiple quantum coherence  HMQC  -heteronuclear multiple quantum coherence  HPLC  -high performance liquid chromatography  CDCI3 -1  Xll  HSPC  -hydrogenated soya phosphatidylcholine  i IPP J m  -signal due to impurity -isopentenyl pyrophosphate -scalar coupling constant -multiplet  m/z Me MeOH mmu NMR nOe PG  -mass to charge ratio -methyl -methanol -millimass units -nuclear magnetic resonance -nuclear Overhauser effect -phosphatidylglycerol  PKS q S s SAE SAM SCUBA  -polyketide synthase -quartet -signal due to solvent -singlet -S-adenosylethionine -S-adenosylmethionine -self-contained underwater breathing apparatus  sp. t IR TLC UV w  -species -triplet -retention time -thin layer chromatography -ultraviolet -signal due to water  Xlll  ACKNOWLEDGMENTS  I am deeply grateful for the unflagging encouragement, guidance and support offered me by my research supervisor, Prof. Raymond Andersen. It has been a privilege to work under the tutelage of such an insightful, dedicated, and enthusiastic scientist. For enhancing my appreciation of the opisthobranchs, I owe much to the many fascinating conversations with Dr. Sandra Millen, who also performed all the taxonomic identifications of the foreign nudibranchs. For assistance in collecting, I must thank Sandra again for the Chilean nudibranchs, Michael LeBlanc and Julie Kubanek for the Norwegian nudibranchs, and my colleagues (Jeff Gerard, Fangming Kong, Yasmin Khan, Todd Barsby, and Paul Haden) for all their help out at Bamfield. The assistance of the staff of the Bamfield Marine Station is also gratefully acknowledged. Chris Hanson and Sarah Halleran in Dr. Theresa Allen's lab must also be acknowledged for their help with the bioassays and liposomes, respectively. I must also credit the support staff at the NMR and Mass Spectrometry facilities at U.B.C. for their assistance. Thanks are also due to Jeff and Julie for their assistance in proofreading this thesis. I am grateful to Ron Long for the color pictures of the British Columbia nudibranchs. remaining color photos I am grateful to M.LeBlanc and S. Millen.  For the  DEDICATION  For Yasmin, who has been a constant source of love and support, and for my parents, who started it all with their stories of Banting and Best...  "Prelog has often said to me that I am a baroque amateur in everything I do." -Ruzicka  1  I. General Introduction LA Introduction to Marine Natural Products  The first published reports 1 of novel secondary metabolites isolated from marine invertebrates and algae marked the beginning of a renaissance in natural products research. With the advent of SCUBA and other methods that made the fruits of the oceans available for study, an abundance of secondary metabolites without precedent from terrestrial sources were described for the first time. This rich new source of secondary metabolites combined with the refinement of NMR techniques- higher field magnets, inverse detection probes, new pulse sequences- has led to the structure elucidation of complex molecules that occur in trace amounts in nature. The invaluable reviews by Faulkner attest to the wealth of structural novelty obtained from the oceans.  1  Technological advances have not only affected the way in which structural determinations are made, but are also providing new sources of marine secondary metabolites. For example, our understanding of how to culture marine microorganisms has expanded rapidly in recent years, and 2  the marine fungi are only just now being investigated. It remains to be seen whether or not our 2  ability to culture new kinds of marine organisms and to collect undescribed species from previously inaccessible places will lead to unprecedented structures, but the challenge is there to be addressed. From asking the question, "Can we find new natural products from marine sources?", other questions began to emerge: "what was the function of these compounds?" (chemical ecology); "how were these compounds being made?" (biosynthesis); "to what therapeutic uses could these compounds be put?" (medicinal chemistry); "how could these new structural types be made in a laboratory?" (synthetic chemistry). As the field expanded, the need for interdisciplinary research became obvious; biologists, pharmacologists, ecologists, and microbiologists were soon to become regular collaborators with the marine natural products chemist.  These new  collaborations brought with them a new appreciation for the function and uses of marine natural products.  It is patently impossible, however, to investigate the synthesis, biosynthesis, chemical ecology, pharmacology, or toxicology of a compound whose structure is unknown. While this may seem a truism, the search for new naturally-occurring compounds is eminently justifiable since the need for new chemical agents in medicine, agriculture, and other applied sciences is without question, and the utility of such agents cannot be assessed until they have been isolated and described. The science of chromatography, when applied to natural products isolation, requires just as much forethought and chemical insight as does the use of well-established techniques in the synthesis of a large organic molecule. The techniques themselves may be commonplace, but it is the efficient, elegant, and unprecedented way in which they are applied that takes natural products isolation from the realm of the routine to the very frontier of chemical inquiry. The intellectual challenges and rewards of spectroscopic data analysis to yield a unique structure will be amply demonstrated in the following chapters. Complementary to the advances being made in isolation and structure determination work, other investigations involving marine natural products are also bearing fruit. The areas of biosynthesis, chemical ecology, chemotaxonomy, synthesis, are all thriving offshoots of the search for new marine natural products. The very definition of a secondary metabolite- that which is not necessary for growth and reproduction in an organism- begs the question as to the origin and purpose of these compounds. Investigations into the biosynthesis of marine natural products are only now catching up to the current levels of sophistication found in terrestrial and microbial natural products, and in this particular regard it is hoped that this thesis will encourage future 3  work on the biosynthesis of natural products from marine invertebrates. The revolution that has taken place in the biological sciences in recent years, specifically the new (chemical) tools that are in the hands of the cell biologists and molecular geneticists, also suggests a possible direction in which the future of marine natural products may be heading. The discovery of new enzymes, new biochemical pathways, new genetic machineries will surely be the result of applying the tools of molecular biology to the organisms that have produced such a staggering array of novel secondary metabolites. The research described herein therefore will illustrate, it is hoped, how the isolation and structure determination of novel secondary metabolites, when assessed in the context of the  chemistry and ecology of the organisms from which they were isolated, can lead to methods of further investigating the origins of these compounds, namely, the use of stable isotopes for investigating biosynthesis in dorid nudibranchs.  LB  Introduction to Dorid Nudibranchs and their Secondary Metabolism  With the exception of the arthropods, the largest phylum of invertebrate animals (in terms of abundance of species) is most certainly the molluscs. The phylum Mollusca consists of seven still-living classes: Monoplacophora, Polyplacophora (chitons), Aplacophora, Bivalvia (clams, oysters, mussels etc.), Scaphopoda, Cephalopoda (squids, octopi etc.), and Gastropoda. The class Gastropoda is the largest class of molluscs, comprising 30,000 living species with an additional 15,000 fossil forms. The abundance of species and penetration of gastropods into a wide variety of environments (marine, fresh-water, land) suggests that this is a very successful group of animals.  4  The class Gastropoda comprises three subclasses: Prosobranchia (i.e. sea snails, limpets etc.), Pulmonata (i.e. land snails and slugs), and Opisthobranchia.  The features (morphological,  anatomical etc.) that distinguish gastropods from other molluscs and opisthobranchs from other gastropods are to a certain extent arcane, at least to the non-specialist. Of particular significance to the opisthobranchs, however, is a general correlation between a loss of shell and the adoption of chemical defenses. Within the opisthobranchs, members of the order Nudibranchia (the so-called "sea-slugs") are perhaps the most spectacular and beautiful. These brightly-colored creatures are often the most noticeable and visually delightful of all invertebrates observed while SCUBA diving in the Northeastern Pacific.  The nudibranchs are further classified into four suborders:  Dendronotacea ("dendronotids"), Arminacea ("arminids"), Aeolidacea ("aeolids"), and Doridacea ("dorids").  This last suborder, the so-called dorid nudibranchs, has been of great interest to  marine natural products chemists in view of the abundance and variety of chemical structures isolated from extracts of these creatures. Numerous reviews will attest to the wealth of novel structures described from these remarkable animals.  5  4  Dorid nudibranchs can be thought of as being of great help to marine natural products chemists in that they often sequester toxic compounds from the sponges, bryozoans etc. on which they feed, "borrowing" the chemical defenses of other animals to serve their own ends. In point of fact, the shell-less and easily distinguished dorids are rarely, if ever, molested by predators, and chemical defenses are often invoked to explain this phenomenon. Perhaps the classic example 6  was the isolation of small amounts of a highly cytotoxic substance from the tropical dorid Chromodoris  elisabethina.  The structure of this compound could not be elucidated, given the  small amount of material, until the same compound was isolated in much larger amounts from the dietary sponge, Latrunculia  magnificat  Latrunculin (1), the compound isolated from this sponge  was identical to that from the nudibranch, as proven later by Scheuer and co-workers.  8  (1)  Thus, the dorid serves as a kind of purification step in its own right since the relatively "clean" (i.e. lacking the complex primary metabolites which are a nuisance to the natural products chemist) extracts obtained from dorids often contain a surprisingly large amount of "pre-filtered" secondary metabolites of great interest and structural novelty. It is no wonder, then, that these are such attractive animals upon which to focus natural products investigations.  RHINOPHORE GANGLIONIC RING  SALIVARY GLAND  REPRODUCTIVE SYSTEM HINDGUT  STOMACH DIGESTIVE GLAND  GILLS  HEART ANUS  Fig. 1: Anatomy of a typical dorid nudibranch  9  Figure 1 shows a generalized anatomy for a typical dorid nudibranch. The outside "skin" or mantle (also called the dorsum), is usually a tough layer which covers and protects the viscera, or inner organs, of the animal. Embedded in the mantle are glands which store the putative defensive compounds. For this reason, dorids are often extracted by immersing them whole in solvent without further maceration, since the vast majority of interesting compounds will leech out from the mantle, uncluttered by the primary metabolites contained in the viscera. Other general features of dorids include gills for respiration, rhinopores (presumed to be sensory organs), a primitive circulatory system, a large digestive gland, and a hermaphroditic reproductive system (thus every other member of the species is a potential mate!).  6  I.C  Research Summary  The research described in this thesis has focussed on chemical investigations of a number of cold-water dwelling dorid nudibranchs resulting in the isolation of two new compounds (and a number of previously reported compounds), for which a dietary origin seemed unlikely. In these cases, it was hypothesized that rather than obtaining the compounds form the diet (as most dorids are wont to do), the animals were in fact biosynthesizing the compounds themselves. A survey of the literature revealed some preliminary evidence for a few examples of de novo biosynthesis, and 5  all experiments involved the use of radioisotopes (these experiments will be discussed in greater detail in the introduction to Chapter IV). The inherent ambiguity in observing the low-levels of radioactivity typically reported in these studies caused some debate as to the validity of some (though not all) such studies, which had the overall effect of discouraging further work in this relatively unexplored field. The need for new experimental methods that provided unambiguous results was clear if such biosynthetic studies were to be successfully renewed using dorid nudibranchs. The primary goal of this study, therefore, became the development of a stable isotope methodology for determining the de  novo  biosynthetic origin of a number of compounds from  dorid nudibranchs.  I.D Endnotes: Chapter I. General Introduction  1. Faulkner, D.J., Nat. Prod.  Rep.,  1995, 12,  223-69, and previous reviews cited therein.  2. Fenical, W.; Jensen, P.R. in Marine Biotechnology. Vol. 1: Pharmaceutical and Bioactive Natural Products. Attaway, D.H.; Zaborsky, O.R., eds., Plenum Press, New York, 1993, Ch. 12. 3. Cane, D.E.; Tandon, M., J. Am. Chem. Soc, 1995, 117 (20), 5602-5603. 4. Ruppert, E.E.; Barnes, R.D., Invertebrate Zoology. Saunders College Publishing, 6th ed., 1994.  5. Karuso, P., "Chemical Ecology of the Nudibranchs", in Bioorganic Marine Chemistry. Vol. I, Springer-Verlag, Berlin, 1987, pp. 31-60; and Avila, C , Oceanography Mar. Biol. Ann. Rev, 1995, 33, 487-559. 6. Faulkner, D.J., in Ecological Roles of Marine Natural Products, ed. Paul, V.J., Cornell University Press, Ithaca, New York, 1992, Ch. 4. 7. Kashman, Y.; Groweiss, A.; Shmueli, U., Tetrahedron Letters, 1980,21,  Ztl^l.  8. Okuda, R.K.; Scheuer, P.J., Experientia, 1985,41, 1355-56. 9. Figure taken from: McDonald, G.R., Nybakken, J.W., Guide to the Nudibranchs of California. American Malcologists Inc., 1980, pg. 12.  II. Isolation and Structure Determination of Lovenone, A Cytotoxic Degraded Triterpenoid from the North Sea Dorid Nudibranch, Adalaria loveni  Fig. 2 Color Plate of Adalaria loveni  IIA. Taxonomy  Adalaria  1  (Alder & Hancock, 1862) is a distinctive and very rare species of dorid  loveni  nudibranch found in the North Sea. While it has previously been identified as Doris (Loven, 1846) and Doris  loveni  (Alder & Hancock, 1862), the current designation as A.  preferred. Ironically, this rare species has served as the type specimen for all Adalaria  muricata loveni  is  species.  Its distinguishing features are a dorsal mantle that bears numerous rounded or club-like tubercles, gills that are simple pinnate, radula, an unarmed penis, and buccal pump pedunculate. A. loveni  appears as a white to pale orange nudibranch whose average length is 1 to 3 cm.  While this species is very similar in appearance to its congener,  A.  it can best be  proxima,  distinguished in the field by the presence of widely spaced (close-packed in A.  proxima  ) large  spiculose tubercles on the dorsal mantle. Generally, nudibranchs are best classified on the basis of their internal anatomy - specifically the oral cavity and the reproductive system. While a critical discussion of these features is beyond the scope of this thesis, identification of specimens of A.  loveni  was based upon dissection and  such an analysis of relevant anatomical features performed by Dr. Sandra Millen at the University of British Columbia. Details of the life-cycle, habits, and feeding behaviour of A. loveni All field reports suggest that A. securifrons  loveni  feeds exclusively on bryozoans, specifically  (Pallas) and Membranipora  observations. Egg masses of A.  loveni  are not well-documented. Securiflustra  spp., which was confirmed by our own collecting have been collected from the kelps upon which its dietary  bryozoans grow and the size of the ova (80 mm) has been used to distinguish them from the much larger eggs of A.  proxima.  The geographical distibution of A.  loveni  extends from Trondheimsfjord, Norway to the Firth  of Forth, Scotland. Reports of collections of  A.  loveni  from the Yorkshire coast require  corroboration and are therefore unreliable. The reported depth range for Norwegian fjords is from 0 to 200 m.  A.  loveni  in the  10  IIB.  Collection and Isolation  Specimens (250 animals) of A. loveni  were collected by hand using SCUBA at depths of 8 to  15 m in surge channels off Flat0y Island near Bergen, Norway in the North Sea. Freshly 2  collected animals were immediately immersed in methanol at the surface and transported back to the laboratory for further work-up. The initial methanol extract was decanted, filtered, and reduced in vacuo.  The animals were subsequently exhaustively extracted with additional methanol and 1:1  MeOH: CH2CI2; these extracts were similarly filtered, reduced in  vacuo,  and combined to yield an  aqueous suspension. This material was further diluted with distilled water and partitioned with EtOAc.  Upon exhaustive extraction with EtOAc, the aqueous layer was set aside, and the  combined EtOAc layers were dried over MgS04 and reduced in  vacuo  to yield an orange oil (400  mg). These non-polar constituents of the extract were fractionated by normal phase silica-gel flash chromatography (step gradient from 100% hexanes to 1:1 hexanes: EtOAc) to yield a fraction of almost pure lovenone (2).  Further purification on normal phase HPLC (65% hexane/ 35%  EtOAc) afforded pure lovenone (2) (1 lmg) as an optically active ([a]rj)= -38°, CHCI3) colorless glass.  IIC.  Structure Determination  The gross structure of lovenone (2) was solved using a number of spectroscopic techniques, primarily two-dimensional NMR experiments (see Appendix A for a discussion of how such experiments are performed and how the data are analyzed). Once a molecular formula had been determined from the EIHRMS, and a number of functional groups had been identified from the IR and l^C spectra, smaller fragments of the molecule were postulated from the 2D NMR connectivity data (COSY, HMBC). Employing a teleological approach to the data as a whole, it was possible to piece together these fragments and arrive at an unambiguous structure for lovenone (2) that was completely consistent with all these data. Furthermore, an analysis of H,H coupling constants in combination with nOe data allowed for the assignment of both relative stereochemical and conformational features of the compound. While the following argument for the structure of  11  lovenone (2) will refer to structure (2) as though it were a fait accompli, it should be noted that such an argument can only be made in retrospect, and every effort will be made to highlight the logic by which structure (2) was obtained and all other possibilities eliminated.  O (2)  Lovenone (2) gave a parent ion at m/z 460.3552 in the EIHRMS from which a molecular formula of C29H48O4 was obtained with acceptable accuracy (AM= -0.6 mmu) employing standard mass spectral algorithms using the standard organic elements (there being no chemical or spectroscopic evidence for halogens, sulfates, phosphates etc.). The  NMR spectrum of (2)  (Table 1, Figure 4) showed resolved signals for all twenty-nine cabon atoms and APT/HMQC (Figures 5,6,7) assignments showed that forty-six of the forty-eight hydrogens were directly attached to carbons (6 X C; 7 X CH; 9 X CH2; 7 X CH3). From the obvious features of the C 1 3  NMR spectrum, it was possible to make the following assignments with a high degree of confidence: a resonance at 6 208.4 (C4) was assigned to a saturated ketone (confirmed by the presence of a strong absorbance at 1707 cm"* in the IR spectrum, (see Fig. 9), typical of a carbonyl stretch for a saturated ketone) and two resonances at 8 125.6 (C24) and 130.9 (C25) 3  were assigned to a trisubstituted olefin.  The absence of further evidence in the  NMR  spectrum for sites of unsaturation indicated that the four sites of unsaturation still unaccounted for were present as rings. Furthermore, a strong OH stretching band in the IR spectrum at 3427 cm'l  15  Fig. 6: HMQC Spectrum of lovenone (2) [500 MHz, C6D6]  16  Fig. 7: Expanded Upfield Region of HMQC Spectrum of lovenone (2) [500 MHz, C DJ t  17  18  1 9  Table 1: *H, C , COSY, HMBC, and nOe NMR Data for lovenone (2) 1 3  c# 1 1' 2 2' 3 3' 4 5 6 7 7' 8 9 10 lleq llax 12  8 C(ppm)  e^ppm)  COSY  HMBC  (125MHz)  (500 MHz)  (500MHz)  (500 MHz)  24.3  1.85 1.45 1.6 1.36 3.83, m 3.53  Hl',2.10 HI,10 H U ^ ' H2,3,3' H2.2',3' H2,2',3  H10  13  30.6 61.1 208.4 68.2 57.4 22.2 44.3 34.6 33.5 40.2 75.7  13  49.7  14  47.6  2.73 d,J=6Hz 2.05 dd, J= 16,7.5 Hz 1.51 dd,J=16,6Hz 1.53 d, J=7.5Hz  H7' H8.7' H6,7 H7  3.67 dd,J=10,2Hz 2.07 dd, J= 16,4 Hz 1.29 dd,J=16,4Hz 3.85 bm  Hl.Hl* Hlla,12 Hlle,12 Hlla,lle  a  nOe  b  (400MHz)  H10  H6.28 H7.10 H7,8 H6.8 H6,7,lle,19,21 H8,10,12,19 Hl,8,l 1,19 H10,12,19 Hlle,17,18  H7',28,29  H29  Hlla,lle,18, 21  Hlle,12,16a,17, 18,29 H7a,8,12,17,18, 29  H16.16' 36.1 0.99 m 15 0.91 m 15' H15,16',17 1.90 m 27.6 16 1.20 H15.16.17 16' H16,16',20 2.03 m 17 44.8 H17 17.4 0.55 s H8,lla,12,15 18 H8.10 0.98 s Hl,7,8 26.8 19 H17,21,22 H17.21 1.39 m 35.6 20 0.93 d,J=6.5Hz H20 H18 17.7 21 H23.24 H20,22',23 1.55 m 36.7 22 H22.23 1.18 m 22' H24 H22,22',23,24 2.18 m 25.4 23 H22.23.24 1.98 m 23" H23.26.27 H23.26.27 5.28 t,J=7Hz H26 24 125.6 H23,26,27 130.9 25 H24.27 1.71 s 26 25.9 H24.26 1.62 s 17.8 27 H6 1.83 s 24.0 28 H8-.15 H6,10,17 1.18 s 21.3 29 a = Proton resonances correlated to carbon resonance in 8 column, b = Proton resonances showing nOe's when resonance in 8 H column was irradiated. 1  20 strongly suggested that the two exchangeable protons as yet unaccounted for in the  NMR  spectrum belonged to alcohol functionalities.  4  The first fragment to emerge from an analysis of the 2D NMR data was the highly substituted cyclohexane ring C of the final structure (see Fig. 10). The strong 2 and 3 bond HMBC correlations to the methyl groups provided the entry point into the solution of this fragment. Since the three *H NMR methyl resonances in question (8 0.55, 0.98 and 1.18) were all singlets, it was postulated  a priori  that each was bonded directly to a quaternary carbon.  Furthermore, since both methyls at 8 0.55 (Me 18) and 8 1.18 (Me 29) showed HMBC correlations ( see Fig. 13) into two quaternary carbons at 8 47.6 (C14) and 8 49.7 (C13), it followed that each of these quaternary carbons bore one of the methyls and were also bonded to one another. At this stage it was impossible to argue stringently as to which quaternary carbon bore which methyl,however this assignment became possible via HMBC correlations from ring B as discussed later. Figure 10 demonstrates pictorially this argument, and those subsequent for this fragment. This fragment was further elucidated as follows: the methyl at 8 1.18 (Me29) showed an HMBC correlation into a methine carbon at 8 44.3 (C8), which in turn also showed an HMBC correlation from the methyl at 8 0.98 (Me 19). This methyl singlet at 8 0.98 (Me 19) correlated into a carbon at 8 26.8 (C19) in the HMQC spectrum. This methyl carbon was bonded directly to a quaternary carbon at 8 34.6, as shown by the HMBC correlation from Mel9 (8 0.98) into C9 (8 34.6). From this it neccessarily followed that the two methyl-bearing quaternary carbons at 8 47.6 (CI4) and 34.6 (C9) were contiguous with the methine carbon at 8 44.3 (C8), as shown in Fig. 10 . In other words, since both methyls shared an HMBC correlation in common to the methine carbon at 8 44.3 (C8), this carbon must lie between the two methyl-bearing quaternaries (8 47.6 and 34.6) as the only possible arrangement that could give rise to these correlations. Having thus established the contiguity of the carbons at 8 49.7 (CI3), 47.6 (CI4), 44.3 (C8), and 34.6 (C9), it remained to "close the ring" by the following argument. The methyl resonance at 8 0.98 (Me 19) showed an HMBC correlation into a methylene carbon resonance at 8 40.2 (CI 1), which in its turn was bonded to two hydrogens at 8 2.07 (HI l ) and 1.29 (HI l ) in the eq  ax  1  2  ' H NMR spectrum, as shown by HMQC.  1  These two proton resonances were doublets of  doublets, and as well as coupling to each other, each showed only one further correlation in the COSY ( see Figs. 11 and 12) spectrum to a downfield proton resonance at 8 3.85 (H12).  Fig. 10: Fragment Describing Ring C of lovenone (2)  This proton was itself directly bonded to a methine carbon at 8 75.7 (CI2); the chemical shift of both the carbon and proton strongly suggested that this was the site of attachment of one of the hydroxyl groups. The carbinol methine at 8 75.7 (C12) showed an HMBC correlation to the methyl at 8 0.55 (Me 18), which has already been shown to be bonded to the quaternary carbon at 8 49.7 (CI3), and thus we have arrived back at where we started. To summarize thus far, it has been possible to prove, using the singlet methyls at 8 0.55 (Me 18), 1.18 (Me29) and 0.98 (Me 19) as HMBC "antennae", and the isolated spin system at 8 2.07 (HI leq)/1.29 (HI lax)/3.85 (H12) in the COSY spectrum, that the carbon resonances at 8 49.7 (CI3), 47.6 (C14), 44.3 (C8), 34.6 (C9), 40.2 (CI 1), 75.7 (CI2) are contiguous in the order listed, which will form ring C of the final structure (2). This is shown graphically in Figure 10 and can be confirmed by analysis of all relevant 2D spectra shown in Figs. 6 to 13, and summarized in Table 1.  o  (2)  (ppm)  o  0  1  H6  Q H6/H7e  0  H37H2.2'  H3'  a  C3-H10/H1  H10  HlZ^HHe  H12  H24/H26 H24/H23 !3 .H24/H27  . 1.  tfH24 "T-T-T T  T J - T T "T I 1  5  H12/H1U H3/H2^'  I I  I  I ' I ' f ' V T T I "1 1 1-  4  T I  )  3  [ "V I* 'I  I  I I  t" 1 T " |  2  T T T I  T'T T T T J  1™t"1  TT—T"  1  Fig. 11: COSY Spectrum of lovenone (2) [500 MHz, QDJ  23  '  1  1  (ppm)  1  1  2.0  1  1  1.8  1  1  1.6  1  1  1.4  1  1  1.2  1  1  1.0  1  r  -  0.8  Fig. 12: Expanded Upfield Region of COSY Spectrum of lovenone (2) [500 MHz, QDJ  24  25 COSY and HMBC correlations also proved critical in elucidating the highly functionalized ring B in lovenone (2). The doublet methine resonance at 8 1.53 (H8) attached to the carbon at 8 44.3 (C8) was correlated in the COSY spectrum to a proton resonance at 8 2.05 (H7), which was itself bonded to the methylene carbon at 8 22.2 (C7) as shown by HMQC. The geminal coupling partner of this proton resonance at 8 1.51 (C7') was in turn coupled to a downfield resonance at 8 2.73 (H6), which proved to be the proton on an oxygen-bearing methine carbon at 8 57.4 (C6). The connectivities for this fragment are shown graphically in Fig. 14.  At this point in the  structural proof, a number of possibilities arose from the remaining connectivity data which had to be carefully thought through before an unambiguous pattern emerged. To follow this logic, we must leave this methine carbon at 8 57.4 (C6), (which we know to be bonded to a proton, an oxygen, the methylene carbon at 8 22.2 (C7) and some as yet unknown group) and approach this site from the other direction.  0.98 1.85 1.45  H  1.85 1.45  H  H H  24.0  1.83  J  C  - HMBC  c-  COSY  Fig. 14: Fragment Describing Ring B of lovenone (2)  An HMBC correlation from a  methine resonance at 8 33.5 (CIO) to the methyl *H  resonance at 8 0.98 (CI9) identified this as the final substituent on the quaternary carbon at 8 34.6 (C9) of ring B. This methine carbon at 8 33.5 (CIO) was directly attached to a proton at 8 3.67 (H10)- a highly deshielded resonance given the carbon's chemical shift. This proton resonance  26  Fig. 15: Expanded Region of HMBC Spectrum of lovenone (2) Showing Correlationsfromthe 'H at 8 3.67 [500 MHz, QDJ  27  Fig. 17: Expanded Downfield Region of HMBC Spectrum of lovenone (2) [500 MHz, C D ] S  6  29 showed HMBC correlations from two carbons: a methylene at 8 24.3 (CI) and a quaternary at 8 68.2 (C5).  The resonance at 8 68.2 (C5) showed only one other HMBC correlation into one of  the protons, 8 2.05 (H7) on the methylene carbon at 8 22.2 (C7). Since this must clearly be a 3 bond HMBC correlation, it follows then that this carbon at 8 68.2 (C5) is the remaining substituent on the methine carbon at 8 57.4 (C6), and that the three carbons at 8 68.2 (C5), 57.4 (C6), and 22.2 (C7) are contiguous. Thus it is possible to trace a pattern of six contiguous carbons- 8 34.6 (C9), 44.3 (C8), 22.2 (C7), 57.4 (C6), 68.2 (C5), and 33.5 (CIO)- which make up the highly functionalized ring B of lovenone (2). See Figures 15 through 17 for pertinent expansions of the HMBC spectrum. There still remains, however, the question of the functionality on this ring. The carbon chemical shifts of the two carbons at 8 68.2 (C5) and 57.4 (C6) strongly suggested that oxygens were directly bonded to these positions. The absence of hydroxyls at these positions was proved by running the COSY spectrum in dg-DMSO. This experiment unambiguously identified the positions of the two hydroxyls at sites other than these two carbons. The only remaining possibility was an epoxide ring, which was confirmed by the chemical shifts of the two carbons, and was neccessary also to provide one of the remaining sites of unsaturation.  2.05  Fig. 18: Fragment Describing Ring D of lovenone (2)  31  Fig. 21: Expanded Region of COSY Spectrum of lovenone (2) in aVDMSO [500 MHz]  32 The remaining substituents on this ring were the methylene at 8 24.3 (CI) attached to the carbon at 8 33.5 (CIO), as previously mentioned, and a substituent on the epoxide carbon at 8 68.2 (C5). This substituent was identified as an acetyl fragment (8 208.4 (C4), 24.0 (Me28)) based upon the HMBC from the carbonyl carbon to the epoxide methine proton at 8 2.73 (H6) (Fig. 14). The l H chemical shift of the methyl group (8 1.83, Me28) was completely consistent with such an assignment, and the methyl protons showed an HMBC correlation into the carbonyl resonance at 8 208.4 (C4).  From the methylene at 8 24.3 (CI) attached to the carbon at 8 33.5 (CIO), it was  possible to trace, through a series of COSY correlations, a three carbon side chain ending in a carbinol methylene (8 61.1, C3) (CI to C3: see Table 1). The^H NMR spectrum of lovenone (2) (see Figs. 19 to 21) recorded in dg-DMSO contained exchangeable resonances at 8 4.31 (t, J= 5 Hz) and 4.33 (d, J= 3 Hz) that showed COSY correlations to these terminal methylene protons at 8 3.31 (H3) and the carbinol methine at 8 3.88 (HI2), respectively, confirming the presence of alcohol functionalities at these positions. It should be noted here that the points of attachment of the two methyl groups on ring C at 8 0.55 (Me 18) and 1.18 (Me29) were determined to be the quaternary carbons at 849.7 (C13) and 47.6 (CI4), respectively (Fig. 10). The HMBC correlation from the quaternary carbon at 8 47.6 (CI4) (see Fig. 16) into the methylene proton at 8 2.05 (H7) (attached to the carbon at 8 22.2 (C7)) of ring B proved that the carbons at 8 22.2 (C7), 44.3 (C8) and 47.6 (C14) must be contiguous, and hence the carbon at 8 47.6 (CI4) must bear the methyl group at 8 21.3 (C29)(8 1.18, Me29). The remaining substituent(s) on ring C of lovenone (2) was determined to be a fivemembered ring from COSY and HMBC data (see Figs. 11,12,13 and 23). An HMBC correlation from a methylene carbon at 8 36.1 (CI5) into the methyl protons at 8 1.18 (Me29) proved that this was the final substituent on the quaternary carbon at 8 47.6 (C14) (see Fig. 18). Similarly, an HMBC correlation from a methine carbon at 8 44.8 (CI7) into the methyl protons at 8 0.55 (Me 18) showed that this carbon was the final substituent on the quaternary carbon at 8 49.7 (CI3). COSY data showed that the methylene protons (8 0.99 (HI5), 0.91 (HI5')) on the carbon at 8 36.1 (CI5) were correlated into two neighboring methylene protons at 8 1.90 (HI6) and 1.20  33  (H16'), which were directly attached to a carbon at 8 27.6 (CI6), as shown by HMQC. One of these methylene protons (8 1.20, HI6') showed a COSY correlation into the methine proton at 8 2.03 (HI7) (attached to the carbon at 8 44.8, CI7), which we have already shown to be the final substituent on the quaternary carbon at 8 49.7 (CI3) of ring C. Thus it has been possible to trace the connectivities from the carbons at 8 47.6 (C14), 36.1 (CI5), 27.6 (C16), 44.8 (C17), and 49.7 (C13) which make up the cyclopentane ring D of lovenone (2) as shown in Fig. 18.  Fig. 22: Fragment Describing Side-Chain of lovenone (2)  The remaining CgHi5 substituent was shown to be the standard lanosterol side-chain by the following arguments (see Fig. 22).  The methine proton at 8 2.03 (HI7) on the carbon at 8  44.8 (CI7) in ring D of lovenone (2) showed a further COSY correlation into a methine proton at 8 1.39 (H20), which was itself attached to a carbon at 8 35.6 (C20), as shown by HMQC. This methine proton showed a COSY correlation into the methyl doublet at 8 0.93 (Me21)(8 17.7, C21; HMQC) and into a proton at 8 1.55 (H22). This resonance proved to be one of two methylene protons (the other being at 8 1.18, H22') on a carbon at 8 36.7 (C22), as shown by COSY and HMQC. The methylene proton at 8 1.55 (H22) at this position showed further COSY correlations into two allylic methylene protons at 8 2.18 (H23) and 1.98 (H23'), which were themselves attached to a carbon at 8 25.4 (C23, HMQC).  From this allylic methylene it was relatively  straightforward to trace the connectivities from the COSY and HMBC data into the trisubstituted  34  21  Me26  22  24  Me27  Fig 23: Expanded Region ('H 6 1.5. to 1.8) of HMBC Spectrum of lovenone (2) [SOOMHz.QDJ  35 double bond which made up the terminal group of the side-chain (see Table 1, and Fig. 22). The assignments of the two allylic methyls were made distinguishable by the strong nOe into the methyl  resonance at 5 1.71 (Me 26) when the olefinic methine at 6 5.28 (H24) was irradiated, proving that this methyl was in fact cis- to the aforementioned olefinic methine. With the gross stucture now established, further difference nOe experiments provided invaluable insight into the relative configuration of lovenone (2) (see Fig. 24). All difference nOes were demonstrated in both directions, wherever possible (see Fig. 25). Irradiation of the methyl resonance at 8 0.55 (CI8) induced nOes in the H8 (8 1.53) and HI l  a x  (8 1.29)  resonances, demonstrating that the the cyclohexane ring C was in a chair conformation, with Me 18 and H8 being in a 1,3- diaxial relation to one another. Also, the Mel8 irradiation induced an nOe in H12 (8 3.85) indicating that H12 was equatorial, and thus that the alcohol at this position was axial. This was in good agreement with the observed 4 Hz coupling between H12 and HI l  a x  Irradiation of the methyl resonance at 8 0.98 (Me 19) induced nOes into the resonances at 8 2.05 (H7), 1.53 (H8) and 1.85 (HI) which demonstrated that the B/C ring junction in lovenone (2) was cis. Irradiation of the methyl resonance at 8 1.18 (Me29) gave some unexpected insight into the conformation of ring B in lovenone (2). Irradiation at this position induced nOes in the protons at  3 (PDfrO  Me 26 H24irr.  H10  H7 HI  Fig. 25: Selected Difference nOe Spectra of lovenone (2) [400 MHz, C D ] 6  6  37 5 2.73 (H6), 3.67 (H10) and 2.03 (H17). The Me29 to H10 nOe showed that CIO was attached to C9 in an axial orientation relative to the cyclohexane ring, and confirmed that the Me 19 and CI to C3 appendages were cis.  Similarly, the Me29 to H6 nOe indicated that the epoxide  functionality was beta to ring B, as shown in Fig.24. Irradiation of the acetyl methyl resonance at 8 1.83 (Me28) induced a large nOe in H6, confirming that their relationship was cis. This strong nOe requires that the acetyl moiety adopt a conformation such that the methyl is close to the epoxide methine; when this occurs, Dreiding models show that the deshielding region of the acetyl carbonyl bond encompasses H10, thus accounting for its anomalously large chemical shift (8 3.67).  The aforementioned nOe from Me29 to H17 suggested that the side-chain at C17 was  beta; this was confirmed by the nOe induced in Me 18 (8 0.55) when Me21 (8 0.93) was irradiated.  IID. Biological Activity  Lovenone (2) was tested for cytotoxic activity against a number of solid tumor cell lines. It 5  showed modest activity against human ovarian carcinoma HEY (EC50= 11 ^.g/mL) and human glioblastoma/astrocytoma (EC50= 11 Jxg/mL) but was inactive against A549 (EC50 = > 25 jxg/ mL).  HE. Origin and Proposed Biogenesis  (2)  (3)  Lovenone (2) represents the first triterpenoid to be isolated from a nudibranch, and only the second example of a triterpenoid from a marine mollusc. Limatulone (3), from the limpet Lottia limatula, was the first and only other example of a triterpenoid isolated from a marine mollusc.  6  38 While the vast majority of terpenoids isolated from nudibranchs have been shown to be dietary in origin, there existed only one report of a triterpenoid having been isolated from a bryozoan. 7  8  Furthermore, re-collection of a number of kelp species from the areas where A.  were  loveni  collected and extraction of the bryozoans growing on these kelp samples (mostly  Membranipora  species) failed to yield even a trace amount of lovenone (2) or related compounds. Given the compound's toxicity, it seems reasonable to postulate that lovenone (2) may serve as a chemical defense for the shell-less mollusc, though this remains to be proved in an ecologically relevant 9  assay.  10  If we accept a defensive role for lovenone (2), then it seems likely that A.  sequesters the metabolite from a dietary source or biosynthesizes the compound de  either  loveni  This  novo.  latter possibility having proven difficult to determine experimentally , we must conclude with 11  equivocation and state merely that the origin of lovenone (2) remains to be determined, though it seems a likely candidate for biosynthetic studies. Regardless of which organism is in fact producing the compound, it is useful to postulate a biogenetic scheme for the biosynthesis of lovenone (2).  It has been well-demonstrated that 12  triterpenoids, and their cousins the steroids, are formed by an epoxide ring-opening initiated cyclization of squalene-2,3-epoxide (4) to yield a carbocation intermediate (5) that is the direct precursor to lanosterol (this reaction and all subsequent are shown in Scheme 1). Taking this carbocation as our starting point, it seems reasonable to postulate a series of suprafacial hydride and methyl shifts such that the intermediate (6) is generated with the appropriate relative stereochemistry for lovenone (2).  Oxidation of the alcohol functionality at C3 to a ketone,  followed by the biosynthetic equivalent of the Baeyer-Villager reaction would lactonize ring A in 13  the direction of C4, as shown. Ring-opening of the lactone by the removal of a proton at one of the methyls would provide intermediate (7).  The biological equivalent of an ozonolysis at the 14  double bond at C4, reduction of the carboxylic acid at C3, oxidation at CI2 and epoxidation of the double bond at C6/7 would then lead directly to lovenone (2), as shown.  Scheme 1: Proposed Biogenesis of lovenone (2)  40 II. F Endnotes: Chapter II. Adalaria  loveni  1. ) All taxonomic data in this section are taken from: Thompson, T.E.; Brown, G.H., Biology of Opisthobranch Molluscs. Vol. II, The Ray Society, 1984, pp. 49-50. 2. ) I am indebted to M. LeBlanc arid J. Kubanek for the collection of A. loveni; in addition, without the assistance of Prof. Ulf Bamstedt, University of Bergen, and the staff of the Institute for Fisheries and Marine Biology, Bergen, Norway, this collection would not have been possible. 3. ) Silverstein, R.M.; Bassler, G.C.; Morrill, T.C., Spectrometric Identification of Organic Compounds. 4th ed., John Wiley & Sons, New York, 1981, pp. 117-19. 4. )  ibid. pp. 112-15.  5. ) Bioassays were performed by Sarah Halleran in the laboratory of Dr. T. Allen, Dept. of Pharmacology, University of Alberta. 6. )  Albizati, K.F.; Pawlik, J.R.; Faulkner, D.J., J. Org. Chem., 1985, 50, 3428-30.  7. ) Karuso, P., "Chemical Ecology of the Nudibranchs", in Bioorganic Marine Chemistry. Vol. I, Springer-Verlag, Berlin, 1987, pp. 31-60. 8. )  Hadjieva, P.; Popov, S.; Budevska, B.; Dyulgerov, A.; Andreev, S., Z. Naturforsch,  1987,42 (C), 1019-22.  9. ) Faulkner, D.J., in Ecological Roles of Marine Natural Products, ed. Paul, V.J., Cornell University Press, Ithaca, New York, 1992, Ch. 4. 10. ) c.f. Harrell, CD.; Fenical W.; Greene, C.H., Mar. Ecol. Prog. Ser., 1988, 49, 287-94; Pawlik, J.R.; Fenical, W., Mar. Ecol. Prog. Ser., 1989, 52, 95-98; and Fenical, W.; Pawlik,  J.R., Mar. Ecol. Prog. Ser., 1992, 75, 1-8.  11. ) Garson, M.J., Nat. Prod. Rep., 1989, 6, 143-70; and Garson, M.J., Chem. Rev., 1993, 93, 1699-1733. 12. )  Mann, J., Secondary Metabolism. Clarendon Press, Oxford, 1978, pp. 120-21.  13. ) 93.  see for example: Watanabe, C.M.H.; Townsend, C.A., J. Org. Chem., 1996, 61, 1990-  14.)  Mann, J., Secondary Metabolism. Clarendon Press, Oxford, 1978.  III. Isolation and Structure Determination of Limaciamine, A Diacylguanidine from the North Sea Dorid Nudibranch, Limacia clavigera  Fig. 26: Color Plate of Limacia clavigera  42  I I I A . Taxonomy *  Limacia clavigera (Miiller, 1776), previously known as Doris clavigera (Miiller), Tergipes pulcher (Johnston, 1834), Euplocamus plumosus (Thompson, 1840), and Triopa lucida  (Stimpson, 1855), is a shallow water dwelling nudibranch, typically 1.5 to 3 cm in length, common to the Western coasts of Europe. Its coloring is white with yellow or orange tips; its rhinopores are yellow-tipped and the dorsal edge bears a number of highly pigmented ceratal processes, making this nudibranch easy to spot and identify in the field. Identification of this nudibranch was based upon dissection and inspection of anatomical features such as radular morphology and the reproductive system, performed by Dr. Sandra Millen, at the University of British Columbia. Also a bryozoan feeder, L. clavigera has been reported to feed on Callopora dumerilii, Cryptosula pallasiana, Electra pillosa, Membranipora membranacea, Porella concinna, Schizoporella unicornis, and Umbonula littoralis. L. clavigera is typically found at depths up to 20  m (though rarely in the intertidal zone), with reports of collections at 80 metres. One report exists for L. clavigera spawning in June, though this may only apply to Britain. The distribution of L. clavigera is extensive; collections have been made beyond the Arctic Circle as far north along the Norwegian coast as Finmarken. It is one of the commonest nudibranchs of the British Isles, and numerous records exist from the French Atlantic coast and the northern Spanish coast. Reports of L. clavigera in the Mediterranean are rare, though it has been spotted off Spain, Marseilles, and in the Bay of Naples. Extending further south from here, records are rather sketchy; there exists one sole report from the Moroccan coast. While L. clavigera is common to the waters off the Cape of Good Hope, there is no good corroboration for it existing in the tropical waters off the west coast of Africa.  43  IIIB.  Collection and Isolation  Specimens of L. clavigera (50 animals) were collected by hand using SCUBA at depths of 10 metres off Tosoy Island, near Bergen, in the North Sea7- Specimens were immediately immersed in methanol (250 mL) at the surface and returned to Vancouver for further study. The methanol extract was decanted, filtered and reduced in vacuo. The animals were subsequently exhaustively extracted with two 250 mL portions of methanol, and two 250 mL portions of 1:1 MeOH: CH2CI2.  All extracts were filtered, combined, and reduced in vacuo to yield an aqueous  suspension. This was diluted up to 500 mL with distilled water and exhaustively extracted with 4 X 500 mL portions of EtOAc. All EtOAc layers were dried over MgS04,filtered,combined, and reduced in vacuo to yield a yellow oil. These EtOAc soluble materials were fractionated by flash silica gel chromatography using a step gradient from 100% hexanes to 100% EtOAc, to yield a fraction eluting with 4:1 hexanes: EtOAc containing mostly limaciamine (8).  This fraction was  further purified on normal phase HPLC (eluent: 15% EtOAc/ hexane) to yield 4.3 mg of pure limaciamine (8) as a colorless glass.  O  N• H  O  11  2  L  6' 8  O  8'  (8)  NH  (9)  2  O  45  46  s  48  N H  O II  I  O  2  *•  H  (8)  (ppm) C8  "10  C6  H5  C6:Me6  r20  C5 C5:H5  C7 C4  C7: H7a  j  €3  "25  H7b  C3: H3a  "30  C4:H4  H3b  "35 "40 "45  C2  J -i  1  (ppm)  :50  C2/H2  «C> 1  p  2.0  i  1.8  1  -i  r  1.6  1.4  1.2  1.0  0.8  1  r  0.6  Fig. 31: HMQC Spectrum of Umaciamine (8) [500 MHz, CDCIJ  Fig. 32: COSY Spectrum of limaciamine (8) [500MHz, CDC1,]  50  (8) H5 H3a H7a., H7b H3b  H2  H4  C8  (ppm)  C6  "15 "20  C5 C7  j  "25  -aaate  C4  "30 C3  "35 "40 :45 C2  "50 ~i  (ppm)  2.0  1.8  r  1.6  1.4  1.2  Fig. 33: Selected Region of HMBC Spectrum of liinaciamine (8) [500 MHz, CDCl]  51  0  N H  1  O  2  1  1  (8) Me6  Me8  (ppm)  "10 "15 "20  C5 _C7_  r25  C4  "30 "35 "40 :45 C2  1  :50 i  i  i . |  (ppm)  i  i  i  i  i  i  i  i  i  i  i  i  i  i  i  1  1  '  '  i  1  ' '  1  i '  '  '  1  i  1  '  '  1  i  1  1  1.00 0.95 0.90 0.85 0.80 0.75 0.70 0.65  Fig. 34: Methyl Region of HMBC Spectrum of Umaciamine (8) [500 MHz, CDClj]  "55  52  IIIC.  Structure Determination  Limaciamine (8) gave a parent ion in the HREJMS at m/z= 311.25789, requiring a molecular formula of C n H ^ C ^ (calculated= 311.25725, Ammu= -0.6). The *H NMR spectrum (Fig. 27) of (8) contained only aliphatic resonances, while the C NMR spectrum (Fig. 30) showed 13  only 8 signals, suggesting a symmetrical molecule. The presence of an intense absorbance at 1700 c m in the IR spectrum (Fig. 28) of (8) indicated that the molecule contained an amide 1  carbonyl,3 though due to a dynamic exchange process, this signal was never observed in the C 13  NMR spectrum, even when run at a number of higher and lower temperatures.  A broad  absorbance at 3340 cm in the IR spectrum suggested the presence of an amine or hydroxy 1 -1  functionality in (8).3 The complete assignment of the acyl portion of limaciamine (8) was based upon 2D NMR data (see Table 2). A methine resonance at 8 2.17 (H2), attached to a carbon at 8 51.5 (C2, from HMQC, Fig. 31), was at the appropriate chemical shift for a proton deshielded by a carbonyl, by analogy to triophamine (9).4 This methine resonance showed extensive COSY and HMBC (see Figs. 32 to 34) correlations into two different neighboring methylene groups. One such set of methylene resonances at 8 1.66 and 1.51 (H7)(attached to a carbon at 8 25.2; C7, HMQC) showed COSY correlations into a methyl triplet at 8 0.89 (Me8) (8 11.9, C8), proving that this methylene was part of a terminal ethyl branch off of the methine alpha to the carbonyl. The other set of diastereotopic methylene resonances at 8 1.62 and 1.41 (H3) showed COSY correlations into a neighboring methylene resonance at 8 1.26 (H4) ( attached to a carbon at 8 29.6; C4, HMQC). It was also relatively straightforward to trace further connectivity from this methylene into another methylene resonance at 8 1.31 (H5)(attached to a carbon at 8 22.7; C5, HMQC) from the COSY data. HMBC and COSY showed unambiguously that this methylene at 8 1.31 (H5) was attached to a terminal methyl triplet at 8 0.86 (Me 6) (8 13.9, C6, HMQC), thus proving the presence of an n-butyl group attached to the methine alpha to the carbonyl.  53 Table 2: H , C , COSY, and HMBC NMR Data for limaciamine (8) 1  c#  1 3  8 C(ppm)  8H  COSY  HMBC  (125MHz)  (500MHz)  (400MHz)  (500MHz)  2,2' 3, 3'  51.5 31.9  H3. 7 H2.4  H3.7 H2.7  4, 4' 5, 5' 6, 6' 7, 7"  29.6 22.7 13.9 25.2  2.17 1.62 1.41 1.26 1.31 0.86 1.66 1.51 0.89  H3, 5 H4, 6 H5 H2 8  H3,6 H3,6 H5 H2.3.8  13  ]  i, r  3  H7,2  H7 11.86 8, 8' 158.9 9 a= Proton resonances correlated to carbon resonance in 813C column.  The remaining C resonance at 5 158.9 was strongly indicative of a guanidyl carbon. 1 3  This fact, in combination with IR and mass spectral fragmentation data, provided the first insight that limaciamine was an analogue of the only previously known naturally-occuring diacylguanidine, triophamine (9)4 High resolution measurements of the fragment ions in the electron impact mass spectrum (see Fig. 29) of limaciamine (8), gave convincing proof for a symmetric structure around a guanidine nucleus, as follows. Scheme 2 outlines the EIMS fragmentation pattern observed for limaciamine (8). The peak at m/z= 57 (EIHRMS = 57.07055, observed; calculated m/z= 57.07043, Ammu= -0.1) could only be rationalized as an n-butyryl ion, resulting from fragmentation pathway (c) as shown. Furthermore, the large peak at m/z= 86 (EIHRMS= 86.03502, observed; calculated m/z= 86.03544, Ammu= 0.4) gave a formula of C2H4ON3 for this fragment, resulting from the pathway a,b shown in Scheme 2. The fragmentation pattern a,b leading to the ion at m/z= 86 was confirmed by the large peaks at m/z= 212 (EIHRMS= 212.14027, observed; calculated m/z= 212.13991, Ammu= -0.4) giving a best-fit formula of C H O N 3 , and m/z= 127 (EIHRMS= 10  18  2  127.11178, observed; calculated m/z= 127.11229, Ammu= 0.5) giving a molecular formula of CgHi 0. The guanidyl nucleus was further confirmed by a high resolution measurement on the 5  peak at m/z= 60 (EIHRMS= 60. 05656, observed; calculated m/z= 60.05617, Ammu= -0.4)  54 representative of the guanidyl cation itself, C H g ^ . It should be noted that all the fragments in Scheme 2 not discussed in detail here were also confirmed by high resolution measurements. Thus, the mass spectral data, in combination with the NMR evidence and by analogy with triophamine (9), present a strong case for the structure of limaciamine (8) as described.  HID.  Origin and Biogensis  Since triophamine (9) became the subject of an experimental biosynthetic investigation, the discussion as to the origin and biogenesis of limaciamine (8) will be reserved for a future chapter.  55  NH  2  NH  Q  2  0  'C, H <W A  s  C  1 3  H  2 5 ° 2  3 *  N  m/z = 282.21844 calc. = 282.21814 AM = -0.3  m/z = 255.19441 calc. = 255.19467 AM = 0.3  NH, O  O. b  N  A JU NH  C H 7  C  I 0  H  1 8 ° 2  N  13  3 *  m/z = 99.11785 calc. = 99.11738 AM = -0.5  m/z = 212.14027 calc. = 212.08826 AM = -0.4 NH,  O  NH  C H, 0 8  C^ONj*  +  5  m/z= 127.11178 calc. = 127.11229 AM = 0.5  m/z = 86.03502 calc. - 86.03544 AM = 0.4  C H * 4  9  m/z = 57.07055 calc. = 57.07043 AM = -0.1  Scheme 2: Proposed EIMS Fragmentation Pattern for limaciamine (8)  56 III.D Endnotes: Chapter III. Limacia clavigera 1. ) All taxonomic details were taken from: a) Thompson, T.E.; Brown, G.H., Biology of Opisthobranch Molluscs. Vol. II, The Ray Society, 1984, pp. 74-75; and b) Cattaneo-Vietti, R.; Chemello, R.; Giannuzzi-Savelli, R.; Perrone, A.; eds., Atlas of Mediterranean Nudibranchs. Editrice La Conchiglia, Roma, 1990, p. 113. 2. ) I am indebted to M. LeBlanc and J. Kubanek for the collection of L. clavigera; in addition, without the assistance of Prof. Ulf Bamstedt, University of Bergen, and the staff of the Institute for Fisheries and Marine Biology, Bergen, Norway, this collection would not have been possible. 3. ) Silverstein, R.M.; Bassler, G.C.; Morrill, T.C., Spectrometric Identification of Organic Compounds. 4th ed., John Wiley & Sons, New York, 1981, pp. 124-125. 4.) Gustafson, K.; Andersen, R.J., J. Org. Chem., 1992,47, 2167-2169.  IV.  Biosynthetic Studies of Isoprenoid Secondary Metabolites from Dorid Nudibranchs Using Stable Isotopes  IV.I.A. Introduction  The marine environment has yielded a large variety of new terpenoid skeletons in recent years.  1  Since these oftentimes unique structural types have been reviewed elsewhere, no effort 2  here will be taken to catalogue all such examples. A cursory glance at such reviews, however, reveals a very active interest in isolating and determining the structures of such compounds. In addition, the biological activities of some marine terpenoids have aroused interest in terms of both the implications of toxicity to the chemical ecology of the organisms in question, as well as active interest into the potential pharmacological uses of such compounds.  (10)  (15)  For example, the cembrane diterpene sarcophytol A (10) isolated from the soft coral Sarcophyton glaucum} inhibited carcinogenesis in various organs, including the colon, liver, breast and thymus.  4  Also, the sesterterpene mycaperoxide A (11), isolated from the sponge  Mycale sp. has shown cytostatic activity in solid tumor assays of renal and ovarian carcinomas.  6  5  In addition to antineoplastic activities, the pseudopterosins (12), diterpenoid glycosides isolated from the Caribbean sea whip Pseudopterogorgia elisabethae, possess potent anti-inflammatory and analgesic properties that exceed those of existing drugs. These are the most recent examples of 7  highly bioactive marine terpenoids that have shown promise as leads for the discovery of new drugs. The discovery of these new structural types has spurred interest into both the total  syntheses of such structures and into structure/activity relation studies in order to improve or determine a mechanism for such activity. For example, the antimicrobial activity and structural novelty of the sesterterpenoid palauolide (13), isolated from a number of sponges found at Palau, Western Caroline Islands, prompted a recent investigation into its total synthesis. Furthermore, 8  9  the anti-inflammatory activity of the sesterterpenoid manoalide (14), isolated from the sponge Luffariella variabilis, has been the subject of an intensive structure activity study. 10  11  Returning  to anticancer therapeutics, the aforementioned carcinogenesis inhibitor, sarcophytol A (10), has been the subject of a similar structure/activity study from which the much simpler and more potent analogue, canventol (15) emerged.  12  With a much lower overall cytotoxicity than sarcophytol A  (10), canventol (15) is currently being promoted as a new cancer chemopreventive agent. O  5 The above examples are in no way intended to suggest that marine terpenoids will exclusively become the anticancer and antiinflammatory drugs of the future. To the contrary, the very thorny issues associated with drug discovery have been largely side-stepped in the above discussion. These examples do, however, illustrate that irrespective of their place in the modern day pharmacopaeia, the remarkable structural diversity of marine terpenoids has generated numerous investigations into their total synthesis and their potential as leads for drug development. An area for potential scholarship that has been largely overlooked involves an investigation into the biosynthesis of structurally novel marine terpenoids.  While authors who present  structural proofs for new terpene skeletons often propose a theoretical biogenetic scheme, there have been only a handful of attempts to back-up such claims with experimental tests of the proposed pathways. The following discussion will attempt to review the problems associated with biosynthetic studies in marine invertebrates, catalogue the successful incorporation studies into marine terpenoids to date, and propose that the unique biology of certain dorid nudibranchs makes them ideal as candidates for biosynthetic feeding studies.  IV.LB. Biosynthetic Studies with Marine Invertebrates: Practical Considerations  In contrast to many microorganisms and higher plants, marine invertebrates often produce secondary metabolites in only trace amounts, making biosynthetic studies difficult.  13  This  problem may be overcome if large numbers of individual organisms can be collected and kept alive for study. This may be feasible for cold-water organisms that are generally present in the environment in large numbers to begin with, but in the tropics it is often difficult to find large numbers of a single organism, nor is it neccessarily desirable to indiscriminately remove large quantities of an organism from an ecologically fragile environment such as a coral reef. Also, it is generally recognized that the rate of de novo synthesis in slow-growing organisms such as  60 sponges or corals is painfully slow, often on the order of months if not years.  14  Clearly, when  simply keeping organisms alive in an aquarium for study can be a challenge, one must therefore expect the "turnover" of secondary metabolites in such studies to be very low. The implications of this are twofold: first, given a limited time-frame, it simply may not be possible to demonstrate de novo  synthesis at all, and secondly, the use of labeled precursors must necessarily be assessed  against a high background level of unlabeled material, with the result that dilution of any label may lead to measurably insignificant levels of incorporation. A further complication that arises in considering biosynthetic studies with marine invertebrates is the strong possibility of symbiotic associations occuring within the organism in question. For example, there are well-documented examples of sponges containing within their tissues large numbers of microalgae and/or bacteria, and similar relationships exist between soft corals and dinoflagellates.  15  Thus, even if de novo  synthesis of a metabolite could be  unambiguously demonstrated in such a system, without effective partitioning of symbionts, there would still remain doubt as to which partner was in fact performing the synthesis. To add further complication, it is not unreasonable to assume, given a potential co-evolution of host and symbiont, that a cooperative mechanism may exist whereby both organisms contribute to biosynthesis. Such a case could only be resolved by rigorous experiments with cell-free extracts such that preparations from the host cells, uncontaminated with symbiont materials, exclusively showed biosynthetic activity with respect to symbiont preparations, or vice-versa.  16  The  limitations of such techniques are many: they are best suited to situations where a reasonable rate of biosynthesis is expected, the disruption of cells may interfere with the spatial arrangement of 17  enzymes responsible for host-symbiont cooperative biosynthesis, and the requirements for specialized apparatus (i.e. refrigerated centrifuges) may preclude the possibility of carrying out these experiments at remote sites where removal of the organism to the laboratory is not feasible.  13  Assuming, then, that a suitable candidate for biosynthetic study may be chosen such that a reasonable life-expectancy for the organism is in agreement with the presumed rate of biosynthesis, the amounts of metabolites obtained at the end of the experiment are large enough for study, and  that complications resulting from symbiont relationships have been ruled out, further experimental details must be considered.  Seasonal variation of secondary metabolite content in marine  invertebrates has often been observed, and therefore every effort must be made to attempt 18  biosynthetic feeding experiments only when there is a good chance that metabolic pathways will be active. A complete understanding of the biology of the organism to be studied is therefore an absolute prerequisite for attempting a biosynthetic study, as has been pointed out by many researchers.  19  The next stage in planning a biosynthetic feeding experiment involves the choice of precursor and the method by which such a precursor will be administered to the animal. The marine environment affords unique challenges in this regard. All marine invertebrates have some mechanism whereby seawater is passed through some portion of the animal's anatomy. This can 20  be a boon for the administration of water-soluble precursors that are dissolved in the aquarium water, though the corollary of this is that precursors may just as easily be washed out of the organism by similar processes. More often than not, injection of a solution of precursor directly into the tissues where biosynthesis is presumed to take place is preferable, since the uptake of nutrients from seawater has been shown to occur against a concentration gradient requiring very low levels of dissolved nutrients.  21  Another alternative is to somehow incorporate labeled  precursor into the organism's natural food source, such as microalgae that have been grown up in labeled media.  13  This may not be possible when the native food source is either unattainable or  difficult to incorporate with label, and although this method has its obvious merits from an ecological point of view, one is still at the mercy of the invertebrate's notoriously finicky appetite. Assuming that a methodology can be developed in order to effectively present the precursor to the tissues of the test organism, there often remains the problem of ensuring that the precursor is able to keep its integrity with the onslaught of digestive enzymes, and is furthermore capable of crossing a cell membrane where it can be taken up by the biosynthetic machinery of the cytoplasm or mitochondrion. With very simple precursors such as acetate or amino acids, the problem of digestion is not generally of great concern, although scrambling or rearrangement of label may  obscure incorporation patterns in the product metabolites, as will be discussed later.  22  The  problem of penetration of the cell membrane has been quite effectively overcome by employing precursors that are iiva form more easily recognized as a biochemical building block, for example, the dibenzylethylenediamine (DBED) salts of acetate or mevalonate, or the N-cysteamine adducts of more advanced precursors.  23  An alternate approach is to encapsulate the precursors in either  slow-release gel capsules or liposomes, in order that their entry into the cell is facilitated.  24  Liposomes may have another advantage for experiments with filter feeders such as sponges or tunicates, in that they may more effectively mimic the bacterial particles that are the natural food source for these organisms.  25  The concentrations at which the precursors are to be used will largely depend upon the kind of label used. The use of radioisotopes has obvious advantages in that the amounts used can be very small, which both more exactly mirrors the concentration of nutrients typically found in seawater and is more economical, since these labeled compounds are often expensive. Perhaps the greatest advantage to using radiolabeled precursors is the enhanced sensitivity of detection of the label in the secondary metabolite being studied. However, use of radioisotopes requires that the product metabolites be rigorously pure to the point where the specific activity of the compound is constant after a number of purifications; the small amounts of metabolite produced by marine invertebrates, as mentioned earlier, often prevents such a stringent analysis. If constant activity cannot be demonstrated, the results become highly suspect, since even a trace amount of contamination from either the labeled precursor itself or a labeled primary metabolite i.e. fat can lead to a significant measure of activity above background.  In addition, while a positive result  with radiolabeled precursors is an effective demonstration of de novo synthesis, it affords no information as to the fate of the label at specific sites within the molecule. Historically, very careful degradative studies of labeled compounds have afforded some information in this regard,  26  but again, the small amounts of compounds typically recovered from marine invertebrates make such procedures difficult at best. For these reasons, the use of stable isotopes such as C , H , 13  2  and N have gained favor, since they provide a wealth of information as to origin of individual 15  bonds in a molecule without the requirement of chemical degradation.  22  On the down side, the  use of stable isotopes requires much higher levels of incorporation for detection; in the case of C, 13  due to its low natural abundance and small gyromagnetic ratio resulting in an overall low sensitivity to detection by NMR, incorporation levels in biosynthetic experiments must be significantly higher than natural abundance. Since precursors labeled with stable isotopes are either very expensive or require some investment of time and effort to synthesize, it is very important that a precursor is chosen with care before a biosynthetic feeding experiment is initiated with a marine invertebrate. Ideally, initial work with radioisotopes should be used to demonstrate de novo synthesis and to work out the protocols by which the highest incorporation levels may be attained. Then, if feasible, experiments with simple precursors labeled with stable isotopes should be attempted in order to gain information as to the fate of individual atoms and to fine tune the methodology such that the precursor is fed at levels high enough to be detected in the final product. Only then should experiments with more advanced precursors be attempted, and only if further mechanistic details may realistically be afforded by such costly experiments  IV.I.C  2 7  Introduction to Isoprenoid Biosynthesis  28  The family of natural products comprising the terpenoids and the steroids have long been exploited by humankind as perfumes, drugs, pigments, and flavorings.  It was not, however,  until the nineteenth century that the structures of the simplest of these "essential oils" became known, largely through painstaking degradative studies. The fact that these simple terpenes (called "monoterpenes", made up of ten carbon atoms) always yielded isoprene (2methylbutadiene) upon pyrolysis, led early investigators to propose a general rule, called the 'isoprene rule'. This rule stated that all compounds classified as terpenes must contain an integral number of isoprene units. limited utility.  Until the structures of higher terpenes were known, this rule had  64  isopentenyl pyrophosphate (IPP)  dimethylallyl pyrophosphate (DMAPP)  Scheme 3: Biosynthesis of IPP/DMAPP via Acetate and Mevalonic Acid  The pioneering work of natural products chemists in the early half of this century, and in particular the work of Ruzicka, led to a refinement of the isoprene rule which became known as the 'biogenetic isoprene rule'. Ruzicka realized that the structures of all terpenes could be rationalized as having come from a 'head-to-tail' coupling of a number of biological equivalents of isoprene, with subsequent modifications arising via rearrangements and functional group interconversions. While a thorough review of all the careful work that laid the foundation for the major breakthroughs in isoprenoid biosynthesis is beyond the scope of this introduction, it must be acknowledged that our present understanding owes a great debt to these early studies.  The discovery of the pathways by which this biogenetic equivalent of isoprene that Ruzicka had proposed was assembled into isoprenoids was accomplished largely by Bloch, Lynene, Cornforth, and Popjak (who shared the Nobel Prize in 1964). This work was the result of a detailed investigation into the biosynthesis of cholesterol (the steroids being a close 'relative' of the triterpenoids, which contain 30 carbons). Using radioisotopically labeled precursors and careful degradative techniques, they were able to trace the origin of all the carbons in the cholesterol molecule and its earlier intermediates. As a result of this work and more recent enzymatic studies, the complete picture as to how small two-carbon units are converted into the biogenetic equivalent of isoprene and subsequently transformed into isoprenes has emerged with utmost clarity. Scheme 3 shows how acetyl coenzyme A (an acetate group derived from primary metabolism i.e. citric acid cycle) condenses in a Claisen-type fashion with a second equivalent of acetyl CoA.  The product of this reaction undergoes an Aldol-type condensation with a third  equivalent of acetyl CoA to produce hydroxymethylglutaryl-CoA (HMG-CoA). Reduction of HMG-CoA with two moles of NADPH yields mevalonic acid (MVA).  Mevalonic acid is  phosphorylated with three equivalents of ATP to yield MVA-(3S)-phospho-5-pyrophosphate, which undergoes decarboxylation to form the first of the 'biogenetic isoprene equivalents', isopentenyl pyrophoshpate (IPP). The second isoprene equivalent, dimethylallyl pyrophosphate (DMAPP), is formed by the stereospecific isomerization (removal of the pro-R hydrogen) of IPP. The higher terpenes are formed by the coupling of IPP and DMAPP to give geranyl pyrophosphate (GPP), which is the immediate precursor to the monoterpenes. GPP can also couple with another unit of IPP to give rise to farnesyl pyrophosphate (FPP), the immediate precursor to the sesquiterpenoids. Coupling of another IPP to FPP gives geranylgeranyl pyrophosphate (GGPP), which goes on to form the diterpenoids. The coupling of two units of GGPP, 'tail-to-tail', gives rise to squalene, which is the precursor to the triterpenoids and the steroids. This hierarchical type of assembly is depicted in Scheme 4.  squalene  Scheme 4: Biosynthesis of Higher Terpenoids From IPP/DMAPP  67  IV.I.D. Previous Studies into the Biosynthesis of Terpenoids by Marine Invertebrates  With the exception of marine steroids and carotenoids, which shall for the purposes of this discussion be considered as primary metabolites, very few biosynthetic studies have been performed to investigate the origin of structurally novel marine terpenoids. All but one of these studies have used radioisotopically labeled precursors, and to date there has not been a single report of successful incorporation of stable isotopes into the carbon skeleton of a marine terpenoid. In fact, the earliest reports of biosynthetic investigations involving terpenoids derived from marine invertebrates focussed on the incorporation of precursors into the functional groups, as opposed to the carbon skeleton, of the terpenoids in question. The earliest such study, by Iengo, Sodano,  (17) and co-workers demonstrated that the sponge Cacospongia mollior, when incubated in seawater 29  containing [2-'4c] ornithine, showed incorporation of radioisotope into the four carbon bridge of molliorin-b (16), after purification to constant activity. Interestingly, no label was detected in the  68  co-metabolite molliorin-a (17). The presence of isocyanoterpenes in a large number of sponges has prompted numerous biosynthetic investigations into the origin of this unusual functional group.  That marine isonitriles may arise from cyanide ions, presumably concentrated from seawater, was first proposed by Herbert and Mann, and then later demonstrated by Garson : 30  31  1.8% incorporation with [^C] cyanide into diisocyanoadociane (18).  More recently, Karuso and  Scheuer have demonstrated that the nitrogen as well as the carbon in the isocyano functionality 32  are derived from an intact cyanide. In their experiment, 500 mg of potassium was fed to a single specimen of the sponge Ciocalypta addition to a signal for the cyano carbon in the '  3c,l S^jcyanide  sp. over a period of nine weeks.  In  NMR spectrum of the isolated compound (19)  that was enhanced by 156% with respect to an unlabeled sample (i.e an overall enrichment of 1.46%), it was also claimed that this signal was a doublet (J= 5.9 Hz) resulting from 15^-13c coupling. To reiterate a point made earlier, this study remains the only example of stable isotope incorporation into a marine terpenoid, and only into a functional group at that.  69 Although not technically from a marine invertebrate, the biosynthesis of the cembrane diterpene crassin acetate (20) by the zooxanthellae symbiont of the gorgonian Pseudoplexaura porosa demonstrated by Papastephnou and Anderson in a cell-free extract, is an elegant example 16  of dealing with the question of symbiosis and biosynthesis.  Cell preparations of the  zooxanthellae, uncontaminated with gorgonian cellular material, when incubated with [2-^C] mevalonate and co-factors, yielded radioisotopically labeled crassin acetate (20) after rigorous purification to constant activity. This pioneering study was the first report of biosynthesis of a terpene by a single-celled algae, so it is surprising that extending this type of approach to sponges has not yet occured. A number of biosynthetic studies involving sesquiterpenes and diterpenes from a number of species of soft coral adequately illustrate both the utility and limitations of attempting labeling studies in marine invertebrates. In the first study, small colonies of the soft coral Sinularia 33  capillosa were incubated with 2- H mevalonolactone and sodium [ ^ C ] bicarbonate. The latter 3  precursor was used to trace the production of metabolites by the symbiotic zooxanthellae that are usually associated with soft corals via photosynthesis- a very intriguing way of circumventing the complications due to symbiotic biosynthesis. Isolation of the sesquiterpene furanoquinol (21) and radioactive counting showed that the compound had a specific activity of 2600 dpm/mg of tritium, and no evidence for 14c in the compound. Chemical degradation of (21) by oxidative cleavage with osmium tetroxide-sodium metaperiodate yielded the aldehyde (22) with a specific activity of 4.6 dpm/mg. This result of essentially zero activity for this fragment demonstrated, as expected, that mevalonolactone was incorporated solely into the terpene portion of the molecule. It is unfortunate that the terpene fragment from this degradation was not similarly isolated and counted, in order to verify this finding. Having thus established that incorporation feeding experiments with soft corals were feasible, Coll and co-workers went on to study two related soft corals, Alcyonium molle and 34  Heteroxenia sp, which contain diterpenes and sesquiterpenes, respectively. The authors comment  For this reason, the colonies were incubated in aquaria containing dissolved precursors. In experiments with A. molle, it was shown that incubation with racemic radiolabeled mevalonolactone resulted in extracts that had little to no activity, suggesting that this precursor was not being taken up by the organism. Furthermore, incubation with radiolabeled butyrate led to noticeable toxicity in the colonies, and since again no activity was detected in the crude extracts, this experiment was discontinued.  Undiscouraged, incubation with [2-^C]sodium acetate  resulted in the isolation of the compound (23) with very low levels of activity (typically 0.002% incorporation).  It was further demonstrated by chemical degradations that this activity was  localized to the two butyrate ester side chains (though curiously, given the precursor, not in the acetate group), while the terpene portion itself contained no activity at all. It was thus concluded that de novo synthesis of terpenes could not be shown to be occuring in A. molle, but rather that esterification of the terpene was the major biosynthetic process. With Heteroxenia  sp. the  researchers were more successful; incubation with both [1-^C] acetate and [2-^C] mevalonate led to a measurably significant level of activity in the sesquiterpenes cubebol (24) and clavukerin A (25). Interestingly, chemical degradations showed that the [2-^C] mevalonolactone was being degraded in vivo to [2-^C] acetate, with subsequent incorporation of label as [1-^C] acetate.  Thus, having overcome some difficulties with precursor rejection and negative results, de novo biosynthesis of terpenes was demonstrated in a number of soft corals.  (25)  (26)  (27)  The case of dorid nudibranchs provides an excellent example of the role of planning before attempting a biosynthetic feeding experiment. The vast majority of terpenoids isolated from dorids can be traced back to other organisms upon which the dorids feed.  35  It has been suggested that 36  these shell-less molluscs, which seem so vulnerable to predation, have adapted to "borrow" defensive chemicals from the sponges, bryozoans, and cnidarians upon which they feed. The case for dietary origin of terpenoids found in dorids rests on the premise that since the same compound can be isolated from a nudibranch and the sponge upon which it is found, then a priori, the compound is made by the sponge and sequestered from this same source by the nudibranch. While this seems a perfectly natural conclusion, it is far from rigorous, since for example polygodial (26), a known plant metabolite, has also been isolated from a number of dorids, 37  38  and no one would argue a terrestrial source in this instance. While the fact that the dorids are often observed feeding on the very sponges that contain the terpenoids in question, this is still a far cry from experimental proof. Recently, Dumdei and Garson have shown experimentally that it is 39  possible to trace a radiolabeled metabolite contained in a sponge into extracts of nudibranchs that were allowed to feed on the sponge in an aquarium. While it would be tedious and ultimately unneccessary to repeat this experiment for every case, it is nonetheless reassuring that the putative dietary origin for most dorid terpenoids is on a firm experimental footing.  While a dietary origin for many terpenoids isolated from dorid nudibranchs exists, there are nonetheless well-documented examples of de novo synthesis of terpenoids by dorids. A very useful hypothesis has been proposed by Faulkner, Andersen and co-workers that has served to 40  identify those nudibranchs that might be good candidates for biosynthetic study. Nudibranchs that sequester compounds from invertebrates in their diet frequently show significant variation in their terpenoid constituents from one collection site to another, reflecting a difference in the terpenoid content of their diet. Those that make terpenoids via de novo biosynthesis would be expected to have similar terpenoid constituents at all collecting sites. Thus, geographic invariance of terpenoid content, while not an absolute predictor of de novo biosynthesis, can be used to identify nudibranchs that are good candidates for biosynthetic investigations. The first example to have been studied is the aforementioned sesquiterpenoid polygodial (26), isolated from the Mediterranean dorids Dendrodoris limbata, D. grandiflora, and from the 41  42  Pacific dorids Dendrodoris nigra, D. tubercolosa, and D. krebsii.  4 3  The fact that polygodial (26)  had never been isolated from any other (marine) source, led Cimino and co-workers to essay a 44  biosynthetic feeding experiment.  Specimens of D. limbata were kept alive in seawater and  injected with [2-^C] mevalonic acid-dibenzylethylenediamine salt into the digestive gland. After 24 hours, the animals were sacrificed and pure polygodial (26) was obtained having significant activity ( 32,400 dpm/mg). Purification to constant activity, reduction to the corresponding diol (27) and repurification afforded a sample of polygodiol (27) of comparable activity (32,500 dpm/mg). It was thus shown that feeding experiments with radiolabeled precursors were feasible in demonstrating de novo synthesis in dorids. Another class of terpenoids isolated from dorid nudibranchs for which no dietary source had ever been found were the diterpenoic acid glycerides (28 to 33), that have been isolated from a large number of geographically distinct species including the British Columbia dorids Archidoris odhneri and A. montereyensis* '^ the 45  6  California dorid Sclerodoristanya, the Mediterranean dorids Doris verrucosa 48  4 9  and Archidoris  tuberculata, the Eastern Atalantic dorid Archidoris pseodoargus, the Southeastern Pacific 50  51  73  dorids Archidoris carvai  and Anisodoris fontaini,  5 2  53  and lastly from the Antarctic dorid  Austrodoris kerguelensis. * 5  Gustafson and Andersen were the first to demonstrate de novo synthesis in this class of 46  compounds.  Injection of specimens of A. odhneri and A. montereyensis with [ 2 - ^ C ]  mevalonic acid dibenzylethylenediamine salt and incubation for 48 hours afforded labeled samples of farnesic acid glyceride (28) and diterpenoic acid glyceride (29) that showed significant levels of activity after a number of purifications and chemical transformations. However, when Cimino et al. attempted similar experiments with terpenoic acid glycerides (31) from D. verrucosa , they 55  were unable to observe significant activity in their compounds. These authors concluded with equivocation, which had the unfortunate effect of casting serious doubt upon the validity of the results of Gustafson and Andersen. It was argued that since the measured activities in these compounds was low, and that since these compounds were extremely difficult to purge of all traces of coeluting fatty acid esters due to their own glycerol moeties, that the possibility of contamination with radiolabeled fatty acid esters could not be completely ruled out. This was an unfounded  criticism, since Gustafson and Andersen had very carefully performed chemical transformations that would allow for rigorous purifications. This controversy had one positive effect in that it clearly showed a need for the reinvestigation of the biosynthesis of these compounds in a manner that gave clear and unambiguous results.  IV. I I . Biosynthesis of Terpenoic Acid Glycerides by the Dorid Nudibranchs Archidoris odhneri and A. montereyensis  IV.II.A  Preamble  The decision to investigate terpenoic acid glyceride (27,28) biosynthesis by Archidoris species with stable isotopes was justified, as the preceeding introduction has alluded to, for a number of reasons. First, some C incorporation had already been demonstrated,^ though this 14  result has been subject to unfounded controversy,55 and thus it was hoped that stable isotope incorporation would put this issue to rest once and for all.  Secondly, the large amounts of  terpenoic acid glycerides present in these animals would afford C spectra of reasonble signal-to,3  noise, thus allowing for clear and visually convincing results. Third, these animals could be found in large numbers all-year round, enabling repeat experiments to be performed if necessary. Fourth, as the radioisotope results had confirmed,^ these nudibranchs conformed well to the geographical invariance hypothesis,^ and were thus good candidates for study. Lastly, it was hoped that if a methodology could be worked out that demonstrated unambiguous de novo synthesis with stable isotopes, that such a methodology could be applied to investigate the biosynthesis of other, more interesting terpenoids from nudibranchs that also fit the geographical invariance hypothesis, and where no previous experimentation had been done- notably, the sesquiterpenes found in Acanthodoris nanaimoensis .56 Thus, the stable isotope incorporation studies detailed herein were a test case to demonstrate that: i) a methodology could be developed such that levels of incorporation of stable isotopes were detctable by NMR; ii) de novo synthesis of terpenoic acid glycerides could be demonstrated unambiguously, providing both a first proof that stable isotopes could be incorporated into a terpene skeleton and also rebutting once and for all the controvery generated from radioisotope incorporations; and iii) that this stable isotope methodology could be applied to other suitable nudibranchs, thus encouraging further research into the biosynthesis of marine invertebrate terpenoids. The initial results of Gustafson and Andersen^6 pointed out that a number of obstacles had to be overcome if stable isotope methodology was to succeed.. Even with the highly sensitive radioistope C, the measured activities of the recovered compounds were low, necessitating that 14  the precursor and mode of delivery for this study be chosen with care. The simplest precursor, 13  C labeled sodium acetate, was chosen for a number of reasons. Previous studies ** with soft 3  77 corals had shown that mevalonate could perhaps be reconverted back to acetate, and moreover, acetate was a more economical choice for a study where the outcome was by no means assured. Deuterium, while having the advantage of being far more sensitive to detection than C, was 13  rejected at this early stage of experimentation since hydrogens of acetate are in the first place removed during mevalonic acid synthesis, and morover are readily "washed-out" by exchange processes and enzymatic action. It was hoped that should studies with C prove successful, only 13  then might deuterium play a role in probing more subtle aspects of terpenoid biosythesis in dorid nudibranchs. Experiments with singly-labeled acetate, as has been admirably demonstrated with plants and microorganisms,^^ rely on the ability to detect significant increases in the peak heights of C 13  resonances in the product spectrum, with reference to a control spectrum of unlabeled material. The levels of incorporation previously reported in studies of marine invertebrates with radioisotopes suggested that such significant increases in peak height could not be expected in this study. It was anticipated that relaxation processes that depend on variables beyond experimental control could give rise to peak height differences of the same order of magnitude as those generated by the anticipated levels of incorporation of single label. The use of doubly labeled [1,2- C ] 13  2  acetate avoids this problem altogether, since instead of peak height increases, one is now able to measure C- C doublet satellites around labeled resonances, provided they are significantly 13  13  above natural abundance (0.55% of the singlet peak area for each satellite). Moreover, the ability to detect the pattern of incorporation of intact acetate units without resorting to chemical degradation was another anticipated benefit of using doubly-labeled acetate. As a biosynthetic wit has remarked,^ ne can never have too many labels in a precursor. 0  The mode of delivery of the precursor was also carefully considered. Previous work using radioisotopes in dorids44,46 involved the injection of precursors into the digestive gland of the organism; in fact, nudibranchs are well suited to this approach since unlike sponges and tunicates etc. their anatomies are well-enclosed by a thick-skinned mantle. Thus, nutrients are unlikely to be taken up from the surrounding seawater by diffusion. Since acetate is water-soluble, it was hypothesized that the use of liposomes might prevent the label from being quickly removed from  7  the organism through the gills, and also might better facilitate the penetration of precursor through cell membranes. In retrospect this turned out to be a poor choice; the liposome results are included here as a demonstration of the difficulties involved in developing a stable isotope methodology for the study of biosynthesis in dorids, and also as a testament to the need for perseverence in investigators undertaking such experiments.  IV.II.B  Preliminary Results Using Liposomes  Specimens of Archidoris odhneri (42 animals) and A.montereyensis (36 animals) were collected^ via SCUBA in and around surge channels off Sanford and Fleming Islands, Barkley Sound, B.C. at depths of 15 and 8 metres, respectively. All animals were kept alive in running seawater tanks at Bamfield Marine Station and were then shipped in a large cooler of Barkley Sound seawater to the laboratory in Vancouver. Upon arrival, all nudibranchs were immediately transferred to insulated containers, one for each species, containing fresh Barkley Sound seawater that had been transported back for this purpose. The containers were maintained in a cold room with an ambient air temperature of 12 °C, and air was constantly bubbled into the seawater through an airstone. All specimens of A. montereyensis were injected through their dorsums on the left side with lOOuX of a 550 mM solution of [1,2- C ] sodium acetate encapsulated in liposomes ^ 13  0  2  and allowed to incubate for 24 hours. The concentration of precursor in solution was chosen to approximate the osmolality of Barkley Sound seawater. All specimens of A.odhneri were similarly injected, but after a 24 hour incubation period, all animals were injected with a second 100 |iL portion of the same solution, and were then allowed to incubate for a further 24 hours. The seawater in the A. odhneri container was changed prior to the second injection. At the end of the incubation periods, all animals were sacrificed by immediate immersion in methanol. Extracts were obtained, and the compounds purified as described in the experimental portion of this thesis. Control samples of all compounds were obtained from nudibranchs extracted immediately upon collection.  79  28 R = R = H  29 R = R = H  34 R!=H, R = A c  36 R j - H ,  35 Rj=Ac,  37 R j s A c ,  1  1  2  2  R =Ac 2  The experiment with A. odhneri  2  R =Ac 2  R =H 2  was successful, although the evidence was not as  overwhelmingly convincing as had been hoped for. The presence of very small flanking doublets around some of the resonances in the C spectrum of farnesic acid glyceride (28) had appropriate ,3  coupling constants for different C-C bonds in the molecule. Integration of these doublets, where possible, revealed that the intensities of these signals were not significantly greater than natural abundance, and moreover, the presence of very small multiplets around some of the resonances was cause for some consternation. Therefore, it was concluded that, although the result was encouraging, irrefutable and visually convincing evidence had yet to be obtained and thus the experiment bore repetition with the hope of increasing incorporation levels. At this stage, an observation that had been made in our group's long experience with nudibranch chemistry suggested a new strategy to achieve higher levels of incorporation. When nudibranchs are collected via SCUBA and transported back to the laboratory alive, a much smaller quantity of metabolite is obtained from their skin extracts than is the case when nudibranchs are extracted in the field immediately following collection. In some extreme cases, it is impossible to recover any quantities of secondary metabolites unless the animals are immersed in methanol within the first few minutes after being brought to the surface by divers. This observation is in good accord with the presumed defensive role that these compounds have in the organism,46,61 since it is likely that the physical act of handling nudibranchs, which is an unavoidable aspect of their capture, simulates an attack by a predator causing them to release the putative defensive compounds. While this facet of nudibranch biology is a problem when new chemistry is being sought, any mechanism whereby a nudibranch is caused to shed its complement of "unlabeled" metabolites  would be a great asset to a biosynthetic study. It was therefore reasoned that a number of smaller injections of precursor over a longer period of time may realize the goal of higher levels of incorporation for two reasons. First, 48 hours may simply not be long enough an incubation time, since the rate of biosynthesis may be quite slow as earlier research has shown.34 Second, and more importantly, by administering many injections instead of one, it was hypothesized that the animals may become stressed at each injection and thus release a portion of stored terpenoic acid glycerides that had been synthesized prior to the beginning of the experiment. In this way, it was hoped that the pool of unlabeled material would be depleted, while the high concentrations of labeled acetate administered would ensure that any subsequent biosynthesis during the experiment would be detected. It was also decided to sacrifice small groups of animals at intervals throughout the course of the injection period, in order to determine if the levels of incorporation were increasing. The only concerns with this strategy were the ability to keep the nudibranchs alive for a long enough period, and the fear that the multiple injections might so stress the animals that they would perish. Specimens of A. odhneri (38 animals) and A. montereyensis (25 animals) were recollected in Barkley Sound, as previously described, and transported back to Vancouver as before. Each animal was injected with 25 uL of a 550 mM solution of [1,2- C ] sodium acetate encapsulated in ,3  2  liposomes,^ with a seawater change accompanying the injection. The animals were allowed to incubate for 24 hours, after which another injection was made in the same manner. Injections continued every 24 hours in exactly the same manner. After 4 days, one third of the A. odhneri specimens (14 animals) were sacrificed (hereafter referred to as the '4 day odhneri' group, 40D). After 8 days, one half of the remaining A. odhneri specimens were sacrificed (10 animals, referred to as 80D) and one half of the original specimens of A. montereyensis (12 animals, referred to as 8MON) were similarly sacrificed. After 12 days, the health of the remaining animals was assesed, and it was decided to terminate the experiment (10 odhneri, 120D; 13 montereyensis, 12MON). All extracts were worked up ort the usual fashion, as described in the experimental. The results from this set of experiments were both encouraging and surprising. While a rigorous discussion of labeling patterns, measured specific incorporations and coupling constants etc. will be reserved for the final series of experiments on these nudibranchs, these results for both  compounds (28) from 40D and 120D and compound (29) from 8MON serve to illustrate a number of conclusions drawn from this study. First, it was noted that the incorporation level in 40D was greatly improved, to the point where the integrated areas of the flanking doublets were significantly greater than natural abundance. Also, flanking doublets were observed for the first time in samples of diterpenoic acid glyceride (29) from A. montereyensis . This suggested that the multiple injection strategy was indeed yielding a greater turnover of labeled compounds. Surprisingly, the level of incorporation in the 12 day samples (120D and 12MON), as measured by the integrated peak areas of the flanking doublets, was significantly reduced compared to the 4 day spectra. This was doubly surprising since the amounts of material recovered from the 12 day experiments were significantly greater than the 4 day experiment, suggesting that biosynthesis was still occuring over the last 8 days of the 12 day period. For example, 40D yielded 18.6 mg (1.3 mg/animal), 80D yielded 99.8 mg (9.9 mg/animal), and 120D yielded 235 mg (23:5 mg/animal) of farnesic acid glyceride (28).  Animal size and number discrepancies in each sample set could  not account for the increase in compound over time, and thus it was concluded that over greater periods of time, greater amounts of the compound were being synthesized. Yet if this were the case, it would be expected that the 120D spectra would show the highest levels of incorporation, and just the opposite was observed. The only way to account for this observation was to assume that over longer periods of time, the labeled precursor was somehow being diluted with unlabeled acetate, such that while biosynthesis was still occuring, greater amounts of compound were produced that had smaller and smaller amounts of labeled acetate incorporated into them, as was observed. It was reasoned that since the animals were provided with no food source during the experiment, at some point primary metabolism required that acetate be diverted from secondary metabolism into providing energy for the animal's survival via the citric acid cycle. However, if the rate of secondary metabolism were slowing with time, we would not expect to recover larger and larger amounts of farnesic acid glyceride over time. Clearly, the precursor was being diluted over time with larger and larger amounts of unlabeled acetate, but from where? While acetate from the digestion of the animals' reserves of fat could explain some portion of this dilution, it seemed unlikely that the animal would sacrifice all its energy reserves in the production of prodigious amounts of defensive compounds.  Upon reflection, it was realized that the liposomes were themselves composed largely of lipids which could very easily be broken down to acetate without depriving the animal of its fat reserves. Thus, while at the outset of these experiments it was believed that liposomes would provide an ingenious method of increasing incorporation levels, this second set of experiments strongly suggested that these same liposomes were in fact diluting the labeled precursor to the point that they hindered the experiment more than they helped. It was thus decided to repeat the experiment one more time, using the multiple injection hypothesis, but dispensing with the liposomes.  IV.II.C  Successful Incorporation Studies  Specimens of A. odhneri and A. montereyensis were collected by hand using SCUBA in Barkley Sound B.C. and transported to Vancouver in refrigerated seawater. The nudibranchs were maintained at 12 °C in Barkley Sound seawater that was changed every second day for the duration of the experiments. Individual specimens of A. odhneri and A. montereyensis were given 100 |iL injections of a 550 mM solution of [l,2- C2]acetate every second day for 16 days. 13  The handling of the animals required for each injection caused the nudibranchs to shed mucus and terpenoids, resulting in a partial turnover of the pool of terpenoic acid glycerides. Oh the eighteenth day, two days after the last injection, the nudibranchs were immediately immersed whole in methanol. The methanol extracts from the injected specimens (20 animals) of A. odhneri were fractionated to give 130 mg of the major farnesic acid glyceride (28). Acetylation of glyceride (28) with acetic anhydride and pyridine at room temperature for 12 h gave the diacetate (35) in essentially quantitative yield. The methyl and carbonyl carbons of the acetate units in (35) provided an internal control of unlabeled carbons. Assignments of all the carbon resonances in the  1 3  C NMR spectrum of (28) were not  established in the literature and therefore had to be experimentally determined before the results of the biosynthetic feeding experiments could be assessed.  Since these assignments did not  necessitate a complete structure determination argument from first principles, the following argument will be based upon the structure of (28) already well-established in the literature.  85  H2  H6/10  HI'  L_JL  H3'  (ppm)  20  h 40 60  C3'  cr  C2'  80 100 C2 C6  4  120  CIO f'T'i i " i i [ i i i i | i i i i  (ppm)  i'i"  r "f i "| ""i i ' i i | i i  I'ff  r'i'i  i \ i11 1 i ' | 1  T  5.0 4.5 4.0 3.5 3.0 25 2.0 1.5  Fig. 38: HMQC Spectrum of farnesic acid glyceride (28) [500 MHz, CDC1 ] 3  Fig. 39: Expanded Upfield Region of HMQC Spectrum of farnesic acid glyceride (28) [500 MHz, CDC1 ] 3  87  H2  HI'  C13-  o o  - 20  *  0  - 40  C4  - 60 - 80 - 100 " 120 CO  o  -140  - 160 CI I  (ppm)  III I  I  II II  I  I  I  I I  I  5.0 4.5 4.0 3.5  IIII  I  I I II  I  II  I  I  I  II  I  I  3.0 2.5 2.0 1.5  Fig. 40: HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDC1 ] 3  I  II  88  Fig. 41: Expanded Upfield Region of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDC1,]  89  OH  (28)  Mei: H5 H4  Mel4 Mel5  Mel2 H9  H8  (ppm)  C2  E-120  C6 CIO  t 130 Cll C7  140  t 150  E- 160  C3 •I  (ppm) 2.1 2.0  1.9  1.8  1.7  1.6  1.5 1.4  Fig. 42: Expanded Region (Upfield 'H, Downfield C) of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDC1 ] 13  3  90  (ppm) 5.6  5.5  5.4  5.3  52  5.1  5.0  Fig. 43: Expanded Region (Downfield 'H, Upfield C) of HMBC Spectrum of farnesic acid glyceride (28) [500 MHz, CDClj] ,3  91 In the original paper, carbons 1,2 and 3 of farnesic acid glyceride (28) had been assigned to resonances at 8 165.9, 114.7, and 161.4, respectively. These were in good agreement with the observed HMQC (Fig. 38) correlation from the olefinic singlet at 8 5.62 (H2) into the carbon at 8 114.7, and the HMBC (Figs. 40-43) correlation from this same olefinic signal at 8 5.62 (H2) into the quaternary carbon at 8 165.9 (CI). H2 also showed an HMBC correlation into a methyl carbon (HMQC) at 8 18.9, which was assigned to C13.  Lastly, H2 showed an HMBC  correlation into a methylene carbon (HMQC) at 8 40.9, which was assigned to C4.  OH 28a  Having established the identity of C4, the next step was to distinguish between the methylene resonances for C5, C8, and C9. Only H8 protons can show an HMBC correlation into a methyl carbon resonance other than CI3 (i.e four bond correlations from H9 or H5 into methyls are not observed in the HMBC experiment). By this reasoning, the HMBC correlation from a well-resolved methylene signal at 8 1.93 into the methyl carbon resonance (HMQC) at 8 15.9 must be assigned to the H8 protons. Thus, the HMQC correlation from this signal at 8 1.93 established  92 that C8 could be assigned to the methylene carbon resonance at 8 39.6, and similarly, that the methyl carbon at 8 15.9 must neccessarily be C14.  Furthermore, the H8 protons showed  additional HMBC correlations into two methine carbons (C6 and CIO) and one quaternary carbon. Again, by neccessity, this quaternary carbon must be C7 (CI 1 is too far away to show an HMBC correlation into H8). This quaternary carbon resonance at 8 136.1 was therefore assigned to C7. By a process of elimination, the remaining quaternary carbon at 8 131.2 must be CI 1. HMBC correlations were observed from C l l (8 131.2) into two methyl singlets (8 1.55 and 1.62, CI2 and CI5) and a methylene resonance at 8 2.00. Since H9/H9' are the only methylene protons that can show an HMBC correlation into C l l , this resonance at 8 2.00 which gave an HMQC correlation into the methylene carbon resonance at 8 26.6, thus established the identity of C9. While the H8 protons showed HMBC correlations into both methine carbons (C6 and CIO), H9 showed a correlation into only one of these carbon resonances, at 8 124.1, which must neccessarily be C10. By elimination, therefore, the remaining methine carbon resonance at 8 122.7 must be assigned to C6, and similarly, the remaining methylene carbon resonance at 8 25.9 must be assigned to C5. It remained only to distinguish between C12 and C15. Using chemical shift arguments, the carbon resonance at CI5 should be further upfield than CI2, due to the shielding effect of C9. Thus, C15 was assigned to the remaining methyl carbon at 8 17.6, and C12 was assigned to the resonance at 8 25.4. The glycerol carbons were more straightforward to assign.  The HMQC experiment  established the identity of the methine carbon (C2 ), while CI' was distinguished from C3' from 1  an HMBC correlation into the carbonyl carbon (CI) of the farnesic acid moiety from the H I ' protons. Shown in Figure 44 are the truncated C NMR resonances, all normalized to the same peak 13  height for the central singlet component, for the terpenoid, glycerol and acetate carbons in the proton noise decoupled C NMR spectrum of the labeled sample of 35. The resonances assigned , 3  to the terpenoid carbons 1, 2, 3, 5, 6, 7, 9, 10, 11, 13, 14 and 15 as well as the resonances assigned to the glycerol carbons 1', 2' and 3' in the spectrum of labeled 35 all appear as central singlets due to natural abundance  1 3  C flanked by clear doublets resulting from specific  93 incorporation of intact acetate units. As expected, the acetate methyls show only a singlet resonance resulting from natural abundance C . The C NMR spectrum of a control sample of 13  13  completely unlabeled diacetate 35 acquired at the same concentration on the same spectrometer and for the same number of scans did not show any evidence for the doublet components (Figure 44). The intensity of the doublet components in the labeled spectrum of 35 indicates an average specific incorporation of intact acetate units of 0.09% (Table 3). Numbers listed for specific incorporations = % enrichments above natural abundance = 1.1% X (combined integrated peak area of enriched satellites minus the combined theoretical peak area for these same satellites resulting from natural abundance coupling)/(peak height of the natural abundance singlet plus the combined theoretical peak area for all satellites resulting from natural abundance coupling). The probability of observing natural abundance coupling between a pair of adjacent carbons is 0.011 X 0.011 = 0.000121. The intensity of this signal would be split between the two doublet components resulting in a predicted abundance of 0.0000605 for each component. Thus the doublet components should each be (0.0000605/(0.011 - n X 0.000121)) X 100 = 0.55% of the intensity of the unenriched central singlet for each coupling interaction, (n is the number of next neighbor carbons and it must have values between 1 and 4. Therefore, 0.011 - n X 0.000121 is always = 0.011.). Analysis of the coupling constants observed for the doublet components in the terpenoid carbon resonances (Table 3) revealed the pattern of intact acetate incorporation shown in 28a (Scheme 5) that was completely consistent with the expected biosynthesis from mevalonic acid. It is interesting to note that the resonances for the terpenoid carbons 4 and 8 in the C , 3  NMR spectrum of the labeled diacetate 35 (Figure 44) also appear as singlets flanked by doublet components. In the case of these two resonances, the doublet components are less intense than those observed for the other terpenoid carbons. Despite their relatively low intensity, a comparison with the control spectrum of unlabeled 35 (Figure 44) indicates that the doublet components flanking the resonances for carbons 4 and 8 in the spectrum of labeled 35 must also be the result of incorporation of C labeled acetate units. C-8 could be coupled to either or both of C-7 and C-9, , 3  and C-4 could be coupled to either or both of C-3 and C-5. Coupling constant data  T TM H M 1 1 1 T  1  1  1  27.2  1'IM'IT  HIMUM'I'MI  27.2  (ppm)  (ppm)  L  J  !Mi|l|i|t|'l'l  16.0  16.0  (ppm)  (ppm)  ••••••••••••••I  1JI.2  (ppm)  (ppm)  C1 •  FVTMM'MIM  17.6  (ppm)  17.6  (ppm)  61.0  (ppm)  Tf'PPPIM'P  MM'I'l MUM'  62.0  62.0  (ppm)  (ppm)  J  C2'  69.0  (ppm)  69.0  (ppm)  OAc C-O  C3'  J L  ni'Mi'i'i'i'i  124.0  C1 5  C14Pl'l'l'l'l'l't  C1 1  CIO  C9  _/ 170.0  61.0  (ppm)  (ppm)  170.0 (ppm)  Fig 44: Normalized and Truncated C Resonances for Labeled Oeft) and Unlabeled (right) Samples of farnesic acid glycende diacetate (35) [125 MHz, CDC1 ] U  3  95 Table 3: Specific incorporation data for the [1,2- diacetate feeding experiments with 1  Archidoris odhneri and A. montereyensis.  Diterpenoic acid glyceride (29)  Farnesic acid glyceride ciacetate (35) c#  8  J (Hz)  % specific incorporation  cc  l3C(ppm)  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 1' 2* 3'  8 13  (125  MHz) 165.9 114.7 161.4 40.9 25.9 122.7 136.1 39.6 26.6 124.1 131.2 25.4 18.9 15.9 17.6  MHz) 39.9 18.7 41.9 33.0 56.5 18.5 41.9 37.4 54.3 36.6 22.7 124.3 128.5 62.6 173.4 21.6 33.2 15.6 21.2 • 15.7 63.5 70.4 65.1  62.4 69.2 61.3  0.15 0.10 0.08 (0.03) 0.07 0.07 0.08 (0.04) 0.08 0.07 0.09 (0.007) 0.10 0.08 0.09 a  a  3  a  42.9 42.9 43.9  0.07 0.09 0.07  a  % specific incorporation  (34.3) 33.6 33.6 35.1 34.3 35.1  (0.03) 0.06 0.08 0.05 0.07 0.10  c<G  C(ppm)  (125 76.3 76.3 40.0 (33.4) 43.9 43.9 42.0 (33.4/41.0) 43.9 44.8 42.0 (44.8) 39.1 42.0 42.0  J (Hz)  a  35.9 34.3 35.9 35.1 (41.2) 44.2 57.2 59.5 35.9 35.9 44.2 35.1 40.4 41.2 41.2  a  0.12 0.06 0.08 0.06 a  0.09 0.05 (0.07) 0.06  a  0.08 0.05 0.05 0.04 0.07 0.05  a) These doublets are very weak making it difficult to obtain accurate coupling constant or specific incorporation measurements.  for the C-4 and C-8 resonances did not clearly indicate which neighboring carbons were responsible for the coupling. The most likely explanation for the coupling observed in the C-4 and C-8 carbons of labeled 35 is that it originates from incorporation of more than one C labeled acetate unit into a 13  single farnesic acid molecule. The nudibranchs were starved during the 16 day injection period 62  and this could well have led to a highly labeled acetate pool that in turn could have resulted in a reasonable probability of more than one labeled acetate unit being incorporated into individual  97  C2/ C6  C1  J L iJL I • I • I' I' I' I • I • I  40.0 (ppm)  MI'Ml'MI'l'  40.0 (ppm)  J  L  • I" I • I' I • I' I' I •  18.2 (ppm)  C5  J  (•••••••••••Ml  (ppm)  56.0 (ppm)  I'MMI'MMI'I  CI 1 J  I'MIII'MI'MI  (ppm)  22.4 (ppm)  173.6  'MMMI'MI'M  173.6  (ppm)  18.2 (ppm)  ••••••••••••MI  41.6 (ppm)  124.6 (ppm)  41.6 (ppm)  'I'l'MI'MPM  33.6 (ppm)  C9  I'MI'MI'MMI  37.3 (ppm)  • MMI'MI'MM  Tnn'i'i'i'i'i  J  • 'MI'MMMI'I  54.4 (ppm)  'MMMI'MI'M  124.6 (ppm)  CIO  I ' I'  IMM ' l M ' l  54.4 (ppm)  I'MMI'MMIM  36.8 (ppm)  CI 3  ll fi l' l' 'l ll 'l I'l'MI  (ppm)  I' I' I' I' I' T • I' I  21.3 (ppm)  21.3 (ppm)  T T 'MMl'i'l'  33.6 (ppm)  f 11 f | M i |i| t| M  128.4  128.4  (ppm)  (ppm)  I'MI'MI'MMI  36.8 (ppm)  C14  I'MMMMMI'I  62.4 (ppm)  C1 6/| 19,  C1 5  • I • I • I • I' I • I • I •  M ' l ' l-TTTTT 'l'  CI 2  MI'MI'I'I'I'I  22.4  J  37.3 (ppm)  C47 C1 7  C7  C8  I • I • I' 1 • I' I • I' I  56.0  C3/  11111' p I' I ' I ' I  62.4 (ppm)  C1 '  ' M111' M Ml IM  (ppm)  15.4 (ppm)  I'MI'MI'MMI  64.0 (ppm)  Fig. 46: Normalized and Truncated "C Resonances for Labeled (left) and Unlabeled (right) Samples of diterpenoic acid glyceride (29) [125 MHz, CDC1,]  I'MI'MI'MMI  64.0 (ppm)  9 8  mevalonic acid molecules and/or more than one labeled mevalonic acid unit being incorporated into a single farnesic acid residue. As a result, the observed C-4 coupling would arise from molecules of farnesic acid glyceride diacetate 35 having a singly labeled carbon at C-4 and an intact acetate unit at either C-3/C-13 or C-5/C-6. Similar clusters of three adjacent C labeled carbon atoms in 13  single molecules of 35 could lead to the coupling observed in C-8. Although the arguments presented above account for all of the features of the weak flanking doublets observed in the C 13  NMR spectrum of 35, alternate origins for these doublets involving different incorporation patterns of intact acetate units into isopentenyl pyrophosphate and dimethylallyl pyrophosphate cannot be completely ruled out with the current data. The presence of doublets flanking the singlet resonances for all three glycerol carbons 1', 2' and 3' indicates that intact acetate units have been incorporated at both positions C-17C-2' and at C-2VC-3' in the glycerol residue. This incorporation pattern would result from esterification at both primary carbons of an achiral glycerol intermediate containing one intact acetate unit. Thus, the specific incorporation of acetate into the glycerol residue is the sum of the incorporations at the C-17C-2' and C-27C-3' positions. The methanol extracts from the injected specimens of A. montereyensis (20 animals) were fractionated as previously described to give a crystalline sample (38 mg) of the major diterpenoic 47  acid glyceride 29. Fortunately, ^ C assignments for this compound were already available in the 3  literature^, and thus no extensive 2D NMR data was required for compound (29). Examination of the proton noise decoupled C NMR spectrum for labeled 29 (Figure 46) revealed that intact 13  acetate units had been incorporated into both the terpenoid and glycerol fragments although at lower levels than were observed for 28. Analysis of the C / C coupling constants in the 13  13  spectrum of labeled 29 (Table 3) demonstrated that the incorporation pattern of intact acetate units in the terpenoid fragment was as shown in 29a in Scheme 5. In summary, the experiments described above provide the first unambiguous proof for the de novo biosynthesis of terpenoic acid glycerides by dorid nudibranchs and they confirm the results of previous radioisotope investigations of the biosynthesis of 28 and 2 9 .  4 6  These  experiments also demonstrate that stable isotope methodology can be used to investigate the biosynthesis of terpenoid metabolites by marine invertebrates.  99 IV.III.  Stable Isotope Incorporation Studies On Sesquiterpenoids from the Dorid Nudibranch Acanthodoris nanimoensis  Fig. 4 7 : Color Plate of Acanthodoris nanaimoensis  100 The dorid nudibranch Acanthodoris nanaimoensis is common on the outer coast of British Columbia. Skin extracts of A. nanaimoensis collected at many sites along the B.C. coast always contain nanaimoal (38), acanthodoral (39) and isoacanthodoral (40), that all have unprecedented sesquiterpenoid skeletons.  56  41 X =  <  42 X = <^  OH  43 X  OH  It has been proposed that the nanaimoane and isoacanthodorane skeletons might arise via opening of the cyclobutane ring in an acanthodorane precursor as shown in Scheme 6 (d and/)-  56b  Alternatively, the nanaimoane skeleton could be formed via direct cyclization of a monocyclofarnesane precursor  (Scheme 6, a, c ) ,  56a  and it in turn could be a precursor to the  acanthodorane skeleton (via e). Interest in the biogenesis of the three new sesquiterpenoid skeletons represented by 38, 39, and 40 prompted the initiation of stable isotope incorporation studies with A. nanaimoensis, as an extension of the successful work undertaken with the terpenoic acid glycerides from Archidoris odhneri and A. montereyensis, as described earlier. The objectives of the investigation were: i) to demonstrate that the aldehydes 38 to 40 were being made de novo by A. nanaimoensis as indicated by the geographic invariance of their occurrence, and ii) to compare the stable isotope incorporation patterns with those predicted by the biogenetic proposals in Scheme 6. Specimens (95 animals) of A. nanaimoensis were collected via SCUBA in Barkley Sound, B.C. and transported back to UBC in refrigerated seawater. The nudibranchs were maintained at 12 °C in an aquarium filled with Barkley Sound seawater that was changed every 2 days. Individual specimens of A. nanaimoensis were given 100 |iL injections of a 550 mM solution of  101  Scheme 6 [l,2- C2]acetate every second day for 16 days. The physical act of handling the nudibranchs for 13  injections caused them to partially shed the terpenoids in their dorsums. Two days after the last injection the specimens of A. nanaimoensis were carefully removed from the aquarium seawater and the intact animals were immediately immersed in methanol (250 mL). The methanol was decanted from the whole animals and evaporated in vacuo to give an aqueous suspension/Dilution of the aqueous suspension with water followed by extraction with EtOAc gave an organic soluble fraction (650 mg) containing the sesquiterpenoid aldehydes. Chromatographic separation of the aldehydes was facilitated by reduction with NaBrL; to the corresponding alcohols as previously described and the alcohols were fractionated via 56  102  104  105  H9b  C9  Jl  Cll  1 Me 13 Mel4 1  1.75 1.70 1.65 1.60 1.55 150 1-45 140 1.35  (ppm)  H2 H6 HI  H12  Mel 5  "",^7  (ppm) C2_ C6  20  C15  25  C14 C13 CI  r  6  L  C3 :  C9  3 0  35  40  -45  Cll  50 55 C12  60  -6  (ppm) 3.5  3.0  2.5  2.0  1.5  1.0  Fig. 51: HMQC Spectrum of nanaimool (41) [500 MHz, CDC1 ] 3  106  Fig. 52: COSY Spectrum of nanaimool (41) [500 MHz, CDC1 ] 3  107  OH  Mel 3 Mel4 H9a H9b  MelS  H3 H7  C2  (ppm)  C6 C15 C14  ^  C13  ^>  C8 CI  C4 C7  3  0  0  "25  £3  "30  ^ 3  F35  C3  -as  40  C9  F45  Cll  50  C12  60 1  (ppm) 1.8  1  r  1.6  1.4  1.2  1.0  0.8  Fig. 53: Expansion of Upfield Region of HMBC Spectrum of nanaimool (41) [500 MHz, CDC1 ] 3  108 O H  H9a HI  H6  H9b H2  Mel 3 Mel4  H7  (ppm)  124  64  CIO  126 128 130  6  C5  -i  (ppm)  fe >  r  1.8  6—«e) -i  1.6  1.4  1.2  134 136  r  1  132  1.0  Fig 54* Expansion of Downfield C Region of HMBC Spectrum of nanaimool (41) [500 MHz, CDC1 ] 13  3  109  OH  (ppm) 3.78 3.76 3.74 3.72 3.70 3.68 3.66 3.64  Fig 55: Expansion ofH12 Region of HMBC Spectrum of nanaimool ( 4 1 ) [500 MHz, CDC1 ] 3  110 reversed phase HPLC (eluent: 4:1 MeOH/rhO) to give pure samples of nanaimool (41) (10 mg: 0.1 mg/animal) and isoacanthodorol (43) (7.0 mg: 0.07 mg/animal). Only trace amounts of acanthodorol (42) were obtained from the reduction mixture. Since the structures of these sesquiterpenoids had been solved largely on the basis of X-ray crystal structures, there were no 13  C assignments in the literature. Presented in Table 4 are these assignments, based upon COSY,  HMQC, and HMBC data.  Unlike an ab initio  structure determination, it was relatively  straightforward with the known structures in hand to assign all the C resonances to carbons in ,3  the molecule.  nanaimool (41) 1 3  C assignments in nanaimool (41) were made with confidence by the following  arguments, based upon spectroscopic data. The proton resonance at 8 3.70, clearly belonging to the H12 carbinol protons, showed an HMQC correlation (see Fig. 51) to a carbon at 8 59.6, which was assigned to CI2. A COSY correlation (see Fig. 52) from H12 (8 3.70) into a resonance at 8 1.50 established the identity of the C l l protons, and HMQC showed that the 8 1.50 proton resonance was correlated to a carbon at 8 44.0 (Cll).  This assignment was confirmed by the  HMBC correlation (see Figs. 53 to 55) observed from the HI 1 proton resonance (8 1.50) into the C12 carbon resonance (8 59.6). The HI 1 proton resonance showed further HMBC correlations into a quaternary carbon (APT, Fig. 50) at 8 30.7, which neccessarily had to be assigned to C8, and into a methyl carbon (APT) at 8 24.8, which similarly could only be assigned to CI5.  HI 1 also showed HMBC  correlations into two.methylene carbons at 8 43.8 and 34.7, which must be C7 and C9. These two resonances were distinguished by the following argument. The protons on the carbon at 8 43.8, 8 1.75 and 1.56 (shown by HMQC), gave HMBC correlations into both olefinic quaternary carbon resonances at 8 125.5 and 133.4. In contrast, the protons on the carbon at 8 34.7,8 1.35  Ill (shown by HMQC), gave an HMBC correlation into only one of these olefinic quaternary carbons, that at 8 133.4. It follows that the resonance at 8 43.8 must be assigned to C9, while that at 8 34.7 must be assigned to C7. Similarly, the carbon resonance at 8 133.4 can unambiguously be assigned to C5, and that at 8 125.5 to CIO.  These assignments are confirmed by the strong  COSY correlation from H7 (8 1.35) into a resonance at 8 1.94 (H6), while the H9 protons (8 1.75 and 1.56) showed only weak allylic coupling into H6. An HMQC correlation from H6 (8 1.94) into a carbon resonance at 8 21.4 allowed for the assignment of this resonance to C6. The methyl singlets at 8 0.94 and 0.95 in the proton spectrum of nanaimool must be assigned to CI3 and C14 (since the previously assigned CI5 methyl shown an HMQC correlation into the methyl singlet at 8 0.85). From HMQC correlations, CI3 and C14 were assigned to resonances at 8 27.8 and 28.0; it was not possible to distinguish in an absolute way between these two methyls, and thus the labels "C13" and "C14" are to a certain extent arbitrary, though this does not affect the biosynthetic feeding results presented here. These methyl protons showed HMBC correlations into a quaternary carbon resonance at 8 33.5 (assigned to C4) and a methylene resonance at 839.8 (assigned to C3). An HMQC correlation from this carbon at 839.8 (C3) into a proton resonance at 8 1.41 revealed the identity of the H3 protons. This H3 resonance at 8 1.41 showed a COSY correlation into a proton resonance at 8 1.57, which must be assigned to H2. HMQC showed these H2 protons to be attached to a carbon at 8 19.4 (C2). The final carbon resonance at 8 31.7 was assigned to C I , which was confirmed by the COSY correlation from the H2 protons (8 1.57) into a resonance at 8 1.78 (HI) which showed an HMQC correlation into 8 31.7 (CI). The assignment of all the carbons in isoacanthodorol (43) was accomplished in a similar fashion.  The H12 protons at 8 3.67, correlated to a carbon at 8 60.1 (C12) in the HMQC  spectrum (Figs. 58,59), showed a COSY correlation (Fig. 60) into two protons at 8 2.03 and 1.39, which must be the HI 1 protons. These HI 1 resonances showed HMQC correlations into the carbon resonance at 8 46.8, which could thus be assigned to CI 1. An HMBC correlation (Figs. 61-63) from H12 (8 3.67) into this carbon at 8 46.8 (CI 1) clearly confirmed this assignment. It was also clear from the chemical shift that the proton resonance at 8 5.08 must be assigned to H9, and from the HMQC it could be shown that C9 could be assigned to a resonance at  OH  (ppm) 5  4  3  2  1  Fig. 58: HMQC Spectrum of isoacanthodorol (43) [500 MHz, CDC1 ] 3  OH  —i  1  (ppm)  1  1  1.8  1  1  1.6  1  T  1.4  1  1  1.2  1  1  1.0  1  1—-  0.8  Fig. 59: Expanded Upfield Region of HMQC Spectrum of isoacanthodorol (43) [500 MHz, CDC1 ] 3  (ppm) 3.5  3.0  2.5  2.0  1.5  Fig. 60: COSY Spectrum of isoacanthodorol (43) [500 MHz, CDCL]  Fig. 61: Expansion of Upfield Region of HMBC Spectrum of isoacanthodorol (43) [500 MHz, CDC1J  118  Fig 62- Expansion of Downfield 1H Region of HMBC Spectrum of isoacanthodorol (43) [500 MHz, CDC1 ] 3  Fig 63: Expansion of Downfield "C Region of HMBC Spectrum of isoacanthodorol (43) [500 MHz, CDC1,]  120 8131.6 in the C spectrum. The proton resonance of H9 showed HMBC correlations into a 13  methyl carbon at 8 23.4, which was assigned to CI5. HMQC showed that the methyl protons of C15 were at 8 1.59. This H15 resonance showed HMBC correlations into carbon resonances at 8 131.6 (C9), 8 134.1 (clearly assigned to C8) and 8 29.2 (which must neccessarilybe CI). The C7 carbon resonance at 8 29.2 gave an HMQC correlation to a proton resonance at 81.90; this H7 resonance showed HMBC correlations into carbons at 8 134.1 (C8) and a methylene carbon at 8 20.1, which was therefore assigned to C6. The H6 protons at 8 1.88 and 1.68, as shown by HMQC, showed COSY correlations into a resonance at 8 1.20, which was assigned to H5. From this H5 resonance (8 1.20) it was straightforward to assign C5 to a resonance at 8 45.5 from the HMQC spectrum. As was argued for nanaimool, the methyl resonances at 8 0.86 and 0.98 for Me 13 and Mel4 were otherwise indistinguishable and were assigned to carbons at 8 26.7 and 32.C, from the HMQC spectrum. Both these methyl resonances at 8 0.98 and 0.86 showed identical HMBC correlations into carbon resonances at 8 45.5 (C5), 34.1 ( assigned to C4, since this is a quaternary carbon i.e no correlations in the HMQC) and 8 45.5 ( assigned to C3, since this is a methylene carbon in the HMQC). It was possible to distinguish CI from C2 based on the HMBC correlation from HI lb into a carbon at 8 37.9, which must therefore be CI. The HMQC spectrum provided the connectivities of these protons to their respective carbons: C2 (8 19.5) and CI (8 37.9). The final carbon at 8 37.4, and the only remaining aliphatic quaternary carbon in isoacanthodorol, was assigned to C10. Shown in Figure 64 are the truncated C NMR resonances, all normalized to the same 13  peak height for the central singlet component, for all the carbon atoms in both a labeled sample and an unlabeled control sample of 41. All of the the resonances shown in the C NMR spectrum ,3  of labeled 41 clearly show flanking doublets resulting from the incorporation of C labeled 13  acetate units. An analysis of the intensity of the flanking doublets (Table 5) indicates that there is one set of relatively intense doublets flanking the resonances assigned to C-2, C-3, C-5, C-6, C8, C-9, C-10, C-11, C-12, C-13 and C-15 (average specific incorporation 0.35 %) and another set of relatively weak doublets flanking the resonances for C-1, C-7 and C-14 (average specific  incorporation 0.11 %). Numbers listed for specific incorporations = % enrichments above natural abundance = 1.1 % X (combined integrated peak area of enriched satellites minus the combined theoretical peak area for these same satellites resulting from natural abundance coupling)/(peak height of the natural abundance singlet plus the combined theoretical peak area for all satellites resulting from natural abundance coupling). The probability of observing natural abundance  Table 4: H , C , COSY, and HMBC NMR Data for nanaimool (41) and isoacanthodorol (43) 1  1 3  nanaimool (41)  isoacanthodorol (43)  c  8 C(  #  ppm)  (500  (125  MHz)  13  8H l  COSY  HMBC  8* C(  8H  COSY  HMBC  (400MHz)  (500MHz)  ppm)  (500  (400MHz)  (500MHz)  (125  MHz)  37.9  H2,3  2  19.5  3  40.1 .  4 5 6  34.1 45.5 20.1  7 8  29.2 134.1  9  131.6  10 11  37.4 46.8  12 13 14 15  60.1 26.7 32.0 23.4  !  MHz)  MHz) 1  3  1.53 1.22 1.43 1.26 1.30 1.14  H2.1'  H2  H2',3,l  31.7  1.78  H2  19.4  1.57  H2,3  1.41  H2  HI,13,14  1.94  H7,l,9  H2,3,13,14 H6,l,9,7,13,14 H7  H2  H13.14  39.8  1.20 1.88 1.68 1.90  H6 H5,6'  H13.14 H13,14 H7  33.5 133.4 21.4  H6  H9 H7.15  34.7 30.7  1.35  H6  5.08  H7,15  H15  43.8  1.75 1.56  H9'  H6.9.9', 11,15 HI 2,6,9,9', 11,7 ,15 Hll.7,15  2.03 1.39 3.67 0.86 0.98 1.59  H12 H12 HI 1,11"  Hll H12  125.5 44.0  1.50  H12  H6,1,9,2 H12,9,9',7,15  59.6 28.0 27.8 24.8  3.70 0.95 0.94 0.85  Hll  H14 H13 H9  H3,14 H3.13 H9,9',ll,7  coupling between a pair of adjacent carbons is 0.011 X 0.011 = 0.000121. The intensity of this signal would be split between the two doublet components resulting in a predicted abundance of 0.0000605 for each component. Thus the doublet components should each be (0.0000605/(0.011 n X 0.000121)) X 100 =0.55% of the intensity of the unenriched central singlet for each coupling interaction, (n is the number of next neighbor carbons and it must have values between 1 and 4.  122  Fig 64: Truncated Normalized "C Resonances for Labeled (left) and Unlabeled (right) samples of nanaimool (41) [125 MHz, CDC1] 3  Therefore, 0.011 - n X 0.000121 is always = 0.011.). In practice the corrections for natural abundance coupling that have been included in the above calculation for specific incorporation are so small that they have a negligible effect on the resulting values. The specific incorporation values for the weak doublets (i.e. C-l, C-7 and C-14 in labeled 41) are only approximate because the contribution to the intensity of the central singlet resulting from enrichment with labeled acetate could not be reliably determined by comparing peak heights or areas in the C NMR spectra of 13  labeled and unlabeled samples of 41 or 43 for these low levels of incorporation. A complete analysis of the C/ C coupling constants (Table 5) for the intense doublets revealed the pattern of 13  13  intact acetate unit incorporation for nanaimoal indicated in 38a, which is consistent with the expected biogenesis from mevalonic acid following either pathway a,c or a,b,d shown in Scheme 6.  Scheme 6  124  C1  ilM'I'I'I'I'M 38.0 38.0 (ppm) (pprn)  J  m  C9  C8  C7  'I'l'I'I'I'I'M  I > I' I' MI' I • I'  ''li'^o''  I'l'I'I'IM' 1  134.0  29.0 29.0  (ppm)  (ppm) (ppm)  MM't'l'  131.6  131.6  (ppm)  (ppm)  (ppm)  • I' I • I • I' I • I • I • ' 37.0 37.0 (ppm)  (ppm)  C14  TT 111111111' i' i  nifliMMifffi  32.0  32.0  (ppm)  (ppm)  ln'i'i'i'm  (iini|ii'PMi  46.4  46.4  (ppm)  j  C1 3  CI 2  C1 1  CIO  (ppm)  //U WimH'i'i'i  ifl'fMU'i'l'i  60.0  60.0  (ppm)  (ppm)  C1 5  Minn'i'i'i'i  «»f • • • • • • • • • > i  23.0  23.0  (ppm)  (ppm)  Fig 65: Truncated Normalized C Resonances for Labeled (left) and Unlabeled (right) samples of isoacanthodorol (43) [125 MHz, CDC1 ] l3  3  26.6  mm 26.6  (ppm) (ppm)  125 Biogenetic pathways a,c or a,b,d (Scheme 6) predict that the C label incorporated at C13  1, C-7, and C-14 in labeled (41) should not be part of intact acetate units. Therefore, if only one labeled acetate unit was incorporated per labeled nanaimool (41) molecule the resonances for carbon atoms C-1, C-7, and C-14 should appear only as enriched singlets. The most likely explanation for the weak flanking doublets observed in the C-1, C-7 and C-14 resonances of labeled 41 is that they originate from incorporation of more than one C labeled acetate unit into a 13  single nanaimoal (38) molecule. A. nanaimoensis specimens were starved during the 16 day 62  injection period and this could have led to a highly labeled acetate pool and a reasonable probability of more than one labeled acetate unit being incorporated into individual mevalonic acid molecules and/or more than one labeled mevalonic acid unit being incorporated into individual molecules of 38. As a result, the C-7 doublet shown in Figure 65 would arise from molecules of 41 having a singly enriched carbon at C-7 and an intact acetate unit at either C-5/C-6 or C-8/C-15. Similar clusters of three adjacent C labeled carbon atoms in single molecules of 41 could lead to the 13  couplings observed in C-1 and C-14. Although the arguments presented above account for all of the features of the weak flanking doublets observed in the C NMR spectrum of 4 1 , alternate 13  origins for these doublets involving different incorporation patterns of intact acetate units into isopentenyl pyrophosphate and dimethylallyl pyrophosphate cannot be completely ruled out with the current data. The C NMR spectrum of labeled isoacanthodorol (43) also showed evidence for 13  significant levels of labeled acetate incorporation (Figure 65). Once again there was one set of relatively intense doublets and another set of relatively weak doublets resulting from incorporationof isolated intact acetate units and incorporation of more than one adjacent labeled acetate unit per molecule, respectively. An analysis of the coupling constants observed for the . complete set of intense doublets (Table 5) revealed the pattern of acetate incorporation shown in 40b in Scheme 6. Of particular significance was the obvious difference in intensity of the doublets flanking the C-11 and C-12 resonances, which provided a clear demonstration that C-11 and C-12 do not arise from an intact acetate unit (i.e. compare C-1 l/C-12 in Figure 66b with C-1 l/C-12 in Figure 66a).  This result was unexpected and it ruled out the proposed pathway / to the  127 isoacanthodorane skeleton shown in Scheme 6. An alternate pathway g,h,i to the rearranged isoacanthdorane skeleton, that is consistent with the observed acetate incorporation pattern, is presented in Scheme 6b. It is interesting to note that the pathway g,h,i to isoacanthodoral also proceeds through a tricyclic cyclobutane containing intermediate (i.e. 44).  The conservation of  the oxygen from farnesyl pyrophosphate via an eneal intermediate results in the correct oxidation level for C12 in isoacanthodorol (43). Table 5: C NMR data for labeled nanaimool (41) and isoacanthodorol (43) recorded in at 125 MHz.. 1 3  Nanaimool (41) c#  8 C(ppm)  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15  (125MHz) 31.7 19.4 39.8 33.5 133.4 21.4 34.7 30.7 43.8 125.5 44.0 59.6 28.0 27.8 24.8  13  Isoacanthodorol (43)  J (Hz)  Specific  8 C(ppm)  33.4 33.4 33.4 35.3 41.0 41.0 34.3 36.2 43.9 42.0 34.3 37.2 35.3 35.3 35.3  Incorporation 0.12 0.33 0.33 0.18 0.51 0.32 0.13 0.51 0.49 0.25 0.26 0.37 0.28 0.08 0.34  (125MHz) 37.9 19.2 40.1 34.1 45.5 19.9 29.0 134.1 131.6 37.4 46.8 60.1 26.7 32.0 23.4  cc  CDCI3  13  J  c c  (Hz)  29.6 33.4 33.4 35.3 34.3 33.4 36.2 42.9 41.0 33.5 34.3 37.2 35.3 35.3 43.9  Specific Incorporation 0.11 0.28 0.27 0.49 0.28 0.31 0.12 0.54 0.10 0.47 0.29 0.07 0.31 0.09 0.46  In light of the weak doublet at C-12, some further discussion as to the origin of the 'weak doublets' is in order. It has been argued that although interacetate coupling seems the most likely explanation for these doublets, in the case of nanimool the data could not completely discount other possibilities (i.e. a lack of fidelity in the JPP isomerase). This explanation cannot account, however, for the doublet at C-12 since it clearly is not part of an intact acetate unit with C-11. That is, in the proposed biogenetic scheme, C-11 and C-12 in isoacanthodoral at no point belong to the same mevalonic acid unit, and thus coupling between them cannot be accounted for as arising from an intact acetate that has been incorporated into mevalonic acid in the other 'direction'. The  128 only way that C-12 in compound 43 can show doublet coupling is if there is an intact acetate unit at C-10/C-11, as indicated by the C-11 flanking doublet, and a singly labeled carbon at C-12 in individual molecules of 43. What, then, is the probability of observing interacetate coupling, and does this match with the doublet intensities observed?  40b  Scheme 6b The probability of two adjacent acetate units originating in the pool of C labeled acetate ,3  would be the product of their independent probabilities for incorporation as given by their specific incorporations. Taking the highest specific incorporation (0.5%), the highest probability of seeing inter-unit coupling would be 0.005 X 0.005 = 0.000025. Dividing this by the intensity of the natural abundance singlet (0.011) gives 0.00227 as the percentage intensity of the doublets. Thus, in the case of inter-acetate unit coupling, each doublet signal should be only 0.114% of the centerline intensity, which is substantially smaller than the intensity observed for the 'weak doublets' in isoacanthodorol.^^ However, this calculation is only correct if one assumes that all of the molecules isolated in the labeling experiment were in fact made during the injection period such that the pool of isolated metabolites was uniformly labeled. While this may be a safe assumption in a microbial stable isotope labeling study, where the initial pool of metabolites is  essentially zero and all metabolites isolated are indeed biosynthesized during the experiment, such is not neccessarily the case in an invertebrate feeding experiment.  In this case, the animal has a  sizable pool of unlabeled metabolites at the outset of the experiment and despite the multiple injection hypothesis for increased turnover, only a small amount of new molecules are made in the presence of the isotopically labeled precursor. In this sense, the products isolated at the end of the injection period are made up of two components: molecules that were present at the outset which are completely unlabeled, and molecules which have been made during the experiment and are therefore labeled. If the concept of 'specific incorporation' is adapted to accomodate a background of unlabeled material, it is possible to calculate the probability of inter-acetate unit couplings that is in good agreement with the observed intensities of the weak doublets in isoacanthodorol. The observed specific incorporation at any particular site, in the case of an invertebrate feeding study, must be a weighted average of the incorporation in the molecules present at the outset of the experiment (i.e. zero) and the specific incorporation in the labeled molecules. This can be expressed by the following equation: (1)  S.I.(z) + 0(1-z) = observed specific incorporation for the intense doublets  where S.I. = the specific incorporation in the newly formed metabolites that were made during the labeling experiment; z = the fraction of molecules made during the labeling experiment. That is, all equation (1) is saying is that the observed specific incorporations of the intense doublets in Table 5 are a measure of the true and as yet unknown specific incorporation, S.I., when multiplied by that (unknown) fraction of molecules that are in fact labeled. Furthermore, in this scenario the expected observed intensity of the weak doublets is given by equation (2): (2)  (S.I.) z + 0(1-z) = observed intensity for the weak doublets 2  where (S.I.) = the probability of having two adjacent C labeled carbons in the molecules 2  formed during the labeling experiment; z = the fraction of labeled molecules;  13  130 (1-z) = fraction of unlabeled molecules present at the outset of the feeding experiment. Equation (2) is merely an extension of the first equation used to calculate the probabilty of the observed interacetate couplings at the outset of this discussion, which now takes into account the presence of unlabeled material. If we solve equations (1) and (2) for S.I. and z, using a value of 0.5% (as before) for the observed specific incorporation of the intense doublets (i.e the right-hand side of equation (1)), and a value of 0.08% for the observed intensity of the weak doublets (this is the value for C-14 in nanaimool), we get S.I. = 0.16 and z = 0.03. That is, the specific incorporation into the labeled nanaimool is actually 16%, but only 3% of the molecules in the entire pool are labeled as such. These numbers are completely consistent with the expectation that the acetate pool in the nudibranchs during the injection period is highly labeled and that molecules made from this pool should also be highly labeled. In such highly labeled molecules there is good literature precedent for observing two labeled acetate units incorporated next to each other in the same molecule.  62  Moreover, the calculted values for S.I. and z are consistent with the expectation that only a relatively small percentage of the metabolite pool is being replaced in these experiments, which is in good accord with the results of the conclusions drawn in previous radioisotope labeling experiments with marine invertebrates. Incorporation of C labeled acetate into nanaimoal (38) and isoacanthodoral (40) has thus 1 3  demonstrated that A. nanaimoensis is capable of de novo terpenoid biosynthesis and a detailed analysis of the incorporation patterns has uncovered an unanticipated rearrangement in the biogenetic pathway to the isoacanthodorane skeleton. A detailed analysis of the specific incorporations of doubly labeled acetate has afforded a better measure of the extent to which biosynthesis may be assesed against a background of unlabeled material. It is hoped that further research in our group, now that the way has been paved for experiments with more advanced precursors, may probe more subtle aspects of the biosynthesis of compounds 38 to 40.  Endnotes: Chapter IV: Biosynthetic Studies of Isoprenoid Secondary Metabolites from Dorid Nudibranchs Using Stable Isotopes 1.)  For example: Pika, J.; Faulkner, D.J., Tetrahedron, 1995, 51(30), 8189-8198.  131 2. )  Faulkner, D.J., Nat. Prod. Rep., 1995,12, 223-269, and previous reviews cited therein.  3. )  Kobayashi, M ; Nakagawa, T.; Mitsuhashi, H., Chem. Pharm. Bull. (Tokyo), 1979, 27,  2382-2387.  4. ) Fujiki, H.; Suganuma, M.; Tagaki, K.; Nishiwaki, S.; Yoshizawa, S.; Okabe, S.; Yatsunami, J.; Frenkel, K.; Troll, W.; Marshall, J.A.; Tius, M.A., in Phenolic Compounds in Food and Their Effects on Health IL eds. Huang, M.-T.; Ho, C.-T.; Lee, C.Y., American Chemical Society, Washington, D C , 1992, pp. 380-387. 5. ) Tanaka, J.; Higa, T., Suwanborirux, U.; Kokpol, G; Bernardinelli, G.; Jefford, /. Org. Chem., 1993, 58, 2999. 6. )  Higa, T., personal communication.  7. ) Look, S.A.; Fenical, W.; Jacobs, R.S.; Clardy, J., Proc. Natl. Acad. Sci. USA, 1986, 83, 6238-6240. 8. )  Sullivan, B.; Faulkner, D.J., Tet. Lett., 1982, 23, 907-10.  9. )  Piers, E.; Wai, J.S.M., Can. J. Chem., 1994, 72, 146-157.  10. )  DeSilva, E.D.; Scheuer, P.J., Tet. Lett., 1980, 21, 1611-1614.  11. ) Glaser, K.B.; DeCarvalho, M.S.; Jacobs, R.S.; Kernan, M.R.; Faulkner, D.J., Molecular Pharmacology, 1989, 36, 782-788. 12. ) Komori, A.; Suganuma, M.; Okabe, S.; Zou, X.; Tius, M.A.; Fujiki, H., Cancer Research, 1993, 53, 3462-3464. 13. )  Garson, M.J., Nat. Prod. Rep., 1989, 6, 143-170.  14. )  ibid.  15. )  Trench, R.K., Annu. Rep. Plant. Physiol, 1979, 30, 485.  16. )  Papastephanou, C ; Anderson, D.G., Comp. Biochem. Physiol., 1982, 73(B), 617-624.  17. )  Cane, D.E., Methods EnzymoL, 1985,110, 383.  18. )  For example: Thompson, J.E.; Murphy, P.T.; Berquist, P.R.; Even, E.A., Biochem Syst.  19. )  Mann, J., Chemical Aspects of Biosyntheis. Oxford University Press, 1994, Ch.l.  Ecol, 1987, 75, 595.  20. ) Ruppert, E.E.; Barnes, R.D., Invertebrate Zoology. Saunders College Publishing, 6th ed., 1994, p. 80. 21. ) Riley, J.P.; Chester, R., Introduction to Marine Chemistry. Academic Press, New York, 1971. 22. )  Vederas, J.C., Nat. Prod. Reports, 1987, 4, 277-337.  23. ) For example: Cimino, G.; DeRosa, S.; DeStefano, S.; Sodano, G.; Villani, G., Science, 1983,2/9, 1237-1238.  132 24. )  Tymiak, A.A.; Rinehart, K.L. Jr., J. Am. Chem. Soc, 1981,103, 6763.  25. )  ibid.  26. ) Spurgeon, S.L.; Porter, J.W., in Biosynthesis of Isoprenoid Compounds Vol.1, cds. Porter, J.W.; Spurgeon, S.L., John Wiley & Sons, New York, 1981, Ch. 1, pp. 1-46. 27. ) I am indebted to Prof. Mary Garson's reviews for much of the material covered in this introduction; c.f. Garson, M.J., Nat. Prod. Rep., 1989, 6, 143-170; Garson, M.J., Chem. Rev., 1993,93, 1699-1733. 28. ) all material in this introductory section is from Spurgeon, S.L.; Porter, J.W., in Biosynthesis of Isoprenoid Compounds Vol.1, eds. Porter, J.W.; Spurgeon, S.L., John Wiley & Sons, New York, 1981. 29. )  Iengo, A.; Pecoraro, C ; Santacroce, C ; Sodano, C , Gazz. Chim. Ital, 1979,109, 701-  30. )  Herbert, R.B.; Mann, J., J. Chem. Soc, Chem. Commun., 1984, 1474-1475.  31. )  Garson, M.J., J. Chem. Soc, Chem. Commun., 1986, 35-36.  32. )  Karuso, P.; Scheuer, P.J., J. Org. Chem., 1989, 54, 2092-2095.  702.  33. ) Coll, J.C.; Bowden, B.F., Tapiolas, D.M.; Willis, R.H.; Djura, P.; Streamer, M; Trott, L, Tetrahedron, 1985,4/(6), 1085-1092. 34. )  Dai, M.C.; Garson, M.J.; Coll, J.C., Comp. Biochem. Physiol., 1991, 99(B), 775-783.  35. ) Faulkner, D.J., in Ecological Roles of Marine Natural Products, ed. Paul, V.J., Cornell University Press, Ithaca, New York, 1992, Ch. 4. 36. ) 37. ) 1976,  38. )  Faulkner, D.J.; Ghiselin, M.T., Mar. Ecol. Prog. Ser., 1983, 13, 295-301. Kubo, I.; Pettei, M.; Pilkiewicz, F.; Nakanishi, K., J. Chem.Soc, Chem. Commun., 1013. see references 40 to 42 below.  39. ) Dumdei, E.J.; Flowers, A.E., Garson, M.J.,Moore, C.J., Comp. Biochem. Physiol. C, 1996, in press. 40. )  Faulkner, D.J.; Molinski, T.F.; Andersen, R.J.; Dumdei, E.J.; DeDilva, E.D., Comp.  Biochem. Physiol., 1990, 97(C), 233-240.  41. ) a) Cimino, G.; DeRosa, S.; DeStefano, S.; Sodano, G., Tet. Lett., 1981, 22, 1271; and b) Cimino, G.; DeRosa, S.; DeStefano, S.; Sodano, G., Pure Appl. Chem., 1986, 58, 375. 42. ) Cimino, G.; DeRosa, S.; DeStefano, S.; Morrone, R.; Sodano, G., Tetrahedron, 1985, 41, 1093. 42. )  Okuda, R.K.; Scheuer, P.J.; Hochlowski, J.E.; Walker, R.P.; Faulkner, D.J., J. Org.  Chem., 1983, 48, 1866.  44.) Cimino, G.; DeRosa, S.; DeStefano, S.; Sodano, G.; Villani, G., Science, 1983, 219, 1237-1238.  133 45. )  Andersen, R.J ; Sum, F.W., Tet, Lett., 1980, 27, 797.  46. )  Gustafson, K.; Andersen, R.J., Tetrahedron, 1985, 41, 1101-1108.  47. ) Gustafson, K.; Andersen, R.J.; Chen, H.M.H.; Clardy, J.; Hochlowski, J., Tet. Lett., 1984,25, 11-14. 48. )  Krug, P.J.; Boyd, K.G.; Faulkner, D.J., Tetrahedron, 1995,57, 11063-11074.  49. ) Cimino, G.; Gavagnin, M; Sodano, G.; Puliti, R.; Mattia, C.A.; Mazzarella, L., Tetrahedron, 1988, 44, 2301-2310. 50. )  Cimino, G.; Crispino, A.; Gavagnin, M.; Trivellone, E.; Zubia, E.; Martinez, E.; Ortea,  J., J. Nat. Prod., 1993,56, 1642-1646.  51. )  Soriente, A.; Sodano, G.; Reed, K.C.; Todd, C , Nat. Prod. Lett., 1993, 3, 31.  52. ) Zubia, E.; Gavagnin, M.; Crispino, A.; Martinez, E.; Ortea, J.; Cimino, G., Experientia, 1993, 49, 268-27'1. 53. )  see Appendix B  54. ) a) Davies-Coleman, M.T.; Faulkner, D.J., Tetrahedron, 1991,47, 9743-9750; and b) Gavagnin, M.; Trivellone, E.; Castelluccio, F.; Cimino, G., Tet. Lett., 1995,56, 7319-7322. 55. )  Avila, C ; Ballesteros, M.; Cimino, G.; Crispino, A.; Gavagnin, M; Sodano, G., Comp.  Biochem. Physiol., 1990, 97(B), 363-368.  56. ) a) Ayer, S.W.; Hellou, J.E.; Tischler, M.; Andersen, R.J., Tet. Lett., 1984, 25, 141-144; and b) Ayer, S.W.; Andersen, R.J.; Cun-heng, H.; Clardy, J, J. Org. Chem., 1984, 49, 26532654. 57. ) For example: Grue-S0rensen, G.; Spenser, I.D., J. Am. Chem. Soc, 1993, 775, 20522054. 58. )  Wright, J., personal communication.  59. ) I am indebted to J. Gerard, J. Kubanek, M. LeBlanc, F.M. Kong, T. Barsby, Y. Khan, and the staff at the Bamfield Marine Station for their assistance in the collection of all the B.C. nudibranchs. 60. ) Liposomes (9:1 HSPC: egg PG multilammelar vesicles, 2.4 u. particle size, trap volume = 1.1 u;mol Na[ ' - C2] acetate/ (xmol lipid) were prepared by Chris Hanson in the laboratory of Prof. T. Allen, Dept. of Pharmacology, University of Alberta. ]  2  ,3  61. )  Cimino, G.; Sodano, G., Chem. Scri., 1989, 29, 389.  62. )  For a precedent see: Needham, J.; Hu, T.; McLachlan, J.; Walter, J.; Wright, J., J.  Chem. Soc, Chem Commun., 1995, 1623.  63. ) This analysis greatly benefitted from a correspondance with an anonymous referee during the review process of this work for publication: cf. Graziani, E.L; Andersen, R.J., J. Am. Chem. Soc, 1996, 118, 4701-2.  134  V.  Stable Isotope Investigations on the Biosynthesis of Triophamine by Triopha  catalinae  Fig. 67: Color Plate of Triopha catalinae  13 V.I. Introduction to Polyketide Biosynthesis  The primary building-block of polyketide biosynthesis is acetate in the form of acetyl-CoA, which is further activated to malonyl CoA via carboxylation carried out by transfer of a carboxyl group from the cofactor biotin and its associated enzyme biotin carboxyl carrier protein (BCCP). The conversion of acetyl CoA to malonyl CoA is shown in Scheme 7. The production of complex natural products from these acetate units is under the direction of a large, multifunctional enzyme called a polyketide synthase (PKS) and proceeds in a manner analogous to fatty acid biosynthesis. O  Scheme 7: Formation of malonyl CoA from acetyl CoA  A unit of malonyl CoA is bound to the active site of a cofactor, the acyl carrier protein (ACP), where it condenses with another acetyl CoA bound to the PKS. The resultant diketide bound to the ACP can then undergo some or all of a number of further reactions, depending upon the oxidation level required at "C3" of the biosynthetic intermediate. Reduction of the C3 carbonyl by  136 a ketoreductase (KR) enzyme (usually a subunit of the PKS itself) via hydride delivery of the pro4S hydrogen of NADPH gives the stereospecific R-hydroxythioester product. (In polyketide synthesis, in contrast to fatty acid biosynthesis, alternate stereochemistries can be generated at C3 if needed). Further reduction via dehydration and .ryrt-elimination gives the (E)-unsaturated thioester. Complete saturation is accomplished by a second reduction via transfer of the pro-4R hydrogen of NADPH. The condensation of malonyl CoA and acetyl CoA, and the subsequent functional group interconversions the resulting polyketide chain can undergo, are shown in Scheme 8.  i  Jo , co 2 A  -£  pks  h  pv  S—PKS  / ^ s r ^ A  \  O O ACP-SH O^^^^SCoA  .  -1+  O\ O O ^ ^ ^ ^ S - A C P  O  O  X X  ^ ^ ^ ^ S - A C P NADPH  P  NADPH ^S-ACP^  H  j?  x ^ X ^ s - A C P  HO H O  ^^N>-ACP  Scheme 8: The Principal Reactions of Polyketide Biosynthesis  At any point in this process, the four carbon intermediate can be transferred back to the KS where another condensation with malonyl CoA occurs, thus extending the growing "polyketide" chain to six carbons. It was originally thought that the polyketide chain was assembledfirstby multiple condensations of acetate, and only then did the various reductions that give rise to the appropriate functionalization in the product molecule occur. Contemporary evidence, however,  13  overwhelmingly favors the model of a processive process, whereby the correct functionalization is established before chain elongation. In cases where an odd number of carbons are present in the polyketide metabolite, the cofactor S-adenosylmethionine often serves as a biogenetic source of a one-carbon electrophile in the formation of carbon-carbon bonds, as well as the formation of OMe and NMe moieties. This reactivity is demonstrated in Scheme 9.  OH OH  Scheme 9: S-Adenosylmethionine (SAM) Methylation  V.II. Precedents for a Polyketide Origin of Triophamine  Skin extracts of the British Columbia dorid nudibranch Triopha catalinae always contain triophamine (9) as the principal secondary metabolite.2>3 Triophamine has also been isolated from the related B.C. species, Polycera tricolor? natural sources of the compound.  Until recently, these were the only known  In the course of the investigations detailed in this thesis,  triophamine (9) was also isolated from a Chilean dorid, Thecacera darwinii (see Appendix B). The structurally similar triophamine analogue, limaciamine (8) was also isolated from the North Sea dorid, Limacia clavigera, as has been discussed in Chapter JJI. In light of the fact that triophamine (9) is consistently present in specimens of T. catalinae collected all the way from  138 Alaska to California , the geographical invariance hypothesis would suggest that T. catalinae is a 3  3  prime candidate for biosynthetic study.  O  NH  2  O  6  8  (8)  o 6  8  10  (9)  Investigations into the biosynthesis of triophamine (9) using stable isotopes affords an interesting extension of the previous work on terpenoids, as described in the preceeding chapters. The only other naturally occuring monoacylguanidine, arenaine (45) from the seeds of the terrestrial plant Plantago arenaria, is presumed to originate from guanidine (46) and the linalool5  derived acid (47), which is itself of isoprenoid origin. Although the acyl portion of triophamine contains ten carbons, an isoprenoid origin for triophamine (9) would involve unprecedented rearrangements of mevalonic acid in order to give rise to such a moiety. While such an origin for the acyl portion of triophamine (9) seems highly improbable, the use of acetate as a precursor could nonetheless provide evidence for such an unlikely pathway. However, it seems much more probable that the C]0 acid of triophamine (9) is polyketide in origin. Stable isotope methodology had demonstrated clearly the de novo synthesis of both sesquiterpenoids and diterpenoids in dorid nudibranchs; could the same methodology be extended to a metabolite of likely polyketide origin? The use of stable isotopes to probe the biosynthesis of polyketide metabolites in microbial extracts is widespread; studies on the biosynthesis of  139 oncorhyncolide^ and okadaic acid? from marine bacteria and dinoflagellates are examples of such studies in a marine context.  (46)  As has been pointed out previously, there is a world of difference in performing biosynthetic experiments with invertebrates as compared with microorganisms. It was hoped that the stable isotope methodologies developed to study terpene biosynthesis in dorid nudibranchs could be extended to provide conclusive results for the de novo biosynthesis of triophamine (9). While there exists no literature precedent for polyketide biosynthesis in dorid nudibranchs, there are a number of studies on the biosynthesis of polyketide metabolites from related molluscs. The first report of de novo polyketide biosynthesis by a marine mollusc was the incorporation of [l-^C]propionate into denticulatins A (48) and B (49) by the pulmonate Siphonaria denticulata.% The authors proposed that the highly methylated skeleton of (48) could arise as a result of condensation of propionate units, or from SAM methylation of an acetate derived backbone. Propionyl CoA can be built either from succinyl CoA via methylmalonyl CoA, or from methylmalonyl CoA via the degradation of L-valine.9 The results of the denticulatin feeding study strongly support the role of propionate as a building block, rather than subsequent methylation of an acetate derived skeleton, consistent with the results of extensive studies into the biosynthesis of the polyketide antibiotic erythromycin (50) from propionate 10. More recent studies* 1 on the biosynthesis of the related compounds, siphonarins A (51) and B (52), from the  Scheme 10 : Biogenetic Routes to the denticulatins (48 and 49)  141  (50) erythromycin A : Ri= OH, R = L-cladinose 2  (51) R = H (52)R = Me  marine pulmonate Siphonaria zelandice have demonstrated the incorporation of both [ l - ^ C ] propionate and [2,3-^^C] succinate into these compounds. Biosynthetic feeding studies^ using [2-^C] propionate have also demonstrated the de novo synthesis of the cyercenes (53 to 59) from the ascoglossan (herbivorous opisthobranch), Cyerce cristallina.  Similarly, incorporation of [ l - l ^ C ] propionate into the polypropionate  metabolite, elysione (60), by the ascoglossan Elysia viridis has also been demonstrated recently.^ Curiously, incorporation studies^ f [l-^c] propionate into polypropionate metabolites (61, 0  62) from a related ascoglossan, Elysia timida, yielded inconclusive results. In addition to this experimental evidence for some polypropionate biosynthesis in Ascoglossan molluscs, these opisthobranchs have also provided a stunning example of symbiotic biosynthestic capabilities. A number of studies^, 16  n a v e  demonstrated the ability of many species of the  family Elysioidea to incorporate functioning chloroplasts from the algae in their diet, giving the mollusc photosynthetic capabilities.  Whether or not the compounds synthesized in these  chloroplasts are of algal or molluscan origin is no longer of any relevance, and these animals provide a unique example of co-operative biosynthesis in marine molluscs. Another herbivorous opisthobranch has been demonstrated to possess biosynthetic capability. Incorporation of [l-l^C] acetate into navenones B and C (63,64) by the anaspidean opisthobranch, Navanax inermis, was demonstrated. 17 These compounds are believed to be trail pheromones.  14  (63) : R = H (64) : R = OH  The only other precedent for non-isoprenoid de novo biosynthesis by a nudibranch is the biosynthesis of prostaglandin derivatives, prostaglandin 1,15-lactones (65), from arachidonic acid (66) by the dendronotid nudibranch, Tethys fimbria.^%  Prostaglandins are eicosanoids  (unsaturated C20 acids) derived from peroxidic intermediates of fatty acid (usually arachidonic acid) oxidation. Arachidonic acid is itself derived from linolenic acid, which is biosynthesized from acetate units in exactly the same manner as long polyketide chains are produced. *9 While an exhaustive discussion of prostaglandin biosynthesis is beyond the scope of this thesis, the  demonstration of prostaglandin synthesis by a nudibranch provides further evidence for the existence of acetate-derived pathways operating in molluscs. Given these precedents for polyketide metabolism in molluscs and hence the probable polyketide origin of triophamine (9), it is useful to postulate the possible pathways by which triophamine may be assembled in order to design the most efficient feeding experiment for proving  144  which such pathways are in operation. Scheme 11 shows the possible origins of the carbons of the acyl portion of triophamine (9), and in particular the ethyl branches at C2 and C4. The ClO moiety of triophamine can result from : i) the condensation of two butyrates and one acetate (pathway a), or ii) the condensation of two propionates and one acetate followed by SAM methylation at the two methyl branches (pathway b). Butyrate derived polyketides are not without precedent in the biosynthetic literature, the seminal examples being the biosynthesis of laslocid A^O (67) from five acetates, four propionates and three butyrates, and the biosynthesis of monensin^l (68) from five acetates, seven propionates, and one butyrate. Butyryl CoA (or rather the more activated 2-ethylmalonyl CoA) is presumed to arise from acetate via 3-hydroxybutyrate in a manner consistent with standard fatty acid  biosynthesis.22  However, evidence exists for transethylation  from S-  adenosylethionine(SAE) occuring in microbial and animal biosynthesis^ (jn a manner analogous to SAM methylation). Thus, the formation of butyrate via transethylation of malonyl CoA is a possibility that cannot be discounted. The biosynthesis of ethyl groups via SAM methylation of an alkene has been effectively demonstrated in steroid side-chain biosynthesis.24 C-methylation is rare in molluscan polyketide biosynthesis, however, suggesting that ii) above may be unlikely. A more detailed analysis of this pathway presented below also reveals the mechanistic improbability of such a pathway in the biosynthesis of triophamine (9). The biogenesis of triophamine (9) can be envisaged as occuring via the polyketide-type pathway shown in Scheme 12. Condensation of two acetate units followed by an acylation via carboxylation by biotin carboxyl carrier protein (BCCP) yields ethyl malonyl CoA (69), the biosynthetic equivalent of an activated butyrate. Condensation of (69) with an additional acetate followed by decarboxylation gives the polyketide intermediate (70); reduction and stereospecific dehydration at C3 (ketoreductase, alcohol dehydrogenase) yields (E)-2-ethyl-2,3-butenyl CoA (71).  Further condensation of (71) with a second equivalent of ethyl malonyl CoA (69) and  subsequent decarboxylation gives the polyketide intermediate (72). Reduction and dehydration at C3 of (72), followed by hydrogenation of the C2,C3 olefin of (73), yields the (E)-2,4-diethyl-4-  14 hexenoic acid moiety of triophamine (9) as its CoA thioester (74). Coupling of two equivalents of (74) with guanidine (46) yields triophamine (9).  j} acetate  >-  \  j *  = acetate  * =SAM  ^ propionate ^ *  SAM • = propionate  Scheme 11: Proposed Biogenesis of Acyl Moiety of Triophamine (9)  pathway b  146  If, however, the butyrate units were derived from SAE ethylation of malonyl CoA, a similar pathway would presumably operate, though the labeling pattern of triophamine (9) yielded from such a pathway would perhaps differ in the levels of specific incorporation in the ethyl branches at C2 and C4 (see Scheme 11, pathway a2). An alternate pathway for the biosynthesis of triophamine (9) is possible (Scheme 13). Condensation of one unit of methyl malonyl CoA (75) (derived from biotin carboxylate acylation of propionate) with acetyl CoA yields 2-methyl 1,3-diketobutyryl CoA (76).  Reduction and  dehydration to the 2,3-unsaturated intermediate (77) via ketoreductase and alcohol dehydrogenase followed by a rearrangement of the double bond gives intermediate (78), as shown in Scheme 13. SAM alkylation at C-5 followed by a hydride shift yields the familiar intermediate (70b), though with quite a different labeling pattern from the all-acetate derived pathway proposed in Scheme 14. As before, condensation of intermediate (70b) with methyl malonyl CoA (75), followed by reduction to the unsaturated intermediate and subsequent SAM methylation gives the (E)-2,4diethyl-4-hexenoic acid portion of triophamine (9) as its CoA thioester (74b).  P|  ^  o  o  >  o  >  o SCoA C0 H 2  (69)  co 2 O  II  1  ?i  n  OH  (70)  OH  O  O  I  0  O  C  ° 2  O  O  H  co 2 (72)  (71)  O  O  "'SCoA  i r  (73)  (74)  NH  O  2  »  HoN'^'NH  NH,  O  (46)  Scheme 12: Biosynthesis of Triophamine from Acetate Derived Butyrate  148  (76)  E n z y m e  _ -  (77)  S  i  Scheme 13: Alternate Biosynthesis of Triophamine via SAM Alkylations  This pathway seems highly unlikely given the proposed methylation of an a , B unsaturated thioester (78).  SAM methylation of a free intermediate (presumably the CI alcohol of  (78) or (74b)) could be envisioned as occuring. Thus it seemed likely that these pathways could be distinguished in a feeding experiment using [l,2-13c] acetate, based on the expectation of observing intact acetate units at C7/C8 and  1 4 9  C9/C10 if the pathway proposed in Scheme 12 was in operation. If these ethyl branches were the product of SAM alkylations, however, no such interacetate coupling would be expected.  V.IH Feeding Experiments with Triopha catalinae  The biggest problem that was made evident in initial studies with T. catalinae using the multiple-injection protocol developed in previous studies was the fact that with repeated handling, T. catalinae shed its full complement of triophamine (9).  Thus, after multiple injections,  absolutely no triophamine (9) was recovered at the end of the incubation period. It was concluded that the use of multiple injections of precursor was impeding any build-up of metabolite that could be analysed at the end of the experiment; that is, any triophamine that was being made de novo during the 48 hours between injections was immediately shed at the next injection. Clearly, the final 48 hour period after the last injection was insufficient time for an appreciable amount of triophamine to be biosynthesized. For this reason, a slightly different approach was taken with T. catalinae. Specimens of Triopha catalinae (30 animals) were collected by hand using SCUBA at depths of 5 to 10 metres in surge channels off Sanford and Fleming Islands, Barkley Sound, B.C. The animals were brought back to the Bamfield Marine Station and kept alive in running sea water tanks with no food sources available. Approximately 4 hours after collection, all animals were injected with 100 |iL each of a freshly prepared solution of 550 mM (1,2-  Cj)-  13  NaOAc in distilled water. Injections were made  through the dorsum on the left side, directly into the large digestive gland. Another series of injections were performed in exactly the same manner 24 hrs. after the first set of injections. The animals were subsequently allowed to incubate, unmolested, for 9 days, after which they were sacrificed by immediate immersion in methanol (250 mL). The methanol extract was returned to the laboratory in Vancouver, where it was decanted, filtered, and reduced in vacuo . The animals were subsequently exhaustively extracted with two further portions of methanol, and two portions of 1:1 MeOH: CH2CI2. All extracts were filtered, combined, and reduced in vacuo to yield an aqueous suspension. This material was further diluted to 500 mL with distilled water and exhaustively extracted with 4 X 500 mL portions of  151  152  153  N H  O  O  2  (9)  Me8 Me6  MelO  H2 H3b / H3a  H5  CIO  (ppm)  C7  20  C9  40  C3 C2  60  80  h C5  100 120  i  (ppm)  1  4.5  1  1  1  i'*  1  4.0  3.5  3.0  '*  i  1 1  2.5  * 'i  2.0  '* 1 1 1  1.5  1.0  Fig. 71: HMQC Spectrum of triopharnine (9) [500 MHz, CDC1 ] 3  154  O n  N H i  O  2  ^  II  Me6 H7  H 2  H3a H3b A -t-,  CIO  —  a  C8 C7  MelO  H9affi H9b  rj  4>  (ppm) b 10  0  20  -e-  C9  C3  Me8  F 30  4  40  4  C2  (ppm)  6  2.0  1.5  b  1.0  Fig 72- Expansion of Upfield Region of HMBC Spectrum of triophamine (9) [500 MHz, CDC1J  50  60  10  8  10  1  8'  (9)  H5  J C6  (ppm)  C7  20  C3  40  60  -  80  - 100  120  C4  140  (ppm)  5.3  5.2  5.1  5.0  4.9  Fig 73: Expansion of H5 Region of HMBC Spectrum of triophamine (9) [500 MHz, CDC1 ] 3  Fig 74: Expanded Downfield C Region of HMBC Spectrum of triophamine (9) [500 MHz, CDC1 ] 13  3  157  EtOAc. The EtOAc layers were dried over MgS04, filtered, combined, and reduced in  vacuo  to  yield an orange oil (260 mg). The EtOAc soluble materials were fractionated by silica gel flash chromatography (eluent: 85:15 hexanes: EtOAc) to yield a sample containing almost pure triophamine (9) (43 mg) contaminated with fat. This fraction was further purified on reversed phase HPLC (eluent: 4:1 MeOH: H2O) to yield 4.6 mg of pure triophamine (9) as a colorless oil. While the original published account^ of the structure determination of triophamine (9) included C assignments, a number of resonances in the acyl portion of the molecule could not be 13  unambiguosly assigned with the NMR methods then available, and thus a more stringent approach was neccessary. The assignments of the carbonyl (CI) and guanidyl (CI 1) carbons, however, were not in doubt. The downfield quartet at 5 5.18 clearly belonged to H5, and an HMQC correlation into a carbon resonance at 8 120.5 clearly established the identity of C5. The proton resonance for H5 showed HMBC correlations into a methyl carbon at 8 13.1 (showing an HMQC correlation into the methyl doublet at 8 1.54) which was clearly C6, and into a quaternary carbon a 8 139.0, which was clearly C4. H5 showed further HMBC correlations into a pair of methylene carbons at 8 39.3 and 22.6, which must be C7 and C3. These two resonances were distinguished in the following manner. Since C3 is four bonds removed from all methyl protons in the molecule, it cannot show HMBC correlations into any of the methyl protons. In contrast, we expect to observe an HMBC correlation from the methyl protons at H8 into C7. Thus, the HMBC correlation from the methyl triplet at 8 0.94 into the carbon resonance at 8 22.6 clearly establishes this carbon as C7 Hence, the resonance at 8 39.3 must be C3, and an HMQC correlation from the aforementioned methyl triplet at 8 0.94 into a methyl carbon at 8 12.7 demonstrates that this must be assigned to C8. It neccessarily follows that the remaining methyl triplet at 8 0.89 must be H10 (which correlates into a carbon resonance at 8 12.1 in the HMQC spectrum, thus establishing the identity of C10). This resonance at H10 shows HMBC correlations into a methylene carbon at 8 25.4 (which must therefore be C9) and into a downfield methine resonance at 8 50.2 (which must be C2). Table 6 shows the specific incorporation data for triophamine (9). An analysis of the coupling constants clearly indicates incorporation of intact acetate units into C5/C6, C7/C8, and C9/C10. The truncated ^ C resonances for labeled and control samples of triophamine (9) are  158  Fig. 75: Truncated Normalized "C Resonances of Labeled (left) and Unlabeled (right) Samples of triophamine (9) [125 MHz, CDC1 ] 3  15 shown in Fig. 75 and provide visually convincing evidence for this conclusion. This supports the biogenesis of triophamine from two units of acetate-derived ethyl malonyl CoA and one unit of acetate (Scheme 12). Formation of the ethyl branches at C2 and C4 via SAM alkylation can definitely be ruled out with this data. The evidence for intact incorporation at C1/C2 and C3/C4, however, is less clear. A flanking doublet at the broad resonance for the CI carbonyl could not be observed, irrespective of solvent and temperature effects. The coupling constant for the strong flanking doublet at C2 (Jc,C = 51.9 Hz), though, was completely consistent with sp - sp2 3  (carbonyl) * C,* C coupling.25 3  3  Table 6: Specific Incorporation Data for Triophamine (9)  c#  d C(ppm) (125MHz) 185.6 (br) 50.2 39.3 139.0 120.5 13.1 22.6 12.7 25.4 12.1 157.4 13  1,1' 2, 2' 3, 3" 4, 4' 5, 5' 6, 6' 7, 7' 8, 8' 9, 9' 10, 10' 11  J (Hz)  specific incorporation n/a n/a 51.88 0.16 • 42.72 0.20 n/a n/a 43.48 0.28 42.72 0.34 33.57 0.15 32.81 0.24 34.33 0.24 35.09 0.29 n/a n/a S.I. (avg) = 0.24  c c  Similarly, although the quaternary resonance for C4 was too weak to observe a clear doublet, the coupling constant of the flanking doublet about C3 (Jc.c = 42.8 Hz) was consistent with an sp 3  sp2 (olefin) * C , 1 C coupling.26 Thus the different coupling constants at C2 and C3 rule out the 3  3  possibilty of an intact acetate at C2/C3, and the coupling constants are consistent with intact acetates at C1/C2 and C3/C4, as predicted by the proposal shown in Scheme 12. The use of stable isotope methodology has thus been extended to probe polyketide biosynthesis by the dorid nudibranch, Triopha catalinae.  This work represents the first  experimental evidence for de novo polyketide biosynthesis by a dorid nudibranch; moreover, the  160 use of doubly-labeled [1,2-' ^Ci] acetate has provided clear evidence in support of one pathway where a number of biosynthetic pathways were possible. The success of this method points the way for future work with more advanced precursors labeled with stable isotopes in order to probe more subtle aspects of polyketide biosynthesis in Triopha catalinae.  Endnotes: V. Biosynthesis of Triophamine 1. introductory material taken from: Luckner, M., Secondary Metabolism in Microorganisms. Plants, and Animals. 2nd ed., Springer-Verlag, Berlin, 1984 2. Gustafson, K.; Andersen, R.J., J. Org. Chem., 1982,47,  2167-2169.  3. Faulkner, D.J.; Molinski, T.F.; Andersen, R.J.; Dumdei, E.J.; DeDilva, E.D., Comp. Biochem. Physiol., 1 9 9 0 , 97(C), 233-240.  4. Gustafson, K.; Andersen, R.J., Tetrahedron, 1985,47, 1101-1108. 5. Rabaron, A.; Koch, M.; Plat, M.; Peyroux, J.; Wenkert, E.; Cochran, D.W., J. Am. Chem. Soc, 1911,93, 6270-71. 6. Needham, J.; Andersen, R.J.; Kelly, M.T., J. Chem. Soc, Chem Commun., 1992,18,  1367-69. 7. Needham, J.; Hu, T.; McLachlan, J.; Walter, J.; Wright, J., J. Chem. Soc, Chem Commun., 1 9 9 5 , 1623.  8. Manker, D.C.; Garson, M.J.; Faulkner, D.J., J. Chem. Soc, Chem. Commun., 1988, 16, 1061-62. 9. Luckner, M., Secondary Metabolism in Microorganisms. Plants, and Animals. 2nd ed., Springer-Verlag, Berlin, 1984, pg. 189. 10. see for example: a) Kaneda, T.; Butte, J.C.; Taubman, S.B.; Corcoran, J.W., J. Biol. Chem., 1962, 237, 322; b) Cane, D.E.; Hasler, H.; Taylor, P.B.; Liang, T.-C, Tetrahedron, 1983, 39, 3449. 11. a) Garson, M.J.; Jones, D.D.; Small, C.J.; Liang, J.; Clardy, J., Tet. Lett., 1994, 35, 6921-24; b) Paterson, I.; Franklin, A.S., Tet. Lett., 1 9 9 4 , 35, 6925-28; c) Garson, M.J.; Goodman, J.M.; Paterson, I., Tet. Lett., 1994, 35, 6929-32. 12. Di Marzo, V.; Vardaro, R.R.; DePetrocellis, L.; Villani, G.; Minei, R.; Cimino, G, Experientia, 1991,47, 1221.  13. Gavagnin, M.; Marin, A.; Mollo, E.; Crispino, A.; Villani, G.; Cimino, G., Comp. Biochem. Physiol, 1 9 9 4 , 108, 107-15.  14. Gavagnin, M.; Spinella, A.; Castelluccio, F.; Cimino, G, J. Nat. Prod., 1994, 57, 298304. 15. Ireland, C ; Scheuer, P.J., Science, 1979, 205, 922.  16  16. Ireland, C ; Faulkner, D.J., Tetrahedron, 1981, 37 (Suppl. I), 233. 17. Fenical, W.; Sleeper, H.L.; Paul, V.J.; Stallard, M.J.; Sun, H.H., Pure Appl. Chem., 1979,57, 1865-74. 18. DiMarzo, V.; Cimino, G.; Crispino, A.; Minardi, C ; Sodano, G.; Spinella, A., Biochem. J., 1991,275, 593-600. 19. Luckner, M., Secondary Metabolism in Microorganisms. Plants, and Animals. 2nd ed., Springer-Verlag, Berlin, 1984, pp. 168-70. 20. Westley, J.W.; Evans, R.H.; Harvey, G.; Pitcher, R.G.; Pruess, D.L.; Stempel, A.; Berger, J, J. Antibiotics, 1974, 27, 288-97. 21. Day, L.E.; Chamberlin, J.W.; Gordee, E.Z.; Chen, S.; Gorman, M.; Hamill, R.L.; Ness, T.; Weeks, R.E.; Stroshane, R., Antimicrob. Agents Chemother., 1973,4, 410-14. 22. Volpe, J.J.; Vagelos, P.R., Ann. Rev. Biochem., 1973, 42, 21-60. 23. Dulaney, E.L.; Putter, 1 ; Drescher, D.; Chaiet, L.; Miller, W.; Wolf, F.; Hendlin, D., Biochim. Biophys. Acta, 1962, 60, 447-49.  24. Lenfant, M.; Ellouz, R.; Das, B.C.; Zissman, E.; Lederer, E, European J. Biochem., 1969, 7, 159-64 and references cited therein. 25. Chaloner, P.A., J. Chem. Soc, Perkin Trans. II, 1980, 1028.  26. Marshall, J.L.; Miiler, D.E., Org. Magn. Reson., 1974, 6, 395.  162 VI. Concluding Remarks  Investigations of the skin extracts from a number of dorid nudibranchs has led to the isolation of two novel compounds, lovenone (2) and limaciamine (8).  Lovenone (2), isolated from the  North Sea dorid Adalaria loveni, represents the first triterpenoid isolated from a nudibranch, and is only the second triterpenoid ever isolated from a marine mollusc; the structure of lovenone (2) was solved using a number of two-dimensional NMR techniques.  Similarly, the isolation of  limaciamine (8) from the North Sea dorid Limacia clavigera, represents the only naturally occuring analogue reported to date of triophamine (9). Triophamine (9) was originally isolated from the British Columbia dorids, Triopha catalinae and Polycera tricolor. 1  HO  (2)  (9)  The isolation and structure determination of these novel compounds led directly to an investigation into the biosynthesis of secondary metabolites by dorid nudibranchs. While the vast majority of dorid natural products can be traced to a dietary source (i.e. a sponge or bryozoan upon which the nudibranch feeds), a number of factors contributed to the possibility of a de novo biosynthetic origin for the two compounds isolated, which in turn prompted experimental studies using stable isotopes with a number of British Columbia dorids.  First, chemical analysis of  extracts of bryozoans upon which A. loveni were known to feed failed to produce any evidence for the dietary origin of lovenone (2).  Second, the isolation of limaciamine (8) from a  geographically distinct, yet closely related, genus of dorid- T. catalinae, which always contains triophamine (9) regardless of where it is collected-lent further credence to the geographic  163 invariance, and by extension de novo biosynthetic origin, of the compounds. (This hypothesis, formulated by Faulkner, Andersen and co-workers^, suggests that variation in dorid secondary metabolite content with collecting site reflects the different diets present at different sites; thus, invariance of dorid secondary metabolite content with respect to collecting site, while not an absolute predictor of de novo origin, at least helps to identify suitable candidates for biosynthetic study.)  It is generally agreed that biosynthetic experiments involving invertebrates are difficult to perform, with slow turn-over rates and low levels of incorporation usually expected.^ In the first successful example with a dorid, the biosynthesis of the polygodial (26) was effectively demonstrated by Cimino et. al. using radioisotopes.3 However, similar experiments on terpenoic acid glycerides (28 to 33), though successful in the reports of Gustafson and Andersen,4 could not be corroborated by Cimino and co-workers.^ In light of this controversy, it was resolved that any biosynthetic experiments undertaken must generate unambiguous and visually convincing results. For this reason, experiments using precursors labeled with stable isotopes were initiated, thus avoiding the potential ambiguity of low level radioisotope results.  The difficulties in  developing a stable isotope methodology for use with dorid nudibranchs were many: choice of precursor, administration of precursor, and ability to detect the stable isotope by NMR being of primary concern. These problems were successfully addressed in preliminary studies on the  164 British Columbia dorids, Archidoris odhneri and A. montereyensis, where the aforementioned work by Gustafson et. al.4 suggested a high expectation of success.  The true test of this  methodology involved extending it to cases where little or no previous radioisotope work existed. Using the geographic invariance hypothesis as a starting point, the biosynthesis of sesquiterpenoid aldehydes (38 to 40) from Acanthodoris  nanaimoensis and triophamine (9) from Triopha  catalinae were investigated with this stable isotope method.  OH  OH  OH  It has been shown for the first time that both isoprenoid and polyketide biosynthesis can be effectively demonstrated as occurring in certain species of dorid nudibranch using precursors labeled with stable isotopes. Moreover, precursors doubly-labeled with C have provided direct 13  evidence for new and hitherto unprecedented rearrangements in standard biosynthetic pathways, most strikingly in the biosynthesis of isoacanthororal (40) by Acanthodoris nanaimoensis. The problems associated with attempting biosynthetic experiments with marine invertebrates, as outlined in the introduction to Chapter IV and in Garson's reviews^, were surmounted in the experiments described in this thesis in a number of ways. First, only nudibranchs exhibiting invariance of natural product content with respect to geographical occurrence, as suggested by the hypothesis of Faulkner, Andersen and co-workers were chosen for study. That all these species were cold-water dwelling organisms was fortuitous (though perhaps not wholly accidental, as will be speculated upon later) in that large numbers of robust individuals could be collected throughout the year. This situation is contrasted with tropical waters where large numbers of any particular  165  species are extremely rare. While maintenance of the organisms in aquaria did not prove difficult, it was shown that long time periods (on the order of weeks as opposed to days) were necessary for successful incorporation. Similarly, it w a s determined t h a t  the  most effective administration of  precursor to the organisms was in small regular "pulses" throughout the incubation period, as opposed to one large dose at the outset of the experiment. Concurrent with this observation w a s the realization that the use of liposomes was disadvantageous for experiments using sodium acetate as precursor, since the lipid framework of the liposomes was easily digested back to acetate, thus effectively diluting the stable isotope label to the point where its detection became impossible. In both these regards, the early experiments  With  Archidoris odhneri and A. montereyensis, where  previous experimental work with radioisotopes^ had demonstrated a high probability of success (though relatively low incorporation), were particularly valuable. Having developed a successful methodology in these test organisms, it was particularly gratifying to apply this method to an organism, Acanthodoris nanaimoensis, where no previous experimental work had been done and where the biogenesis of the natural products in question was not trivial, with such unqualified success. Furthermore, the surprising result in isoacanthodoral (40) strongly demonstrates the need for biosynthetic experimentation on marine natural products; that previous biogenetic speculation on the origin of isoacanthodoral was in need of revision is now clear, but even more interestingly, the utility of stable isotope experimentation demonstrates not only that future work with more advanced precursors is possible, but that it is of fundamental importance to our understanding of this pathway. Finally, a detailed and rigorous analysis of the smaller couplings present in the spectra from these studies has led to a greater understanding of how to assess high levels of incorporation into a small number of molecules against a high background of unlabeled material, a point which Garson^ represents as being of primary importance to such experiments with marine invertebrates. In light of the above mentioned conclusions of this work, it is of potential interest to revisit some of the fundamental assumptions of dorid nudibranch secondary metabolism and speculate as to the significance of de novo biosynthesis in both ecological and perhaps evolutionary terms. It must be stressed, however, that the following discussion represents mere speculation.  166 It seems perhaps strange that those nudibranchs suitable for biosynthetic study as suggested by the geographical invariance hypothesis and born out by this study are all strictly cold-water, as opposed to tropical, species. The original study by Faulkner et. al.* did in fact only consider California and British Columbia dorids, and the hypothesis has never been experimentally extended to tropical species. However, an examination of the literature^ reveals that every single natural product isolated from a tropical dorid has an earlier (or in a few cases, subsequent) precedent as occurring in an organism that itself forms the diet of the dorid in question. While this fact does not a priori  prove that the compound(s) must be coming from the diet, the empirical  evidence is overwhelming and there is an experimental precedent,? as has already been discussed (see Section F/.I.D). In light of this singular observation, and coupled with the observation that the seemingly defenseless shell-less molluscs are rarely molested by predators, it has been proposed** that the nudibranchs' shell-bearing molluscan ancestor first adapted to overcome and ultimately "borrow" the toxins present in its diet, and then evolved to the point where its shell was no longer necessary for protection. The generally accepted evolutionary genealogy of the opisthobranch molluscs posits that the most primitive ancestor was similar to the still living cephalaspidean opisthobranch, Acteon, which has a well-developed shell.9 This family of primitive opisthobranchs are generally encountered gliding just under the surface of the sediment on sandy bottoms.  As the early  opisthobranchs began to exploit this new environment by burrowing into the sand, an evolutionary pressure arose for the loss of the shell, since this structure is a hindrance to burrowing. Thus we observe a trend towards a vestigial or internal shell in a number of cephalospidians, more likely as a result of burrowing as opposed to the pre-adaptation of a borrowed chemical defense. In order for these primitive shell-less opisthobranchs to radiate back out into more exposed habitats, they were pressured to develop new defenses against predators. In the tropics, it could be presumed that one of the best defenses would be to remain small and have a cryptic coloration (that is, "well-camouflaged").  There are many contemporary examples of small cryptic  nudibranchs that contain the same pigmentation as the sponge upon which they feed, presumably deriving these pigments directly from the sponge.  Assuming a co-evolution of sponges and  nudibranchs in the tropics, as sponges developed chemical defenses in response to increased  predation so too would those nudibranchs capable of "borrowing" these defenses (just as they were able to borrow the encrypting pigments) thrive on the sponge. Thus, these small nudibranchs were able to live their entire lives on one sponge patch (provided they were conservative grazers) and remain unmolested by borrowing either or both pigments and chemical defenses from this sponge. The situation in cold-water environments is markedly different.  The early burrowing  shell-less opisthobranchs faced quite different evolutionary pressures as they emerged into more exposed environments than their cousins in the tropics.  The general trend towards increased  body mass for survival in cold water coupled with the relative paucity of sponge species in these environments made encryption a less attractive adaptation for the carnivorous proto-nudibranchs. Similarly, cold water sponges do not face the same predatory pressures as do tropical species, and thus relatively few species of exclusively cold water sponges have developed chemical defenses. Unlike environments such as coral reefs where the density of species in a small area is high, cold water encrusting species are often found in small isolated areas, well separated by areas containing little or no life.  With encryption of limited utility, the need for an increased body size, and the  need to be able to move from one micro environment to another in search of food, the early coldwater nudibranchs faced different predatory pressures than did the cold-water sponges. Most cold water fish prey on moving creatures, leaving sessile and encrusting species unmolested. Thus, since cold-water sponges did not develop chemical defenses in the first place, the cold-water nudibranchs were forced to develop their own chemical defenses by de novo biosynthesis. Thus we have a speculative argument as to why biosynthetic capabilities seem to be more common in cold-water dwelling species of dorid nudibranchs. The loss of the shell in the primitive opisthobranchs was a result of exploiting a new environment by burrowing into the sandy bottom, where a shell is a hindrance. The shell-less opisthobranchs that radiated back to more exposed environments were only able to do so by adapting to increased predation. In the tropics, the early nudibranchs remained small and cryptic, sometimes borrowing pigments from the sponges upon which they fed.  As the sponges in the tropics developed chemical defenses in response to  increased predatory pressures, so too did the nudibranchs co-evolve to borrow these chemical defenses. In cold-water environments, sponges did not face the same pressures, and thus very  168 few developed chemical defenses of their own. The shell-less nudibranchs, forced to move to a larger body mass, became less able to be cryptic (it's easier to hide when one is small) and no longer had the option of borrowing chemical defenses. The need for these cold-water nudibranchs to be able to move from place to place (far more than would be necessary in the tropics) exposed them to predatory pressures not faced by sessile animals such as sponges. De novo biosynthetic capabilities for the production of chemical defenses were a direct response to these predatory pressures unique to the cold-water environment.10 There are, however, a number of examples of cold water dorids whose secondary metabolites can be traced back to the sponges etc. upon which they feed- Cadlina luteomarginata , which contains spongian diterpenes also isolated from the dietary sponge Aplysilla glacialis is perhaps the most-studied example. ^ In addition to the spongian diterpenes, the nature of which vary from collecting site to collecting site which supports the well-established dietary origin of these compounds, a number of other terpenes are always found in extracts of C. luteomarginata, notably, albicanol acetate (79), luteone (80), and cadlinaldehyde (81). ^  The geographical  invariance of these three compounds suggests that they are of biosynthetic origin, and thus if this  (79)  (80)  (81)  could be demonstrated with the stable isotope methodology herein developed, an example of a dorid with both borrowing and biosynthetic capabilities would be available for argument. Significantly, for unlike all the other dorids studied in this thesis (i.e. Archidoris,  Triopha,  Acanthodoris, Limacia, etc.), genera closely related to Cadlina are found in both tropical and cold-water environments. Thus an argument could be made that ancestors to luteomarginata  Cadlina  originally evolved in the tropics where they developed the ability to borrow  169 chemical defenses. As the dietary sponge moved into colder waters, some chromodorids closely related to C. luteomarginata moved too, where they faced different pressures than did the sponge (i.e. Aplysilla).  Thus, the emergent species.C. luteomarginata,  is able to borrow chemical  defenses as did its tropical ancestors, but in times when either the sponge is unavailable or due to the need to roam further in search of the sponge as it develops, it has evolved biosynthetic capabilities unlike its tropical cousins. Unfortunately, repeated experiments with stable isotopes on Cadlina luteomarginata failed to demonstrate any de novo biosynthesis of terpene carbon skeletons. This negative result may have arisen from a failure of the precursor to reach the site of biosynthesis in the nudibranch (i.e. an inherent problem with the experiment that neither proves nor disproves biosynthesis), or it could be that biosynthesis occurs only at certain times in the year, when the food source is scarce or in the case when the compounds are only needed for protection of eggs, for example (experiments on Cadlina were performed in the fall and the winter, and unlike all the other dorids studied, Cadlina  never laid any eggs during captivity). The experiments may also have failed  simply because Cadlina luteomarginata  does not possess the ability to synthesize terpenoids de  novo, and is thus an example of where the geographical invariance hypothesis breaks down. There does exist an unpublished account^ of incorporation of C radioisotope into the terpene 14  skeleton of albicanol acetate (67), and it is hoped that this result is corroborated by future experiments with stable isotopes at some point in the future, perhaps at a different time of year than the experiments essayed as part of this study. Shown in Scheme 14 is a phylogenetic tree of all genera of opisthobranch and marine pulmonate molluscs that are either known to produce secondary metabolites de novo, or are strongly suspected of such a capability (based upon the geographic invariance of metabolite content reported for species in these genera). Whether a larger pattern can .be extrapolated from the relationships between families with respect to this biosynthestic capability is not clear. That the study of biosynthesis by dorid nudibranchs is eminently feasible using stable isotopes is now clear, and it is hoped that as more experimental evidence becomes available, our understanding of the chemistry, ecology, and evolutionary history of these fascinating creatures will only continue to grow.  170  Polyceratidae f Limacia Polycera Thecacera,  Triophididae  Phanerobranchs  Cyerce Elysia  Sacoglossa Anaspidea  Pulmonaia  ( Siphonari^  Cephalaspidea  prosobranch mesogastropod LEGEND:  polyketide biosynthesis  isoprenoid biosynthesis * = biosynthesis & borrowed defense  Scheme 14: Phylogenetic Tree for Selected Opisthobranch and Pulmonate Genera with Known or Suspected Biosynthetic Capabilities  171 Endnotes: Chapter V I : General Conclusions 1.  Faulkner, D.J.; Molinski, T.F.; Andersen, R.J.; Dumdei, E.J.; DeDilva, E.D., Comp.  Biochem. Physiol., 1990, 97(C), 233-240.  2. Garson, M.J., Nat. Prod. Rep., 1989, 6, 143-170. 3. Cimino, G.; DeRosa, S.; DeStefano, S.; Sodano, G.; Villani, G., Science, 1983,279, 1237-1238. 4. Gustafson, K.; Andersen, R.J., Tetrahedron, 1985,41, 1101-1108. 5. Avila, C ; Ballesteros, M.; Cimino, G.; Crispino, A.; Gavagnin, M; Sodano, G., Comp. Biochem. Physiol., 1990, 97(B), 363-368.  6. Faulkner, D.J., Nat. Prod. Rep., 1995, 72, 223-269, arid previous reviews cited therein; and Avila, C , Oceanography Mar. Biol. Ann. Rev, 1995, 33, 487-559. 7. Dumdei, E.; Flowers, A.E., Garson, M.J.,Moore, C.J., Comp. Biochem. Physiol. C, 1996, in press. 8. Faulkner, D.J.; Ghiselin, M.T.; Mar. Ecol. Prog. Ser., 1983,13, 295-301. 9. Ruppert, E.E.; Barnes, R.D., Invertebrate Zoology. Saunders College Publishing, 6th ed., 1994, p. 80. 10. The evolutionary arguments presented here are the fruit of numerous helpful discussions with Dr. Sandra Millen, University of British Columbia. 11. Tischler, M.; Andersen, R.J.; Choudhary, M.I.; Clardy, J., J.Org. Chem., 1991, 56, 427. 12. Hellou, J.; Andersen, R.J.; Rafii, S.; Arnold, E.; Clardy, J., Tet. Lett., 1981,22, 41736; Hellou, J.; Andersen, R.J.; Thompson, J.E., Tetrahedron, 1982, 38, 1875-9; Gustafson, K.; Andersen, R.J., Tetrahedron, 1985, 41, 1101-1108; Dumdei, E., Ph.D. Thesis, 1993, University of British Columbia. 13. Gustafson, K., Ph.D. Thesis, 1985, University of British Columbia.  172 VII. Experimental  All NMR spectra were recorded on  Bruker AMX-500, AM-400, and WH-400  spectrometers. All spectra were referenced to residual solvent peaks: CDCI3: 5 7.24 (*H), 5 77.0 ( C ) ; C6D6: 57.15 (*H), 8 128.0 ( C ) . Spectra were processed using both Bruker UXNMR 13  13  version 930601.3 software and Bruker Windows™ compatible WIN-NMR software. With the exception of zero-filling, FJDs and/ or spectra were not in any way edited prior or subsequent to Fourier transformation. Low and high resolution electron impact mass spectra were performed on a Kratos MS-50 spectrometer, by the staff of the Mass Spectrometry Facility, University of British Columbia. Infra-red spectra were recorded on a Perkin-Elmer 1600 FT-IR spectrophotometer, using sodium chloride plates. Optical rotations were measured on a Jasco J-710 spectropolarimeter ( 1 cm quartz cell). Melting points were taken on a Fisher-Johns melting point apparatus, and were corrected. Normal and reversed phase thin layer chromatography employed Merck type 5554 aluminium-backed Kieselgel 60 F254 and Whatman MKC18F reversed phase TLC plates, respectively. Plates were visualized by UV (X= 254 nm) or by a vanillin/ H2SO4/ EtOH spray reagent.* Normal phase flash chromatography was carried out with BDH silica gel ( 230- 400 (i mesh) and Sigma type H TLC grade silica (10-40 |i particle diameter, no binder).^ HPLC separations involved one of three possible systems: a) Waters 501 HPLC pump equipped with a Waters 440 absorbance detector and a Perkin-Elmer LC-25 refractive index detector, b) Waters 600E HPLC pump/ system controller with a Waters 486 tunable absorbance detector, and c) Waters 600E HPLC pump/ system controller with a Waters 996 photodiode array detector.  System a) was run with a standard chart recorder, while systems b) and c) could  alternately use a chart recorder or could be interfaced with a personal computer using Millenium™ 2010 chromatography software. Normal phase HPLC separations used a Waters Rad-Pak™ silica column, while reversed-phase separations employed a Whatman Partisil 10 ODS-3 magnum column. All HPLC solvents were Fischer HPLC grade, and were filtered and degassed prior to use.  Biosynthetic experiments employed Isotech [1,2- C] sodium acetate, 99.9% isotopic 13  17 purity. All other solvents and reagents were reagent or commercial grade and were used without further purification. As previously mentioned, taxonomic identifications were performed by Dr. Sandra Millen at the University of British Columbia, solid tumor cytotoxicity assays were performed by Sarah Halleran under the supervision of Dr. Theresa Allen, Dept. of Pharmacology, University of Alberta, and liposomes were prepared by Chris Hanson, also in Dr. Theresa Allen's laboratory.  i) Adalaria  loveni  Specimens (250 animals) of A. loveni were collected by hand using SCUBA at depths of 8 to 15 m in surge channels off Flat0y Island near Bergen, Norway in the North Sea. Freshly collected animals were immediately immersed in methanol (250 mL) at the surface, and transported back to the laboratory as such for further work-up. The initial methanol extract was decanted, filtered, and reduced in vacuo . The animals were subsequently exhaustively extracted with two 250 mL portions of methanol and two 250 mL portions of 1:1 MeOH: CH2CI2; these extracts were similarly filtered, reduced in vacuo, and combined to yield an aqueous suspension. This material was further diluted with 500 mL distilled water and partitioned with 4 X 500 mL EtOAc in a separatory funnel. Upon exhaustive extraction with EtOAc, the aqueous layer was set aside, and the organic soluble fractions were dried over MgS04, reduced in vacuo, and combined to yield an orange oil (400 mg). These non-polar constituents of the extract were fractionated by normal phase silica-gel flash chromatography (step gradient from 100% hexanes to 1:1 hexanes: EtOAc) to yield a fraction of almost pure lovenone (2) eluting with 40 % EtOAc/ hexanes. This fraction was further puified on a small open column employing TLC grade silica using a step gradient from 20% EtOAc/ hexanes to 40% EtOAc/ hexanes. Further purification on normal phase HPLC (65% hexane/ 35% EtOAc, flow = 1 mL/ min, RI detection) afforded pure lovenone (2) (tR = 21.2 min., llmg) Lovenone (2) was isolated as an optically active colorless glass: [a]rj>= -38°, (c = 0.11, CHCI3) IR 3403, 2934, 1707 cm"l; H NMR (C6D6, 500 MHz) 5 0.55 (s, 3H), 8 0.93 (d, J = l  6.5 Hz, 3H), 8 0.98 (s, 3H), 8 0.99 (m, 2H), 8 1.18 (s, 3H), 8 1.18 (m, 1H), 8 1.20 (m, 1H),  174 8 1.29 (dd, J = 16, 4 Hz, 1H), 8 1.36 (m, 1H), 8 1.39 (m, 1H), 8 1.45 (m, 1H), 8 1.51 (dd, J = 16, 6 Hz, 1H), 8 1.53 (d, J = 7.5 Hz), 8 1.55 (m, 1H), .8.1.60 (m, 1H), 8 1.62 (s, 3H), 8 1.71  (s, 3H), 8 1.83 (s, 3H), 8 1.85 (m, 1H), 8 1.90 (m, 1H), 8 1.98 (m, 1H), 8 2.03 (m, 1H), 8 2.05 (dd, J = 16, 7.5 Hz, 1H), 8 2.07 (bd, J = 16 Hz, 1H), 8 2.18 (m, 1H), 8 2.73 (d, J = 6 Hz, 1H), 8 3.53 (m, 1H), 8 3.67 (dd, J = 10, 2 Hz, 1H), 8 3.83 (m, 1H), 8 3.85 (bm, 1H), 8 5.28  (t, J = 7 Hz); C NMR ( 125 MHz, Q>D ) 8 17.4 (q), 8 17.7 (q), 8 17.8 (q), 8 21.3 (q), 8 22.2 13  6  (t), 8 24.0 (q), 8 24.3 (t), 8 25.4 (t), 8 25.9 (q), 8 26.8 (q), 8 27.6 (t), 8 30.6 (t), 8 33.5 (d), 8 34.6 (s), 8 35.6 (d), 8 36.1 (t), 8 36.7 (t), 8 40.2 (t), 8 44.3 (d), 8 44.8 (d), 8 47.6 (s), 8 49.7 (s), 8 57.4 (d), 8 61.1 (t), 8 68.2 (s), 8 75.7 (d), 8 125.6 (d), 8 130.9 (s), 8 208.4 (s); LREIMS: m/z = 460, 442, 399, 381, 343, 331, 313, 109, 95, 82, 69, 55, 43, 41; HREIMS: m/z = 460.3552 (C29H48O4 calculated = 460.3558, AM = -0.6); solid tumor cytotoxicity: HEY ED50 = 11.296 |ig/mL, U373 ED50 = 11.222 (ig/mL, A549 ED50 = > 25 jig/mL.  ii) Limacia  clavigera  Specimens of L. clavigera (50 animals) were collected by hand using SCUBA at depths of. 10 metres off Tosoy Island, near Bergen, Norway, in the North Sea. Specimens were immediately immersed in methanol (250 mL) at the surface and returned to Vancouver for further study. The methanol extract was decanted, filtered and reduced in vacuo. The animals were subsequently exhaustively extracted with two 250 mL portions of methanol, and two 250 mL portions of 1:1 MeOH: CH2CI2. All extracts were filtered, combined, and reduced in vacuo to yield an aqueous suspension. This was diluted up to 500 mL with distilled water and exhaustively extracted with 4 X 500 mL portions of EtOAc.  All EtOAc layers were dried over MgSC»4,  filtered, combined, and reduced in vacuo, to yield a yellow oil. These EtOAc soluble materials were fractionated by flash silica gel chromatography using a step gradient from 100% hexanes to 100% EtOAc, to yield a fraction eluting with 4:1 hexanes: EtOAc containing mostly limaciamine (8).  This fraction was further purified on normal phase HPLC (eluent: 15% EtOAc/ hexane) to  yield 4.3 mg of pure limaciamine (8).  175 Limaciamine (8) was isolated as a colorless glass: IR: 3340, 2960, 2933, 2873, 2859, 1739, 1700,  1641  cm" ; 1  ]  H  NMR  (500  MHz,  C D C I 3 ) 8 0.86  (t, J = 7.5  Hz,  3H),  8 0.89  (t, J = 7.5  Hz,  3H), 8 1.26 (m, 2H), 8 1.31 (m, 2H), 8 1.45 (m, 1H), 8 1.51 (m, 1H), 8 1.62 (m, 1H), 8 1.62 (m,  1H),  8 1.66  (m,  1H),  8 2.37  (m,  1H);  1 3  C  NMR  (125  MHz,  C D C I 3 ) 8 11.86  (q)  8 13.9  (q),  8 22.7 (t), 8 25.6 (t), 8 29.6 (t), 8 31.9 (t), 8 51.5 (d), 8 158.9 (s); LREIMS m/z = 311, 282, 255, 240, 212, 99, 86, 69, 57, 43; HREIMS m/z = 311.25789  (C17H33Q2N3  calculated =  311.25725, AM = -0.6). iii) Biosynthetic Studies on Terpenoic Acid Glycerides from odhneri and A.  a)  Archidoris  montereyensis.  2 Day Experiments Using Liposomes (April 1995)  Specimens of Archidoris odhneri (42 animals) and A.montereyensis (36 animals) were collected via SCUBA in and around surge channels off Sanford and Fleming Islands, Barkley Sound, B.C. at depths of 15 and 8 metres, respectively. All animals were kept alive in running seawater tanks at Bamfield Marine Station and were then shipped in a large cooler of Barkley Sound seawater to the laboratory in Vancouver. Upon arrival, all nudibranchs were immediately transferred to insulated containers, one for each species, containing fresh Barkley Sound seawater that had been transported back for this purpose. The containers were maintained in a cold room with an ambient air temperature of 12 °C, and air was constantly bubbled into the seawater through an airstone. All specimens of A. montereyensis were injected through their dorsums on the left side with 100(iL of a 550 mM solution of [1,2- C ] sodium acetate encapsulated in liposomes 13  3  2  and allowed to incubate for 24 hours. All specimens of A.odhneri were similarly injected, but after a 24 hour incubation period, all animals were injected with a second 100 \iL portion of the same solution, and were then allowed to incubate for a further 24 hours. The seawater in the A. odhneri container was changed prior to the second injection. At the end of the incubation periods, all animals were sacrificed by immediate immersion in methanol. Extracts were obtained, and the compounds purified and characterized as described in the final set of experiments for this section. Control samples of all compounds were obtained from nudibranchs extracted immediately upon collection.  176  b)  2 Week Experiments Using Liposomes (July 1995)  Specimens of A. odhneri (38 animals) and A. motereyensis (25 animals) were recollected in Barkley Sound, as previously described, and transported back to Vancouver as before. Each animal was injected with 25 (iL of a 550 mM solution of [1,2- C ] sodium acetate encapsulated in 13  2  liposomes, with a seawater change accompanying the injection. The animals were allowed to 3  incubate for 24 hours, after which another injection was made in the same manner. Injections continued every 24 hours in exactly the same manner. After 4 days, one third of the A. odhneri specimens (14 animals) were sacrificed . After 8 days, one half of the remaining A . odhneri specimens were sacrificed (10 animals) and one half of the original specimens of A . montereyensis  (12 animals) were similarly sacrificed. After 12 days, the health of the remaining  animals was assesed, and it was decided to terminate the experiment (10 odhneri, montereyensis).  13  All extracts were worked up on the usual fashion, as described in the final set of  experiments in this section.  c)  3 Week Experiments Without Liposomes (September 1995)  60 specimens each of Archidoris  odhneri  and A. montereyensis  were collected in  Barkley Sound, B.C. at depths of -70 to -100 ft. for the former and 0 to -30 ft, for the latter. 20 animals of each species were immediately immersed in 500 mL MeOH at the surface and used as control samples. The remaining animals were transported back to the laboratory in 12 °C Barkley Sound sea water, and each species was stored in separate insulated containers in a cold room equilibrated to an air temperature of 12 °C.  Sixteen hours after arrival in the laboratory, the sea  water in each container was changed with fresh Barkley Sound sea water that had also been preequilibrated to 12 °C. Each animal was then injected through the dorsum on the left side between the rhinopores and gills with 100 (iL of a freshly prepared 550 mM solution of l,2- C-NaOAc in 13  doubly distilled water. Seven subsequent injections were performed in 48 hour intervals, with a sea water change accompanying each injection. Prior to the fifth injection, 20 animals of each  177 species were removed and immediately immersed in 500 mL MeOH; these samples were subsequently referred to as the "eight-day" samples.  Forty-eight hours following the final  injection, the remaining 20 animals of each group were sacrificed by immediate immersion in 500 mL MeOH; these samples were subsequently referred to as the "sixteen day" samples.  Archidoris odhneri isolation  Each group of A. odhneri specimens was dealt with in exactly the same fashion, as outlined below. After 24 hrs., the original 500 mL MeOH extract was decanted and filtered. Four more 500 mL extracts were made from the animals in the following order: MeOH, 2 X 1 : 1 MeOH: CH2C12, MeOH. All extracts were decanted, filtered, combined, and reduced in vacuo, to yield an aqueous suspension. This suspension was diluted up to 500 mL with distilled water and extracted with four portions of 500mL EtOAc.  Each EtOAc extract was dried over MgS04,  filtered, combined, and reduced in vacuo to yield an orange oil. This oil was chromatographed on silica gel (230 -400 \x mesh) under medium N flash in 1:1 hexanes: EtOAc to yield a fraction 2  enriched in farnesic acid glyceride (28).  This fraction was chromatographed repeatedly on  reversed phase HPLC in 4:1 MeOH: H20 to yield a pure sample of (28): control= 71 mg; 8 day= 85 mg; 16 day= 130 mg. To each sample of (28) was added 1 mL of pyridine and 2mL of acetic anhydride, and this reaction mixture was allowed to stir for 16 hrs. Purification of each diacetate of (28) on silica gel (TLC grade) in 85: 15 hexane: EtOAc (HPLC grade) yielded pure samples of (35) in essentially quantitative yield. Farnesic acid glyceride diacetate (35) was isolated as a colorless oil: *H NMR (500 MHz, CDC1 ) 8 1.55 (s, 3H), 5 1.56 (s, 3H), 8 1.62 (s, 3H), 8 1.93 (m, 2H), 8 2.00 (m, 2H), 8 2.01 3  (s, 3H), 8 2.03 (s, 3H), 8 2.09 - 2.14 (m, 7H),8 4.13 (m, 2H), 8 4.25 (m, 2H), 8 5.03 (bs, 2H), 8 5.22 (m, 1H),8 5.62 (s, 1H); C NMR (125 MHz, CDCI3) 8 15.9 (q), 8 17.6 (q), 8 1 3  18.9 (q), 8 25.4 (q), 8 25.9 (t), 8 26.6 (t), 8 39.6 (t), 8 40.9 (t), 8 61.3 (t), 8 62.4 (t), 8 69.2 (d), 8114.7 (d), 8 122.7 (d), 8 124.1 (d), 8 131.2 (s), 8 136.1 (s), 8 161.4 (s), 8 165.9 (s), 8  178 169.9 (s), 8 173.5 (s); HREIMS m/z = 394.23498 (C22H34O6 calculated = 394.23553, AM = 0.6).  Archidoris montereyensis isolation  Extracts were made of each group of A. montereyensis specimens in exactly the same manner as for A. odhneri, as outlined above. The oil that resulted from the EtOAc/ H20 partitioning of the combined crude extracts was chromatographed on silica gel (230 -400 m mesh) under medium N2 flash in 1:1 hexanes: EtOAc to yield a fraction that was highly enriched in the diterpenoic acid glyceride (29). This fraction was recrystallized from hexane QHPLC grade) to yield pure (29) as white needles: control= 52.6 mg; 8 day= 46.0 mg; 16 day= 38.0 mg. Diterpenoic acid glyceride (29) was isolated as white needles: m.p. = 139-141 "C; *H NMR (400 MHz, CDCI3) 5 0.79 (s, 3H), 6 0.80 (m, 1H), 0.83 (m, 1H), 0.84 (s, 3H), 8 0.89  (s, 3H), 80.92 (s, 3H), 8 1.11 (m, 1H), 8 1.14 (m, 1H), 1.27-1.42 (m, 4H), 8 1.51-1.55 (m, 2H), 8 1.58 (s, 3H), 8 d 1.6.1 (m, 1), 8 1.69 (m, 1H), 8 1.93 (m, 1H), 5 2.14 (bt, J = 6 Hz.exchangeable, 1H), 8 2.55 (bd, J = 5 Hz, exchangeable, 1H), 8 2.94 (bs, 1H), 8 3.62 (dd, J = 12, 5 Hz, 1H), 8 3.68 (m, 1H), 8 3.92 (m, 1H), 8 4.14 (dd, J = 11, 6 Hz, 1H), 8 4.19 (dd, J = 11, 4.5 Hz, 1H), 8 5.51 (bs, 1H); C NMR (100 MHz, CDCI3) 8 15.6 (q), 8 15.7 (q), 8 18.5 1 3  (t), 8 18.7 (t), 8 21.2 (q), 8 21.6 (q), 8 22.7 (t), 8 33.0 (s), 8 33.2 (q), 8 36.6 (s), 8 37.4 (s), c 39.9 ( t ) , 8 41.9 ( t ) , 8 41.9 (t), d 54.3 (d), 8 56.5 (d), 8 62.6 (d), 8 63.5 (0,8 65.1 (0,8 70.4 (d), 8 124.3 (d), 8 128.5 (s), 8 173.4 (s); HREIMS m/z = 378.27640 (C23H38O4 calculated = 378.27701, AM = 0.6).  iv)  Biosynthesis of Sesquiterpenoids from Acanthodoris  nanaimoensis  Specimens (95 animals) of A. nanaimoensis were collected via SCUBA in Barkley Sound, B.C. and transported back to UBC in refrigerated seawater. The nudibranchs were maintained at 12 °C in an aquarium filled with Barkley Sound seawater that was changed every 2  179 days. Individual specimens of A. nanaimoensis were given 100 jiL injections of a 550 mM solution of [l,2- C2]acetate every second day for 16 days. Two days after the last injection the 13  specimens of A. nanaimoensis were carefully removed from the aquarium seawater and the intact animals were immediately immersed in methanol (250 mL). The methanol was decanted from the whole animals and evaporated in vacuo . The animals were further extracted with 2 X 250 mL portions of MeOH and 2 X 250 mL portions of 1:1 CH2CI2. All extracts were filtered, combined and reduced in vacuo to yield an aqueous suspension. Dilution of the aqueous suspension with 500 mL distilled water followed by extraction with 4 X 500 mL EtOAc gave an organic soluble fraction (650 mg) containing the sesquiterpenoid aldehydes. This material was fractionated by silica gel chromatography (230 -400 \i mesh, BDH) using 5% EtOAc/ hexanes as the eluent to yield an early eluting fraction enriched in the sesquiterpenoid aldehydes (40 mg). This material was dissolved in 5 mL isopropanol to which was added, dropwise, a solution of 60 mg NaBH4 in 20 mL isopropanol. The reaction mixture was allowed to sit at room temperature, with magnetic stirring, for 24 hours.  After this period, the reaction mixture was quenched with 20 mL of  distilled water and allowed to stir for a further 3 hours. The mixture was then diluted up to 50 mL with distilled water and extracted in a separatory funnel with 4 X 50 mL CHCI3. All the chloroform layers were dried over MgS04, filtered, combined, and reduced in vacuo to yield a mixture of sesquiterpenoid alcohols (41 to 43) (45.6 mg). Care was taken not to expose the relatively volatile sesquiterpenoids to high vacuum conditions, and thus all weights reported in this section may reflect traces of residual solvents. The alcohols were fractionated via reversed phase HPLC (eluent: 4:1 MeOH/H 0) to give pure samples of nanaimool (41) (10 mg: 0.1 mg/animal) 2  and isoacanthodorol (43) (7.0 mg: 0.07 mg/animal). Only trace amounts of acanthodorol (42) were obtained from the reduction mixture. Nanaimool (41) was obtained as a colorless oil: H NMR (500 MHz, CDCI3) 5 0.85 (s, . l  3H), 8 0.94 (s, 3H), 8 0.95 (s, 3H), 8 1.35 (t, J = 6.5 Hz, 2H), 8 1.41 (m, 2H), 8 1.50 (m, 2H), 8 1.56 (m, 1H), 8 1.57 (m, 2H), 8 1.75 (m, 1H), 8 1.78 (m, 2H), 8 1.94 (bm, 2H), 8 3.70 (m, 2H); C NMR (125 MHz, CDCI3) 8 19.4 (t), 8 21.4 (t), 8 24.8 (q), 8 27.8 (q), 8 28.0 1 3  (q), 5 30.7 (s), 8 3.1.7 (t), 8 33.5 (s), 8 34.7 (t), 8 39.8 (t), 8 43.8 (t), 8 44.0 (t), 8 59.6 (t), I  1  80  125.5 (s), 8 133.4 (s); HREIMS m/z = 222.19867 (C15H26O calculated = 222.19836, AM = -0.3). Isoacanthodorol (43) was obtained as white needles: H NMR (500 MHz, CDCI3) 8 0.86 J  (s, 3H), 8 0.98 (s, 3H), 8 1.14 (m, 1H), 8 1.20 (m, 1H), 8 1.22 (m, 1H), 8 1.26 (m, 1H), 8 1.30 (m, 1H), 8 1.39 (m, 1H), 8 1.43 (m, 1H), 8 1.53 (m, 1H), 1.68 (m, 1H), 8 1.88 (m, 1H), 8  1.90 (m, 2H), 8 2.03 (dt, J = 13,7 Hz, 1H), 8 3.67 (t, J = 7 Hz, 2H), 8 5.08 (s, 1H); C 1 3  NMR (125 MHz, CDCI3) 8 19.5 (t), 8 20.1 (t), 8 23.4 (q), 8 26.7 (q), 8 29.2 (t), 8 32.0 (q), 8 34.1 (s), 8 37.4 (s), 8 37.9 (t), 8 40.1 (t), 6 45.5 (d), 8 46.8 (t), 8 60.1 (t), 6 131.6'(d), 8 134.1 (s); HREIMS m/z = 222.19866 (C15H26O calculated = 222.19836, AM = -0.3).  v)  Biosynthesis of Triophamine from Triopha  catalinae  Specimens of Triopha catalinae (30 animals) were collected by hand using SCUBA at depths of 5 to 10 metres in surge channels off Sanford andFleming Islands, Barkely Sound, B.C. The animals were brought back to the Bamfield Marine Station and kept alive in running sea water tanks with no food sources available. Approximately 4 hours after collection, all animals were injected with 100 |iL each of a freshly prepared solution of 550 mM (1,2- C2)- NaOAc in 13  distilled water. Injections were made through the dorsum on the ventral side, directly into the large digestive gland. Another series of injections were performed in exactly the same manner 24 hrs. after the first set of injections.  The animals were subsequently allowed to incubate,  unmolested, for 9 days, after which they were sacrificed by immediate immersion in methanol (250 mL). The methanol extract was returned to the laboratory in Vancouver, where it was decanted, filtered, and reduced in vacuo . The animals were subsequently exhaustively extracted with two further 250 mL portions of methanol, and two 250 mL portions of 1:1 MeOH: CH2CI2. All extracts were filtered, combined, and reduced in vacuo to yield an aqueous suspension. This material was further diluted to 500 mL with distilled water and exhaustively extracted with 4 X 500 mL portions of EtOAc.  The EtOAc layers were dried over MgS04, filtered, combined, and  181  reduced in vacuo to yield an orange oil (260 mg). The EtOAc soluble materials were fractionated by silica gel flash chromatography (eluent: 85:15 hexanes: EtOAc) to yield a sample containing almost pure triophamine (8) (43 mg) contaminated with fat. This fraction was further purified on reversed phase HPLC (eluent: 4:1 MeOH: H2O) to yield 4.6 mg of pure triophamine (9). Triophamine (9) was isolated as a colorless oil: *H NMR (500 MHz, CDCI3) 8 0.89 (t, J = 7.5 Hz, 3H), 8 0.94 (t, J = 7.5 Hz, 3H), 1.50 (m, 1H), 1.54 (d, J = 6.8 Hz, 3H), 8 1.59 (m, 1H), 8 2.0 (q, J = 7.5 Hz, 2H), 8 2.11 (m, 1H), 8 2.29 (m, 1H), 8 2.31 (m, 1H), 8 5.18 (q, J = 6.8 Hz);  1 3  C NMR (125 MHz, CDCI3) 8 12.1 (q), 8 12.7 (q), 8 13.1 (q), 8 22.6 (t), 8 25.4 (t), 8  39.3 (t), 8 50.2 (d), 8 120.5 (d), 8 139.0 (s), 8 157.4 (s), 8 185.6 (bs); HREIMS m/z = 363.28857 (C21H37O2N3 calculated = 363.28857, AM = -0.7).  Endnotes: Experimental 1. Krebs, K.G.; Heusser, D.; Wimmer, H.; in Thin-Layer Chromatography: A Laboratory Handbook. Stahl, E., trans. Ashworth, M.R.F., Springer-Verlag, Berlin, 1969, pg. 904. 2.  Kuhler, T.C.; Lindsten, G.R., J. Org. Chem., 1983,48, 3589-91.  3. Liposomes (9:1 HSPC: egg PG multilammelar vesicles, 2.4 (i particle size, trap volume = 1.1 (imol Na[l>2-13c2] acetate/ |imol lipid) were prepared by Chris Hanson in the laboratory of Prof. T. Allen, Dept. of Pharmacology, University of Alberta.  182 Appendix A: Nuclear Magnetic Resonance Techniques  Nuclear magnetic resonance (NMR) techniques arc of fundamental importance for the structure elucidation of natural products. The detailed information regarding connectivities of protons and carbons in organic molecules provided by such NMR experiments gives an incredible insight into the gross structure, conformation, and relative stereochemistry of molecules.  Moreover, the  advent of superconducting magnets, new pulse programs, and inverse detection probes has allowed for the structure determinations of small amounts of complex natural products, where extensive chemical degradations would have precluded such a possibility in the past. Although complete structural proofs still rely on high resolution mass spectrometry, chemical derivatization, and often X-ray crystallography and total synthesis, NMR techniques are undoubtedly the primary spectroscopic tool for organic structure determination, and thus these techniques deserve a brief introduction here. The following discussion is meant to serve as a very simple description of the experiments employed in this thesis; for a more detailed description of the theoretical basis of these techniques, the reader is referred to the numerous and excellent reviews on this subject.  1  i) One-dimensional NMR experiments  In a static magnetic field, Ho, nuclei of spin = 1/2 precess at a characteristic frequency, called the Larmor frequency, Vrj. The energy of transition for a given spin state (i.e from going "with the field" to "against the field") is given by the Planck equation, and is proportional to Ho:  AE = hv  0  = Y/2TC(H ) 0  where yis the gyromagnetic ratio of the nucleus.  When a radiofrequency pulse, o)i, is applied "on resonance" (i.e. coi = 27tvo), an energy transition from one spin state to another is effected, giving rise to a detectable signal that decays over time. The sum of all these decaying signals generated by all the spin 1/2 nuclei of a given isotope in the molecule, called a free-induction decay (FID), yields a frequency domain spectrum  (i.e. a 'classic' one-dimensional NMR spectrum) via Fourier transformation.  18 The shielding  effects of electrons cause nuclei in different bonding environments to experience slightly different 'effective fields' (i.e. Ho + Heiec = H ff). The x-axis of a one-dimensional NMR spectrum, called e  the chemical shift (8), is a measure of the resonance frequencies of nuclei in different chemical environments with respect to the applied field, Ho. Moreover, NMR signals can be "split" via the interaction of nuclei in different bonding environments. This interaction, called scalar coupling, is the result of through-bond energy tranfer. In quantum mechanical terms, the behaviour of nuclear spins can be described by operators which have well-defined effects upon the magnetization of any given nucleus.  Thus,  magnetization can evolve under the influence of radiofrequency pulse operators, chemical shift operators, and scalar coupling operators. This predictable evolution is exploited in designing experiments that transfer energy from one nucleus to another, for it is this energy transfer that ultimately provides the 'connectivity' information in the processed spectrum.  2  i.A) l H Spectrum  The most common pulse experiment used to acquire a one-dimensional proton spectrum consists of a preparation time, di, (to allow the sample to reach thermal equilibrium), followed by a radiofrequency pulse, 6, (where 0 = (Ojt). Generally, a short delay follows the pulse, allowing the transmitter coil to return to equilibrium. The FID is acquired during this time, and then the procedure is repeated. The sum of all FIDs and subsequent Fourier transformation, produces a spectrum. e  Fig. A l : Pulse sequence for the acquisition of a *H spectrum3  184  i.B) Difference nOe ( nuclear Overhauser effect) Spectra  The pulse sequence for the difference nOe experiment consists of the sequential acquisition of *H spectra with gated decoupling. Gated decoupling refers to a second transmitter coil used to irradite a specific resonant frequency, with the effect of generating an nOe enhancement in those proton resonances of close proximity in the molecule to the irradiated proton. Subtraction of the off-resonance spectrum from the spectrum with nOe enhancement gives a spectrum where all but the enhanced peaks are nulled. Thus, the observation of nOe enhancement at a given proton resonance when another resonance is irradiated suggests a close proximity of the two protons, and the set of all such nOe's in a molecule can provide conformational and relative stereochemical features of the compound.  FID  OBSERVE  DECOUPLE | Decoupler on  Fig. A2: Pulse sequence for the acquisition of the nOe difference spectrum  i.C)  1 3  C Spectrum  Compared with protons,  1 3  C is a relatively insensitive nucleus, due to its low natural  abundance and small gyromagnetic ratio. Typically, C spectra are acquired for a much longer 13  time, typically from 8 to 16 hours depending upon the sample size. The pulse sequence is similar to the one-dimensional proton spectrum, though the entire proton spectral window is irradiated during acquisition ("broad-band decoupling"). This eliminates all the proton-carbon couplings, resulting in a carbon spectrum with all the signals appearing as singlets.  18 6  t>  d OBSERVE  1 3  C  FID  Decoupler on  DECOUPLE H l  Fig. A3: Pulse sequence for the acquisition of a ^ C spectrum 3  i.D) The Attached Proton Test (APT) Spectrum  The APT experiment employs gated H broad band decoupling to achieve a modulation of the l  13  C transverse magnetization due to scalar coupling with protons.  After the first pulse, the  decoupler is switched off, allowing the C magnetization to evolve under the influence of both 13  chemical shift and sclalar coupling with protons. Thus, in this so-called "mixing time", d2 (set to 1/JCH.  typically 7 ms) carbon resonances with different numbers of attached protons will precess  at different "rates". The second 180° pulse and subsequent d delay manipulates the individual 2  magnetizations such that carbons with even numbers of attached protons are out of phase with those attached to an even number.  n__n 0  1 3  C OBSERVE  1  H  d  BB DEC. I decouple  180  80  d +d 2  3  DECOUPLER ON  Fig. A4: Pulse sequence for the APT experiment  6  186 ii) Two-dimensional experiments  A generalized pulse-sequence for a typical two-dimensional (2D) NMR experiment is shown below. The three stages outlined above for the typical ID experiment (preparation-evolutiondetection) are also present in the 2D experiment. However, the evolution time, t i , is incremented over a series of individual experiments (rather than being kept constant as in a ID experiment), during which time the magnetization "evolves" under the influence of chemical shift, scalar coupling, or radiofrequency pulses. Fourier transformation of the FIDs generated from these experiments yields a series of spectra that are modulated with respect to these evolution phenomena of the second time (ti) variable.  A second Fourier transform over ti results in a "two-  dimensional" spectrum as a function of two frequencies. e  preparation  evolution and mixing  e  n  detection  FID ^  tj + T  Fig. A5: Pulse sequence for a typical 2D NMR experiment  ii.A) ^ - i H COSY Spectrum  Homonuclear Correlation SpectroscopY is one of the most widely used 2D NMR techniques. The pulse sequence, shown below, consists of two pulses separated by an incremented delay. In very basic terms, the second pulse modulates the magnetization generated by the first pulse (which has undergone evolution under the influence of scalar coupling) such that in the final spectrum (after the 2nd Fourier transform) off-diagonal peaks correlate spins (i.e. protons) that are coupled to each other. The power of this technique is self-evident: the ability to trace correlations from protons that are coupled to each other in a complex natural product is of great help in establishing the final structure.  187 90°  9° detection  Fig. A6: Pulse sequence for the COSY experiment7  ii.B) HMQC Experiment  The inverse-detected Heteronuclear Multiple Quantum Coherence experiment establishes onebond !H- C connectivities via correlations from *H resonances on one axis into C resonances 13  13  on the other. The experiment provides the same information as the C-detected HETeronuclear 13  CORrelation (HETCOR) experiment, with the advantage of improved sensitivity gained from acquiring the FID in the *H domain. The pulse sequence for the HMQC experiment is shown below. The initial sequence of pulses (BIRD) removes signals from protons not directly coupled to C nuclei (i.e. they are unaffected 1 3  by the 180° C pulse, and are thus made undetectable after the 9 0 \ *H pulse). The remainder 13  x  of the pulse sequence is optimized for polarization transfer from one-bond *H- C scalar coupling. 13  Typical delays used were: di = 1.5 -2.0 s (relaxation delay); d2 = 3.5 ms (1/2JCH);  d3  = 0.7 s  (optimized to eliminate *H signals bonded to C); 04 = 3 ^.s (compensation factor), and ti = 3 u,s 12  (normal incremented factor).  Fourier transformation in the t2 domain gives a series of FIDs  modulated by H - C scalar coupling, and a second FT with respect to ti results in a twoJ  13  dimensional plot of H (f2 axis) vs. C (f\ axis). !  1 3  The correlations between H and C  resonances in each dimension represent direct 'H-'^C connectivity in the molecule.  l  1 3  18  a-iLP 90°x  1H Observe  180"x  90°-x  90°x  IL  C dec. d  2  a  o 90'x  180'x  1 3  180°x  d  2  d3  90'x  IU1 d  tj/2  2  90'x  tj/2  n  B.B. dec. d4  d  2  I  Fig. A7: Pulse sequence for the HMQC experiment8  ii.C) HMBC experiment  The ^-detected Heteronuclear Multiple Bond multiple quantum Coherence experiment is one of the most powerful NMR techniques for structure elucidation.  Similar to the HMQC  experiment, the HMBC is optimized, however, for long range (usually two and three bond) H J  13  C connectivity. The resultant spectrum after Fourier transformation is like the HMQC (i.e. *H  is the f2 axis, C is the fi axis), except that the correlations represent two and three bond H- C 13  J  13  connections. Thus, the connectivity data generated in the HMBC experiment, in combination with the COSY and HMQC data, allow one to trace all the proton-carbon bonds in the molecule. The pulse sequence for the HMBC experiment is shown below. All the one-bond ^ - ^ C magnetization is removed by the initial sequence: H : 90°, d2, C: 90°. !  13  The remainder of the  pulse sequence is optimized for the evolution of 'H magnetization under the influence of two and three bond *H- C scalar couplings, i.e. d3 = l/2JcH-long range- The initial delays (di and d2) are 13  the same as those in the HMQC spectrum, while the value of d3 typically used was 60 ms, and ti was typically 2 \xs.  18 90"x lH nh«Tvi»  180"x  P  di  d  2  d3  tj/2  tj/2  t2  Fig. A8: Pulse Sequence for the HMBC Experiment^  Endnotes: Appendix A: NMR Experiments 1. a) Silverstein, R.M.; Bassler, G.C.; Morrill, T.C., Spectrometric Identification of Organic Compounds. 4th ed., John Wiley and Sons, New York, 1981; b) Derome, A.E., Modern NMR Techniques for Chemistry Research. Pergamon Press, Oxford, 1987; c) Nakanishi, K., ed. OneDimensional and Two-dimensional NMR Spectra by Modern Pulse Techniques. University Science Books, Mill Valley, CA, 1990; d) Kessler, H.; Gehrke, M.; Griesinger, C , Angew. Chem. Int. Ed. Engl., 1988,27, 490; e) Sadler, I.H., Nat. Prod. Rep., 1988, 101.  2. see ref. lb) above. 3.  Benn, R.; Gunther, H, Angew. Chem. Int. Ed. Engl., 1983, 22, 350.  4. Sanders, J.K.M.; Mersh, J.D., Prog. Nucl. Magn. Res., 1982, 25, 353. 5. see ref. Id) above. 6.  Patt, S.L.; Shoolery, J.N., J. Magn. Reson., 1982,46, 535.  7. a) Jeener, J. Ampere International Summer School, Basko Polje, Yugoslavia,. 1971; b) Aue, W.P.; Bartholdi, E.; Ernst, R.R.; J. Chem. Phys., 1976, 64, 2229; c) Bax, A.; Freeman, R., J. Magn. Reson., 1981,44, 542. 8.  Bax, A.; Subramanian, S., J. Magn. Reson., 1986, 67, 565.  9. a) Bax, A.; Summers, M.F., J. Am. Chem. Soc, 1986, 108, 2093; b) Summers, M.F.; Marzilli, L.G.; Bax, A., J. Am. Chem. Soc, 1986,108, 4285.  190 Appendix B: Isolation of Known Metabolites from New Sources  Fig. B l : Color Plates of Anisodoris fontaini (above) and Thecacera darwinii (below)  191 i) Anisodoris  fontaini  Specimens (15 animals) of Anisodoris fontaini  were collected by hand using SCUBA at  depths of 5 to 7 metres, in the Bay of Coliumo near Dichato, Chile. Taxonomic identification was performed by Dr. Sandra Millen at the University of British Columbia.  The animals were  immediately immersed in MeOH at the surface and transported back to Vancouver for further study.  The original 500 mL MeOH extract was decanted and filtered.  Four more 500 mL  extracts were made from the animals in the following order: MeOH, 2 X 1:1 MeOH: CH2CI2, MeOH. All extracts were decanted,filtered,combined, and reduced in vacuo, to yield an aqueous suspension. This suspension was diluted up to 500 mL with distilled water and extracted with four portions of 500mL EtOAc. Each EtOAc extract was dried over MgS04,filtered,combined, and reduced in vacuo to yield an orange oil.  The oil that resulted from the EtOAc/ H20  partitioning of the combined crude extracts was chromatographed on silica gel (230 -400 m mesh) under medium N2 flash in 1:1 hexanes: EtOAc to yield a fraction that was highly enriched in the diterpenoic acid glyceride (28).  This fraction was recrystallized from hexane (HPLC grade) to  yield pure (28) (25 mg). Diterpenoic acid glyceride (28) was isolated as white needles: m.p. = 137 - 139 °C; H NMR l  (400 MHz, CDCI3) 5 0.79 (s, 3H), 5 0.80 (m, 1H), 0.83 (m, 1H), 0.84 (s, 3H), 6 0.89 (s, 3H),  6 0.92 (s, 3H), 51.11 (m, 1H), 6 1.14 (m, 1H), 1.27-1.42 (m, 4H), 6 1.51-1.55 (m, 2H), 6 1.58 (s, 3H), 6 d 1.61 (m, 1), 6 1.69 (m, 1H), 5 1.93 (m, 1H), 5 2.14 (bt, J = 6 Hz.exchangeable, 1H), 6 2.55 (bd, J = 5 Hz, exchangeable, 1H), 6 2.94 (bs, 1H), 6 3.62 (dd, J = 12, 5 Hz, 1H), 6 3.68 (m, 1H), 6 3.92 (m, 1H), 5 4.14 (dd, J = 11, 6 Hz, 1H), 5 4.19 (dd, J = 11, 4.5 Hz, 1H), 5 5.51 (bs, 1H);  13  C NMR (100 MHz, CDCI3) 6 15.6 (q), 6 15.7 (q), 6 18.5  (t), 5 18.7 (t), 6 21.2 (q), 6 21.6 (q), 6 22.7 (t), 6 33.0 (s), 5 33.2 (q), 6 36.6 (s), 6 37.4 (s), c 39.9 (t), 6 41.9 (t), 5 41.9 (t), d 54.3 (d), 5 56.5 (d), 5 62.6 (d), 5 63.5 (t), 6 65.1 (t), 8 70.4 (d), 6 124.3 (d), 5 128.5 (s), 5 173.4 (s); HREIMS m/z = 378.27640 (C23H38O4 calculated = 378.27701, AM = 0.6).  192  ii) Thecacera  darwinii  Specimens (32 animals) of Thecacera darwinii were collected by hand using SCUBA at depths of 2 to 4 metres, in the Bay of Coliumo near Dichato, Chile. Taxonomic identification was performed by Dr. Sandra Millen at the University of British Columbia. immediately immersed in MeOH at the surface.  The animals were  The methanol extract was returned to the  laboratory in Vancouver, where it was decanted, filtered, and reduced in vacuo . The animals were subsequently exhaustively extracted with two further 250 mL portions of methanol, and two 250 mL portions of 1:1 MeOH: CH2C12- All extracts were filtered, combined, and reduced in vacuo to yield an aqueous suspension. This material was further diluted to 500 mL with distilled water and exhaustively extracted with 4 X 500 mL portions of EtOAc. The EtOAc layers were dried over MgS04, filtered, combined, and reduced in vacuo to yield an orange oil (260 mg). The EtOAc soluble materials were fractionated by silica gel flash chromatography (eluent: 85:15 hexanes: EtOAc) to yield a sample containing almost pure triophamine (8) (12 mg). This fraction was further purified on reversed phase HPLC (eluent: 4:1 MeOH: H 2 O ) to yield 5.3 mg of pure triophamine (8). Triophamine (8) was isolated as a colorless oil: *H NMR (500 MHz, CDCI3) 6 0.89 (t, J = 7.5 Hz, 3H), 6 0.94 (t, J = 7.5 Hz, 3H), 1.50 (m, 1H), 1.54 (d, J = 6.8 Hz, 3H), 8 1.59 (m, 1H), 8 2.0 (q, J = 7.5 Hz, 2H), 8 2.11 (m, 1H), 8 2.29 (m, 1H), 8 2.31 (m, 1H), 8 5.18 (q, J = 6.8 Hz);  1 3  C  NMR  (125  MHz,  C D C I 3 ) 8 12.1  (q), 8 12.7  (q), 8 13.1  (q), 8 22.6  (t), 8 25.4  (t), 8  39.3 (t), 8 50.2 (d), 8 120.5 (d), 8 139.0 (s), 8 157.4 (s), 8 185.6 (bs); HREIMS m/z = 363.28857 (C21H37O2N3 calculated = 363.28857, AM = -0.7).  

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.831.1-0059575/manifest

Comment

Related Items