CHEMICAL STUDIES ON THE ORIGIN OF SECONDARY METABOLITES IN SELECTED MARINE INVERTEBRATES by JULIA MAPJE KUBANEK B.Sc, Queen's University, 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 April 1998 © Julia Kubanek, 1998 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) A B S T R A C T Precursor incorporation experiments with several widespread species of Northeastern Pacific Ocean dorid nudibranchs have led to the conclusion that each species studied is capable of de novo biosynthesis. Incorporation studies with [1,2- 1 3C2] acetate, [l,2-13C2,18Oi]acetate, and [2-13C]mevalonate were used to demonstrate the biosynthesis of the sesquiterpenoid albicanyl acetate (18), and the sesterterpenoids cadlinaldehyde (19) and luteone (20) by Cadlina luteomarginata. Quantitative analysis of C NMR data showed that only a small turnover of metabolite took place but that the newly formed molecules had extremely high levels of label incorporation. As a by-product of the examination of biosynthesized metabolites of C. luteomarginata, a novel diterpenoid metabolite, seco-spongian 22, was isolated and identified. 18 19 20 22 A study of Californian specimens of C. luteomarginata indicated that the three biosynthesized metabolites are present across its geographic range. Examination of individual C. luteomarginata specimens from one B.C. population revealed significant variation in metabolite concentration between individuals and between tissues of each individual. Incorporation experiments with [1,2- 1 3C 2] acetate were also used to demonstrate the de novo biosynthesis of acanthodoral (28) by Acanthodoris hudsoni, a metabolite previously known only from the related nudibranch A. nanaimoensis. Attempts to incorporate 1 8 0 labels into the aldehyde oxygen atoms of nanaimoal (24), isoacanthodoral (27), and 28 in these two species revealed that the aldehyde oxygens are not retained in the biosynthesis. o De novo biosynthesis of the 3-hydroxybutyrate side chain of diaulusterol A (34) in Diaulula sandiegensis was proven by incorporation of [1,2- C2]acetate. None of [1,2-13C2]acetate, [2-14C]mevalonate, and [4-14C]cholesterol was successfully incorporated in the steroid portion of 34. Synthesis of [2,l'-13C2]disodium 2-ethylmalonate and [2,3-13C2]sodium butyrate followed by injection into the nudibranch Triopha catalinae demonstrated that T. catalinae utilizes butyrate in the processive polyketide biosynthesis of triophamine (43). This represented the first proven use of butyrate in polyketide biosynthesis by a marine invertebrate. o OH 34 43 Investigation of the extracts of the North Sea flatworm Prostheceraeus vUlatus and its prey, the ascidian Clavelina lepadiformis, revealed the presence of one known and four novel secondary metabolites: the decahydroquinoline alkaloid lepadins A, B, and C and pyrrolidine alkaloid villatamines A and B. P. villatus concentrates the alkaloids from its diet, analogous to the relationship between many shell-less molluscs and their algal or sponge dietary organisms. iv T A B L E O F C O N T E N T S Abstract ii Table of contents iv List of tables..., viii List of figures x List of schemes xiv List of abbreviations xv Acknowledgements x viii 1. General introduction to marine chemical ecology 1 2. Biosynthetic investigations of nudibranch secondary metabolism 11 2.1 Introduction to nudibranch secondary metabolism including biosynthesis 11 2.2 Investigations of biosynthesis by Northeastern Pacific Ocean nudibranchs 15 2.2.1 Terpenoid biosynthesis in Cadlina luteomarginata 15 2.2.1.1 Review of known metabolites of Cadlina luteomarginata 15 2.2.1.2 A new metabolite of Cadlina luteomarginata 21 2.2.1.2.1 Collection of specimens and isolation of a novel seco-spongian metabolite. 21 2.2.1.2.2 Structure determination of seco-spongian 22 22 2.2.1.3 Results of biosynthetic experiments with Cadlina luteomarginata 28 2.2.1.3.1 Collection, incubation of precursors and isolation of metabolites 28 2.2.1.3.2 Biosynthesis of albicanyl acetate (18) 32 2.2.1.3.3 Cadlinaldehyde (19) and luteone (20) sesterterpenoid skeletons 40 2:2.1.3.4 Conclusions of biosynthetic study 53 2.2.1.4 Geographic and anatomical distribution of metabolites in C. luteomarginata 57 2.2.1.4.1 Collection, dissection, and method of analysis 57 2.2.1.4.2 Results of quantitative analysis of metabolites 60 V 2.2.1.4.3 Conclusions of distribution study 67 2.2.2 Terpenoid biosynthesis in Acanthodoris nanaimoensis and A. hudsoni 71 2.2.2.1 Review of known metabolites and terpenoid biosynthesis by A. nanaimoensis... 71 2.2.2.2 Isolation of sesquiterpenoids from Acanthodoris hudsoni 75 2.2.2.3 NMR assignments for nanaimoal (24) and isoacanthodoral (27) 81 2.2.2.4 Results of biosynthetic experiments with Acanthodoris specimens 86 1.2.2AA Collection of specimens, synthesis and incubation of precursors, and isolation of metabolites 86 2.2.2.4.2 Results of the [l,2-l3C2,180i]acetate incorporation experiment 91 2.2.2.4.3 Results of [l,2-13C2]farnesal and [l,2-13C2]farnesyl pyrophosphate incorporation experiments 96 2.2.2.5 Conclusions of biosynthetic study 96 2.2.3 Investigations of diaulusterol biosynthesis by Diaulula sandiegensis 98 2.2.3.1 Review of known metabolites of D. sandiegensis 98 2.2.3.2 Review of steroid biosynthesis in molluscs 101 2.2.3.3 Assignment of the NMR data for diaulusterol A (34) 103 2.2.3.4 Results of biosynthetic experiments with D. sandiegensis 107 2.2.3.4.1 Collection, incubation of precursors and isolation of diaulusterol A (34)... 107 2.2.3.4.2 Polyketide origin of 3-hydroxybutyrate moiety of diaulusterol A (34) 108 2.2.3.4.3 On the assembly of the steroid portion of 34 111 2.2.3.4.4 Conclusions of the biosynthetic study 113 2.2.3.5 Geographic variation of metabolites in D. sandiegensis 115 2.2.4 Investigations of polyketide biosynthesis by the Northeastern Pacific nudibranch Triopha catalinae 118 2.2.4.1 Review of skin chemistry of T. catalinae 118 2.2.4.2 Results of biosynthetic experiments with Triopha catalinae 124 vi 2.2.4.2.1 Collection of specimens, incubation of precursors, and isolation of triophamine (43) 124 2.2.4.2.2 Evidence for butyrate as polyketide intermediate in the biosynthesis of triophamine (43) 127 2.2.4.2.3 On the origin of the guanidine function of triophamine (43) 134 2.2.4.2.4 On the existence of isobutyryl-CoA mutase in T. catalinae 134 2.2.4.3 Conclusions of the biosynthetic study 137 3. New metabolites of the North Sea flatworm Prostheceraeus villatus and its ascidian prey Clavelina lepadiformis 139 3.1 Review of known metabolites of flatworms 139 3.2 Review of known metabolites of Clavelina lepadiformis and other ascidians 140 3.3 New metabolites of Prostheceraeus villatus and Clavelina lepadiformis 145 3.3.1 Collection of specimens 145 3.3.2 Isolation of lepadins and villatamines 147 3.3.3 Structure determination of the lepadins and villatamines 149 3.3.3.1 Lepadin A (66) 149 3.3.3.2 Lepadin B (73) 151 3.3.3.3 Lepadin C (74) 155 3.3.3.4 Villatamine A (75) 159 3.3.3.5 Villatamine B (76) 164 3.4 Conclusions of flatworm-ascidian study 168 4. Conclusions 169 5. Experimental section 172 5.1 Materials and Methods 172 5.2 Investigations of secondary metabolism of C. luteomarginata 174 5.2.1 New metabolite of C. luteomarginata 174 5.2.2 Biosynthetic experiments with C. luteomarginata 175 vii 5.2.3 Investigations into distribution of metabolites in C. luteomarginata 177 5.3 Biosynthetic experiments with A. nanaimoensis and A. hudsoni 180 5.4 Biosynthetic experiments with D. sandiegensis 186 5.5 Biosynthetic experiments with T. catalinae 189 5.6 New metabolites of P. villatus and C. lepadiformis 194 6. Appendix 197 References and notes 215 L I S T O F T A B L E S viii Page Table 2.1. NMR data for seco-spongian 22 recorded in CDC13 25 Table 2.2. Yields of terpenoids 18,19, and 20 obtained from C. luteomarginata 32 Table 2.3. 1 3 C NMR incorporation data for albicanyl acetate (18) labelled with [ 1,2-13C2] acetate recorded in CDC13 34 Table 2.4. Specific incorporation of [2-13C]mevalonolactone into cadlinaldehyde (19) and luteone (20) 51 Table 2.5. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in California and B.C. specimens of C. luteomarginata 60 Table 2.6. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in individual B.C. specimens of C. luteomarginata 63 Table 2.7. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in dissected tissues of 10 specimens of C. luteomarginata 66 Table 2.8. NMR data for acanthodorol (32) recorded in CDC13 80 Table 2.9. [ H and 1 3 C NMR data for nanaimoal (24) and isoacanthodoral (27) recorded in CDC13 86 Table 2.10. 1 3 C NMR incorporation data for acanthodorol (32) labelled with [ l ,2- 1 3 C 2 , 1 8 Oi] acetate recorded in CDCI3 91 Table 2.11. NMR data for diaulusterol A (34) recorded in C 6D 6 106 Table 2.12. 1 3 C NMR incorporation data for diaulusterol A (34) labelled with [ 1,2-13C2]sodium acetate recorded in CeD6 108 Table 2.13 Specific incorporation of [2,3-13C2]sodium butyrate in triophamine (43) 133 Table 3.1. In vitro ED 5 0 values (p:g/mL) of metabolites from P. villatus and C. lepadiformis 148 Table 3.2. NMR spectra for lepadin A (66) in CDC13 149 Table 3.3. NMR data for lepadin B (73) recorded in CDC13 154 Table 3.4. NMR data for lepadin C (74) recorded in CDC13 158 IX Table 3.5. NMR data for villatamine A (75) in CD2C12 164 Table 3.6. NMR data for villatamine B (76) recorded in CDC13 165 X L I S T O F F I G U R E S Page Figure 2.1. Cadlina luteomarginata (photo by R. Long) 15 Figure 2.2. Sites along the western coast of North America where the terpenoid constituents of Cadlina luteomarginata have been studied 17 Figure 2.3. Carbon skeletons represented in terpenoids isolated to date from C. luteomarginata 18 Figure 2.4. Metabolites of Southern California Cadlina luteomarginata (and their sponge source where known) 19 Figure 2.5. Metabolites of British Columbia Cadlina luteomarginata 20 Figure 2.6. 'H NMR spectrum of seco-spongian 22 in CDC13 23 Figure 2.7. 1 3C NMR spectrum of seco-spongian 22 in CDC13 24 Figure 2.8. Selected HMBC (left) and COSY (right) correlations of seco-spongian 22 26 Figure 2.9. Selected NOE resonances of seco-spongian 22 27 Figure 2.10. *H and 1 3 C NMR spectra of albicanyl acetate (18) in CDC13 29 Figure 2.11. *H and 1 3 C NMR spectra of cadlinaldehyde (19) in CDC13 30 Figure 2.12. ! H and 1 3 C NMR spectra of luteone (20) in CDC13 31 Figure 2.13. C NMR resonances of selected carbons of albicanyl acetate (18) labelled with [ 1.2-13C2]acetate and unlabelled 33 Figure 2.14. C NMR resonances of selected carbons of albicanyl acetate (18) labelled with [1,2-13C2]acetate and with [l,2-13C2,180i]acetate 39 Figure 2.15. 1 3 CNMR resonances from the spectrum of cadlinaldehyde (19) incorporated with [ 1,2-13C2] acetate 41 Figure 2.16. 1 3C NMR resonances from the spectrum of luteone (20) incorporated with [1,2-13C2] acetate 42 Figure 2.17. Expanded 1 3 C NMR resonance of C-18 from the spectrum of luteone (19) incorporated with [1,2-l3C2] acetate 46 XI Figure 2.18. 1 3 C NMR spectrum of cadlinaldehyde (19) incorporated with [2-13C]mevalonolactone 50 Figure 2.19. 1 3 C NMR spectrum of luteone (20) incorporated with [2-13C]mevalonolactone 52 Figure 2.20. Some dissected tissues of Cadlina luteomarginata 59 Figure 2.21. GC traces of extracts of C. luteomarginata individuals # 4 and #5 64 Figure 2.22. Acanthodoris nanaimoensis (photo by R. Long) 72 Figure 2.23. Acanthodoris hudsoni (photo by R. Long) 75 Figure 2.24. 'PI NMR spectrum of acanthodorol (32) in CDC13 78 Figure 2.25. 1 3 C NMR spectrum of acanthodorol (32) in CDC13 79 Figure 2.26. *H NMR spectrum of nanaimoal (24) in CDC13 82 Figure 2.27. 1 3 C NMR spectrum of nanaimoal (24) in CDC13 83 Figure 2.28. *H NMR spectrum of isoacanthodoral (27) in CDC13 84 Figure 2.29. 1 3 C NMR spectrum of isoacanthodoral (27) in CDC13 85 Figure 2.30 'H and 1 3C NMR spectra of nanaimool (33) in CDC13 89 Figure 2.31. ! H and 1 3C NMR spectra of isoacanthodoral (31) in CDC13 90 Figure 2.32. 1 3 C NMR resonances of acanthodorol (32) labelled with [ 1,2-13C2,I80i] sodium acetate and control 92 Figure 2.33. Labelling pattern of intact acetate units observed in acanthodorol (32) incorporated with [ 1,2-13C2) 1 8 01 ] acetate 93 Figure 2.34. Diaulula sandiegensis (photo by R. Long) 98 Figure 2.35. 'H NMR spectrum of diaulusterol A (34) in C 6D 6 104 Figure 2.36. 1 3 C NMR spectrum of diaulusterol A (34) in C 6D 6 105 Figure 2.37. 1 3C NMR resonances of selected carbons of diaulusterol A (34) labelled with [1,2- 1 3C 2] sodium acetate 109 Figure 2.38. ! H NMR spectrum of diaulusterol B (35) in C 6D 6 116 X l l Figure 2.39. Triopha catalinae (photo by R. Long) 119 Figure 2.40. Incorporation of butyrate fragments in lasalocid A (44) and monensin (45) 121 Figure 2.41. Marine natural products other than triophamine (43) that may utilize butyrate as a biosynthetic precursor 122 Figure 2.42. ! H NMR spectrum of triophamine (43) in CDC13 125 Figure 2.43. 1 3 C NMR spectrum of triophamine (43) in CDC13 126 Figure 2.44. 1 3 C NMR spectrum of triophamine (43) in CDC13 from [l-13C]butyrate incorporation experiment 128 Figure 2.45. Expected incorporation of [l-t3C]butyrate into triophamine (43) 127 Figure 2.46. 1 3 C NMR resonances from the spectra of triophamine (43) obtained from incorporation experiments with [2,3-13C2]sodium butyrate and [2,l'-13C2]disodium 2-ethylmalonate and from a control sample 132 Figure 3.1. The flatworm Prostheceraeus villatus feeding on the ascidian Clavelina lepadiformis; two specimens of P. villatus (photos by M. Leblanc) 146 Figure 3.2. 'H and 1 3 C NMR spectra of lepadin A (66) in CDC13 150 Figure 3.3. *H NMR spectrum of lepadin B (73) in CDC13 152 Figure 3.4. 1 3C NMR spectrum of lepadin B (73) in CDC13 15 3 Figure 3.5. ! H NMR spectrum of lepadin C (74) CDC13 156 Figure 3.6. 1 3C NMR spectrum of lepadin C (74) CDC13 157 Figure 3.7. Selected COSY and HMBC correlations of lepadin C (74) 159 Figure 3.8. 1 3 C NMR spectrum of villatamine A (75) in CD2C12 161 Figure 3.9. ! H NMR spectrum of villatamine A (75) in CD2C12 162 Figure 3.10. Selected COSY and HMBC correlations of villatamine A (75) 160 Figure 3.11. More selected COSY and HMBC correlations of villatamine A (75) 163 Figure 3.12. *H NMR spectrum of villatamine B in CDC13 166 xiii Figure 3.13. 1 3 C NMR spectrum of villatamine B in CDC13 167 Figure A. 1. 1 3 C NMR resonances from the spectrum of albicanyl acetate (18) incorporated with [1,2- 1 3C2] acetate 197 Figure A.2. HMQC spectrum of seco-spongian 22 in CDCI3 198 Figure A.3. HMBC spectrum of seco-spongian 22 in CDCI3 199 Figure A.4. HMBC spectrum of seco-spongian 22 in CDCI3 200 Figure A.5. COSY spectrum of seco-spongian 22 in CDCI3 201 Figure A. 6. COSY spectrum of lepadin B (73) in CDCI3 202 Figure A.7. HMQC spectrum of lepadin B (73) in CDC13 203 Figure A. 8. HMBC spectrum of lepadin B (73) in CDC13 204 Figure A.9. COSY spectrum of lepadin C (74) in CDC13 205 Figure A. 10. HMQC spectrum of lepadin C (74) in CDC13 206 Figure A. 11. HMBC spectrum of lepadin C (74) in CDC13 207 Figure A. 12. HMQC spectrum of villatamine A (75) in CD2C12 208 Figure A. 13. COSY spectrum of villatamine A (75) in CD2C12 209 Figure A. 14. HMBC spectrum of villatamine A (75) in CD2C12 210 Figure A. 15. HMBC spectrum of villatamine A (75) in CD2C12 211 Figure A. 16. COSY spectrum of villatamine B (76) in CDCI3 212 Figure A. 17. HMQC spectrum of villatamine B (76) in CDC13 213 Figure A. 18. HMBC spectrum of villatamine B (76) in CDC13 214 XIV L I S T O F S C H E M E S Page Scheme 2.1. Proposed biogenesis of albicanyl acetate (18) from [1,2-l3C2]acetate 35 Scheme 2.2. Proposed biogenesis of albicanyl acetate (18) from [ 1,2-13C2,1801 ] acetate via two possible mechanisms 37 Scheme 2.3. Proposed biogenesis of cadlinaldehyde (19) and luteone (20) from [1,2-'3C2] acetate 43 Scheme 2.4. Proposed biogenesis of cadlinaldehyde (19) and luteone (20) from [2- C]mevalonolactone 53 Scheme 2.5. Biogenetic proposal of nanaimoal (24), isoacanthodoral (27), and acanthodoral (28) according to Graziani and Andersen 73 Scheme 2.6. Proposed biosynthetic pathway to nanaimoal (24) via an aldehyde intermediate 73 Scheme 2.7. Synthesis of (£,£)-[l,2-13C2]farnesal and (£,£)-[l,2-13C2]farnesyl pyrophosphate 88 Scheme 2.8. Proposed biosynthesis of nanaimoal (24), isoacanthodoral (27), and acanthodoral (28) via imine intermediates 95 Scheme 2.9. Biosynthesis of triophamine (43) from acetate 119 Scheme 2.10. Proposed biosynthesis of triophamine (43) from [l,2-13C2]2-ethylmalonateor [2,3-13C2]butyrate 120 Scheme 2.11. The syntheses of [2,l'-13C2]disodium 2-ethylmalonate and [2,3-13C2]sodium butyrate 131 Scheme 2.12. Interconversion of isobutyrate and butyrate catalyzed by isobutyryl-CoA mutase 135 Scheme 2.13. Expected C labelling pattern in triophamine (43) due to incorporation of [l,2-13C2]butyrate from [l,3-13C2]isobutyrate via isobutyryl-CoA mutase 136 Scheme 2.14. The synthesis of [1,3- C2]sodium isobutyrate 137 Scheme 3.1. Proposed biosynthesis of eudistomins H (62) and I (63) from tryptophan 142 Scheme 3.2. Biosynthetic proposal of shermilamine B (65) 144 LIST O F ABBREVIATIONS xv [OC]D specific rotation at wavelength of sodium D line at 25 °C Ac acetyl APT attached proton test B.C. British Columbia br broad c concentration CA California CeD6 deuterated benzene CDCI3 deuterated chloroform CD2CI2 deuterated dichloromethane CoA coenzyme A COSY correlation spectroscopy 8 chemical shift in parts per million d doublet D 20 deuterated water DIBAL diisobutylaluminium hydride DMSO dimethyl sulfoxide DMSO-d6 deuterated dimethyl sulfoxide DPM decays per minute e molar absorptivity coefficient ED50 effective dose resulting in 50% response ELMS electron impact mass spectrometry Enz enzyme Et ethyl EtOAc ethyl acetate FTIR Fourier transform infrared GC gas chromatography GCMS gas chromatography mass spectrometry HMBC heteronuclear multiple bond multiple quantum coherence HMG 3-hydroxy-3-methylglutaryl HMPA hexamethylphosphoramide HMQC heteronuclear multiple quantum coherence HO Ac acetic acid HPLC high performance liquid chromatography HRDCIMS high resolution desorption chemical ionization mass spectrometry HREIMS high resolution electron impact mass spectrometry HRFABMS high resolution fast atom bombardment mass spectrometry i signal due to impurity iPrOH isopropyl alcohol IR infrared J scalar coupling constant X,m a x wavelength of maximum absorption LDA lithium diisopropylamide m multiplet M + molecular ion xvii MDF mantle dermal formation Me methyl MeOH methanol m/z mass to charge ratio NMO A -^methylmorpholine-N-oxide NMR nuclear magnetic resonance NOE nuclear Overhauser effect OPP pyrophosphate q quartet Rf retardation factor ROESY rotating frame Overhauser effect spectroscopy s singlet or signal due to solvent SCUBA self-contained underwater breathing apparatus sp. species t triplet TFA trifluoroacetic acid TFfF tetrahydrofuran TLC thin layer chromatography TPAP tetrapropylammonium perruthenate U universal UBC University of British Columbia UV ultraviolet w signal due to water ACKNOWLEDGEMENTS X V l l l My most profound appreciation is extended to my research supervisor, Professor Raymond J. Andersen. He demonstrates serious attention to scientific methodology as well as rapt engagement in the subject matter, communicating his understanding of research to students and associates with duly high expectations of our contributions to science. Regards and appreciation are also due to the members of Professor Andersen's research group, who provided much scientific insight and technical assistance. Many students and staff including Michael Leblanc were instrumental in collecting marine specimens in the name of this research. Dr. Sandra Millen provided much needed information on nudibranch life cycles and habitats. The staff of the Bamfield Marine Station are thanked for hosting our collection expeditions and for assisting with animal maintenance, particularly Joelle Harris who handled specimens on my behalf during several experiments. The members of the Marine Biological Station at the University of Bergen, Norway, are similarly remembered for their hospitality and assistance. Professor D. John Faulkner of Scripps Institution of Oceanography (SIO) is thanked for inviting me to La Jolla, California for the comparative study on nudibranch metabolism and for providing the resources of his laboratory during that project. Ron McConnaughey of SIO assisted in the collection of specimens in La Jolla. 1 1. General introduction to marine chemical ecology Marine animals and plants are known to contain chemical compounds with strong and often very specific biological activities. Analogous to the chemical wealth of terrestrial plants, many of these marine natural products1 are currently being adapted for therapeutic use in humans.2 Scientists have accumulated a significant body of knowledge on the chemical structures and possible human applications of these compounds, and have even gained some understanding of their biological modes of action against human pathogens and diseases.3 However, progress in this area has not been matched by a similar growth in understanding about the origins, functions, and consequences of these compounds in the organisms and ecosystems in which they are found.4 Many coastal marine environments are resplendent with a diversity of both microscopic and macroscopic life forms. Competition for survival between individuals of a species and between species of an ecosystem occurs in relation to several factors such as space and food.5 It is beginning to be evident that natural products, which have been studied for potential human applications, also play a significant role in the interactions between organisms in marine ecosystems. Bioactive secondary metabolites can act as predator deterrents,6 as anti-fouling (preventing settling of other organisms on top of chemically-protected ones) and anti-overgrowth agents (preventing encroachment of other organisms),7 and as cues in the larval settling and metamorphosis of marine organisms.8'9 These compounds contribute not only to the success of the organisms that deploy them, but also to the biodiversity of the entire environment.10 For example, the chemically-defended green alga Chlorodesmis fastigiata of tropical Australian waters was found to provide refuge and associational defences to approximately 16 smaller 2 organisms that congregate near the algae to benefit from its chemical repugnance towards potential predators. The organisms indirectly protected by the algae's secondary metabolites include species of crabs, molluscs, and other algae.11 It is also becoming apparent that compounds exuded by bacteria, microalgae, and fungi are able to dramatically affect species' abundance and ecosystem structure.10 2,3-Indolinedione (1), a metabolite of an Alteromonas sp. bacterium, was found to protect embryos of the shrimp Palaemon macrodactylus from a pathogenic fungus. The shrimp embryos are normally covered with a layer of Alteromonas sp. that exudes a protective halo of 1. Removal of the bacteria resulted in the death of the embryos, while removal of the bacteria but reapplication of the pure compound 1 to the embryos kept them viable.12 O 1 As the above examples illustrate, marine ecosystems are complex and unique. Therefore, theories held by ecologists can benefit from the scrutiny of testing them in this understudied environment. For various reasons including accessibility, most currently held ecological theories were formulated from studies on terrestrial ecosystems.13'14 These theories clearly require testing on marine ecosystems in view of the fact that some animal phyla abundantly represented in oceanic environments, such as sponges (Phylum Porifera) and corals (Phylum Cnidaria), are 3 uniquely aquatic. Even marine organisms and communities which share qualities with terrestrial counterparts cannot be assumed to function according to the same principles. Marine chemical ecologists are revisiting many established theories, asserting, for example, that "seaweeds are not wet trees".13 Recent collaborations between ecologists and chemists have resulted in the development of some preliminary understanding of marine chemical ecology and some agreement as to how experimentation in this fresh area of research should be conducted.4 One of the major goals in developing natural products as drugs or agrichemicals is finding abundant and renewable sources of the compounds of interest. An understanding of the ecology of metabolite-producing organisms can assist in solving this problem. For example, the question, "Where does this compound come from?" must be answered before one can plan mass-production based upon harvesting or biotechnological methods. Thus, from both academic and industrial points of view, the true origin of marine natural products forms one of the most basic problems in marine chemical ecology. This problem is particularly relevant in the study of marine invertebrate metabolites, due to the many and varied chemical structures represented and the complex relationships between organisms. Secondary metabolites found in a certain invertebrate can be of two general origins, de novo biosynthesis by the subject invertebrate or acquisition from another organism. Those that have been acquired from another organism may have been sequestered from a prey species or acquired through an associational (e.g. symbiotic) relationship between two organisms, one living in another's tissue. Examples of each of these sources of secondary metabolties are described below. Predator-prey sequestration has been proven to occur in many instances between shell-less or shell-diminished molluscs and their prey. Anaspidean molluscs such as the sea hare Aplysia californica are well known for sequestering halogenated mono- and sesquiterpenoid metabolites from the algae upon which they feed.15 Herbivorous sacoglossan molluscs also 4 concentrate algal natural products and have even been reported to sequester entire functional chloroplasts from dietary algae which continue to be capable of photosynthesis for a period of weeks or months.16'17 The general phenomenon of predator-prey sequestration is also common with nudibranch molluscs, which commonly feed on chemically-rich sponges.18 For example, the sesquiterpenoids nakafuran-8 (2) and -9 (3) were isolated from the nudibranchs Hypselodoris godeffroyana and Chromodoris maridadilus and also from their prey, the sponge Dysidea fragilis.'* These compounds were found in laboratory assays to deter feeding by fish, indicating a defensive role for 2 and 3. Similarly, the Spanish dancer, Hexabranchus sanguineus, deploys a series of oxazole macrolides (e.g. 5,6-dihydrohalichondramide (4)) on its mantle and egg masses that were shown to deter feeding of reef fishes, and that were traced to two dietary sponges in the genus Halichondria. In fact, so many instances of nudibranch-sponge sequestration have been 18 21 documented that reviews on this specific subject have been written. ' It has been noted that the absolute and relative concentrations of metabolites in prey and predator tissues are often quite different, suggesting that predators possess mechanisms to deliberately select and concentrate 22 23 24 potent antifeedant compounds from their prey. ' ' Although the ecological roles of these 2 3 5 compounds may often be in question, their consistent presence, despite the evident toxicity of many such metabolites, has led researchers to conclude that they serve some adaptive purpose to the organisms that sequester them.6 Several natural products with strong biological activities found in marine sponges are now suspected to be products of associated bacteria or microalgae living in sponge tissues.25 In some cases, culturing microorganisms from fresh sponge tissue has resulted in proof of the microbial origin of the compounds. One successful example was the isolation of andrimid (5) 5 6 9 f i from the sponge Hyatella sp. and from a Vibrio sp. bacterium obtained from the sponge tissue. However, the culture of associated microorganisms from sponges remains extremely difficult due to the unique and mostly unknown conditions required for microbial growth. Proof of microbial origin has also come from the flow-cytometric separation of sponge and microbial cells, followed by chemical analysis of each extract. In this way, 13-demethylisodysidenin (6) was identified as a product of the symbiotic cyanobacterium Oscillatoria spongeliae and not of the sponge Dysidea herbacea which harboured the symbiont.28 In yet other cases, the occurrence of specific compounds from many taxonomically-remote organisms has demonstrated a strong probability that the compounds are of microbial origin. For example, members of the extremely cytotoxic spongistatin family of polyether macrolides, including spongistatin-1 (7), have been found in several sponges including 9Q Spirastrella spinispirulifera, Hyrtios ahum, and Cinachyra sp., which each represent different sponge orders. The premise stated above has held especially true in the case of tetrodotoxin (8), the notorious pufferfish toxin. Tetrodotoxin (8) has been found not only in pufferfish, but also in crabs, octopus, and even in a terrestrial amphibian. Recently, 8 was traced to bacterial strains of Listonella pelagia, Alteromonas tetraodonis, and Shewanella alga which are undoubtably producing 8 from within the tissues of higher organisms.31 8 The biosynthetic origins of some marine natural products have been determined directly using precursor incorporation methodology. A steady groundwork is being laid in which de novo biosynthesis across a broad spectrum of organism and compound types is gradually being mapped.32'33 The paucity of published biosynthetic studies in this area is a testament to the difficulties inherent in applying current experimental techniques to metabolism that is little understood, in an environment to which humans are not adapted. The investigation of de novo biosynthesis in slow-growing organisms such as marine invertebrates is hindered by low levels of incorporation of biosynthetic precursors. The low incorporation rates have been attributed to slow in vivo biosynthesis and slow turnover of metabolite pools, resulting in high proportions of unlabelled material against which incorporation must be measured.32 This problem in detecting incorporation has been compounded by very small amounts of metabolite upon which entire studies must often be based. Complicated experimental conditions also contribute to failure of biosynthetic studies. Many marine organisms are difficult to maintain in aquaria and the alternative, field experimentation, requires specialized equipment and ready access over a long period of time. As well, precursors that are water-soluble may disperse in the surrounding seawater during the incubation period, further lowering incorporation levels. Overall, a host of difficulties associated with experimental design have delayed progress in the study of marine invertebrate biosynthesis. The potential for low detection thresholds has historically made biosynthetic investigators favour radiolabelling over stable isotope labelling. However, improvements in l 3 C and deuterium NMR have made stable isotopes increasingly popular because of their ease of handling and the fact that NMR detection provides information beyond simple confirmation of incorporation. The major advantage is that NMR-active stable isotope incorporation can lead to an understanding of the positioning of atoms from precursors in the natural product and of bonds retained through biosynthetic processes, without having to conduct labour-intensive chemical 1 ^ degradation experiments. The use of multiply C labelled precursors has become increasingly popular, enabling detection of 1 3 C- 1 3 C couplings in 1 3C NMR spectra which are visible as satellite peaks flanking the natural abundance 1 3C NMR singlet. This technique permits detection at lower levels of incorporation than does the quantification of slight increases in overall 1 3 C signal height from single label incorporation. The difficulties in managing biosynthetic experiments in marine invertebrates have not curtailed interest in this area, because of the valuable information that can be gleaned from positive results. Knowing the origin of secondary metabolites leads to improvements in biotechnology and ecological theories. But perhaps more importantly, what can be learned about the biosynthesis of marine natural products is extremely valuable to chemistry. This is because many classes of marine natural products are unique.2 Carbon skeletons without terrestrial precedent are to be found amongst terpenoid, polyketide, and alkaloid products from marine organisms. There are also functional groups (e.g. isonitriles) that are well represented in marine metabolites, while being extremely rare in terrestrial metabolites. Therefore, classical biosynthetic routes beg examination in order to account for the wealth of unique marine natural products. As well, the possibility exists that altogether new biosynthetic routes could be discovered. Even in well-studied terrestrial environments, previously unknown pathways have recently emerged. For example, in the last ten years Rohmer et al.M have identified a novel non-mevalonate isoprenoid pathway utilizing pyruvate and glyceraldehyde that appears to operate in many microorganisms and has even been implicated in the biosynthesis of paclitaxel (Taxol®) by a cell culture of the higher plant Taxus chinensis.35 The broad goal of the research presented in this thesis was to gain an understanding of the origin of secondary metabolites in a series of marine invertebrates and to describe their 10 biosyntheses where possible. De novo biosynthesis was investigated in several species of nudibranch mollusc. In these investigations, precursors labelled with stable and radioactive isotopes were used in incorporation studies aimed at deciphering biosynthetic pathways leading to secondary metabolites. Predator-prey sequestration of metabolites was studied in a flatworm known to feed on a colonial ascidian. Identification and comparison of metabolite content in the two species permitted conclusions to be drawn about the relationship between these two organisms. The investigations reported herein necessitated a knowledge of the molecular structure of each metabolite studied. Therefore, the isolation and structure elucidation of secondary metabolites forms an integral part of this thesis. 11 2. Biosynthetic investigations of nudibranch secondary metabolism 2.1 Introduction to nudibranch secondary metabolism including biosynthesis Nudibranchs are the marine molluscs that have been the most widely studied by chemists.36 Their apparent vulnerability to predation, including a complete lack of shell in the adult state, a slow creeping mode of locomotion, and typically bright colouration, has made nudibranchs ideal subjects for testing hypotheses of chemical defence.37 The first indication that nudibranchs were protected from predation by some chemical means came from Thompson,38 who described the acid secretions of several species of molluscs, including some nudibranchs. Thompson also observed a general phenomenon of unpalatability of nudibranchs to fish, and described the bitter taste (to his own tongue) of their dorsal mantles.39 The first evidence of a nudibranch chemical defence attributable to organic natural products came from the investigation of the nudibranch Phyllidia varicosa and its sponge prey Hymeniacidon sp., which yielded the volatile and odoriferous 9-isocyanopupukeanane (9).40 The compound 9 and its isomer 2-isocyanopupukeanane (10),41 which were both found to be 9 R1 = NC R2 = H 10 R1 = H R2 = NC 12 constituents of the nudibranch mucus, are believed to account for its toxicity42 to fish and crustaceans. Since that time, nudibranchs of the sub-order Doridacea have proven themselves to be rich sources of complex natural products. As a follow-up to the extensive chemical investigations of dorid nudibranchs, it was hypothesized that chemical defences permitted the ancestral nudibranchs to shed their shells, thereby reducing metabolic waste. According to Faulkner and Ghiselin,24 chemical repugnance came to replace the weighty and cumbersome protection of the shell. The majority of nudibranch compounds have been isolated from dietary sources as well as from the nudibranchs, thereby providing direct proof for sequestration. Many other nudibranch compounds are believed to originate from dietary sources due to their structural similarities with other dietary metabolites.37 However, the lack of proposed dietary origins for some compounds, as well as their structural novelty, quickly led to interest in the possibility of de novo biosynthesis by dorid nudibranchs. A pioneering study in this area by Cimino et al.43, revealed that the Mediterranean nudibranch Dendrodoris limbata produced the drimane sesquiterpenoid polygodial (11) by de novo biosynthesis, as indicated by incorporation of l 4 C labelled mevalonate. Subsequently, other analogues of 11 were also found44 to result from de novo biosynthesis by D. limbata and other Dendrodoris species. One of the most interesting biosynthetic products was olepupuane (12), which was thought to be a protected form of the extremely cytotoxic l l . 4 5 ' 4 6 D. grandiflora, one of the nudibranchs investigated by Cimino and co-workers, was found to sequester sponge metabolites as well as biosynthesizing sesquiterpenoids de novo.44'46 This example represents an overlap between the evolutionary adaptations of benefiting from an available dietary source of useful compounds and possessing the ability to make one's own defensive metabolites. AcO. Since the initial discoveries of de novo biosynthesis by some species of Mediterranean nudibranchs, temperate Pacific Ocean specimens have also been investigated for this capability. Sesqui- and diterpenoid constituents of the widely distributed nudibranchs Archidoris montereyensis and A. odhneri incorporated 1 4 C labelled mevalonate within detectable levels.47 More recently, a stable isotope incorporation study48 confirmed the 1 4 C results, demonstrating with 1 3C NMR that doubly 1 3 C labelled acetate was incorporated into the farnesic acid glyceride 13 by A. odhneri. The same methodology was used to prove that acetate was incorporated into the diterpenoic acid glyceride 14 by A. montereyensis. 13 14 14 Other studies have proven that acetate is incorporated by Acanthodoris nanaimoensis into sesquiterpenoid aldehydes49 and also by Triopha catalinae into a polyketide metabolite.50 These investigations will each be described in more detail in subsequent sections. Of particular note has been the hypothesis of geographic variation put forward by Faulkner, Andersen and co-workers,51 who postulated that determination of metabolite content within one species of nudibranch across its geographic range could be used to predict the origin of its metabolites. It was proposed that compounds found to be present across a nudibranch's geographic range are likely to be products of de novo biosynthesis, while compounds present in some locations and absent in others are likely to be sequestered from a dietary organism whose availability varies across the geographic range. To date, all nudibranch secondary metabolites proven to be derived by de novo biosynthesis have satisfied this hypothesis. It was considered of interest to determine whether this hypothesis could be extended to more species of nudibranchs and to include further ecological features of nudibranch secondary metabolism. 15 2.2 Investigations of biosynthesis by Northeastern Pacific Ocean nudibranchs 2.2.1 Terpenoid biosynthesis in Cadlina luteomarginata 2.2.1.1 Review of known metabolites of Cadlina luteomarginata Cadlina luteomarginata is a dorid nudibranch of the chromodorid family that is found in rocky intertidal and shallow subtidal habitats along the west coast of North America from Alaska to Mexico (Figure 2.1). Skin extracts of C. luteomarginata from many locations have proven Figure 2.1. Cadlina luteomarginata (photo by R. Long) to be a rich source of terpenoid secondary metabolites that in some instances have been shown to play a role in defending these shell-less molluscs from predation.53'54'55'5 ' 5 7 ' 5 8 ' 5 9 ' 6 0 Many of 16 these compounds have also been isolated from marine sponges on which C. luteomarginata feeds, demonstrating a dietary origin for these metabolites.54'57'58'59 C. luteomarginata is known to be a generalist sponge grazer, sampling different sponge species across many habitats. It has been shown that the terpenoid content of skin extracts of specimens of C. luteomarginata varies dramatically from location to location, reflecting for the most part the different terpenoid constituents present in the dietary sponges available at each site. The initial discovery of geographic variation in C. luteomarginata terpenoid constituents prompted similar investigations into metabolite content of other species and led ultimately to the general proposal that geographic invariance of metabolite content in nudibranch skin extracts was a good predictor of de novo biosynthesis of the metabolites.51 The terpenoid constituents present in skin extracts of C. luteomarginata have been identified for specimens collected at both the southern and northern extremes of their geographic range (Figure 2.2). Shown in Figure 2.3 are the carbon skeletons of all known compounds from extracts of C. luteomarginata. The Southern California specimens yielded sesquiterpenoid furans and a sesterterpenoid furan along with a number of sesquiterpenoid isonitriles, isothiocyanates and formamides (Figure 2.4).54 Compounds isolated from British Columbia specimens of C. luteomarginata are shown in Figure 2 .5 . 5 3 ' 5 5 ' 5 6 ' 5 7 ' 5 8 ' 5 9 ' 6 0 There are some similarities and some clear differences between the terpenoids reported from the B.C. and California populations of C. luteomarginata. Both groups of nudibranchs yielded sesquiterpenoid furans, isonitriles, isothiocyanates, and formamides; however, only three specific compounds, furodysinin (15), isonitrile 16, and formamide 17 had been reported from both general areas prior to this study. The B.C. specimens have been considered unique in that they contain the drimane sesquiterpenoid albicanyl acetate (18) as the major constituent and the degraded sesterterpenoids cadlinaldehyde (19) and luteone (20) in extracts from nearly all 17 . Sanford Island, Barkley Sound. BC 11. Tasu Sound, BC 2. Passage Island. Howe Sound, BC 12. RenncH Sound BC 3. Kelvin Grove, Howe Sound, BC 13. Emilia Island. BC 4. Agamemnon Channel, BC 14. Langara Island, BC 5. East Redonda Island. BC 15. Mo.ra Sound. Alaska 6. Pendrell Sound, BC 16. Helta Inlet, Alaska 7 Rivers Inlet, Bull Island. BC 17. Point bugema. Mexico 8. Quatsino Sound. BC 18. La Jolla. California 9. Anthony Island, BC 19. Auke Bay. Alaska Narrows, BC ^ | | H H H H H H H H H H R Figure 2.2. Sites along the western coast of North America where the terpenoid constituents of Cadlina luteomarginata have been studied locations in British Columbia. Several degraded and rearranged spongian diterpenoids and two marginatane diterpenoids, whose occurrences are very site dependent, have also been found in B.C. specimens but have not been reported from California. The biological origins of most metabolites from C. luteomarginata are not in question; they have been sequestered unchanged from sponges on which the nudibranchs feed. However, the partial geographic invariance of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20), coupled with the fact that the carbon skeletons of these compounds had not been 18 Figure 2.3. Carbon skeletons represented in terpenoids isolated to date from C. luteomarginata traced to a dietary source (the latter two being altogether unprecedented), encouraged the question of whether C. luteomarginata might be producing these metabolites by de novo biosynthesis. Supporting this hypothesis was the recent discovery of a derivative of 18, la,2a-diacetoxyalbicanyl acetate (21), from an egg mass of C. luteomarginata™ This compound has never been found in skin or whole body extracts of C. luteomarginata, nor in any dietary organism of the nudibranch, and therefore it seems likely that it is being made in whole or at least 19 H furodysinin (15) dendrolasin (Dysidea atnblia) idiadione (Spongia idia) X=-NC (16) X=-NC X=-NC X=-NCS X=-NCS X=-NCS X=-NHCHO (17) X=-NHCHO X=NHCHO {Acanthella sp.) (Acanthella sp.) (Acanthella sp.) Figure 2.4. Metabolites of Southern California Cadlina luteomarginata (and their sponge source where known) in part by the nudibranch. Albicanyl acetate (18) has been shown to deter predation by fish. Therefore, the presence of 21 in the nudibranch's egg mass suggests a protective role for these compounds. Moreover, from an evolutionary perspective, it may be suggested that C. luteomarginata produces these compounds de novo in order to have complete control over the critical issue of protection of its egg masses. If C. luteomarginata were shown to be capable of de novo biosynthesis of secondary metabolites, it would be only the second nudibranch species cadlinaldehyde (19) luteone (20) marginatafuran (25) (Aplysilla sp.) 20-acetoxymarginatone glaciolide (Aplysilla glacialis) 9,11-dihydrogracillin A (Aplysilla sp.) R \ R 2 = 0 cadlinolide A (26) (A. glacialis) R 1 =-0 2 CC 3 H7, R 2 =P0H lutenolide R'=H, R 2=OAC tetrahydroaplysulfurin-1 R'=H, R2=CCOAC violacene X=-NC (16) X=-NHCHO (17) (Acanthella sp.) (Acanthella sp.) X=-NC X=NHCHO (Acanthene C) X=-NCS (Acanthella sp.) (Acanthella sp.) Figure 2.5. Metabolites of British Columbia Cadlina luteomarginata 21 known to both sequester and biosynthesize such compounds. In order to determine the biological origin and biosynthetic pathways of 18 and of the unique sesterterpenoids 19 and 20, and to further explore the ecological significance of the presence of all these compounds, a biosynthetic study was conducted. 2.2.1.2 A new metabolite of Cadlina luteomarginata 2.2.1.2.1 Collection of specimens and isolation of a novel seco-spongian metabolite The first step in the biosynthetic investigation of 18,19, and 20 involved the re-isolation of authentic samples that could serve as chromatography and NMR standards. During the course of the re-isolation of 18,19, and 20, a new terpenoid 22 was encountered. Thus, 34 specimens of C. luteomarginata were collected by hand using SCUBA in Barkley Sound, B.C., and extracted with methanol and methanol/dichloromethane (1:1). The solvents were removed in vacuo and the residue was partitioned between water and ethyl acetate. The ethyl acetate soluble materials were obtained as a brown, fragrant oil that was fractionated using silica flash column chromatography with a gradient elution system of ethyl acetate in hexanes. Fractions obtained from the silica column were further purified using reversed phase 22 silica HPLC. In addition to pure samples of albicanyl acetate (18), cadlinaldehyde (19), luteone (20), and other known compounds, the unknown seco-spongian 22 was isolated from the most polar fraction. i 2.2.1.2.2 Structure determination of seco-spongian 22 Seco-spongian 22 was isolated as a colourless oil that gave a parent ion in the HRDCIMS at m/z 422.26805 appropriate for a molecular formula of C24H38O6. The 1 3C NMR spectrum of 22 contained resonances for all 24 carbon atoms (Table 2.1) and the HMQC data indicated that all of the hydrogen atoms were bonded to carbon. Only two sites of unsaturation, represented by ester carbonyls (5 170.3 and 175.7), could be identified from the 1 3C NMR data, indicating that 22 was tetracyclic. Comparison of the *H and 1 3C NMR data obtained for the seco-spongian 22 (Table 2.1; Figures 2.6 and 2.7) with the literature data for the spongian derivative 2361 suggested that the two molecules were closely related, even though 23 was pentacyclic and 22 was tetracyclic. The HMQC, HMBC, and COSY data62 for 22 confirmed that the A, B, and C rings in 22 were identical to those in 23 and they identified the regions 25 Table 2.1. NMR data for seco-spongian 22 recorded in CDCI3 Carbon # 8 1 3 C a 8'H b COSYc HMBC b NOEc 1 39.0d HI a: 0.93 (m) Hip,2p H2p, 3P, 5, Me20 Hip: 1.66 (m) Hloc, 3p 2 18.5 H2a: 1.65 (m) H3a H2p: 1.43 (m) Hla 3 41.8 H3a: 1.16 (m) H2a, 3p, Hip, 2a, Mel9 Mel 8 H3p: 1.40 (m) Hip, 3a 4 32.6 - - H5,Mel8, Mel9 5 48.5 1.30 (m) H6a, 6p, Me 18, Me20 Me20 6 24.9 H6a: 1.80 (m) H5, 6p, 7 H5,9 H6p: 1.57 (m) H5, 6a, 7 7 73.6 5.28 (t, 7=5) H6a, 6p H6a, 14 8 52.6 - - H7, 9, 15a, 17 9 44.8 1.67 (m) Me20 H5, 13 10 38.2 - - H6P, 9, Me20 11 14.7 HI la: 1.65 (m) Hllp H9 Hllp: 1.40 (m) H l l a 12 19.0 H12a: 1.81 (m) H13 H14 H12P: 1.81 (m) H13 13 39.1d 2.71 (m) H12a/p, 14 H12o/p, 15a, H9, 14 15b 14 44.4 2.46 (m) H13, 15a, 15b H7, 12a/p, 17 H7, 13, 15b 15 66.1 HI5a: 3.64 (dd, 7=10, 2) H14, 15b H17 H15b: 3.80 (dd,7=10, 7) H14, 15a, 17 H14, 15a 16 175.7 - - Me23 17 105.2 5.16 (s) H15b H9, 15a, Me20, Me24 Me24 18 33.3 0.75 (s) Mel9 19 21.3e 0.77 (s) H3a H5,Mel8 20 14.7 0.87 (s) H5 H5 21 170.3 - - Me22 22 21.4e 2.06 (s) 23 51.6 3.62 (s) 24 55.0 3.30 (s) H17 H17 "recorded at 125 M H z Recorded at 500 M H z "recorded at 400 M H z Resonances may be interchanged Resonances may be interchanged 26 of difference in the two molecules. HMBC correlations between a methyl resonance at 8 2.06 (Me-22) and a carbonyl resonance at 8 170.3 (C-21), and between a deshielded methyl resonance at 8 3.62 (Me-23) and a carbonyl resonance at 8 175.7 (C-16), identified acetate and methyl ester functional groups in 22. Additional HMBC correlations (Figure 2.8) between a second deshielded methyl resonance Figure 2.8. Selected HMBC (left) and COSY (right) correlations of seco-spongian 22. at 8 3.30 (Me-24) and a methine carbon resonance at 8 105.2 (C-17), and between the H-9 (8 1.67) resonance and the same methine carbon resonance (8 105.2: C-17) identified a methyl acetal function and tentatively located it at C-17. The acetal methine proton resonance (8 5.16: H-17) showed an HMBC correlation to an oxygen-bearing methylene carbon resonance at 8 66.1 assigned to C-15, which identified the second alkoxy component of the acetal function. HMQC data showed that C-15 (8 66.1) was attached to a pair of protons that gave resonances at 8 3.64 (H-15a) and 3.80 (H-15b). In the COSY spectrum, the H-15 resonances gave correlations to a methine resonance at 8 2.46, assigned to H-14 (HMQC to 8 44.4), which was in turn correlated to a methine resonance at 8 2.71, assigned to H-13 (HMQC to 8 39.1), thereby identifying an -O-CH2-CH-CH- fragment. The methine resonance assigned to H-13 (8 2.71) and the carbon to 27 which it was attached (C-13: 8 39.1) had chemical shifts nearly identical to those assigned to H-13 (8 2.74) and C-13 (8 37.6) in 23, suggesting that the methyl ester function in 22 was attached toC-13. A series of difference NOE experiments established the relative stereochemistry in the seco-spongian 22 (Figure 2.9). Irradiation of H-7 (8 5.28) induced an NOE in H-14 (8 2.46) indicating that H-14 was a and H-7 was P; irradiation of H-17 (8 5.16) induced an NOE in Me-20 (8 0.87) demonstrating that C-17 was axial and that H-17 was (3; and irradiation of H-13 (8 2.71) induced an NOE in H-9 (8 1.67) demonstrating that ring C was in a boat conformation with H-13 a. The downfield shift of H-7 (8 5.28) in 22 relative to H-7 in 23 (8 4.75) was consistent with the expected deshielding influence of the a-oriented C-17 methoxy substituent in 22 relative to the influence on H-7 of the P-oriented C-17 hydroxyl substituent in 23. Thus, these results defined the relative stereochemistry of the seco-spongian 22 as depicted in Figure 2.9. Figure 2.9. Selected NOE resonances of seco-spongian 22 (arrow heads indicate direction of observed NOE effect) 28 2.2.1.3 Results of biosynthetic experiments with Cadlina luteomarginata 2.2.1.3.1 Collection, incubation of precursors and isolation of metabolites Incorporation experiments utilizing stable isotope-labelled precursors with C. luteomarginata followed the general protocol used in recent investigations of terpenoid48'49 and polyketide50 biosynthesis in dorid nudibranchs. Three isoprenoid precursors, [1,2- 13C2Jacetate, [l,2-13C2,180i]acetate, and [2-13C]mevalonate were used in the current study. Specimens of C. luteomarginata were collected by hand using SCUBA in Barkley Sound and Jervis Inlet, B.C. and transported to UBC in refrigerated seawater. The nudibranchs were maintained at 12 °C in an aquarium filled with seawater brought from the collection site that was changed every two days. In the first experiment, individual specimens of C. luteomarginata were given 100 -200 U.L injections of a 0.55 mol/L solution of [l,2-13C2]sodium acetate every second day for 15 days (seven injections total). The physical act of handling the nudibranchs for injections caused them to partially shed the terpenoids, as evidenced by the fruity odour emitted during handling that matched the odour of pure cadlinaldehyde (19) and luteone (20). Two days after the last injection, the C. luteomarginata specimens were carefully removed from the aquarium seawater and the intact animals were immediately immersed in methanol. Fractionation of the methanol extract as described in section 2.2.1.2 gave pure samples of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20). Figures 2.10, 2.11, and 2.12 show 'H and 1 3 C NMR spectra of 18,19, and 20, respectively. Table 2.2 gives a comparison of the yields of 18,19, and 20 obtained from the different experiments. It is interesting to note that in the July, 1996 incorporation experiment with Barkley Sound animals, the injection of labelled acetate led to a nearly ten-fold increase in the isolated yield of 18 M e l 7 (ppm) 12 16 1 13 10 7 f 14 15 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 (ppm) Figure 2.10. NMR spectra of albicanyl acetate (18) in CDC13 top: 'H spectrum recorded at 500 MHz bottom: l 3C spectrum recorded at 125 MHz 20 17 18 (ppm) 20 13 200 190 180 170 160 ISO 140 130 120 110 100 90 80 70 60 50 40 30 20 10 (ppm) Figure 2.11. NMR spectra of cadlinaldehyde (19) in CDCL, top: 'H spectrum recorded at 500 MHz bottom: 13C spectrum recorded at 125 MHz w (ppm) S 210 200 190 180 170 160 150 140 130 120 l i b 1 0 0 9 0 8 0 7 0 6 0 5 0 4 0 3 0 2 0 1 0 (ppm) Figure 2.12. NMR spectra of luteone (20) in CDCI top: 'H spectrum recorded at 500 MHz bottom: 1 3C spectrum recorded at 125 MHz 32 18 19 20 Table 2.2. Yields of terpenoids 18,19, 20 obtained from C. luteomarginata Collection site Date Experiment Yield of metabolite (mg) per 20 specimens 18 19 20 Jervis Inlet May 2, 1996 [13C2] acetate 12.2 3.2 1.8 Barkley Sound May 22, 1996 [13C]mevalonate 7.8 3.8 7.1 Barkley Sound July 14, 1996 control 9.0 1.8 4.1 Barkley Sound July 14, 1996 [13C2]acetate 78.4 2.3 4.1 Barkley Sound Sept 11, 1996 control 4.1 <1.0 1.6 Barkley Sound Sept 11, 1996 [13C2-180]acetate 34.7 <1.0 2.4 (9 mg/20 animals for the control vs. 78 mg/20 animals for the incorporation experiment). It appears possible that either the availability of abnormally large quantities of acetate or the unnatural living conditions for the animals throughout the experiment prompted this increase in concentration of 18. 2.2.1.3.2 Biosynthesis of albicanyl acetate (18) 2.2.1.3.2.1 Terpenoid skeleton and acetate functional group The 1 3C NMR spectrum obtained for albicanyl acetate (18) isolated from animals injected with [ 1,2-l3C2] sodium acetate showed clear evidence for the de novo biosynthesis of 18 by C. luteomarginata. Figure 2.13 shows selected resonances from the C NMR spectrum of 18 33 obtained from the incorporation experiment along with the corresponding resonances from an unlabelled control sample.63 The weak doublets flanking the strong natural abundance central singlet in the C-2 (5 19.0), C-3 (41.8), C - l l (61.4), C-12 (107.0), C-16 (171.0), and C-17 (20.9) resonances resulted from incorporation of intact [1,2- 1 3C2] acetate units into 18 i i I i i i i 19.0 C2 i i i i i I i i i i 19.0 C3 I I I I I I 41.8 i i I I i | i i i i 41.8 i i i i I i i 61.4 Cl I I I I I I I I I I 61.4 C12 I i u 111111 107.0 i i i i i | i i i i 107.0 C16 I I I I I | I I I I 171.0 1 3 , 111 I I 11 i I I 171.0 C17 1111111111 20.9 i i i I I | i I I i 20.9 Figure 2.13. C NMR resonances of selected carbons of albicanyl acetate (18) labelled with [1,2-13C2]acetate (left) and unlabelled (right). All expansions in this and other figures are normalized to a common central peak height and then truncated for ease of viewing, and span a 1.00 ppm range. 34 Table 2.3. 1 3C NMR incorporation data for albicanyl acetate (18) labelled with [1,2- 13C2]acetate recorded in CDCI3 Carbon # 5 1 3 C a /(c-o (Hz) % specific incorporation 1 38.9 - -2 19.0 33.4 0.23 3 41.8 33.4 0.24 4 33.3 38.2 b 5 55.0 33.4 0.15 6 23.8 33.8 0.28 7 37.5 - -8 146.6 72.5 0.16 9 54.6 40.1 0.12 10 38.8 38.1 b 11 61.4 42.0 0.20 12 107.0 72.5 0.17 13 33.5 - -14 21.6 36.2 0.18 15 15.0 37.2 0.16 16 171.0 60.1 0.81 17 20.9 59.1 0.83 'recorded at 125 M H z incorporation too low to measure Analysis of the coupling constants for all the resonances in the 1 3 C NMR spectrum (Table 2.3; Figure A.5 (Appendix)) confirmed that the pattern of acetate incorporation was identical to that predicted by the biogenetic pathway shown in Scheme 2.1. The data also indicated that there was a significant difference in the efficiency of acetate incorporation into the terpenoid skeleton (average specific incorporation 0.19 %) and into the acetyl group (specific incorporation 0.82 %). This was calculated from the integrated singlets and flanking doublets of each carbon in the 1 3C NMR spectrum of 18, according to the equation, %si= 1.1 % X {(peak area of flanking doublets) - (n X 0.011 X peak area of central singlet)} / {(peak area of central singlet) + (n X 0.011 X peak area of central singlet)} 35 18 Scheme 2.1. Proposed biogenesis of albicanyl acetate (18) from [1,2- 13C2Jacetate (heavy lines indicate incorporation of intact acetate units; dots indicate incorporation of single carbon of acetate) where si is the specific incorporation at any given carbon and n is the number of carbons bonded to that carbon. This analysis provides a means of comparing the size of doublets due to incorporation of labelled precursor with the singlet originating from natural abundance 1 3 C labelling present in any sample (1.1 %). The correction involving n is necessary to account for random occurrence of two adjacent 1 3 C labels in any sample (1.1 % of the natural abundance singlet multiplied by the number of adjacent carbons). In several replicates of this experiment it was found that the level of incorporation into the acetyl group was high, whereas incorporation into the terpenoid skeleton was low or undetectable using l 3 C NMR methodology. Interestingly, the levels of incorporation into the terpenoid skeleton appeared to be seasonal and the highest levels were observed in animals collected during egg laying season (July). Overall, the results clearly show that albicanyl acetate (18) is biosynthesized by the standard isoprenoid biogenetic pathway from mevalonate. 36 2.2.1.3.2.2 O n the origin of the oxygen atoms A central issue in a concurrent investigation of the biosynthesis of sesquiterpenoids (e.g. nanaimoal (24)) by the dorid nudibranch Acanthodoris nanaimoensis (see Section 2.2.2), was the origin of the oxygen atoms present in the aldehyde functional group. One possibility was that the alcohol oxygen atom in farnesyl pyrophosphate was retained in the aldehyde functional groups of these three secondary metabolites. In order to test this possibility, it seemed reasonable to feed [ 1,2-13C2,18Oi] acetate and then look for the 1 8 0 isotope-induced shift in the doublets arising from intact incorporation of acetate units. Albicanyl acetate (18) biosynthesis by C. luteomarginata appeared to offer an ideal model to test the feasibility of this approach since the farnesyl pyrophosphate alcohol oxygen atom could reasonably be retained at C-11 in this molecule according to Scheme 2.2. Two distinct mechanisms of nucleophilic substitution could lead to O labelling of the oxygen atoms of 18: (1) hydrolysis of the pyrophosphate to the alcohol64 followed by nucleophilic attack by the alcohol oxygen at the carbonyl of acetate, or (2) nucleophilic attack by acetate onto the sesquiterpenoid at the methylene carbon adjacent to the 37 Scheme 2.2. Proposed biogenesis of albicanyl acetate (18) from [ 1,2-13C2,180i]acetate via two possible mechanisms pyrophosphate group, with departure of pyrophosphate. Using [1,2- C 2, Oi]acetate incorporation, the two possibilities were expected be differentiated by examination of the C-l 1 13 C NMR resonance of 18. Retention of the farnesyl pyrophosphate oxygen at C-l 1 (mechanism (1) in Scheme 2.2) would be observed as a 1 3 C- 1 3 C doublet shifted upfield due to 1 8 0 incorporation, relative to the normal 1 3 C- 1 3 C doublet. Observation of a single 1 3 C- 1 3 C doublet 38 would indicate that no 1 8 0 was retained, represented by mechanism (2) in Scheme 2.2. Thus, [l,2-13C2,18Oi]sodium acetate was injected into specimens of C. luteomarginata following the protocol described for the [1,2- 1 3C2] acetate incorporation experiment. The [1,2-13C2,18Oi]sodium acetate was prepared by reaction of [l,2-13C2Jacetyl chloride with H2 1 8 0 (>99 % 1 80) to give [ 1,2-13C2,180i]acetate that was -50 % labelled with 1 8 0. Figure 2.14 (B) shows the acetate carbonyl (C-16) and the alkoxy (C-l 1) resonances in the 1 3C NMR spectrum of albicanyl acetate (18) obtained from the [ 1,2- 1 3C 2 , 1 80i] acetate incorporation experiment. The carbonyl resonance at 8 171.0 (bottom right) has two sets of essentially equal intensity doublets flanking the central natural abundance singlet. Comparison with the carbonyl resonance in the spectrum of 18 obtained from the [1,2- 1 3C2] acetate incorporation experiment (Figure 2.14, top right) shows that in the 1 8 0 labelling experiment, 50% of the 1 3 C doublet was shift upfield. The observed isotope shift in the 8 171.0 resonance was 0.03 ppm which is within the reported range of 0.030-0.55 ppm for 1 8 0 labelled carbonyl carbons.65 It is interesting to note that the 1:1 ratio of the two doublets demonstrates that there was virtually no exchange of the acetate oxygen atoms during the course of the incorporation experiment. Thus, it was reasonably concluded that the two sets of doublets at C-16 in the [1,2-1 3 1 8 1 ^ 1 ^ 1 8 C2, Oi]acetate incorporation experiment corresponded to C- C- O incorporation (upfield doublet), and 1 3 C- 1 3 C- 1 6 0 incorporation (downfield doublet), with the oxygen label referring to the carbonyl oxygen. However, it was possible that the labelling pattern observed at C-16 might have corresponded to 13C-13C-180(carbonyl) incorporation (upfield doublet), and 1 3 C- 1 3 C-18 O(alkoxy) incorporation (downfield doublet), resulting from incorporation of whole acetate, not acetyl, at C-16 and C-17 (Scheme 2.2 mechanism (2)). The different isotope shifts derived from carbon-oxygen double and single bonds could have accounted for the 0.03 ppm difference in 39 CM C16 A: B: Figure 2.14. 1 3C NMR resonances of selected carbons of albicanyl acetate (18) labelled with [1,2-I3C2]acetate (top; A), labelled with [l,2-13C2,180i]acetate (bottom; B). V i—i—i—i—i—i—i—i—r 61.4 C l l I 1 I I I I I I I i 61.4 i — i — i — i — i — i — i — i — r 170.0 C16 i—i—i—i—i—i—i—i—i—r 170.0 A + B c 6 u V { I M I M I M I 170.0 their chemical shifts. In order to confidently reject this possibility, small amounts of 18 isolated from this experiment and the one incorporating [1,2-13C2] acetate were combined, and a 1 3C NMR spectrum was run. Because the latter experiment involved no 1 8 0 incorporation, the doublet observed at C-16 for that portion of 18 material had to originate from 1 3 C- 1 3 C- 1 6 0 labelling. Thus, it was expected that the combination of the two samples would either yield a 1 3C NMR resonance for C-16 involving three doublets, if whole acetate was incorporated at C-16 and C-17, corresponding 13C-13C-180(carbonyl), 13C-13C-180(alkoxy), and 1 3 C- 1 3 C- 1 6 0 incorporation, or two doublets, if acetyl was incorporated, corresponding to 13C-13C-180(carbonyl) and 1 3C- 1 3C-1 60, the latter doublet possessing enhanced intensity from the combination of two samples. This second situation was the one observed in the 1 3C NMR spectrum, as shown in the box in Figure 2.14, confirming that acetyl, not acetate, was incorporated at C-16 and C-17 of 18, and therefore the alkoxy oxygen could yet be derived from farnesyl pyrophosphate. Examination of the C-11 resonance (8 61.4) in the sample of 18 obtained from the [1,2-40 13C2,180i]acetate incorporation experiment (Figure 2.14 bottom left) revealed the presence of two doublets flanking the central natural abundance singlet, confirming that the farnesyl pyrophosphate alcohol oxygen atom was retained in the biosynthesis. Once again, the observed isotope shift of 0.025 ppm in the C-11 resonance was within the reported range of 0.010-0.035 ppm for C-O carbons.65 In the case of C-11, the upfield flanking doublet resulting from molecules that had the alkoxy oxygen labelled with O was less intense than the downfield doublet arising from molecules with 1 6 0 at the alkoxy oxygen. This difference in intensity indicates that there was some limited exchange of the oxygen atoms during the processing of acetate in the terpenoid pathway. Thus, it was concluded that the oxygen atom of farnesyl pyrophosphate was partially retained in the natural product albicanyl acetate (18). 2.2.1.3.3 Cadlinaldehyde (19) and luteone (20) sesterterpenoid skeletons An objective of the biosynthetic study was to verify that the unique cadlinalane and luteane carbon skeletons found in cadlinaldehyde (19) and luteone (20), respectively, arose from degradation of a sesterterpenoid precursor as proposed in Scheme 2.3. Several secondary metabolites of the general structural type of 19 and 20 have been isolated from B.C. specimens of C. luteomarginata (Figure 2.5), including marginatafuran (25) and cadlinolide A (26). These all possess three of four fused rings, along with one or two carbonyl functions, oxygens present in other functional groups, and carbon-carbon double bonds. However, none isolated to date other than 19 and 20 has been proposed to be derived from a sesterterpenoid biosynthetic precursor. Figures 2.15 and 2.16 show l 3 C resonances from the NMR spectra of cadlinaldehyde (19) and luteone (20) obtained from the [1,2- 1 3Ci] acetate incorporation experiment. Three resonances at 8 16.6 (C-19), 20.8 (C-18 and 21), and 206.4 (C-20) in the 1 3C NMR spectrum of 19 41 Figure 2.15. I 3C NMR resonances from the spectrum of cadlinaldehyde (19) (CDC13, 125 MHz) incorporated with [l,2-13C2]acetate 43 19 20 Scheme 2.3. Proposed biogenesis of cadlinaldehyde (19) and luteone (20) from [1,2- 13C2]acetate H ,0 (Figure 2.15) showed relatively intense doublets flanking a central natural abundance singlet indicating incorporation of intact acetate units into the molecule. A number of other resonances showed relatively weak doublets (e.g. 8 31.9 (C-17) and 64.9 (C-16)) or more complex multiplets (e.g. 8 22.9 (C-15) and 41.7 (C-3)) flanking the central singlet that were also indications of 44 incorporation of l 3 C labelled acetate units. The low signal to noise ratio in the remaining peaks 13 in the spectrum made it impossible to detect conclusive evidence for incorporation of C labelled acetate into the carbons assigned to these resonances. Scheme 2.3 shows that the methyl appendages C-18, C-19, C-20, and C-21 should all have been part of intact acetate units if 19 was formed via mevalonate-derived isoprenoid biosynthesis. Indeed, the clear flanking doublets in the C-19 (8 16.6) and C-20 (206.4) resonances were consistent with incorporation of intact doubly labelled acetate units at C-8/C-19 and C-10/C-20 in 19. The C-18 and C-21 methyl carbon resonances in the spectrum of 19 coincidentally had identical chemical shifts of 8 20.8 making it impossible to be absolutely certain that intact acetate units were incorporated at both these carbons. However, the flanking doublet in the 8 20.8 resonance had essentially the same intensity relative to the central singlet as did the flanking doublets in the C-19 and C-20 resonances, consistent with both C-18 and C-21 being derived from intact acetate units as expected. The biogenetic proposal in Scheme 2.3 predicts that C-16 and C-17 in 19 were both derived from C-2 of cleaved acetate units. Thus, if during the incorporation of acetate into 19 there was high dilution of labelled precursor, the resonances for C-16 and C-17 should have appeared only as enriched singlets. However, inspection of the resonances for C-16 and C-17 in Figure 2.15 shows that there were weak doublets flanking the central singlets. The most likely explanation for this is that they originated from incorporation of more than one 1 3 C labelled acetate into a single molecule of 19. C. luteomarginata specimens were starved during the 16 day injection period and this could have led to a highly labelled acetate pool and a reasonable probability of more than one labelled acetate unit being incorporated into individual mevalonic acid molecules and/or more than one labelled mevalonic acid unit being incorporated into individual molecules of 19. As a result, the C-16 doublet shown in Figure 2.15 would arise from 45 molecules of 19 having a singly 1 3C enriched carbon at C-16 and a labelled intact acetate unit at C-14/C-15. Similarly, the C-17 weak doublets would result from having a singly labelled carbon at C-17 and a labelled intact acetate unit at C-4/C-18. Figure 2.16 shows resonances from the 1 3C NMR spectrum of luteone (20) isolated from the [1,2-13C2]acetate incorporation experiment. Once again, there were a series of resonances which had relatively intense doublets flanking the central singlet (8 16.2 (C-21), 20.7 (C-20), 206.3 (C-22), and 106.9 (C-23)), one that showed a relatively weak doublet flanking the central singlet (8 31.9 (C-l9)), a group that had complex multiplets flanking the central singlet (e.g. 8 30.1 (C-l8), 41.7 (C-3), and 23.5 (C-l 1)), and several in which the signal to noise ratio was so low that it was impossible to detect conclusive evidence for incorporation of labelled acetate. The biogenetic proposal presented in Scheme 2.3 required that the C-20, C-21, C-22, and C-23 methyl appendages in 20 all be formed from intact acetate units. This was supported by the observation of relatively intense flanking doublets in the resonances assigned to these carbons in the labelled sample of 20. The average apparent specific incorporation at these four sites was 3.6 % in luteone (20), roughly four times the apparent specific incorporation observed for the methyl appendage carbons in cadlinaldehyde (19), which was 0.86 %. It has been stated that biosynthetic experiments with marine invertebrates suffer from at least two difficulties: slow biosynthesis of each metabolite molecule and slow turnover of pools of metabolite, resulting in large amounts of unlabelled material against which incorporation must be assessed.32 Using the [1,2- 13C2]acetate incorporation data for luteone (20) it was possible to get a quantitative measurement of both the percentage of molecules that were formed during the feeding experiment and the extent to which these molecules incorporated the labelled precursor. The basis for this calculation was the fortuitous observation of a well-resolved multiplet structure in the C-l8 resonance of 20 (Figure 2.17). 46 By numbering the component lines in the C-18 resonance (Figure 2.17) from 1 to 7 starting with the highest frequency component, it was possible to assign the individual peaks to the various forms of labelled luteone (20) present in the sample obtained from the [1,2-1 3C2] acetate incorporation experiment. The central component (line #4) in the multiplet was assigned to a natural abundance 1 3 C singlet arising from three possible sources of 20: (1) unlabelled luteone molecules that were present in C. luteomarginata at the outset of the experiment, (2) molecules formed during the experiment without the use of [l,2-13C2]acetate, and Figure 2.17. Expanded 1 3 C NMR resonance of C-18 from the spectrum of luteone (20) incorporated with [1,2- l 3C 2] acetate (3) molecules that incorporated [l,2-I3C2]acetate during the experiment but not at positions C-16, C-17, or C-18. Component lines # 2 and # 6 were assigned to a doublet (7 = 41) arising from molecules that incorporated an intact [1,2- 1 3C 2] acetate unit at C-17 and C-18 but had a 1 2C atom at C-16. The remaining component lines (#1,3, 5,7) were assigned to a doublet of doublets (7 = 41, 14) arising from molecules that incorporated an intact [1,2- 1 3C 2] acetate unit at C-17/C-18 and a single 1 3 C label from C-2 of a [1,2- 13C2]acetate unit at C-16. This latter assignment is in good agreement with the observation that geminal 1 3 C- 1 3 C couplings are typically -15 Hz when the central carbon is a ketone as is the case in the C-16/C-18 coupling of the triply labelled 30.1 47 luteone (20) molecules.66 In order to calculate the percentage of molecules formed during the incorporation experiment and the extent to which these molecules incorporated the labelled precursor, it was assumed that the nudibranchs had a sizable pool of unlabelled metabolites at the outset of the experiment and only a small pool of new molecules were made in the presence of the isotopically labelled precursor. Therefore, the isolated products at the end of the experiment were really made up of two significant components: (1) molecules which were present at the outset and are completely unlabelled, and, (2) labelled molecules that were made during the experiment. Consequently, the "apparent specific incorporation" of 1 3C label at any particular site was a weighted average of the incorporation in the original molecules (zero) and the specific incorporation in the labelled molecules. This can be expressed as follows if it is assumed that there is uniform incorporation of acetate at all positions in 20: sa = si (x) + 0 (1 -x) where sa is the apparent specific incorporation observed in the whole sample (overall degree of labelling), si is the specific incorporation in the newly formed metabolites that were made during the labelling experiment (the degree of labelling among the molecules that have at least one label), and x is the fraction of molecules made during the labelling experiment (the fraction of molecules showing any labelling at all). In the [1,2-13C2] acetate incorporation experiment the total intensity of the doublet and the doublet of doublet components present in the C-l8 resonance of 20 was used to directly calculate sa at C-l8 as follows: sa = % enrichment above natural abundance = 1.1 % X (combined integrated peak area of enriched satellites: lines # 1, 2, 3, 5, 6, 7) / 48 (peak area of the natural abundance singlet: line # 4) Equation 1. A correction for the random chance of two adjacent 1 3C atoms by natural abundance was also included analogous to the calculation for specific incorporation in albicanyl acetate (18). Using the same argument presented above, the apparent specific incorporation of an intact [1,2- 1 3C2J acetate unit at C-17/C-18 as well as a singlet acetate label at C-16 in a singlet molecule of 20 was calculated as follows: sa,2 = si (x) + 0 (1-JC) where sa2 is the apparent specific incorporation for the doublet of doublets, si2 represents the probability of having two adjacent 1 3 C labelled carbons arising from different labelled acetate units in a singlet molecule of 20 formed during the labelling experiment, x represents the fraction of labelled molecules as earlier, and (\-x) represents the fraction of unlabelled molecules present at the outset. The calculation of sa.2 would also use the equation: sa2 = % enrichment above natural abundance = 1.1 % X (combined peak area of doublet of doublet components: lines # 1, 3, 5, 7) / (peak area of the natural abundance singlet: line # 4) Equation 2. The measured relative intensities of the component lines in the C-18 multiplet were as follows: line # 1: 0.38; line # 2: 0.68; line # 3: 0.48; line # 4: 1.0; line # 5: 0.52; line # 6: 0.65; line # 7: 0.58. Using the values for sa and sa2 obtained from the component line relative intensity measurements, it was possible to solve Equation 1 and 2 for si and x. The calculated values were: si = 0.60 and x = 0.060. Therefore, the specific incorporation within the group of labelled luteone (20) molecules was actually 60 %, but only 6.0 % of the molecules in the entire 49 pool showed any labelling at all. The overall apparent specific incorporation (sa) was 60 % X 6.0 %, or 3.6 %. These numbers are completely consistent with the argument that the acetate pool in the nudibranchs during the injection experiment was highly labelled, and therefore, molecules made from this pool should be highly labelled. The si and x values are also consistent with the expectation that only a relatively small percentage of the metabolite pool was turned over in this experiment. The [1,2-13C2] acetate incorporation experiment provided conclusive evidence for the de novo biosynthesis of 19 and 20 by C. luteomarginata. However, the complexity of the multiplet patterns in the 1 3 C NMR spectra of labelled 19 and 20 made it difficult to use coupling constant information to establish the complete pattern of acetate incorporation in either molecule. Scheme 2.3 proposes a degraded sesterterpenoid biogenetic origin for both cadlinaldehyde (19) and luteone (20). In order to test this hypothesis, a second incorporation experiment was conducted utilizing [2-13C]mevalonolactone (a stabilized derivative of mevalonic acid) as a precursor and the same methodology of precursor administration, incubation, and metabolite isolation as earlier. Figure 2.18 shows the 1 3C NMR spectrum of cadlinaldehyde (19) isolated from the [2-13 C]mevalonolactone incorporation experiment. Five resonances in the spectrum, assigned to C-1 (5 34.3), C-l (40.4), C-12 (39.2), C-16 (64.9), and C-17 (31.9) show clear enrichment above natural abundance (Table 2.4). This labelling pattern is in complete agreement with the biogenetic proposal that 19 is a degraded sesterterpenoid as shown in Scheme 2.4 which depicts the assembly of five molecules of mevalonate to form a 25-carbon precursor. Of particular significance was the enrichment of the C-16 resonance, which clearly demonstrated that this carbon originated from C-2 of mevalonic acid, ruling out any alternate type of homologation of a diterpenoid to give the 21-carbon cadlinalane skeleton. From comparing the integrated peak 50 51 Table 2.4. Specific incorporation of [2- C]mevalonolactone into cadlinaldehyde (19) and luteone (20) cadlinaldehyde (19) luteone (20) Carbon # 8 1 3 C a % specific incorporation 5 1 3 C a % specific incorporation 1 34.3 6.6 34.4 14 2 19.3 19.4 3 41.7 41.7 4 33.6 33.6 5 55.6 55.1 6 19.9 18.4 7 40.4 6.3 39.5 15 8 36.5 39.8 9 61.4 60.2 10 53.4 53.8 11 17.4 23.5 12 39.2 6.5 37.7 15 13 79.5 147.1 14 59.5 55.2 15 22.9 17.6 16 64.9 5.1 42.6 18 17 31.9 5.8 209.2 18 20.8 30.1 19 16.6 31.9 13 20 206.4 20.7 21 20.8 16.2 22 - 206.3 23 - 106.9 "recorded in CDC1 3 at 125 M H z areas of the labelled 19 with a control sample and averaging over the five sites of incorporation, the specific incorporation of [2-l3C]mevalonolactone in 19 from this experiment was found to be 6.1 %. The 1 3C NMR spectrum of luteone (20) (Figure 2.19) obtained from the [2-13 CJmevalonolactone incorporation experiment also showed clear enrichment above natural abundance at C-l (5 34.4), C-7 (39.5), C-12 (37.7), C-16 (42.6), and C-19 (31.9) (Table 2.4), once again in complete agreement with the proposed biogenesis of the luteane skeleton from a 1 52 53 19 20 Scheme 2.4. Proposed biogenesis of cadlinaldehyde (19) and luteone (20) from [2-13C]mevalonolactone (dots indicate positions of 1 3C labels) sesterterpenoid precursor as shown in Scheme 2.4. In this case the specific incorporation in 20 was calculated to be 15 %, significantly greater than in 19. 2.2.1.3.4 Conclusions of biosynthetic study The results presented above have clearly demonstrated that albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) are biosynthesized de novo by Cadlina luteomarginata as predicted by the geographic invariance of their occurrence in British Columbian specimens of the nudibranch. In all the experiments that were attempted, there was efficient incorporation of either labelled acetate or mevalonate into 19 and 20, but a number of experiments failed to show incorporation of labelled precursors into the terpenoid portion of albicanyl acetate (18). The 54 highest levels of incorporation of labelled acetate into the drimane portion of 18 were obtained with animals that were collected during egg laying season suggesting some reproductive cycle control on the biosynthesis of this compound. This observation is consistent with the recent discovery of la,2oc-diacetoxyalbicanyl acetate (21), a diacetoxy derivative of 18, in egg masses of C. luteomarginata, where it is presumed to be playing a defensive role.60 The different incorporation rates of acetate and mevalonate into the sesquiterpenoid 18 compared to the sesterterpenoids 19 and 20 suggest that 18 is produced in a different metabolic cycle than 19 and 20. The mevalonate incorporation results support a sesterterpenoid origin for the unique degraded cadlinalane and luteane carbon skeletons as proposed in Schemes 2.3 and 2.4. This study represents the first demonstration of de novo sesterterpenoid biosynthesis in marine molluscs, which have previously been shown to be capable of sesquiterpenoid and diterpenoid biosynthesis.43'48'49' 5 0 There is only one previous report of successful incorporation of a stable isotope labelled mevalonic acid equivalent into a marine invertebrate terpenoid in an incorporation study.48 In that example, the levels of incorporation were low enough that the labelling pattern could only be discerned by statistical analysis of 1 3C NMR peak heights in the labelled sample. The levels of incorporation of [2-13C]mevalonolactone into cadlinaldehyde (19) and luteone (20) in the current study with C. luteomarginata were much higher (6.1 % and 15 %, respectively), allowing for visual confirmation of the labelling pattern. The higher incorporation rate in 20 than 19 observed from both the mevalonate and acetate incorporation experiments suggests that the biosynthesis of 20 is favoured over that of 19 if they indeed originate from a common precursor as predicted in Schemes 2.3 and 2.4. Quantitative analysis of the [1,2- 1 3C2] acetate incorporation into 20 showed that only 6.0 % of the isolated molecules were made during the incorporation experiment with labelled 55 precursors, but that this small pool of 20 was exceptionally highly labelled (60 %). This result suggests that it is probably not the effectiveness of incorporation of precursors into biosynthetic products that is the limiting factor in the use of stable isotopes to investigate the biosynthesis of marine invertebrate terpenoids, but rather the large pool of unlabelled compound against which incorporation must be assessed. C. luteomarginata is only the second marine mollusc that is known to be capable of both sequestering skin extract terpenoids from dietary sources and producing them via de novo biosynthesis. The other known example is the Mediterranean dorid nudibranch Dendrodoris grandiflora which also makes drimane sesquiterpenoids that it deploys on its dorsum and in its egg masses much like C. luteomarginata!^ The previously reported54 absence of 18,19, and 20 from California specimens of C. luteomarginata put into question the hypothesis of geographic variation, because these compounds were indeed being biosynthesized de novo by B.C. specimens of the same species. Were these compounds present in California specimens after all, having been missed in previous studies? Or was it possible that populations from the two habitats were genetically distinct, only one being capable of de novo biosynthesis of these compounds? Do these two populations then represent two separate species or sub-species (even though classical taxonomic analysis52 based on morphology indicates that there is only a single species all along the western coast of North America)? Or could the biosynthetic pathways in question remain dormant if they were not needed, resulting in some populations containing different biosynthesized compounds than others? Could these pathways be controlled by the dietary availability of other secondary metabolites, for instance, stopping 'domestic' production of secondary metabolites if an abundant 'foreign' source was available? Could these pathways then be activated by a change in diet or habitat? 56 It was deemed necessary to re-visit the chemical composition of Californian C. luteomarginata in order to explore further the nature of secondary metabolism in marine molluscs, specifically biosynthetic capability and regulation. This work is presented in the next section. 57 2.2.1 A Geographic and anatomical distribution of metabolites in C. luteomarginata 2.2.1.4.1 Collection, dissection, and method of analysis Three investigations were undertaken: (1) a comparison of metabolite content between specimens of two C. luteomarginata populations, one in British Columbia and one in California, (2) a comparison of metabolite content between individuals of one population in British Columbia, and (3) a comparison of metabolite content between tissues of individual B.C. specimens. Each involved a common collection and analysis methodology. A collaboration with Professor D. J. Faulkner at Scripps Institution of Oceanography (SIO) in La Jolla, California, facilitated a field expedition aimed at collecting specimens of C. luteomarginata native to Southern California. As stated above, it was deemed possible that the terpenoids found to be biosynthesized de novo by C. luteomarginata in B.C. might also have been present in the California specimens but missed in earlier studies. Alternatively, it seemed possible that if these compounds were indeed not present in the California nudibranchs (while still being members of the same species as their B.C. counterparts), that their production by de novo biosynthesis might be triggered by the administration of suitable biosynthetic precursors. The scarcity of nutrients in oceanic environments may limit non-essential metabolism. Therefore, it was imagined that an unusual abundance of biosynthetic precursors may permit the re-emergence of dormant pathways. Collection by hand using SCUBA resulted in the acquisition of 18 individual nudibranchs which were transferred to aquaria at SIO. Of the 18 collected, nine were immersed in methanol soon after collection in order to analyze their extract for the presence of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20). The other nine were subjected to injections with either of two known biosynthetic precursors to C. luteomarginata terpenoids, sodium acetate or 58 mevalonolactone. Because C labelled mevalonolactone was very costly, unlabelled mevalonolactone was used, expecting that "triggering" of biosynthesis could be detected by the presence of 18,19, and/or 20 in the extracts of these four nudibranchs, compared to the absence 13 of these compounds in extracts of untreated animals. Because doubly C labelled sodium acetate was readily available, it was chosen for injection into five specimens of C. luteomarginata, so that not only could the presence of compounds demonstrate de novo biosynthesis, but further confirmation would be possible by NMR detection of incorporated acetate, assuming the metabolites were isolated in sufficient quantities to perform 1 3C NMR spectroscopy. Each nudibranch subjected to injections of putative precursor was delivered 200 -300 pL injections of 0.55 mol/L acetate or mevalonolactone on four consecutive days, starting seven days after collection.69 On the day after the last injection, the specimens were immersed in methanol. Specimens and their extracts were returned to UBC for chemical analysis. For the study on variation of metabolite content within a population of C. luteomarginata nudibranchs, 19 specimens of C. luteomarginata were collected in Barkley Sound, British Columbia and immersed separately in methanol in individual vials. Ten specimens collected in Barkley Sound were dissected and their common organs combined in methanol to form extracts of different tissues of C. luteomarginata. Figure 2.20 shows some of the tissue types dissected from C. luteomarginata. In total, extracts were each made of: (1) mucus swabbings of mantles of nudibranchs, (2) mantle dermal formations (MDFs: yellow glands found on the dorsum that were excised with a scalpel), (3) yellow margins of mantles, (4) dorsal mantles minus MDFs and yellow margins, (5) gut (representing primarily the hepatopancreas but also smaller visceral organs) (6) feet, and (7) water in which the nudibranchs were dissected that may have contained shed metabolites. 59 M a r g i n \ Dorsal mantle • Mantle dermal formation 4> Foot Figure 2.20. Some dissected tissues of Cadlina luteomarginata 70 Extracts were each analyzed by the same chemical method. Because of the small number of specimens involved in each investigation, the isolation and characterization methods used in the earlier biosynthetic experiments were not chosen. Instead, a method of analysis using gas chromatography and mass spectrometry was devised for the identification and quantification of metabolites in each extract. The nudibranch(s) or nudibranch tissues were extracted in methanol for at least one week. The wet weight of animal material and the volume of extracting liquid were measured for each sample. Then, an aliquot of extract was removed and concentrated to a known volume. A smaller known volume of extract was then subjected to GC analysis using a column and temperature gradient shown to adequately separate constituents of a standardized mixture of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20). Four dilutions of 60 standard solutions of each of these three metabolites were made and used as calibration for the quantification of metabolites. GC peaks for each extract were compared to the calibration data and the concentration of metabolite was calculated as milligram metabolite per gram of nudibranch tissue. 2.2.1.4.2 Results of quantitative analysis of metabolites 2.2.1.4.2.1 Variation between California and B.C. specimens GC spectra were recorded for three extracts of La Jolla, California specimens of C. luteomarginata: (1) control specimens, (2) specimens injected with [1,2- 13C2jsodium acetate, and (3) specimens injected with mevalonolactone. These were compared with the GC spectrum corresponding to an extract of 14 specimens of C. luteomarginata collected in Barkley Sound, British Columbia. The GC peaks were calibrated from standards made up of pure samples of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20), and compound identification was confirmed by HREIMS. The results of this comparison study are found in Table 2.5. Table 2.5. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in California and British Columbia specimens of C. luteomarginata Origin of specimens {number of total wet weight Concentration (mg specimens] of specimens (g) metabolite / g nudibranch) 18 19 20 California control {9} 31.1 0.075 0.0022 ~0 California acetate-fed {5} 25.1 0.13 0.0066 0.01 l a California mevalonolactone-fed {4} 30.2 0.19 ~0 ~0 B.C. control {14} 5.90 0.15 0.063 0.096 anot confirmed by mass spectrometry 61 It was found that California specimens of C. luteomarginata did indeed contain the biosynthesized metabolites that had previously only been identified from B.C. specimens. The presence of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in California nudibranchs was demonstrated by three GC peaks that corresponded with GC peaks of the three pure compounds run separately as standards. The positive identification of 18 and 19 in the California extracts was further confirmed by mass spectrometry. The GC peak matching the retention time for 18 was found to possess a parent ion in the HREIMS at m/z 264.20859, corresponding to the molecular formula of C17H28O2, that of albicanyl acetate (18). The GC peak with the same retention time as 19 was found to give a [M-Me] ion (as pure cadlinaldehyde (19) often did) with m/z 303.23206, corresponding to a molecular formula of C20H31O2, representing the loss of a methyl group from cadlinaldehyde (19). Luteone (20) appeared to be present in sufficient quantity for GC detection in the California specimens that were injected with acetate, but not in the control or mevalonate-fed specimens. The presence of 20 in California specimens was not confirmed by the MS measurements taken (the peak not being strong enough for a mass measurement). The small quantities of 20 in California specimens of C. luteomarginata made absolute confirmation of this metabolite difficult. Because the biosynthesized metabolites were detected in the control sample specimens of Californian C. luteomarginata, it was irrelevant whether or not administration of biosynthetic precursors could trigger their biosynthesis. Indeed, comparison of the three sample sets of California specimens (control, acetate-fed, and mevalonate-fed) indicated no significant difference in metabolite concentration between these experiments. The concentration of albicanyl acetate (18) in California nudibranchs was sufficient to use separation methodology described in earlier sections to isolate 5.6 mg of 18 that was characterized by 'H and 1 3C NMR. 62 The 1 3C NMR spectrum had small doublets around the carbon singlets representing C-16 and C-17 indicating incorporation of labelled acetate into the acetyl function of 18. Other carbons did not display visible labelling, which was not surprising due to the short incubation period of that experiment (4 days). Comparison of metabolite concentrations between California and B.C. specimens of C. luteomarginata led to the realization that two of the three metabolites (cadlinaldehyde (19) and luteone (20)) were present in lower concentrations in California nudibranchs than in B.C. ones, while the third metabolite (albicanyl acetate (18)) was present in approximately the same concentration in both populations. Although similar relative differences were also observed amongst individuals of one population (see next section), the observed difference of approximately one order of magnitude in the concentrations of 19 and 20 (see Table 2.5) over an average of 18 (California) and 14 (B.C.) specimens represented a significant geographic variation. It was deemed possible that metabolite concentration might be controlled by necessity. Thus, in an environment where a metabolite is not needed, possibly due to an abundant dietary source of a similar metabolite, individuals possessing the genetic capability to biosynthesize metabolites might reduce their output in the interest of metabolic efficiency. 2.2.1.4.2.2 Variation between individuals of one B.C. population Nineteen specimens of C. luteomarginata were collected in Barkley Sound, B.C. and individually extracted and subjected to GC analysis. Table 2.6 shows the results of the quantification calculations on biosynthesized metabolites in individual nudibranchs from this experiment. 63 Table 2.6. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in individual B.C. specimens of C. luteomarginata Specimen # wet weight of Concentration (mg specimen (g) metabolite / g nudibranch) 18 19 20 1 8.67 0.19 0.14 0.23 2 9.61 0.012 0.048 0.012 3 0.52 0.021 0.032 0.44 4 1.28 0.20 0.16 0.27 5 2.64 0.014 0.033 0.042 6 3.53 0.011 ~0 0.11 7 1.15 0.14 0.19 0.19 8 0.64 0.018 0.024 ~0 9 3.59 0.023 0.021 0.0077 10 2.12 0.17 0.12 0.097 11 2.68 0.018 0.059 0.11 12 1.05 0.017 ~0 ~0 13 1.90 0.083 0.043 ~0 14 1.93 0.027 0.010 ~0 15 1.59 0.46 ~0 ~0 16 3.06 0.041 0.015 ~0 17 2.19 0.029 ~0 ~0 18 1.62 0.24 ~0 ~0 19 1.72 0.020 ~0 ~0 Figure 2.21 shows two representative GC traces of individual nudibranch extracts. Trace (a), belonging to individual # 4, gave intense GC peaks at retention times 7.13, 11.38, and 13.39 minutes, corresponding to albicanyl acetate (18), cadlinaldehyde (19), and luteone (20), respectively (the strong peak at 10.74 minutes was found in many samples and identified as an ester of phthalic acid). In this spectrum, GC signals corresponding to most other compounds were much less intense than of the three biosynthesized metabolites. In trace (b), representing individual # 5, the signals pertaining to 18,19, and 20 were each significantly smaller than in trace (a), and in fact smaller than peaks belonging to many other compounds. Sixteen of the 17 64 Figure 2.21. GC traces of extracts of C. luteomarginata individuals (a) # 4, and (b) # 5 traces belonging to the individual nudibranchs fell into one of these two types: (a) higher concentration of 18,19, and 20; lower concentration of other metabolites, or (b) lower concentration of 18,19, and 20; equal or higher concentration of other metabolites. The extracts of individuals # 1,4, 7, 9, 10, 15, 16, and 18 approximately resembled type (a) and those of individuals # 3, 5, 6, 8, 11, 12,13, 14, 17, and 19 resembled type (b). In order to draw conclusions as to what these results might signify, it was considered advantageous to know the identity of the other compounds present. Samples were submitted for GC-MS in an attempt to identify the other metabolites found in the individual extracts, by comparing parent ions of GC peaks with molecular masses of known metabolites60 of B.C. specimens of C. luteomarginata. However, the low concentrations of metabolites in these extracts meant that not all peaks could be subjected to mass spectrometric measurements, and none except albicanol (GC retention time 6.72 minutes; identification by comparison with authentic sample) corresponded with the 65 molecular weights of compounds known to have been isolated from this nudibranch. From the comparison of metabolite concentrations in individual nudibranchs (Table 2.6), it is clear that individual variation in metabolite content is very significant within populations of C. luteomarginata. In fact, the individual variation within this population is similar to the difference between the average concentrations in B.C. and California populations. While on average, B.C. specimens of C. luteomarginata contain greater quantities of biosynthesized metabolites than do California specimens, B.C. specimens either contain 18,19, and 20 in high concentrations with little else, or contain 18,19, and 20 in low concentrations (similar to the California specimens) along with many other undetermined constituents. As was postulated in the section above from the results of the two-population study, it is possible that concentration of biosynthesized metabolites is regulated within each individual and/or each population by the availability of other, dietary metabolites that serve the same ecological function. Determination of the identity of the other metabolites involved and the ecological roles of these compounds is necessary in order to further understanding in this area. 2.2.1.4.2.3 Anatomical distribution of biosynthesized metabolites It has been stated that defensive allomones are likely to be found in peripheral parts of nudibranchs (such as in the mantle or gills) from where they can be readily excreted.71 An earlier report of dissection of Californian C. luteomarginata indicated that the mantle contained most of the terpenoid constituents, at that time known to be isonitrile, isothiocyanate and furanoterpenoids.54 A study of anatomical distribution in Mediterranean nudibranchs showed that biosynthesized terpenoid metabolites were located in the border of the mantle, where there were specific glands called mantle dermal formations, MDFs, as well as in the hermaphrodite 66 gland and in the egg masses.46 Other reports have suggested that metabolites are commonly found in the MDFs and digestive glands (as would be expected for metabolites acquired by dietary sequestration).72 An earlier study with C. luteomarginata had already resulted in the discovery of a derivative of albicanyl acetate (18), loc,2oc-diacetoxyalbicanyl acetate (21) in an egg mass of this nudibranch,60 suggesting a defensive role for 21 and corroborating another previous study which found 18 to act as a fish antifeedant.55 Thus, ten specimens of C. luteomarginata collected in Barkley Sound, B.C., were subjected to dissection and GC analysis of tissue extracts in order to determine the distribution of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) within tissues of C. luteomarginata. Table 2.7 shows the results of the anatomical distribution analysis of C. luteomarginata. By considering the percentage of the total quantity of metabolite found in any given tissue along with the proportion of body weight that that tissue represented, it was possible Table 2.7. Quantification of albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) in dissected tissues of 10 specimens of C. luteomarginata Tissue body % body amount metabolite (mg) % metabolite content weight (g) weight 18 19 20 18 19 20 MDF - < 1 2.6 1.2 0.32 26 31 10 margin 1.4 3.8 1.6 0.32 0.46 16 8.1 15 remaining 18.6 51 3.9 0.86 1.1 38 22 35 mantle hepatopancreas 9.9 27 0.46 0.41 0.58 4.5 10 19 feet 6.7 18 0.61 0.60 0.27 6.0 15 8.7 mucus - < 1 0.67 0.24 0.12 6.6 6.0 4.0 swabbings dissection water - <1 0.39 0.29 0.29 3.8 7.4 9.2 total mantle 20.0 55 8.1 2.4 1.9 79 61 60 total 36.6 100 10.2 3.9 3.1 100 100 100 67 to conclude that the greatest concentration of all three biosynthesized metabolites was found in the mantle, including the MDFs and mantle margins. The combined mantle sections represented 79, 61, and 60 %, respectively, of the total amount of 18,19, and 20 found in C. luteomarginata. Specifically, the MDFs which accounted for less that 1 % of the total body weight, contained 26, 31 and 10 % of 18,19, and 20, respectively. Significant amounts of all three metabolites were found in the water in which the nudibranchs were dissected and in the mucus collected by swabbing prior to dissection, indicating that these metabolites were indeed shed through handling of the animals. Because the MDFs were difficult to locate and excise from the remaining mantle, it is believed that some of the metabolite content attributed to the remaining mantle was indeed derived from MDF tissues. These small organs are believed to be glands responsible for the excretion of defensive allomones in many nudibranchs,71 strengthening the hypothesis that the biosynthesized metabolites of C. luteomarginata fulfill some protective role to the nudibranch. At this time, it is unknown whether the MDFs are also the sites of biosynthesis of terpenoids in C. luteomarginata, or merely storage sites. Further biochemical investigation is needed to identify the location of enzymes responsible for de novo biosynthesis of sesquiterpenoid and sesterterpenoid metabolites. 2.2.1.4.3 Conclusions of distribution study The geographic and anatomical variation study revealed substantial variation in metabolite content between populations, between individuals of one population, and between tissues of Cadlina luteomarginata specimens. When comparing the average of several specimens, representatives of the Barkley Sound region of British Columbia, Canada were found 68 to possess greater concentrations of two out of three biosynthesized metabolites than did specimens from La Jolla, California, U.S.A. The much lower concentrations of cadlinaldehyde (19) and luteone (20) found in California nudibranchs account for their absence in previous reports. The discovery that 18,19, and 20 are indeed present in C. luteomarginata at both the northern and southern extremes of its geographic range reasserts a previous hypothesis that de novo biosynthesis is predicted by the constancy of biosynthetic metabolites across the entire range.51 Individual variation of biosynthesized metabolite content within the Barkley Sound population was found to be of similar magnitude to the average difference between B.C. and California populations. Two patterns of metabolite content emerged within the B.C. population: almost all individuals either possessed higher concentrations of 18,19, and 20 (along with lower concentrations of other constituents), or lower concentrations of 18,19, and 20 (along with equal or higher concentrations of other constituents). Some dietary metabolites of C. luteomarginata as well as albicanyl acetate (18) have been demonstrated to possess antifeedant activity.54,55 Thus, it is not unreasonable to propose that other metabolites including cadlinaldehyde (19) and luteone (20) share this ecological function. The presence of all biosynthesized metabolites in the exuded mucus of C. luteomarginata and in tissues on the exposed mantles support defensive ecological roles for these compounds. If one accepts this premise, it then follows that individual specimens of C. luteomarginata might regulate biosynthesis of defensive metabolites according to need, increasing biosynthetic output when dietary metabolites are less available. This would account for B.C. specimens having greater concentrations of 19 and 20 than Californian specimens, since the more northern sponge communities are believed to be poorer chemically. It would also account for the observed variation in metabolite content between individual specimens of B.C. C. 69 luteomarginata: individuals having recently fed upon chemically-rich sponges should contain lower concentrations of biosynthesized metabolites, while those without internal stocks of dietary metabolites should possess more biosynthesized metabolites. This hypothesis is supported by the structural similarities of the many biosynthesized and sequestered terpenoid metabolites of C. luteomarginata (Figures 2.4, 2.5), despite their different biogenetic origins. One possible rationale for the hypothesis above is that nudibranchs evolved the ability to biosynthesize compounds after they had achieved the ability to sequester dietary allomones. The metabolic processes they evolved replaced the sequestered compounds with biosynthesized compounds of similar structures that could fulfill the same ecological functions. The nudibranchs' ability to produce important metabolites by de novo biosynthesis increased their autonomy and therefore their evolutionary success by permitting them to expand into chemically-poor habitats from which they were previously excluded because of lack of dietary sources of defensive metabolites. If this hypothesis were true, then one would expect negligeable biosynthetic activity amongst nudibranchs in tropical environments, where a great richness and diversity of dietary allomones can be found. Indeed, no species of tropical nudibranchs are known to be capable of de novo biosynthesis of secondary metabolites.71 Despite the dearth of certainty regarding the ecological roles of nudibranch secondary metabolites, one clear pattern has emerged. All seven Northeastern Pacific Ocean species of nudibranchs known to be capable of de novo biosynthesis (Cadlina luteomarginata,61 Acanthodoris nanaimoensis49 and A. hudsoni (section 2.2.2), Triopha catalinae50'73 (section 2.2.4), Archidoris montereyensis and A. odhneri,4* and Diaulula sandiegensis (section 2.2.3)) inhabit broad geographic ranges. Therefore, in the examples studied to date there is a well-defined correlation between biosynthetic capability and wide geographic distribution. This 70 observed correlation lends support to the hypothesis stated above that de novo biosynthesis has permitted expansion of species into previously unwelcome territory, where dietary allomones may not be available. The results of this study in context with previous understanding contributes to the prediction that de novo biosynthesis in nudibranchs is an adaptation of temperate species that permits them independence from scarce or unreliable dietary sources of important metabolites. 71 2.2.2 Terpenoid biosynthesis in Acanthodoris nanaimoensis and A. hudsoni 2.2.2.1 Review of known metabolites and terpenoid biosynthesis by A. nanaimoensis The skin extracts of the Pacific dorid nudibranch Acanthodoris nanaimoensis (Figure 2.22) have been found to consistently yield the sesquiterpenoid metabolites nanaimoal (24), isoacanthodoral (27), and acanthodoral (28).74'75 Each of these compounds possesses an unprecedented terpene skeleton and none has been identified from any dietary organism of A. nanaimoensis. Recently, stable isotope incorporation experiments utilizing C labelled acetate successfully proved that these metabolites are produced by de novo biosynthesis in A. nanaimoensis.49 The patterns of intact acetate incorporation in 24 and 27 were consistent with the biogenetic proposal shown in Scheme 2.5, although 28 could not be examined due to low yields. The biosynthetic experiments performed were useful in eliminating the pathway leading to 27b among the potential explanations for the biosynthesis of 27. The results 72 Figure 2.22. Acanthodoris nanaimoensis (photo by R. Long) suggested a eight-membered ring intermediate followed by a cyclobutane-containing intermediate as shown in the pathway leading to 27a. The proposal in Scheme 2.5 includes only one suggestion for the expected pattern of intact acetate incorporation in acanthodoral (28), for which there were no experimental results. It was the goal of this investigation to examine the incorporation of acetate into 28, in order to fortify the above biogenetic proposal. As well, the nature of the putative carbocation intermediates was of interest, especially relating to the origin of the oxygen atoms in the Acanthodoris sesquiterpenoids. We hypothesized that an oxygen of farnesyl pyrophosphate might be retained through the cyclizations to the aldehyde products, as it was with albicanyl acetate (18) in Cadlina luteomarginata. According to this hypothesis (Scheme 2.6), monocyclofarnesal might cyclize Scheme 2.6. Proposed biosynthetic pathway to nanaimoal (24) via an aldehyde intermediate 74 via carbocations with identical carbon skeletons to those in Scheme 2.5, but with full retention of the farnesyl pyrophosphate oxygen throughout the biosynthesis of 24, 27, and 28. This hypothesis has precedent in the biomimetic synthesis of nanaimoal (24) by Engler et al.76 Intramolecular Lewis acid-promoted reactions of the monocyclized aldehydes 29 and 30 yielded mixtures of products that included 24. The substrate 29 is identical to one proposed intermediate in the biosynthetic hypothesis shown in Scheme 2.6. Incorporation of O labelled precursors was identified as a means to probe the origin of oxygen atoms in 24, 27, and 28. Thus, feeding studies were conducted using [1,2-13C2,180i]acetate as reported in section 2.2.1. In order to identify intermediates in the biosynthetic pathway of Acanthodoris metabolites, we decided to attempt incorporation of more advanced biosynthetic precursors which had never been achieved with marine invertebrates. Thus, we chose to feed two relatively simple C 1 5 precursors, [l,2-13C2jfarnesal and [1,2-13 C2]farnesyl pyrophosphate, with a plan to develop techniques leading to the incorporation of more complex precursors such as labelled 29. 75 2.2.2.2 Isolation of sesquiterpenoids from Acanthodoris hudsoni The Pacific dorid nudibranch Acanthodoris hudsoni is not as abundant in Barkley Sound, B.C. as is its close relative, A. nanaimoensis. It is also generally smaller and does not possess the red-tipped rhinophores and branchial plumes that make A. nanaimoensis easily distinguishable from other nudibranchs (Figure 2.23).52 However, upon removal from seawater Figure 2.23. Acanthodoris hudsoni (photo by R. Long) it was noted that A. hudsoni emitted the same fruity odour that had been observed with A. nanaimoensis, and indeed that the aldehydes 24,27, and 28 themselves possessed. A. nanaimoensis had not been an efficient source of acanthodoral (28), possessing 24, 27, and 28 in 76 an approximate ratio of 40:10:1. Therefore it was decided to explore the chemical components of A. hudsoni skin extracts, in the hopes that a better source of 28 or possibly even new metabolites might be found. Ongoing collections of A. hudsoni over several expeditions to Barkley Sound, B.C. eventually yielded 50 specimens for chemical analysis. The same isolation scheme as previously used49 for A. nanaimoensis was employed. Thus, solvent partitioning followed by flash silica chromatography yielded a fraction containing the fragrant materials as a yellow oil. Reduction of the mixture of aldehydes to their corresponding alcohols by reaction with sodium borohydride was then accomplished in order to ease the difficulty of handling the volatile and reactive aldehydes. Finally, reversed phase HPLC yielded two pure compounds, identified by comparison with literature data75 as isoacanthodorol (31) (0.5 mg) and acanthodorol (32) (1.7 mg). From the results of this analysis, it appeared that two species of nudibranchs belonging to the same genus contained the same secondary metabolites, albeit in different quantities. Because it had been demonstrated that 24 and 27 were biosynthesized de novo by A. nanaimoensis,49 it appeared extremely probable that A. hudsoni produced 27 and 28 by the same means. 77 The original structure determinations of 24, 27, and 28 were mostly performed on (p-bromophenyl)urethane derivatives of the parent compounds.74'75 In order to conduct biosynthetic studies on these compounds using C NMR, it was necessary to have a full assignment of the 1 3 C NMR resonances. Although complete assignment of NMR data was accomplished for nanaimool (33) and isoacanthodorol (31) in the previous biosynthetic study,49'77 the small quantities obtained had prevented any such rigour with acanthodorol (32). The isolation of 1.7 mg of 32 from A. hudsoni in this study therefore permitted a full assignment of its NMR data. Acanthodorol (32) isolated from reduced extracts of A. hudsoni was characterized by NMR and mass spectral data. It possessed a parent ion in the HRDCIMS corresponding to a molecular formula of C15H26O, isomeric with both nanaimool (33) and isoacanthodorol (31). Its •H, 1 3C, HMQC, HMBC, and COSY NMR data are presented in Table 2.8. The *H and 1 3C NMR resonances were assigned from the two-dimensional NMR data and by comparison with previously published values for the (p-bromophenyl)urethane derivative of 32,75 and are shown in Figures 2.24 and 2.25, respectively. All two-dimensional NMR data were found to be consistent with the structure and lH and 1 3 C NMR assignments. Many nudibranch secondary metabolites have been shown to benefit the organism that contains them by chemically repelling potential predators.37 The ecological roles of other nudibranch metabolites are unknown. The sesquiterpenoid metabolites of Acanthodoris nudibranchs are among those whose ecological functions have remained unstudied. The two Acanthodoris species have been found to each contain 24, 27, and 28, but in opposite relative abundances. If these metabolites function as defensive allomones, it would seem uneconomical for the two species to produce biosynthetically related compounds in different relative concentrations, since both species presumably share the same potential predators. Therefore, it 79 Table 2.8. NMR data for acanthodorol (32) recorded in CDC13 Carbon # 5 1 3 C a 8'H b COSYb HMBC b 1 39.1 HI: 1.44 (m) H9 HI': 1.46 (m) H l l ' 2 19.6 H2: 1.33 (m) HI H2': 1.43 (m) H3, 3' 3 42.0 H3: 1.06 (m) H2',3' Me 13, Me 14 H3': 1.32 (m) H2',3 4 34.6 - - H6,Mel3,Mel4 5 43.3 1.33 (m) H6, 6' H6, 9, 11, H',Mel3,Mel4 6 17.8 H6: 1.57 (m) H5,7 H5, 7, 7' H6': 1.64 (m) H5,7' 7 30.8 H7: 1.48 (m) H6 H6',9, 11, l l ' ,Mel5 H7': 1.56 (m) H6' 8 37.5 - - H9, 11, 11', 12, 12',Mel5 9 55.9 1.47 (m) H12, 12' H7, 11', 12, 12',Mel5, OH 10 41.8 - - HI', 11, 11', 12, 12' 11 40.6 HI 1: 1.03 (d, 7=9.2) HI', 11' Hl,H5,Mel5 H l l : 1.79 (d, 7=9.1) H l l 12 60.1 H12: 3.63 (d, 7=7.4) H9 H9, 11' H12': 3.63 (d, 7=7.4) H9 13 20.6 0.87 (s) H5 14 31.2 0.79 (s) Mel 3 15 27.2 0.95 (s) H9, 11 OH - 1.24 (br s) 'recorded at 125 M H z "recorded at 500 M H z appears plausible that 24, 27, and 28 may serve a different purpose. One possible function of these metabolites could be chemical attraction. In this scenario, an aldehyde mixture containing 24, 27, and 28 in a 40:10:1 ratio would attract A. nanaimoensis, while a mixture in a 1:10:40 ratio would attract A. hudsoni. The aldehydes are found in the mucus exudate of these nudibranchs, permitting identification of potential mates by following the mucus trails of members of one's own species. The possibility that these compounds may function as attractant mixtures is supported by several field observations. The distribution of A. nanaimoensis and A. hudsoni on the rocky reefs of British Columbia is sparse (at most, approximately one specimen per 20 m2), and their 81 locomotion is slow, similar to that of snails on land. Also, similar to other nudibranchs, Acanthodoris nudibranchs are solitary during most of their lives. Nevertheless, during the brief mating season they successfully locate each other to mate and lay eggs.78 The existence of chemical cues to mating and aggregation is well documented in terrestrial environments, particularly with respect to insects.5 Some evidence also points to this phenomenon amongst marine molluscs, particularly regarding the anaspidean Navanax inermis which is known to follow the slime trails of members of its own species.79 Experimental evidence is required to establish the validity of this tentative hypothesis in relation to Acanthodoris nudibranchs. 2.2.2.3 NMR assignments for nanaimoal (24) and isoacanthodoral (27) As mentioned above, the natural products 24, 27, and 28 had been initially characterized as their (p-bromophenyl)urethane derivatives74'75 and then as their respective alcohols49'77 but never fully as the aldehyde natural products. It was decided that ! H and 1 3 C NMR data for the aldehydes 24, 27, and 28 should be made available for consultation by future researchers. Pure samples of the alcohols 33,31, and 32 were each oxidized to the natural products 24, 27, and 28, respectively, by treatment with A/-methylmorpholine-A/-oxide and TPAP in dichloromethane, as previously described.80 The products were confirmed by HRDCIMS measurements. NMR data was collected for 24 and 27, although 28 decomposed before this was possible. ! H and 1 3 C NMR resonances were assigned by HMQC, HMBC, and COSY NMR experiments and by comparison with similar data collected on the respective alcohols. These data are reported in Table 2.9. The *H and , 3 C NMR spectra are shown in Figures 2.26 and 2.27 (nanaimoal (24)), and Figure 2.28 and 2.29 (isoacanthodoral (27)). 83 85 86 Table 2.9. 'H and l 3C NMR data for nanaimoal (24) and isoacanthodoral (27) recorded in CDC13. Carbon # Nanaimoal (24) Isoacanthodoral (27) 5 1 3 C a 8 'H b 8 1 3 C a 8 'H b 1 31.5 1.77 (m, 2H) 38.4 1.39 (m); 1.64 (m) 2 19.3 1.40 (m); 1.57 (m) 19.1 1.45 (m, 2H) 3 39.7 1.41 (m, 2H) 40.0 1.17 (m); 1.32 (m) 4 33.6 - 34.2 -5 133.7 - 46.2 1.25 (t, 7=5.2) 6 21.3 1.99 (m, 2H) 19.9 1.70 (m); 1.87 (m) 7 34.7 1.45 (m); 1.57 (m) 28.9 1.88 (m); 1.93 (m) 8 32.1 - 135.5 -9 43.6 1.74 (m); 1.80 (m) 129.9 5.21 (s) 10 125.3 - 38.1 -11 53.6 2.21 (dd, 7=14.5, 57.1 2.12 (dd, 7=14.9, 3.2); 2.25 (dd, 3.3); 2.68 (dd, 7=14.6,3.3) 7=14.9,3.3) 12 203.8 9.83 (d, 7=3.2) 204.8 9.71 (d, 7=3.3) 13 27.9 0.96 26.3 0.89 (s) 14 27.8 0.95 32.3 0.97 (s) 15 25.9 1.03 23.4 1.63 (s) "recorded at 125 M H z "recorded at 500 M H z 2.2.2.4 Results of biosynthetic experiments with Acanthodoris specimens 2.2.2.4.1 Collection of specimens, synthesis and incubation of precursors, and isolation of metabolites In each biosynthetic experiment, specimens of A. nanaimoensis or A. hudsoni were collected, maintained, injected with precursor, and extracted in the same way. Those specimens of A. hudsoni used for the control experiment were immersed in methanol upon removal from the ocean and extracted identically to the incorporation experiment specimens. Details of collection, incubation of precursor, and isolation of nanaimool (33), isoacanthodorol (31), and acanthodorol (32) are similar to those previously reported.49 Thus, Acanthodoris nudibranchs were collected 87 from Barkley Sound, B.C. using SCUBA and transported to UBC in refrigerated seawater, where they were maintained for the duration of the biosynthetic experiment. Three 1 3 C labelled precursors, [l,2-13C2,18Oi]sodium acetate, (E,£M1.2-13C2]farnesal, and (£,£)-[! >2-13C2]farnesyl pyrophosphate, were synthesized for administration to specimens of A. nanaimoensis and/or A. hudsoni. Each was synthesized from commercially available 1 3C labelled starting materials. [l,2-13C2,18Oi]Sodium acetate was prepared as described in section 2.2.1. (E,E)-[\,2- C2]Farnesal and (E,E)-[1,2- C2]farnesyl pyrophosphate were synthesized as shown in Scheme 2.7. The first step involved a modified Wittig reaction between [l,2-13C2]ethyl 2-(diethylphosphono)acetate and geranylacetone. This reaction was adapted from previously 81 published methodology. The modified Wittig reaction was stereoselective, producing overwhelmingly the desired (E) product. The resulting ester was reduced to (E,E)-[ 1,2-13C2]farnesol with DLBAL. A portion of this material was then used in the final step towards (£:,£)-[l,2-13C2]farnesal, by oxidation using NMO and TPAP, while the rest of (£,£)-[ 1,2-13C2]farnesol was directed towards the synthesis of (£',£)-[l,2-13C2]farnesyl pyrophosphate. In order to obtain the desired putative biosynthetic precursor (E,E)-[\,2-13C2]farnesyl pyrophosphate, (£,,£)-[l,2-13C2]farnesol was first converted to (£,,£')-[l,2-13C2]farnesyl chloride, using A/-chlorosuccinimide. The procedure used in the synthesis of (£,£)-[l,2-13C2]farnesyl pyrophosphate from its corresponding chloride was taken from Davisson et al.8 2 It involved the coupling of the chloride with tetrabutylammonium hydrogenpyrophosphate ((NBu4)3HP207), prepared from Na 2H 2P207 using a cation exchange column and then titration with tetrabutylammonium hydroxide. The product, (E,E)-[ 1,2- 13C2]farnesyl pyrophosphate as its ammonium salt, was obtained as an extremely labile white solid that was administered immediately to specimens of A. nanaimoensis. 88 ( N B u 4 ) 3 H P 2 0 7 C H 3 C N Scheme 2.7. Synthesis of (£,,£T)-[l,2-13C2]farnesal and (£,,£)-[l,2-13C2]farnesyl pyrophosphate The three putative biosynthetic precursors, [ 1,2-13C2,180i] sodium acetate, (£,£)-[ 1,2-13C2]farnesal, or (£',£)-[l,2-13C2]farnesyl pyrophosphate, were administered to specimens of A. nanaimoensis or A. hudsoni by injection into the hepatopancreas. Injections with [1,2-13C2,18Oi]sodium acetate were performed seven times on alternating days beginning on the third day after collection. Injections with (E,E)-[l,2-13C2]farnesal and with (£',£)-[l,2-13C2]farnesyl pyrophosphate were each performed on the third and fifth days following collection. All animals *. were extracted two days after the last injection. Isolation, purification and characterization of metabolites from Acanthodoris nudibranchs 1 1 *% was accomplished as described in section 2.2.2.2. H and C NMR spectra can be found in Figure 2.30 (for 33), Figure 2.31 (for 31) and Figures 2.24 and 2.25 (for 32). 89 .OH 12 Mel3 Mcl4 12' 6 i 1 9' s i I 1 I Mel5 u I 1 ' ' I I I I I I I 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 (ppm) 10 12 9 11 13 14 156 160 150 140 130 120 110 100 90 70 60 50 40 30 20 10 (ppm) Figure 2.30. NMR spectra of nanaimool (33) in CDC13 top: 'H spectrum recorded at 500 MHz bottom: l 3C spectrum recorded at 125 MHz 90 14 13 12 12' Mel4 Mcl5 11 l j Mel3 ' I *—1 1 1 I 1 1 1 1 I 1 1 1 1 I ' '—r-T—i—i i i i | i i • • | i 3.5 3.0 2.5 2.0 1.5 1.0 0.5 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 (ppm) I ' • 1 I I 11 1 I I I I I I 1 I I I T l M T ITI I T 17 I 1TI I T I H M M TTI n T i l Tl M l 1 T p I I TTfT 1 1 I I I I II I I I I I I T I I I I T I T I I I I I I T I I I t T T 1 I I T • • T T 1 r T T • I I I • I' I I T T T T T I I T I I T T T IT r I I I II 12 14 15 13 160 150 140 130 120 110 100 90 70 60 50 40 30 20 10 (ppm) Figure 2.31. NMR spectra of isoacanthodorol (31) in CDC13 top: 'H spectrum at 500 MHz bottom: 13C spectrum at 125 MHz 91 2.2.2.4.2 Results of the [l,2-13C2,18Oi]acetate incorporation experiment The experiment incorporating [ 1,2-13C2,180i] acetate into metabolites of A. hudsoni yielded new information about the biosynthesis of acanthodoral (28). Doublets flanking the 1 ^ natural abundance singlets were visible in the C NMR spectrum of acanthodorol (32) isolated from this experiment (Figure 2.32). An analysis of the coupling constants (Table 2.10) resulted in the assignment of intact acetate units incorporated into 32 as shown in Figure 2.33. The average specific incorporation at those carbons labelled with intact acetate units was 0.36 %. Table 2.10. "C NMR incorporation data for acanthodorol (32) labelled with [l,2-13C2,18Oi]acetate recorded in CDC13 Carbon # 8 1 3 C a J(c-o (Hz) % specific incorporation 1 39.1 - -2 19.6 33.7 0.33 3 42.0 33.3 0.31 4 34.6 weak 5 43.3 34.1 0.33 6 17.8 34.1 0.27 7 30.8 - -8 37.5 weak 9 55.9 39.4 0.48 10 41.8 weak 11 40.6 28.1 0.36 12 60.1 39.4 0.42 13 20.6 35.7 0.44 14 31.2 - -15 27.1 37.8 0.26 "recorded at 125 M H z The 1 3C scalar couplings observed for acanthodorol (32) from the [ 1,2-13C2-180i]acetate feeding experiment were consistent with the biogenetic proposal in Scheme 2.5. It was not 92 Cl I I I I I I I I 39.1 i ' ' ' ' 11 39.1 C2 r ~i—T—i—r— r~r~TT~r~) (~r-TT~r~i~|~r~t~r~ | i i 1 i | i i i 19.6 19.6 C3 42.0 1 1 1 ' ' i ' 42.0 C4 34.6 ' 34.6 C5 I I I | I 1 I I ] 43.3 1 1 1 1 1 1 ' 43.3 C6 i i i 11 17.8 i 1 * 1 1 i 1 17.8 C7 i i i i i • i i i — i i i i i i i i i i 30.8 ' 30.8 C8 l l l | i 37.5 37.5 C9 Ll ^m^ WW ¥ r i i i i i i i : . 1 i i i i i i 1 1 1 i | i i 1 i i i i i | 55.9 55.9 CIO 41.8 ' 41.8 C l l w JJ i I i i i i 1 i i i I i i 1 r •]• T 40.6 40.6 C12 i I i i i 'i I * i i—i" r'| r vi" 60.1 60.1 C13 i I i i i i I i i I ri~I—r—i—i™I—i- -j— I 1 1 1 1 I 20.6 i 1 ' 111 20.6 C14 i i | i i i i | i i r'i~r ,f ,T" ,r ,t"i i11 "• 31.2 31.2 C15 I M M I r i i I I 111 I I 27.2 27.2 Figure 2.32. 13C NMR resonances of acanthodorol (32) labelled with [l,2-13C2,'801]sodium acetate (left; control, right) recorded in CDC13 at 125 MHz. Figure 2.33. Labelling pattern of intact actetate units observed in acanthodorol (32) incorporated with [l,2-13C2,180i]acetate obvious at this time whether nanaimoal (24) biogenetically precedes acanthodoral (28), or vice versa. One way to determine this would have been to incorporate a derivative of 24 into 28, or of 28 into 24. However, before attempting the incorporation of such large biosynthetic precursors, it was decided that conditions should be optimized by the feeding of derivatives of farnesol, presented in the next section. From the labelling patterns in nanaimool (33), isoacanthodorol (31), and acanthodorol (32) isolated from the [l,2-13C2,180i]acetate incorporation experiments with A. nanaimoensis and A. hudsoni, it was possible to determine whether an oxygen of acetate had been retained in the C-12 position of each metabolite. The acetate fed to the Acanthodoris nudibranchs was 50 % labelled with 1 8 0. If an oxygen atom of acetate was retained through the putative intermediate farnesyl pyrophosphate all the way to the aldehyde products, one would expect 50 % of the C-l2 atoms that were labelled with an intact acetate unit to also possess 1 8 0 labelling. This would be visible in the 1 3 C NMR spectrum as two sets of doublets flanking the natural abundance singlet, as was observed for albicanyl acetate (18) in biosynthetic experiments with Cadlina luteomarginata. The doublet set corresponding to the 1 8 0 labelled intact acetate unit would be 94 expected to be shifted upfield relative to the other doublet by 8 0.010-0.035 ppm,65 readily observable at 125 MHz. The C-l2 signals in the 1 3 C NMR spectra of 33, 31, and 32 isolated from the [1,2-13C2,18Oi]acetate incorporation experiments contained no second set of doublets offset from the one due to 1 3 C- 1 3 C coupling. The incorporation patterns of 33 and 31 were identical to that previously reported from the [l,2-13C2]acetate incorporation experiments,49 indicating that the oxygen atom of acetate was not retained through the biosynthesis to the natural products. It was possible that the aldehyde oxygens were simply exchanged by hydration during the biosynthetic process, eliminating all l sO labels. However, a different pathway than that shown in Scheme 2.6 could be in effect, whereby the aldehydes 24, 27, and 28 are formed without retention of the oxygen atom from acetate/farnesyl pyrophosphate. Scheme 2.8 depicts a reasonable pathway in which monocyclofarnesal is attacked at C-l2 by an amine functional group attached to an enzyme. This leads to enzyme-bound imine intermediates which undergo cyclization yielding each of the carbon skeletons of 24, 27, and 28. The aldehyde natural products are then released from the enzyme by hydrolysis with water. The chemical reaction of aldehydes or ketones with primary amines to form imines is well known.83 Total syntheses of natural products have utilized the reactivity of aldehydes to ammonia, in introducing nitrogen atoms into alkaloid compounds.84 Thus, the proposal in Scheme 2.6 has precedence in laboratory chemistry. This proposal is also supported by the prevalence of imine intermediates in enzymatic reactions. While there have been no proven instances of imine involvement in terpene cyclizations, evidence of imine intermediates in a variety of other biochemical processes, such as lysine biosynthesis,85 DNA polymerase activity,86 reductive animation,87 and oxidative deamination,88 lends credibility to the possibility that such a mechanism could be operating this system. 96 2.2.2.4.3 Results of [l,2-13C2]farnesal and [l,2-13C2]farnesyl pyrophosphate incorporation experiments In order to probe the nature of the biosynthetic intermediates in the pathways to nanaimoal (24), isoacanthodoral (27), and acanthodoral (28), the putative biosynthetic precursors {E,E)-[l,2-13C2]farnesal and (E,E)-[l,2-l3C2]farnesyl pyrophosphate were injected into specimens of A. nanaimoensis. The low solubility of farnesal in water was a problem overcome by dissolving the putative precursor in a minimal amount of DMSO and then diluting the solution with water. Nevertheless, nudibranchs injected with labelled farnesal did not survive beyond the second injection and the experiment was terminated. Thirty-three nudibranchs injected with 1 3C labelled farnesyl pyrophosphate yielded nanaimool (33) in modest quantity (4.9 mg) which did not display any labelling visible by 1 3C NMR. Thus it was concluded that we were unsuccessful at incorporating farnesyl pyrophosphate using the established methodology. No further attempts were made at incorporating large precursors into metabolites of A. nanaimoensis or A. hudsoni. Because farnesyl pyrophosphate is undoubtably a precursor to the sesquiterpenoids 24, 27, and 28, the failure of incorporation must be related to the methodology used. Further research in this area will have to address this problem. 2.2.2.5 Conclusions of biosynthetic study It was found that Acanthodoris hudsoni produces that same sesquiterpenoid metabolites found in A. nanaimoensis, but in different relative quantities. The biosynthesis of acanthodoral (28) by A. hudsoni from acetate by mevalonate-derived isoprenoid assembly was determined to 97 be consistent with a previous proposal put forward for the biosynthesis of nanaimoal (24) and isoacanthodoral (27) by A. nanaimoensis.49 Therefore, it would appear that the biosynthesis of these three metabolites follows the same pathway in two species of Acanthodoris nudibranchs, one that likely involves common intermediates. The identities of the sesquiterpenoid intermediates in the biosynthetic pathways to 24, 27, and 28 were not determined, because experiments aimed at incorporating farnesol-related precursors failed and 1 8 0 labels of acetate were not incorporated. It was determined, however, that the oxygen atoms in the aldehyde functions of 24, 27, and 28 do not come from acetate. It is supposed that the loss of 1 8 0 label from acetate might be explained by the existence of enzyme-bound imine intermediate. 98 2.2.3 Investigations of diaulusterol biosynthesis by Diaulula sandiegensis 2.2.3.1 Review of known metabolites of D. sandiegensis The dorid nudibranch Diaulula sandiegensis is common to the rocky shores and subtidal reefs of Pacific North America, from Alaska to Mexico (Figure 2.34).52 Specimens collected in British Columbia and Alaska at many different sites have consistently yielded two steroid 89 metabolites, diaulusterols A (34) and B (35). These sterols contain an unusual 2a,3oc-diol functional group which has only previously been encountered in steroids isolated from the hydroid Eudendrium glomeratum (e.g. cholest-5-ene-2a,3oc,7p\15(3,18-pentol 2,7,15,18-tetraacetate (36)).90 Diaulusterol A (34) and B (35) are structurally related to pinnasterol Figure 2.34. Diaulula sandiegensis (photo by R. Long) 99 O 37 R=H 38 R=Ac (37) and acetylpinnasterol (38), isolated from the red alga Laurencia pinnata.91 In addition to the unusual 2a,3a-diol array, 34 possesses a 3-hydroxybutyrate extension to the sterol side chain at C-25. We postulated that the 3-hydroxybutyrate fragment would be derived from acetate via standard polyketide biosynthesis, making 34 a metabolite of mixed steroid and polyketide biosynthetic origin. 100 Californian specimens of D. sandiegensis were not known to contain either 34 or 35. Instead, extracts of Californian specimens yielded a series of chlorinated acetylenes (e.g. 39) not 92 found in B.C. specimens. ClCH=CHCH=CHC=CC=CCH=CH(CH2)4COCH3 39 (1Z, 3E, 9Z) The constancy of diaulusterols A (34) and B (35) over a restricted but still substantial geographic range of D. sandiegensis partly satisfied the conditions for de novo biosynthesis predicted by Faulkner, Andersen, and co-workers.51 As was demonstrated in the case of Cadlina luteomarginata (section 2.2.1), it was possible that biosynthesized metabolites had been missed in the study of Californian specimens. The fact that no dietary source for 34 or 35 was known supported the possibility of de novo biosynthesis. Alternatively, we supposed that B.C. and California groups of D. sandiegensis may represent two species with different biosynthetic capabilities, rather than two populations of a single species. We decided to attempt a biosynthetic study of this nudibranch in order to establish the biogenesis of the diaulusterols. If successful, our study was to demonstrate de novo biosynthesis of 34 and/or 35 by D. sandiegensis and to describe the biosynthetic pathway of these compounds by stable isotope incorporation patterns elucidated with l 3 C NMR. 101 2.2.3.2 Review of steroid biosynthesis in molluscs In has been demonstrated on several occasions that gastropod molluscs are capable of de novo biosynthesis of sterols.93'94 The most commonly reported sterol in molluscs is cholesterol, particularly in the higher orders of molluscs (e.g. the cephalopods, which include octopuses). It is generally believed that there is an evolutionary trend towards increasing simplicity of steroid content in molluscs, and, therefore, that the presence of complex non-cholesterol steroid metabolites such as 34 and 35 indicates the primitive nature of their nudibranch source.94 There are believed to be a number of possible sources of sterols in marine invertebrates: (1) straight de novo biosynthesis, (2) sequestration of sterols from dietary organisms, (3) modification of dietary sterols, and (4) passage of sterols from associational microorganisms.95 Sequestration of sponge steroids by nudibranchs has been shown to be the origin of petrosterol (40), found both in the nudibranch Peltodoris atromaculata and in its prey, the sponge Petrosia HO 40 102 ficiformis.96 Evidence has also suggested that nudibranchs modify sponge sterols acquired in their diets. For example, cholestanone (41), is found in the sponge Anthoarcuata graceae, and is structurally related to the acid 42 isolated from the extracts of its nudibranch predator Aldisa sanguinea. Interestingly, 42 possesses greater antifeedant activity than cholestanone (41).97 De novo biosynthesis of sterols by molluscs has been demonstrated using radiolabeled precursors such as [14C]acetate and [14C]mevalonate.94 In an early study,93 [l-14C]acetate was successfully incorporated into mixtures of 3fJ-sterols in four species of nudibranchs. In each of these, cholesterol was found to be the main component of the biosynthesized sterols. The difficulty of tracing steroid biosynthesis in molluscs has been intensified by the possibility that internal regulation of biosynthetic pathways may occur. Steroid biosynthesis requires the sequential condensation of three units of acetate in a pathway leading towards mevalonate. 3-Hydroxy-3-methylglutaryl CoA (HMG-CoA) is an intermediate along this pathway which is converted to mevalonate by the action of the enzyme HMG-CoA reductase. This step is known to be subject to feedback inhibition in vertebrates, slowing biosynthesis from any precursors up to, and including, 3-hydroxy-3-methylglutarate. This internal regulation is believed to operate in invertebrates as well.98 Therefore, the uptake of early biosynthetic precursors such as acetate may be inefficient in biosynthetic studies of sterols.32 Additionally, Djerassi and Silva99 have stated that acetate and mevalonate are generally unsatisfactory precursors in the study of sterol biosynthesis, at least in sponges. The reasons proposed were poor uptake due to precursor water solubility and re-direction of precursors away from sterol biosynthesis to other more efficient biosynthetic pathways. Using lipid precursors such as squalene and lanosterol, Silva etal.100 demonstrated that sponges are abundantly capable of de novo sterol biosynthesis. It is possible, although unproven, that these factors may be significant in nudibranch sterol biosynthesis as they are with sponges. 41 42 2.2.3.3 Assignment of the NMR data for diaulusterol A (34) Although the structure of 34 was previously determined using spectroscopic techniques, a complete assignment of the 'H and 1 3C NMR data was not possible at the time of the original work.89 In order to use 1 3C NMR to detect incorporation of l 3 C labelled precursors in our biosynthetic study, it was necessary to assign all of the 1 3C NMR resonances. 'H, 1 3C, COSY, HMQC, and HMBC experiments were performed on 34 in CeD6 and the resulting data are reported in Table 2.11 (for *H and 1 3 C NMR spectra see Figure 2.35 and 2.36, respectively). 104 105 Table 2.11. NMR data for diaulusterol A (34) recorded in C 6D6 Carbon # 6 1 3 C a 8'H b COSYb HMBC b 1 38.8 HI: 1.53 (m) HI', 2 Mel9 HI': 1.63 (m) HI, 2, Mel9 2 66.6 3.58 (dt, 7=11.2, 3.6) HI, 1', 3 H l , l ' , 4 3 65.5 3.83 (t, 7=3.6) H2,4 HI, 1' 4 128.5 6.68 (d, 7=5.1) H3 H3 5 145.8 - - Hl,3,4,7,Mel9 6 187.5 - - H4 7 124.3 5.98 (t, 7=2.2) H9, 14 H14 8 164.5 - - H9, 14, 15/15' 9 47.4 2.06 (tm, 7=7.4) H7 HI', 7, l l ,Mel9 10 40.3 - - HI', 4, Mel9 11 21.8 H l l : 1.33 (m) H12' H l l ' : 1.43 (m) 12 38.8 HI2: 1.00 (m) H12' Mel8 H12': 1.86 (m) H l l , 12 13 44.2 - - H l l , 11', 14, Mel8 14 55.7 1.61 (m) H7 H7, 15/15', Mel8 15 22.6 H15: 1.23 (m) HI 6/16' H15': 1.23 (m) HI 6/16' 16 20.8 H16: 1.73 (m) H15/15' H16: 1.73 (m) HI 5/15' 17 36.1c 1.00 (m) 18 12.3 0.42 (s) H12/17, 14 19 21.0 0.90 (s) HI' HI' 20 56.3 1.04 (m) Me21 21 18.8 0.92 (d, 7=6.6) H22/22' 22 36.3C H22: 1.32 (m) Me21 H22': 1.32 (m) Me21 23 27.8 H23: 1.08 (m) H24 H23': 1.08 (m) H24 24 41.3 H24: 1.69 (m) H23/23', 24' Me26, Me27 H24': 1.79 (m) H24 25 82.8 - - Me26, Me27 26 26.2 1.40 (s) H24, 24', Me27 27 26.2 1.41 (s) Me26 28 172.2 - - H29' 29 44.2 H29: 2.16 (dd, 7= 16.1,3.8) H30 Me31 H29': 2.21 (dd, 7= = 16.1,8.4) H30 30 64.5 4.05 (m) H29, 29', Me31 H29', Me31 31 22.6 1.02 (d, 7=6.3) H30 H29, 29' 'recorded at 125 M H z Recorded at 500 M H z 'assignments may be interchanged 107 2.2.3 .4 Results of biosynthetic experiments with D. sandiegensis 2.2.3.4.1 Collection, incubation of precursors and isolation of diaulusterol A (34) Three different biosynthetic studies were attempted: the incorporation of acetate, mevalonate, and cholesterol. Although radiolabelling (e.g. 14C) generally permits a greater 1 ^ sensitivity of detection, we chose to use C labelled precursors whenever possible, due to the valuable information that can be gleaned using 1 3C NMR in locating the position of labels in the natural products. As described in earlier sections, we had already devised experimental protocols for increasing incorporation rates to NMR-detectable levels, and so we decided to make use of this protocol in examining the biosynthesis of the diaulusterols. Therefore, for the acetate incorporation experiment, we chose [l,2-13C2]acetate as precursor. Because acetate is used in both standard polyketide and isoprenoid biosynthetic pathways, we judged this to be an appropriate initial tool for examining the biosynthesis of 34. All D. sandiegensis specimens used in the biosynthetic experiments were treated in a similar manner. Animals were collected by SCUBA in Barkley Sound, British Columbia, with seasonal variations in population abundance dictating the size of each experiment. Nudibranchs were kept alive in an aquarium at the Bamfield Marine Station prior to transport in refrigerated seawater to UBC, where the incubation of precursors and the isolation of metabolites were conducted. In each experiment, putative precursors were administered by injection as described in earlier sections. Multiple injections and extended incubation periods totaling 17 days for the acetate incorporation experiments, 11 days for the mevalonate incorporation experiment, and 9 days for the cholesterol incorporation experiment were used to maximize precursor incorporation. 108 The isolation of diaulusterols proceeded as reported previously, using solvent partitioning, flash silica column chromatography, and reversed phase silica HPLC. 8 9 The small number of animals available for each biosynthetic experiment (3-37 specimens per experiment), meant that the minor metabolite diaulusterol B (35) was not isolated in sufficient quantities for characterization. This was expected, because 35 is reportedly present in approximately 15 % yield of 34,89 of which we obtained only 2-10 mg per experiment. Therefore, all biosynthetic results relate only to diaulusterol A (34). 2.2.3.4.2 Polyketide origin of 3-hydroxybutyrate moiety of diaulusterol A (34) Incorporation of 1 3 C labelled acetate into 34 was attempted on three different occasions with D. sandiegensis, all yielding the same results. On each occasion changes to the 1 3C NMR spectrum of 34 were observed, indicating incorporation of 1 3C labels from [l,2-13C2]sodium acetate. Four carbon resonances in the 1 3C NMR spectrum of 34, identified as C-28, 29, 30, and 31, were affected by the labelling studies. Expanded signals for these four carbons are shown in Figure 2.37. The observed coupling constants, patterns, and the calculated incorporation of acetate into C-28 are listed in Table 2.12. Table 2.12. 1 3C NMR incorporation data for diaulusterol A (34) labelled with [ 1,2-13C2]sodium acetate recorded in CeD6. Carbon # 8 1 3 C a Coupling pattern /(C-c) (Hz) % specific -•• incorporation 28 172.2 d 57.0 3.0 29 44.2 d 57.0 b dd 57.0,37.4 30 64.5 d 39.4 c t 38.0 31 22J6 d 394 b "recorded at 125 M H z "integration disrupted by overlapping l 3 C resonances integration disrupted by triplet 109 C28 C29 C 3 0 C31 i i i U U L J I I I I I I I I I I I I II II I I I I I I I I I I 44.2 64.5 22.6 172.2 O OH 34 Figure 2.37. liC NMR resonances of selected carbons of diaulusterol A (34) labelled with [1,2-C2]sodium acetate. In the 1 3 C NMR spectrum of labelled diaulusterol A (34) (Figure 2.37), flanking doublets and more complex multiplets were observed around the usual 1 3C singlet resonance of carbons in the 3-hydroxybutyrate moiety. Matching coupling constants for C-28 and C-29 (7 = 57.0), and C-30 and C-31 (7 = 39.4) indicated that incorporation of intact acetate units did occur at these two sites. Complex multiplet patterns around C-29 and C-30 suggested that there was a small concentration of very highly labelled molecules resulting in some molecules containing two intact acetate units. The expected random incidence of two adjacent intact acetate units would be below 1 3 C NMR detection limits. However, clear doublet of doublets were visible around C-29 (7 = 57.0, 37.4) in addition to the central singlet and the flanking doublet (7 = 57.0) expected from coupling to C-28. Similarly, the 1 3C NMR resonance for C-30 (singlet plus doublet plus triplet) indicated a high degree of coupling not only to C-31 but to C-29 as well. The overall specific incorporation of intact acetate units for this molecular fragment was calculated from the integrated singlet and flanking doublet assigned to C-28. Although it would 110 have been more precise to calculate specific incorporation at all carbons found to be labelled, integration of the other signals was not possible due to overlapping NMR resonances and complex multiplet patterns. Thus, the 1 3C NMR resonance for C-29 overlapped with that of C-13 at 5 44.2, and the resonance of C-31 overlapped with that of C-15 at 5 22.6. The 1 3C NMR resonance for C-30 was complicated by the presence of a triplet in its multiplet structure, which overlapped with the central singlet. Therefore, the overall specific incorporation of acetate into the 3-hydroxybutyrate moiety was based upon the calculation for C-28, as described in Section 2.2.1. Using this formula, the integrated peak areas for C-28 in labelled 34 yielded an overall calculated specific incorporation rate of 3.0 %. The patterns observed in the 1 3 C spectrum of labelled 34 were similar to that of cadlinaldehyde (19) and luteone (20) described in section 2.2.1.3, where it was concluded that there was a large number of unlabelled molecules and a small number of highly labelled molecules resulting from the incorporation of 1 3C labelled acetate. In the case of 20, the presence of a doublet of doublets at one carbon made possible a calculation of the local incorporation amongst labelled molecules and of the percentage of molecules biosynthesized during the labelling experiment. From inspection of the 1 3C NMR signals in Figure 2.37, it would appear that the doublet of doublets at C-29 of diaulusterol A (34) could also facilitate such a calculation. However, the overlap of another carbon at 8 44.2 (C-l3) made this determination impossible since the area of the 1 3C NMR singlet of C-29 could not be measured independently. 13 The [1,2- C2]acetate incorporation experiment showed that the 3-hydroxybutyrate side chain extension of the sterol 34 is produced by de novo biosynthesis in D. sandiegensis in a manner consistent with standard polyketide biosynthesis from two units of acetate. While our use of labelled acetate did not reveal which carbons of this moeity came from C-l or C-2 of I l l acetate, it is most likely that C-28 and C-30 came from C-1 of acetate, due to their oxidized state, and that C-29 and C-31 came from C-2 of acetate. 2.2.3.4.3 On the assembly of the steroid portion of 34 2.2.3.4.3.1 From acetate The incorporation experiments described above utilizing [1,2- 1 3C2] acetate did not yield any specific information about the biogenesis of the steroid portion of diaulusterol A (34). While an overall incorporation rate of 3.0 % was achieved in the 3-hydroxybutyrate fragment of 34, no incorporation of labelled acetate was visible by l 3 C NMR in any of the 27 carbons assigned to the steroid moiety. Two possible explanations for the lack of incorporation are: (1) that D. sandiegensis does not produce the steroid portion of 34 by de novo biosynthesis but simply modifies it by adding the butyrate fragment to an exogenous steroid, or (2) that D. sandiegensis does produce the steroid portion of 34 by de novo biosynthesis but the uptake of acetate in the steroid biosynthetic pathways was too inefficient to give detectable incorporation. We decided to test the hypothesis (2) by attempting to incorporate other precursors and with different methodology than that used in the acetate experiments. The failure to observed acetate incorporation was not altogether unexpected based on the knowledge gained through previous research on the biosynthesis of steroids by marine molluscs94 and sponges.99 As discussed earlier, it has been noted in the study of vertebrate metabolism that steroid biosynthesis is controlled and limited at the step of conversion of HMG-CoA into no mevalonate. Since acetate is used to biosynthesize mevalonate via HMG-CoA, this regulation would slow the incorporation of acetate into natural products that proceed through HMG-CoA. 112 Alternatively, acetate utilization could have been favoured in other non-steroidal lipid biosynthetic pathways such as fatty acid biosynthesis, which might have diverted labelled acetate away from steroid biosynthesis. In view of the possibility of HMG-CoA reductase regulation, we next chose to administer a mevalonate precursor. 2.2.3.4.3.2 From mevalonate In order to benefit from the sensitivity of radiolabelling methodology we chose to administer [2-14C]mevalonate to Diaulula sandiegensis. Thus, [2-'4C]mevalonate (1.85 MBq) was administered in four injections to three specimens of D. sandiegensis on days three, five, seven, and nine after collection. On the 11th day the animals were immersed whole in methanol and extracted. The extract of D. sandiegensis was spiked with 0.8 mg of unlabelled 34 in order to facilitate the isolation of small quantities of material from the radiolabelling experiment. Diaulusterol A (34) was isolated by normal phase silica gel column chromatography and then purified by reversed phase silica HPLC. The pure compound (1.2 mg) was identified as diaulusterol A (34) by comparison of its TLC and HPLC data with data from an authentic sample. An aliquot of purified 34 was subjected to 1 4C scintillation counting and was found to possess radioactivity similar to the background level. Thus, it was concluded that mevalonate administered during the experiment was not incorporated into 34 within the detection limits of the experiment. Although a very sensitive technique (radiolabelling) had been employed, and a reasonable period of administration and incubation of precursor had been allowed, we were unsuccessful in demonstrating mevalonate incorporation into 34. It remained possible that D. sandiegensis was producing 34 by de novo biosynthesis but that the experimental techniques used were either too 113 insensitive or incompatible with the biosynthetic process employed by the organism. The limitations encountered in previous studies" on sponge sterol biosynthesis (poor uptake/alternate metabolism) could also have been playing a role. However, [14C]mevalonate had been successfully incorporated into sterols of nudibranchs on other occasions using similar methodology. It was also possible that D. sandiegensis was producing 34 and/or 35 by chemical modification of a dietary sterol, as has been suggested for the nudibranch Aldisa sanguinea. Thus, it was decided to probe the possibility that 34 was being biosynthesized by modification of cholesterol, the most common sterol found in molluscs.93 2.2.3.4.3.3 From cholesterol As with the mevalonate incorporation study, radiolabeled precursor was chosen for the incorporation experiment with cholesterol. [4-14C]Cholesterol (1.85 MBq) was administered in DMSO to three specimens of D. sandiegensis on days three, five and seven following collection. Due to the death of one specimen, the incubation period was reduced to nine days (from 11 in the mevalonate study and 17 in the acetate study) after which time the two surviving specimens were immersed in methanol and extracted. The isolation procedure used previously (including spiking the extract with 0.8 mg of unlabelled 34) resulted in a yield of ~ 0.5 mg of diaulusterol A (34), identified by comparison of TLC Rf and HPLC retention times with that of authentic 34. The radioactivity of 34 from this experiment was determined by scintillation counting to be within the range of other blank samples measured. From this result, it was concluded that radiolabeled cholesterol was not incorporated into diaulusterol A (34) in this experiment. 114 2.2.3.4.4 Conclusions of the biosynthetic study Incorporation studies using [1,2- 13C2]acetate conclusively demonstrated that the 3-hydroxybutyrate extension of the sterol side chain of diaulusterol A (34) did originate from acetate by de novo biosynthesis in D. sandiegensis. The pattern of incorporation of intact acetate units dictated that this four carbon fragment was produced by standard polyketide biosynthesis from two acetate molecules. 1 3 C NMR was used to demonstrate and measure the incorporation and an overall incorporation rate of 3.0 % was achieved. Coupling patterns of the resonances assigned to the 3-hydroxybutyrate group (Figure 2.37) indicated a high level of incorporation of molecules produced during the feeding experiment. Although we have shown that the polyketide side chain of 34 is produced by de novo biosynthesis in D. sandiegensis, we were unsuccessful in determining the origin of the steroid portion of 34. Attempts were made to incorporate acetate, mevalonate, and cholesterol. It remains possible that our experimental conditions were not adequately optimized to faciliate incorporation of any of the three precursors into 34. Alternatively, the steroid portion of 34 might be produced by a dietary organism of D. sandiegensis, or by D. sandiegensis using a biosynthetic process other than mevalonate-derived isoprenoid biosynthesis. Future work might include attempts to incorporate the steroid precursors squalene or lanosterol in order to prove de novo biosynthesis of 34. 115 2.2.3.5 Geographic variation of metabolites in D. sandiegensis The Pacific nudibranch Diaulula sandiegensis is common in both the coastal waters of British Columbia and California. Although the two populations appear to be different in size and colouration, taxonomists have consistently reported them to be one species.52 Previously, others had reported that La Jolla, California specimens of D. sandiegensis contained a series of chlorinated acetylenes but not diaulusterol A (34) or B (35) . 9 2 Because isolation and structure determination techniques have improved since this original work, we decided to analyze La Jolla specimens in order to ascertain whether or not 34 or 35 had been missed in the original study. Despite taxonomists' reports, we believed it possible that the two populations of D. sandiegensis were two different species which would account for the difference in metabolite content. We hoped to use the identification of secondary metabolites from these two populations to assist in a revision of the taxonomic classification of D. sandiegensis. During a field expedition to La Jolla, California, seven specimens of Diaulula sandiegensis were collected and immersed in methanol. The same isolation scheme as for B.C. D. sandiegensis was employed. A compound with similar TLC Rf and HPLC retention time to 34 was isolated and purified. The clear, colourless oil gave an [M+H] ion in the HRDCIMS at m/z 431.31611, consistent with a molecular formula of C27H43O4. This result and the *H NMR (Figure 2.38) data for this compound were consistent with that of diaulusterol B (35) reported in RO the literature. The amount of metabolite isolated (less than 0.5 mg) was insufficient to obtain a 1 3C NMR spectrum. The finding that diaulusterol B (35) is indeed present in California specimens of D. sandiegensis suggests that the populations of D. sandiegensis in British Columbia and California are closely related. In British Columbia populations, diaulusterols A (34) and B (35) are 116 117 consistently present in D. sandiegensis in proportions of approximately 17:3. However, California specimens appear to contain only 35 (or possibly 34 below current NMR detection limits). Because the ecological or physiological function of the diaulusterols is yet unknown, it is difficult to account for the different distribution of these compounds across the geographic range of D. sandiegensis. The presence of diaulusterol B (35) in both British Columbia and California specimens, in two distinct ecosystems, leads to the expectation that 35 is produced by de novo biosynthesis in D. sandiegensis, according to Faulkner's and Andersen's hypothesis of geographic variation.51 Although we were ultimately not successful in proving this, it is expected that further experiments using radiolabeled precursors as described earlier will establish the de novo biosynthetic origin of diaulusterols A (34) and B (35). 118 2.2.4 Investigations of polyketide biosynthesis by the Northeastern Pacific nudibranch Triopha catalinae 2.2.4.1 Review of skin chemistry of T. catalinae Triophamine (43) is a unique diacylguanidine metabolite that has been isolated from skin extracts of the Northeastern Pacific dorid nudibranch Triopha catalinae (Figure 2.39).101 Its structure was originally determined by analysis of spectroscopic data as well as by degradative chemical experiments.101 The structure was confirmed by total synthesis.102 Recently, Graziani and Andersen showed that triophamine (43) is biosynthesized de novo by T. catalinae and that the two identical ten carbon acyl residues are each derived from five intact acetate units in a pattern consistent with polyketide biosynthesis (see Scheme 2.9). The presence of two ethyl appendages in the acyl residues of 43 suggested that acetate incorporation into these fragments might occur in a processive manner via intact butyrate units through the intermediacy of the putative polyketide building block 2-ethylmalonate (Scheme 2.10). We decided to test this hypothesis by administering the putative biosynthetic precursors 119 O O NH 2 O Scheme 2.9. Biosynthesis of triophamine (43) from acetate 5 0 butyrate and 2-ethylmalonate to Triopha catalinae using the C labelling methodology developed previously.50 Although ethyl branches derived from butyrate are only rarely encountered in polyketide biosynthesis, precedent can be found in the biosynthesis of the microbial metabolites lasalocid A (44)103 and monensin (45).104 Biosynthetic experiments utilizing 1 3C and 1 4C labelled precursors demonstrated that butyrate is incorporated into 44 and 45 as shown in Figure 2.40. 120 In addition to triophamine (43), there are numerous other marine natural products (Figure 2.41) whose biogenesis would appear to involve butyrate. Sponges of the genus Plakortis have yielded many such metabolites: plakortin (46), the ketone 47,105 3-epiplakortin, 9,10-dihydro-3-epiplakortin, the ester 48, the lactone 49, and two closely related peroxides (e.g. 50)106 all isolated from P. halichondrioid.es; and two closely related peroxide acids (e.g. 51)107 isolated from P. angulospiculatus. Structurally similar to 48 is cladocrocin A (52), isolated from the sponge Cladocroce incurvata, and a series of esters and alcohols (e.g. ketal 53) from 121 Chondrosia collectrix. C. collectrix also contained the diol 54 and several peroxides similar to plakortin (46).109 Amongst dorid nudibranchs, triophamine (43) has not only been isolated from T. catalinae, but also from the morphologically similar Polycera tricolor51 and from the Chilean 77 dorid Thecacera darwinii. A related compound, limaciamine (55) was found in the dorid Limacia clavigera from the North Sea.77 However, previous to the work described herein there had been no experimental demonstration in marine invertebrates of polyketide biosynthesis involving intact butyrate or 2-ethylmalonate units. The biogenesis of the guanidine function of triophamine (43) also presented a challenge to the study described herein. Guanidine-containing natural products are quite common in the marine environment, often being noted for their strong biological activities.110 The best known y ^ C 0 2 M e 46 47 examples of biologically active guanidine derivatives are saxitoxin111 (56) and tetrodotoxin112 (8), famous as paralytic shellfish poison and pufferfish toxin, respectively. The guanidine functional group in most natural products is generally believed to come from L-arginine (57), although this has been demonstrated in only a handful of cases. Several classes of bacterially-produced antibiotics, including the streptothricins, streptomycin, viomycin, the arphamenines, and cinodine, have been proven to be derived directly from L-arginine, as have higher plant metabolites galegine and the hordatines.110b Saxitoxin (56), found in a variety of organisms including dinoflagellates, cyanobacteria, and red algae, was shown by 1 3C and l 5 N incorporation methodology to be biosynthesized from arginine, acetate, and methionine.113 Guanidine itself has been suggested as a biosynthetic precursor to other bacterial and marine invertebrate natural products such as ptilocaulin and isoptilocaulin from the sponge Ptilocaulis spiculifer.xu' 1 1 0 b Therefore, one goal of this biosynthetic study was to determine the origin of the guanidine group of triophamine (43). 124 2.2.4.2 Results of biosynthetic experiments with Triopha catalinae 2.2.4.2.1 Collection of specimens, incubation of precursors, and isolation of triophamine (43) In each biosynthetic experiment, specimens of Triopha catalinae were collected, maintained, injected with precursor, and extracted in the same way. Those specimens used for the control experiment were immersed in methanol upon removal from the ocean and extracted identically to the incorporation experiment specimens. Thirty specimens of T. catalinae were collected from Barkley Sound, B.C. by hand using SCUBA for each experiment. The animals were transferred to an aquarium at the Bamfield Marine Station. Those specimens that were fed 1 3C labelled precursors were injected with a 0.55 mol/L solution of the precursor on two occasions, four and twenty hours following collection. Seven days after administration of the precursor, the animals were carefully removed from the seawater and placed in methanol. Isolation of triophamine (43) for each experiment was accomplished as previously reported.50 Thus, exhaustive extraction followed by solvent partitioning and solvent removal in vacuo yielded an ethyl acetate-soluble oil that was further fractionated using silica gel column chromatography. Fractions containing 43 were identified by TLC, combined, and subjected to purification by reversed phase silica HPLC. Triophamine (43) was obtained as a white, crystalline solid that was characterized by comparison of physical data with previous reports.50'101 'H and 1 3 C NMR spectra of 43 are found in Figures 2.42 and 2.43, respectively. 126 127 2.2.4.2.2 Evidence for butyrate as polyketide intermediate in the biosynthesis of triophamine (43) The first step taken in confirming butyrate as a biosynthetic precursor to the acyl residues of triophamine (43) was the feeding of commercially available [l-13C]sodium butyrate. I 3 C NMR was used to determine the extent of incorporation of [l-13C]butyrate into 43 by T. catalinae. Figures 2.43 and 2.44 show the 1 3C NMR spectra of a control and labelled sample, respectively. According to the biosynthetic proposal in Scheme 2.10, four units of butyrate should have been incorporated into the two acyl fragments if butyrate was an intermediate along the polyketide pathway. Therefore, it was anticipated that carbons 1, 1', 3, and 3' would show labelling in the [l-13C]butyrate incorporation experiment (Figure 2.45). \ 8 ' \ 1 0 ' %y Figure 2.45. Expected incorporation of [l-13C]butyrate into triophamine (43) While the use of [l-1JC]sodium butyrate was convenient due its low cost and availability, it was understood that the information it might provide regarding the assembly of 43 was limited. First of all, C-l and C- l ' of 43 provided poor l 3 C NMR signals, being very broad due to tautomerism of the guanidine function. Thus, specific incorporation above natural abundance would only be expected to be visible at C-3/3'. Second, if [l-13C]butyrate underwent degradation 128 129 to acetate, 1 3 C labels might misleadingly be incorporated into 43. Third, incorporation of [1-13C]butyrate would not yield any information about whether a butyrate unit was incorporated at C-3, 4, 5, and 6, or at C-3,4, 7, and 8. And finally, the low levels of incorporation commonly achieved with biosynthetic experiments on marine invertebrates32 would mean that specific incorporation above natural abundance at C-3/3' would be difficult to establish simply by an increase in l 3 C peak height compared to a control sample. Indeed, these limitations proved to restrict the usefulness of results of the [l-13C]butyrate incorporation experiment. On comparison of the peak heights of 1 3 C NMR signals for C-3/3' between the control (Figure 2.43) and incorporation (Figure 2.44) experiments, it is clear that the slight peak height enhancement of C3/3' in the labelled sample is far less significant than the relative peak height differences within carbons of either sample. Therefore, it could not be concluded with confidence that there was specific incorporation of [l-13C]butyrate in 43 above natural abundance. It was necessary to attempt a different approach. Doubly labelled precursors [2,l'-13C2]2-ethylmalonate and [2,3-13C2]butyrate were chosen for the next experiments. These doubly labelled precursors offered two potential advantages. First, retention of the one bond 1 3 C- 1 3 C scalar coupling in triophamine (43) isolated from these incorporation experiments would provide evidence that the C-2/C-3 bond in the butyrate equivalents had not been cleaved during incorporation, and second, the doublet resonances for enriched carbons in the labelled samples would give visually convincing incorporation data that did not depend on measuring what were anticipated to be small peak height differences of singlet resonances resulting from low level incorporation of singly labelled precursors. An interesting side issue in the biogenesis of 43 concerned the pseudo-symmetry of the intermediate 58 (Scheme 2.10), which is a 2,2-diethylacetic acid derivative. The processive biogenetic pathway proposed in Scheme 2.10 predicts that [2,3-13C2]butyrate would label the 130 saturated ethyl appendage (i.e. C-4-C-7) in 43. However, given the pseudo-symmetry in the carbon skeleton of intermediate 58, the alternate labelling pattern (i.e. C-4-C-5) could also have been possible. The [2,l'-13C2]2-ethylmalonate and [2,3-13C2]butyrate incorporation experiments were expected to distinguish between the two possibilities. The syntheses of [2,l'-13C2]disodium 2-ethylmalonate and [2,3-13C2]sodium butyrate were performed according to the methods shown in Scheme 2.11. Thus, the enolate of [2-13C]diethyl malonate was generated by treatment with sodium hydride in THF, and then reacted with [l-13C]ethyl iodide to yield [2,l'-13C2]diethyl 2-ethylmalonate in 81.6 % yield after purification by flash silica chromatography. The use of no more than 0.95 equivalents of [1-13C]ethyl iodide was necessary to avoid a dialkylated malonate product. It was found that the mono- and dialkylated products were difficult to separate using conventional chromatography. Saponification of the ester functions of [2,l'-13C2]diethyl 2-ethylmalonate was achieved by reaction with sodium hydroxide in methanol maintained at reflux for 16 hours. The resulting [2,l'-13C2]disodium 2-ethylmalonate was isolated with a slight amount of residual sodium hydroxide and characterized without further purification. [2,l'-13C2]Disodium 2-ethylmalonate was then either used for the incorporation experiment with T. catalinae, or carried on to the next synthetic step towards [2,3-13C2]sodium butyrate, which was decarboxylation. Reflux for three days in dilute hydrochloric acid successfully accomplished the desired decarboxylation to yield [2,3-13C2]butyric acid. Finally, [2,3-I3C2]butyric acid was converted to its sodium salt by titration to pH 7-8 with a solution of sodium bicarbonate. The intended biosynthetic precursor 13 [2,3- Cysodium butyrate was isolated by removal of water by lyophilization. The first of these experiments performed was the attempted incorporation of [2,1'-1 3C2jdisodium 2-ethylmalonate. The 1 3C NMR signals of all carbons of 43 from this experiment 131 EtO O O l.NaH THF, 0°C 0 0 OEt 2. [l-13C]EtI EtO reflux, 16 hours OEt NaOH, MeOH reflux, 16 hours O O O 1. Titrate, NaHCQ3 ONa 2. Lyophilize O NaO 0.4N HC1 reflux, 72 hours ONa OH Scheme 2.11. The syntheses of [2,l'-13C2]disodium 2-ethylmalonate and [2,3-13C2]sodium butyrate are shown in expanded, normalized, and truncated form in Figure 2.46 column (b). Comparison of the , J C NMR resonances for the sample from the feeding experiment with the corresponding resonances from a control sample (column (c)) revealed new doublets flanking the singlet peaks for C-4/4', C-111', and C-9/9' in the labelled sample. However, the doublets were not very intense and therefore the results were not conclusive. Therefore, it was decided that an incorporation experiment with [2,3-13C2]butyrate was necessary. [2,3- C2]Sodium butyrate was synthesized as described above and administered via injection as a putative biosynthetic precursor to specimens of T. catalinae. Triophamine (43) isolated from this experiment was found from its 1 3C NMR spectrum to possess significant 1 3C specific incorporation as shown in column (a) of Figure 2.46. There were visible 1 3C NMR doublets flanking the 1 3C NMR singlets at C-2/2', C-4/4', C-7/7', and C-9/9', with scalar coupling constants indicating that C2/2' (7 = 32.8) was coupled to C9/9' (7 = 33.6), and C4/4' (7 = 41.2) was coupled to C7/7' (7= 42.0). Table 2.13 offers the specific incorporation results for this experiment, including 1 3 C NMR shifts, coupling constants, and calculated specific incorporation 132 C2,2': I I i i 11 i i I I 50.1 I I i i i I i i i i r r 50.1 1 1 1 1 1 1 1 1 50.1 C3,3': 1 1 i i 11 1 1 i i 39.2 i i i i i | i 1 1 i 39.2 i i i i i I i i 39.2 C4, 4': i i i 1 1 I i i i i 138.9 1 1 i i i | i i i i 138.9 i I I 1 1 1 1 1 1 1 138.9 C5, 5': 11111 1111 120.4 i 1 1 1 1 1 1111 120.4 I I i I I | i I I 120.5 C6, 6': i i i i i I i i i i 13.1 1 1 1 1 1 1 1 1 1 13.1 1 1 1 1 1 I 1 1 1 1 13.1 Cl, T: i 11 i i I i i i i 22.6 i i i i i I i i i i 22.6 I I I I I | I I I I 22.6 C8, 8': I I I I I I I I I I 12.7 W W ' I I I I I | I I I I 12.7 1 1 1 1 i I i i 1 1 12.7 C9, 9': 1 1 1 1 1 1 1 1 25.4 1 1 1 1 1 l 1 1 1 1 1 1 1 1 1 1 i 1 1 1 25.4 25.4 Figure 2.46. 13C NMR resonances from the spectra of triophamine (43) obtained from (a) incorporation experiment with [2,3-13C2]sodium butyrate, (b) incorporation experiment with [2,l'-'3C2]disodium 2-ethylmalonate, and (c) unlabelled control experiment. 133 into 43. Specific incorporation above natural abundance was calculated as described in section 2.2.1, using the integrated peak areas from the 1 3C NMR spectrum of 43. From these calculations at C-2/2', C-4/4', C-111', and C-9/9' and averaging the specific incorporation at these four sites, the specific incorporation of [2,3-l3C2]butyrate into 43 was found to be 0.17 %. Table 2.13. Carbon # 5 1 3 C /(C-C) Specific incorporation (%) 1,1' 185.5 (broad) 2, 2' 50.1 32.8 0.18 3,3' 39.2 4,4' 138.9 41.2 0.14 5,5' 120.4 6, 6' 13.1 7, 7' 22.6 42.0 0.17 8,8' 12.7 9, 9' 25.4 33.6 0.20 10, 10' 11.9 11 157.4 "recorded at 125 M H z in CDC1 3 From the positive incorporation of [2,3- C2]sodium butyrate into triophamine (43), it was shown that butyrate is indeed an intermediate in the processive polyketide biosynthetic pathway from acetate to the acyl fragments of 43. Moreover, the pattern of incorporation demonstrated that the four butyrate units incorporated into 43 are assembled as shown in Scheme 2.10. The two ethyl branches of each acyl portion of 43 were found to be derived from butyrate, while carbons 5/5' and 6/6' come from acetate. Thus, four intact units of butyrate and two of acetate constitute biosynthetic precursors to triophamine (43). 134 2.2.4.2.3 On the origin of the guanidine function of triophamine (43) In an attempt to identify the biosynthetic origin of the guanidine function of 43, we decided to administer radiolabeled arginine (57). Thus, according to the protocol used in the butyrate experiments, 1.85 MBq of L-[U-14Ce]arginine was administered to specimens of T. catalinae twice, four and twenty hours after collection and permitted to incubate for seven days. After extraction and isolation of triophamine (43) from this experiment, a measurement of 1 4C radioactivity comparable to background activity was determined following two rounds of HPLC purification. Thus, it was concluded that no 1 4C label from [U-14C6]arginine was incorporated into the guanidine function of 43. Since labelled guanidine itself was not available from any commercial source known to us, no further attempt was made at determining the biogenesis of guanidine in 43. 2.2.4.2.4 On the existence of isobutyryl-CoA mutase in T. catalinae It has recently been reported in the literature115 that butyrate and isobutyrate can be interconverted in certain bacteria by the action of an enzyme termed isobutyryl-CoA mutase. This phenomenon was first detected by precursor rearrangement during biosynthetic incorporation in natural products such as lasalocid A (44), monensin A(45),117andtylosin.118 The organisms involved were antibiotic-producing streptomycetes, and the natural products involved each contained at least one butyrate subunit that had shown incorporation of rearranged isobutyrate or valine. NMR methods were used to show interconversion of labelled butyrate and isobutyrate and to identify the mechanism of reaction (shown in Scheme 2.12).115 These labelling studies 135 indicated partial loss of stereocontrol (approximately 2:1 selectivity towards the pathway shown in Scheme 2.12). H' Me SCoA Isobutyryl-CoA mutase Me'"/ \ ^ SCoA Scheme 2.12. Interconversion of isobutyrate and butyrate catalyzed by isobutyryl-CoA mutase 115 Previous to our biosynthetic study demonstrating butyrate to be a biosynthetic precursor to triophamine (43), no marine invertebrate natural product had been proven to be butyrate-derived. Therefore, there had been no experimental testing of whether marine invertebrates possessed a similar isobutyryl-CoA mutase enzyme. It was decided that such an experiment should be undertaken. Because of the success experienced in the incorporation of [2,3-13C2]sodium butyrate into 43, a similar experiment was planned using a 1 3C labelled isobutyrate precursor. The precursor chosen was racemic [l,3-13C2]sodium isobutyrate. It was rationalized that if T. catalinae possessed a similar enzyme to the streptomycetes bacteria, that [l,3-13C2]isobutyrate would isomerize to [l,2-13C2]butyrate and [l,4-13C2]butyrate (see Scheme 2.13). The mechanism yielding these two products would depend upon the stereoselectivity of the enzyme-catalyzed reaction, but because a racemic mixture of C labelled isobutyrate was employed, both [1,2] and [1,4] labelled butyrates would be produced. 1 3C NMR would then be used to measure 136 incorporation of [l,2-13C2]butyrate in 43, by observing scalar coupling between carbons derived from intact butyrate units. SCoA O CoA Scheme 2.13. Expected C labelling pattern in triophamine (43) due to incorporation of [1,2-13C2]butyrate from [l,3-13C2]isobutyrate via isobutyryl-CoA mutase reaction [l,3-13C2]Sodium isobutyrate was synthesized by enolate alkylation of [l-13C]propionic acid with [ C]methyl iodide. The synthetic steps that were employed are depicted in Scheme 2.14. The base used for enolate dianion formation was LDA, generated by treatment of diisopropylamine with n-butyllithium. HMPA was added to increase the nucleophilicity of the enolate anion. [l-13C]Propionic acid was thus doubly deprotonated with two equivalents of LDA and reacted with [13C]methyl iodide at the more reactive anion to yield racemic [1,3-13C2]isobutyric acid after titration with aqueous HC1. The product was characterized using HREIMS, 'H, and 1 3C NMR. The desired biosynthetic precursor, racemic [l,3-13C2]sodium isobutyrate, was generated by titration of [1,3- Cyisobutyric acid to neutral pH with sodium hydroxide solution followed by lyophilization. 137 O 1. LDA, HMPA THF O Titrate, 1.ON NaOH O OH 2. [13C]MeI 3. Titrate, 1.0NHC1 OH ONa Scheme 2.14. The synthesis of [ 1,3- 13C2]sodium isobutyrate The biosynthetic experiment testing the hypothesis that T. catalinae may possess a similar mutase enzyme to streptomycetes bacteria was scheduled to occur in September of 1997. A field expedition was undertaken to Barkley Sound, B.C. where T. catalinae had been collected for all previous biosynthetic experiments. Unfortunately, no specimens of T. catalinae were found during this expedition, even though trips undertaken at the same time of year on different years had always yielded at least 30 specimens. Upon consultation with nudibranch experts Dr. Sandra Millen of UBC and Dr. Dennis Willows of Friday Harbor Marine Station in Washington State, U.S.A., it was concluded that unusual weather patterns known as "El Nino" may have been responsible for the low numbers of some Pacific coast nudibranchs in the autumn of 1997. Due to the lack of specimens with which to conduct the biosynthetic experiment, the hypothesis outlined above could not be tested at that time. The experiment is scheduled for a later date. 2.2.4.3 Conclusions of the biosynthetic study 1 o Administration of [2,3- C2]butyrate to T. catalinae has provided clear evidence for the intact incorporation of four butyrate units in the biosynthesis of triophamine (43). The incorporation pattern was consistent with a processive polyketide biosynthetic pathway in which butyrate was incorporated via the putative intermediate 2-ethylmalonate as shown in Scheme 138 2.10. [2,r-13C2]2-Ethylmalonate itself was not as effectively incorporated into 43 in a companion biosynthetic experiment. The overall study represents the first experimental 73 demonstration of polyketide biosynthesis involving butyrate units in a marine invertebrate. The chemical structures of other marine natural products, such as plakortin (46) and cladocrocin A (52) from marine sponges, suggest that they have a similar biogenesis. Therefore, butyrate may be an important polyketide precursor in marine invertebrate biosynthesis. A biosynthetic experiment attempting to incorporate the guanidine functional group of arginine (57) into triophamine (43) was not successful. As a result, it is still not clear whether the guanidine function in 43 is derived from arginine (57), from free guanidine, or from another biosynthetic precursor. Finally, a biosynthetic experiment designed to explore the possibility that T. catalinae contains an enzyme similar to the isobutyryl-CoA mutase found in streptomycetes bacteria was not completed. Because triophamine (43) is the first marine invertebrate natural product proven to use butyrate as an intermediate in polyketide biosynthesis, it would be very interesting to determine whether T. catalinae is capable of other transformations associated with butyrate-assembling organisms. This work will be continued and reported elsewhere. 139 3. New metabolites of the North Sea flatworm Prostheceraeus villatus and its ascidian prey Clavelina lepadiformis 3.1 Review of known metabolites of flatworms Free living flatworms are the most primitive animals to show true bilateral symmetry.119 They are distinguished from other animals by having thin leaflike bodies that are usually ciliated, the cilia moving them smoothly over the substratum. Many flatworms are brightly coloured and strikingly visible against the sea bottom. Flatworms in the class Turbellaria, order Polycladida are carnivores that feed on a variety of small animals. The physical attributes of some flatworm species, namely the bright colouration, soft unprotected body, and relatively slow rate of locomotion resemble the attributes of many soft bodied gastropod molluscs. It is well documented that shell-less molluscs utilize defensive allomones, often sequestered from dietary organisms, to thwart predators.37 Therefore, it is reasonable to hypothesize that flatworms might also use defensive compounds as one method to deter predation. At the time of this study's inception, the only secondary metabolites reported in the literature to be isolated from flatworms came from cultures of dinoflagellates of the genera Amphidinium and Symbiodinium that were isolated from the inner tissues of Okinawan flatworms belonging to the genus Amphiscolops. The eight known amphidinolides (e.g. amphidinolide J 1 2 1 (59)) are macrocyclic lactones with ring sizes ranging from 15 to 27 members. These macrolides were all found to possess antineoplastic activity. The successful acquisition of amphidinolides by the culturing of dinoflagellates established the origin of these compounds. It remains to be shown what, if any, purpose the amphidinolides may serve the flatworm hosts of the dinoflagellates. 140 In order to test the hypothesis that marine flatworms share the ecological traits of marine molluscs, and in order to investigate the potential chemical wealth of an unexamined species and phylum of marine invertebrate, we undertook a chemical study of Prostheceraeus villatus, a marine flatworm found feeding on the ascidian Clavelina lepadiformis in the North Sea. 3.2 Review of known metabolites of Clavelina lepadiformis and other ascidians Ascidians, also known as sea squirts, are sessile in their adult forms. They are members of the phylum Chordata due to the presence of a primitive spinal chord in the larval state.119 The adult animals possess circulatory systems including blood and a heart, as well as a mouth and full digestive system. Ascidians have shown themselves to be rich sources of novel secondary metabolites, most often yielding complex nitrogenous compounds. Natural products chemists' interest in ascidians as sources of bioactive compounds has increased dramatically in the last ten years. From 1988 to 1992 approximately 165 new ascidian metabolites were discovered, representing a three-fold increase over the previous ten year period.122 141 123 Didemnin B (60), from the Caribbean ascidian Trididemnum solidum, has been the most successful drug candidate to be isolated from an ascidian to date. This cyclic depsipeptide exhibited extremely potent biological activity against a host of infectious viruses and also against a number of human cancers.122 The ecteinacidins (e.g. ecteinacidin 743 (61)), tyrosine-derived alkaloids isolated from the Caribbean ascidian Ecteinascidia turbinata, have also shown promising activity against several cancer cell lines.124 There has been ongoing interest in the origin and biosynthetic pathways leading to ascidian alkaloids, in particular towards the eudistomins, tunichromes, and shermilamines. In vivo radioactive incorporation experiments were used to demonstrate the chemical origin of 19^ eudistomin H (62) and I (63) in the Atlantic ascidian Eudistoma olivaceum. Tryptophan and proline were found to be precursors to both of these (3-carboline derivatives. However, since tryptamine was only incorporated into 63 and not 62, it was concluded that in the pathway to 62, halogenation must precede decarboxylation (see Scheme 3.1). 142 The tunichromes are polyhydroxyphenyl peptide pigments found in the blood cells of several species of ascidians, including Ascidia ceratodes, A. nigra, Molgula manhattensis, and 143 Perophora viridis. It is believed that the principle purpose of tunichromes in ascidians is as reducing agents, forming green tunichrome-vanadium complexes analogous to the hemoglobin-iron complex in erythrocyte red blood cells. The biosynthesis of tunichrome An-1 (64) was recently examined using radiolabeled phenylalanine. A. ceratodes exposed to [14C]phenylalanine maximally incorporated this precursor into 64 after 20 days incubation. Investigations were also made into the biosynthesis of shermilamine B (65), natural product of the colonial ascidian Cystodytes dellachiajei.128 Radiolabelling experiments with tritium on intact organisms and on cell-free extracts of C. dellachiajei confirmed the incorporation of tryptophan in 65. Clarifying experiments with [oc-13C]tryptophan indicated the position of tryptophan in 65, leading to a hypothesis that tryptophan, dopa and cysteine are all precursors to 65, via intermediates shown in Scheme 3.2. Clavelina lepadiformis is a social ascidian (Class: Ascidiacea; Order: Aplousobranchiata; Family: Clavelinidae) common to the shallow waters of the North Sea. Previous examination of this species yielded the novel decahydroquinoline alkaloid lepadin A (66) and the structurally-unique lepadiformine (67). Other species of the genus Clavelina collected in many different locations have also been found to contain complex alkaloids, namely the pyrroloiminoquinone wakayin (68) from an undescribed species of Clavelina,131 the quinolizidines pictamine (69), clavepictines A (70), B (71), and the indolizidines piclavines A, B and, C (72)134 all isolated from Clavelina picta. The piclavines represent the first indolizidines found in the marine environment. Most of the metabolites isolated possess modest cytotoxicity and antimicrobial activity. 145 71 R=H 3.3 New metabolites of Prostheceraeus villatus and Clavelina lepadiformis 3.3.1 Collection of specimens Prostheceraeus villatus is a polyclad flatworm (Phylum: Platyhelminthes; Class: Turbellaria; Order: Polycladida; Family: Pseudoceridae) found in the Northeastern Atlantic Ocean and North Sea at depths of 3 to 20 m (Figure 3.1). It is oval in shape and relatively large compared to other flatworms, with a pair of sensory tentacles at the front and an elongated Figure 3.1. Top: the flatworm Prostheceraeus villatus feeding on the ascidian Clavelina lepadiformis, and bottom: two specimens of P. villatus (photos by M. Leblanc) and centrally located intestine. It is a bottom and ledge dweller, creeping across rocks and biota by the use of cilia. Its pale colour and regular brown stripes running from front to rear contrast with the dark colour of most other features of the temperate substratum. In the collection activities reported herein, P. villatus was often observed feeding on the ascidian Clavelina 147 lepadiformis, and never observed feeding on any other organism. Specimens of P. villatus and C. lepadiformis were found in the subtidal waters near Bergen, Norway. Multiple collections over two years afforded approximately 200 specimens of P. villatus and 12 g dry weight of C. lepadiformis. Fresh specimens were immersed in methanol and transported to UBC for laboratory analysis. 3.3.2 Isolation of lepadins and villatamines Following exhaustive extraction of P. villatus with methanol and dichloromethane, a chloroform soluble fraction was obtained by multiple solvent partitions using a modified method of extraction by Kupchan et al.135 The chloroform-soluble materials were fractionated twice with Sephadex LH-20, using methanol and then EtOAc/methanol/water as eluents. Fractions pooled according to common TLC Rf s were next subjected to reversed phase silica column chromatography, and finally reversed phase silica HPLC to yield pure samples of the known alkaloid lepadin A (66)129 (0.2 % dry weight), and the new alkaloids lepadin B (73) (0.05 % dry weight), lepadin C (74) (0.02 % dry weight), villatamine A (75) (0.02 % dry weight), and villatamine B (76) (0.09 % dry weight). Fractionation of the C. lepadiformis extract according to the same methods resulted in the isolation of trace amounts of all five metabolites 66, 73, 74, 75, and 76. In vitro biological testing of these compounds found that 66, 73, and 76 all possess modest cytotoxicity towards several types of cancer cells (Table 3.1). Compound 74 possesses no significant activity towards any of the seven cell lines tested, and 76 was not tested due to chemical instability. Table 3.1. In vitro ED50 values (pig/mL) of metabolites from P. villatus and C. lepadiformis* Compound Murine Human Human Human Human Human leukemia breast glioblastom ovarian colon lung cancer P388 cancer a/ carcinoma cancer A549 MCF7 astrocytoma HEY LOVO U373 Lepadin A (66) 1.2 2.3 3.7 2.6 1.1 0.84 Lepadin B (73) 2.7 17 10 15 7.5 5.2 Lepadin C (74) >50 >50 >50 >50 >50 >50 Villatamine B 11 2.8 1.9 2.8 1.7 2.8 (76) "Villatamine A (75) not tested due to rapid decomposition 149 3.3.3 Structure determination of the lepadins and villatamines 3.3.3.1 Lepadin A (66) Lepadin A (66) was identified by comparison of its NMR spectral data (Table 3.2) and mass spectrometric data with literature values.129 The 'H NMR and 1 3 C NMR spectra of 66 are shown in Figure 3.2 for the purpose of comparison with the new alkaloids. Table 3.2. NMR data for lepadin A (66) recorded in CDC13 Carbon # 5 1 3 C a 5 1 H b 1(N) - 8.15 (br s);9.15(br s) 2 56.5 3.49 (br m) 3 68.5 5.22 (br s) 4 29.6 1.75 (m);2.41 (d, 7=15.3) 4a 36.8 1.70 (m) 5 39.3 2.71 (qm, 7=10.1) 6 28.9 1.70 (m);2.10(m) 7 33.2 1.15 (m); 1.75 (m) 8 19.4 1.70 (m); 1.70 (m) 8a 57.2 3.53 (br m) 9 14.6 1.30 (d, 7=6.7) 1' 132.9 5.18 (dd, 7=15.1,8.8) 2' 132.8 6.15 (dd, 7=15.1,10.1) 3' 129.5 5.95 (dd, 7=15.0, 10.5) 4' 134.7 5.65 (dt, 7=15.0, 7.7) 5' 32.2 2.10 (m); 2.02 (m) 6' 31.3 1.39 (m); 1.39 (m) 7' 22.2 1.39 (m); 1.39 (m) 8' 13.9 0.95 (t, 7=7.0) 1" 171.1 -2" 60.9 4.37 (br s); 4.37 (br s) (OH) - 2.10 (br s) 'recorded at 125 MHz "recorded at 400 MHz Figure 3.2. NMR spectra of lepadin A (66) in CDC13; top: 'Ff spectrum run at 400 MHz bottom: 1 3C spectrum run at 125 MHz 151 3.3.3.2 Lepadin B (73) Lepadin B (73) was isolated as an optically active, clear, colourless oil that gave a parent ion in the HREIMS at m/z 277.2407 appropriate for a molecular formula of C18H31NO. Comparison of the 'H and 1 3 C NMR data of 73 (Table 3.3, Figures 3.3, 3.4) with that of 66 indicated that the two compounds had very similar structures. Two carbons at 8 171.1 and 60.9 and two protons at 8 4.33 were missing in 73, indicating the loss of the glycolic acid residue found in 66. Also, the 'H NMR resonance for H-3 shifted from 8 5.22 in 66 to 8 3.86 in 73, as expected for loss of the acyl shift. The COSY, HMQC, HMBC, and NOE data136 collected for 73 (Table 3.3) were all consistent with the proposed structure. 153 154 Table 3.3. NMR data for lepadin B (73) recorded in CDC1, Carbon # 8 1 3 C 8 'H b COSY" HMBCb NOEc 1(N) - HI: 7.50 (br s) Hl*,2 - Me9 - HI*: 9.94 (br s) HI - H2, 8a, Me9 2 57.4 3.29 (br m) Hl,Me9 Me9 H3, Me9 3 66.6 3.86 (br d) H4*, 4a H4*, Me9 H2, 4* 4 32.3 H4: 1.56 (brm) H4* H4*: 2.29 (dd, 7=13.2,3.0) H3,4 H3, 1', 4a 4a 37.0 1.56 (br m) H3, 8a, 5 H3,4*,6*, Y H4* 5 39.6 2.78 (dd, 7=8.7,3.1) HI', 4a, 7, 7* H4a, 1', 2' H7*, 1', 2' 6 29.0 H6: 1.65 (brm) H6* H6*: 2.09 (d, 7=9.7) H6 7 33.0 H7: 1.05 (dm, 7=12.8) H7* H7*: 1.72 (d, 7=12.8) H5,7 8 19.5 H8: 1.60 (brm) H8*: 1.60 (brm) 8a 56.7 3.47 (br m) H4a H4* 9 14.9 1.39 (d, 7=6.7) H2 1' 134.3 5.27 (dd, 7=15.1,8.7) H5,2' H3' H5,6 2' 132.3 6.11 (dd 7=15.1, 10.4) HI', 3' H3', 4' H5 3' 129.9 5.97 (dd 7=15.2, 10.4) H2', 4' HI', 2', 5', 5'* H5', 5'* 4' 134.0 5.62 (dt, 7=15.2, 7.2) H3', 5', 5'* H2', 5', 5'* H5', 5'* 5' 32.3 H5': 2.04 (q, 7=7.2) H4', 6', 6'* H3', 4' H3' ,4' H5'*: 2.04 (q, 7=7.2) H4', 6', 6'* H3' ,4' 6' 31.4 H6': 1.35 (brm) H5', 5'* H5', 5'*, 7', 7'*, 8' H6'*: 1.35 (brm) H5', 5'* 7' 22.3 HT: 1.33 (brm) H8' H5', 5'*, 6', 6'*, 8' H7'*: 1.33 (brm) H8' 8' 13.9 0.87 (t, 7=7.0) H7', 7'* H7', 7'* (OH) - 3.30 (br s) "recorded at 125 M H z "recorded at 500 M H z 'recorded as NOE difference at 400 M H z or ROESY at 500 M H z 155 3.3.3.3 Lepadin C (74) Lepadin C (74) was isolated as an optically active, clear, colourless oil that gave a parent ion in the HREIMS at m/z 349.2257 corresponding to a molecular formula of C20H32NO4. Comparison of the NMR data for 74137 (Table 3.4; Figures 3.5, 3.6) with the NMR data for 66 revealed that the two molecules had identical perhydroquinoline ring systems with the 2-hydroxyacetate function intact at C-3. COSY correlations between H-5 and H-l', H-l ' and H-2', H-2' and H-3', H-3' and H-4', H-4' and 2H-5', and 2H-5' and 2H-6' (Table 3.4) showed that the A 1 ' 2 ' and A3''4' diene system and the C-5' and C-6' aliphatic methylene carbons of 66 were also present in 74. This was supported by HMBC correlations connecting all protons on the side chain to carbons through two and/or three bonds. The presence of 1 3 C NMR resonances at 8 208.1 and 30.0 (correlated in the 157 158 Table 3.4. NMR data for lepadin C (74) recorded in CDC1, Carbon # 8 1 3C a 8 'H b c o s r HMBC" 1(N) - HI: 8.64 (br s) HI*, 8a - HI*: 9.39 (br s) HI 2 57.0 3.41 (br s) H3, Me9 Me9 3 68.2 5.20 (br s) H2, 4*, 4a, Me9 Me9 4 29.7 H4: 1.70 (m) H4*: 2.30 (m) H4*, 4a H3,4 4a 37.1 1.65 (m) H3,4, 5, 6*, 8a H3 5 39.3 2.76 (q, 7=11.5) HI', 4a, 7 HI', 2' 6 28.9 H6: 1.55 (m) H6*: 2.04 (m) H4a 7 33.4 H7: 1.05 (m) H7*: 1.75 (m) H5,7* H7 8 19.7 H8: 1.56 (m) H8*: 1.68 (m) 8a 57.5 3.48 (br s) HI, 4a, 5 9 15.2 1.26 (d, 7=6.6) H2, 3 1' 134.6 5.20 (dd, 7=13.5, 11.5) H2',5 H3' 2' 132.4 6.16 (dd, 7=13.5, 10.6) Hl',3' H4' 3' 130.9 5.94 (dd, 7=14.9, 10.6) H2', 4' HI', 2', 5', 5'* 4' 131.5 5.56 (dt, 7=14.9,7.3) H3', 5', 5'* H3\ 5', 5'*, 6', 6'* 5' 26.7 H5': 2.30 (m) H4', 6', 6'* H3\ 4', 6', 6'* H5'*: 2.30 (m) H4', 6', 6'* H3', 4', 6', 6'* 6' 43.0 H6': 2.50 (t, 7=7.3) H5',*5',8' H4',5',5'* H6'*: 2.50 (t, 7=7.3) H5', 5'*, 8' H4', 5', 5'* 7' 208.1 - - H5', 5'*, 6', 6'*, 8' 8' 30.0 2.12 (s) H6', 6'* 1" 171.3 - - H2", 2"*, 3 2" 61.5 H2":4.26 (d, 7=17.6) H2"*:4.32 (d, 7=17.6) (OH) - 3.81 (brs) "recorded at 125 MHz "recorded at 500 MHz 'recorded at 400 MHz HMQC to a methyl singlet at 8 2.12), suggested a methyl ketone. HMBC correlations between 2H-6' (a triplet at 8 2.50) and the carbon resonance at 8 208.1, and between 2H-5' at 8 2.30 and the same carbon resonance, showed that the methyl ketone carbonyl was attached to C-6'. An HMBC correlation observed between the methyl singlet at 8 2.12 (Me-8') and both the carbon resonances at 8 208.1 and 43.0 (C-6') confirmed that 74 differed from 66 only by the oxidation state ofC-7' (Figure 3.7). 159 Figure 3.7. Selected COSY (left) and H M B C (right) correlations of lepadin C (74) 3.3.3.4 Villatamine A (75) Villatamine A (75) was isolated an as optically active, clear, colourless oil that decomposed rapidly upon dissolution in chloroform. Its HRDCIMS spectrum exhibited an [M+H] ion at m/z 260.2372 corresponding to a molecular formula for 75 of C 1 8 H 2 9 N , requiring five sites of unsaturation. 75 The l 3 C (Figure 3.8) and APT N M R spectra of 75 in deuterated dichloromethane showed resolved resonances for all 18 carbon atoms (2 X C H 3 , 9 X C H 2 , 5 X C H , 2 X C) and that all 29 hydrogen atoms were attached to carbons (Table 3.5). 1 3 C N M R resonances at 8 108.8 (CH), 160 129.8 (CH), 138.5 (CH) and 142.2 (CH) were assigned to two disubstituted alkenes and a pair of resonances at 8 81.8 (C) and 89.0 (C) were assigned to a disubstituted alkyne. COSY correlations138 confirmed that the two olefins were conjugated and the coupling patterns observed for the two external olefinic protons (H-13: 8 5.81, dt, 7=15.1, 7.3; H-10: 8 5.46, d, 7=15.6) indicated that one end of the diene was bonded to a methylene group and the other end was attached to a non-protonated atom ( !H NMR spectrum: Figure 3.9). An HMBC correlation from the olefinic proton resonance at 8 6.49 (H-11) to the alkyne carbon at 8 81.8 showed that the alkyne carbon represented the non-protonated atom attached to one end of the diene (Figure 3.10). The UV absorbance at 266 nm (e = 18 700) was consistent with a proposed conjugated diene-yne chromophore. Scalar couplings of 15.6 Hz and 15.1 Hz observed between H-10/H-11 and H-12/H-13, respectively, showed that both alkenes had the E geometry. An absence of 1 3C NMR evidence for additional unsaturated functional groups indicated that the fifth site of unsaturation demanded by the molecular formula of 75 was present in a ring. Figure 3.10. Selected COSY (left) and HMBC (right) correlations of villatamine A (75) Ozonolysis of 75 followed by reductive workup with dimethyl sulfide gave pentanal that was identified by GCMS and by comparison of GC retention times with an authentic sample of 161 163 pentanal. The generation of pentanal located the diene-yne chromophore at carbons C-8 to C-l3 in the linear alkyl appendage attached to C-5 of the iV-ethylpyrrolidine. COSY correlations between a methyl triplet at 8 1.37 (H-19) and a pair of geminal methylene protons at 8 2.90 (H-18) and 3.37 (H-18'), that showed no further COSY correlations, identified an isolated ethyl fragment. The chemical shift of the methylene protons and the carbon to which they were correlated in the HMQC spectrum (C-l8: 8 49.3) indicated that the ethyl fragment was attached to the nitrogen atom. A methylene carbon at 8 53.3 (C-2) and a methine carbon at 8 67.0 (C-5) also had chemical shifts consistent with attachment to the nitrogen. The base peak in the mass spectrum of 75, which had a m/z of 98, was attributed to a fragment ion resulting from a-cleavage of an alkyl appendage from an /V-ethylpyrrolidine ring. The connectivity of the alkyne moiety to the pyrrolidine ring via the alkyl appendage was confirmed by HMBC correlations between two geminal methylene protons at 8 2.00 and 2.20 (H-6 and H-6*) and C-8 at 8 89.0; between H-7* at 8 2.61 and C-9 at 8 81.8; and between H-7* and C-5 in the ring at 8 67.0. COSY correlations justified the same conclusion, by demonstrating a long range coupling between the two geminal methylene protons H-7/H-7* and the olefinic H-10 at 8 5.46; and by multiple COSY correlations amongst all four protons of the ethyl appendage (positions 6 and 7) and from them to protons in the ring (at positions 4 and 5) (see Figure 3.11). 164 Table 3.5. NMR data for villatamine A (75) in CD,C1, Carbon # 8' 3C 8'H b COST HMBC" 1(N) - - - -2 53.3 H2: 2.85 (br m) H2* H2*:3.87 (br m) H2 3 29.8 H3: 1.90 (brm) H4 H3*: 2.15 (br m) H4, 4* 4 29.9 H4: 2.27 (br m) H3, 3*, 5 H4*: 2.40 (br m) H3* 5 67.0 3.30 (br m) H4,6 H2*, 3, 7*,18 6 22.2 H6: 2.00 (br m) H5, 6*, 7* H6*: 2.20 (br m) H6,7 7 17.3 H7: 2.40 (br m) H6*, 7* H7*: 2.61(dt, 7=15.6,3.9) H6,7 8 89.0 -9 81.8 - H7*, 11 10 108.8 5.46 (d, 7=15.6) H l l H7*, 12 11 142.2 6.49 (dd 7=15.6, 10.8) H10, 12 H12, 13 12 129.8 6.08 (dd, 7=15.1, 10.8) H l l , 13, 14, 14* H10, 14, 14* 13 138.5 5.81 (dt, 7=15.1,7.3) H12, 14, 14* H14, 14* 14 32.8 H14: 2.10 (dm, 7=7.3) H13,15,*15 H14*:2.10(dm, 7=7.3) H13,15,*15 15 31.6 H15: 1.32 (m) H12, 14, 14*, 17 H14, 14*, 16, 16*, 17 H15*: 1.32 (m) H12, 14, 14*, 17 H14, 14*, 16, 16*, 17 16 22.6 HI6: 1.27 (m) H17 H14, 14* H16*: 1.27 (m) H17 H14, 14* 17 14.0 0.89 (t, 7=7.1) H15, 16 18 49.3 H18: 2.90 (m) H18*,Mel9 H18*: 3.37 (m) H18,Mel9 19 10.4 1.37 (t, 7=7.2) H18, 18* H18, 18* recorded at 125 MHz recorded at 500 MHz recorded at 400 MHz 3.3.3.5 Villatamine B (76) Villatamine B (76) was isolated as an optically active, clear, colourless oil with an [M+H] ion in the HRDCIMS at m/z 264.2688 corresponding to a molecular formula for 76 of C i 8 H 3 3 N , requiring three sites of unsaturation. The UV (tanax 230 nm (e = 19500), 13C/APT and COSY data (Table 3.6) all showed that 76 contained a conjugated E,E-&\en& that was flanked by two 165 methylene carbons ('H and 1 3C NMR spectra: Figures 3.12 and 3.13). The DCIMS had a base peak at m/z 98, attributed to the same N-ethylpyrrolidine fragment formed by oc-cleavage of an alkyl appendage of the villatamine ring system, observed with 75. Ozonolysis of 76 followed by reductive workup gave heptanal that was identified by GCMS and by comparison with an authentic sample. Thus the structure of villatamine B was formulated as shown for 76. The remaining COSY, HMQC, and HMBC data139 were all consistent with this structure. Table 3.6. NMR data for villatamine B (76) recorded in CDC1, Carbon # 5 , 3 C 8'H b COST HMBCh 1(N) - - - -2 52.6 H2: 2.80 (br m) H2*, 3, 3* H2*: 3.88 (br m) H2, 3, 3* 3 21.5 H3: 1.93 (m) H2, 2*3* H3*:2.18(m) H2, 2*, 3 4 29.4 H4: 1.86 (m) H4*,5 H4*:2.24 (m) H4,5 5 67.3 2.98 (br m) H4, 4* H4, 4* 6 29.9 H6: 2.00 (m) H7* H6*:2.19(m) 7 29.4 H7: 2.00 (m) H7*, 8 H8, 9 H7*: 2.25 (m) H6, 7, 8 8 128.4 5.41 (dt, 7=14.6,7.1) H7, 7*, 9 9 132.4 6.00 (dd, 7=14.6, 10.2) H8, 10 H l l 10 129.5 5.94 (dd, 7=14.6, 10.2) H9, 11, 12, 12* H8,H12,H12* 11 134.2 5.57 (dt, 7=14.6,7.2) H10, 12, 12* H9,H12,H12* 12 32.5 H12: 2.02 (dm, 7=7.2) H10, 11, 13, 13* H10, 11 H12*: 2.02 (dm, 7=7.2) H10, 11, 13, 13* 13 29.2 H13: 1.34 (m) H12, 12* H l l , 12, 12* H13*: 1.34 (m) H12, 12* 14 28.8 HI4: 1.23 (m) H13, 13* H14*: 1.23 (m) 15 31.7 H15: 1.24 (m) Mel7 H15*: 1.24 (m) 16 22.5 HI6: 1.23 (m) Mel7 Mel7 H16*: 1.23 (m) Mel7 17 14.0 0.85 (t, 7=6.9) H16, 16* 18 48.6 H18:2.82 (m) H18*,Mel9 Mel9 H18*:3.38 (m) H18,Mel9 19 10.2 1.34 (t, 7=7.3) H18, 18* "recorded at 125 M H z "recorded at 500 MHz 'recorded at 400 MHz 167 168 3.4 Conclusions of flatworm-ascidian study One known and four novel alkaloid compounds were isolated from a free living flatworm, Prostheceraeus villatus, and its prey, a colonial ascidian, Clavelina lepadiformis. The lepadins (66, 73, 74) contained a common decahydroquinoline system with an eight carbon conjugated unsaturated side chain. The villatamines (75, 76) each possessed an A/-ethylpyrrolidine ring system connected to a twelve carbon conjugated unsaturated side chain. The structure of villatamine A (75) included an unusual carbon-carbon triple bond conjugated with a diene. Three of these alkaloids demonstrated biological activity against human cancer cell lines. While the ecological functions of the lepadins and villatamines in either the ascidian C. lepadiformis or the flatworm P. villatus have not yet been determined, it is clear that P. villatus obtained these alkaloids as a result of feeding on C. lepadiformis. The quantities found in P. villatus greatly exceeded those found in the ascidian prey, indicating that the flatworm concentrated the alkaloids from its diet. Because of the large number of alkaloids found in 122 ascidians, it is likely that C. lepadiformis is responsible for the biosynthesis of the lepadins and villatamines. Thus, it is plausible that the flatworm P. villatus sequesters alkaloids from its ascidian prey in a manner analogous to many nudibranch-sponge and anaspidean-alga/sacoglossan-alga ecological systems, in which molluscs have been known to benefit from chemical defences obtained by dietary sequestration from prey species. These results represent the first evidence that worms of the phylum Platyhelminthes share this ecological characteristic with members of the phylum Mollusca. In this case, the observation that the life history and physical attributes of the flatworm P. villatus resembled that of many shell-less molluscs lead to the discovery of an undescribed ecological phenomenon and four new alkaloid natural products.140 4. Conclusions The biological and chemical origins of secondary metabolites from several species of marine invertebrates were investigated. The biosynthetic capabilities of five species of Northeastern Pacific dorid nudibranchs were studied using precursor incorporation methodology. The extracts of a North Sea flatworm and its ascidian prey were analyzed and their metabolite constituents determined, establishing a predator-prey sequestration model for the origin of metabolites in this system. The dorid nudibranch Cadlina luteomarginata was found to produce albicanyl acetate (18), cadlinaldehyde (19), and luteone (20) by de novo biosynthesis, as evidenced by incorporation of [l,2-13C2]acetate and [2-13C]mevalonolactone. The unique cadlinalane and luteane skeletons were determined to be derived from a sesterterpenoid biosynthetic precursor, representing the first proven instance of sesterterpenoid biosynthesis in a marine mollusc. Incorporation of an oxygen label from [l,2-13C2,180i]acetate into the alkoxy oxygen of 18 indicated that the oxygen atom was retained from farnesyl pyrophosphate. C. luteomarginata is only the second nudibranch to be shown capable of de novo biosynthesis and dietary sequestration of metabolites. Examination of Southern Californian specimens of C. luteomarginata (as well as B.C. specimens) demonstrated that the biosynthesized metabolites 18,19, and 20 are present across the entire geographic range of this species. However, the sesterterpenoids 19 and 20 were found to be present in significantly lower concentrations at the southern end of its range than in the north. Variation of metabolite concentration amongst individuals from one population in B.C. was also found to be significant, leading to a hypothesis that metabolite concentration may be 170 regulated according to need. According to this hypothesis, the ready availability of dietary metabolites may reduce the production of biosynthesized metabolites. Stable-isotope incorporation studies were also used to demonstrate de novo biosynthesis of acanthodoral (28) by Acanthodoris hudsoni. An oxygen incorporation study with A. nanaimoensis showed that the oxygen atoms of the aldehyde functions nanaimoal (24), isoacanthodoral (27), and 28 are not derived from farnesyl pyrophosphate. It was postulated that the biosynthesis might proceed via an imine intermediate, and that the oxygen in the natural products could come from hydrolysis of the imine with water. Similar methodology to that used above was applied to the study of diaulusterol A (34) biosynthesis by Diaulula sandiegensis. Incorporation of [1,2- 1 3C2] acetate demonstrated that D. sandiegensis produced the 3-hydroxybutyrate side chain extension of 34 by standard polyketide biosynthesis. Feeding of labelled acetate, [2-14C]mevalonate, and [4-14C]cholesterol all failed to demonstrate biosynthesis of the sterol portion of 34. Incorporation of [2,3- C2]butyrate conclusively showed that the diacylguanidine triophamine (43) is assembled by processive polyketide biosynthesis using butyrate in the dorid nudibranch Triopha catalinae, presumably via 2-ethylmalonate. Labelled 2-ethylmalonate was not as successfully incorporated in a companion feeding experiment. This study represents the first proven instance of marine invertebrate biosynthesis utilizing butyrate. The significant number of other marine polyketide metabolites with ethyl branches implies that butyrate may be an important intermediate in marine invertebrate metabolism. From the above study of de novo biosynthesis in five species of Northeastern Pacific Ocean dorid nudibranchs, several hypotheses have emerged. It was observed that tropical dorid nudibranchs benefit from access to diets rich in defensive metabolites and are not known to be capable of de novo biosynthesis. In contrast, the diets of temperate dorid nudibranchs are 171 chemically poorer and many temperate nudibranchs produce secondary metabolites by de novo biosynthesis. Thus, we propose that the ability to biosynthesize one's own defensive allomones has permitted the expansion of several nudibranch species into previously unwelcoming environments, where dietary allmones may be scarce or unavailable. Northern populations C. luteomarginata and D. sandiegensis were found to contain more biosynthesized metabolites than southern populations. We suggest that this is because northern temperate sponge communities are chemically poorer than southern temperate ones, creating a greater ecological need in the north for de novo biosynthesis. Substantial variation in concentration of biosynthesized metabolites within individuals of one population of C. luteomarginata was also observed. This may be caused by internal regulation of biosynthetic output. For instance, individuals that have recently fed upon organisms containing ecologically-important metabolites require and produce small amounts of biosynthesized metabolites. Finally, examination of extracts of the flatworm Prostheceraeus villatus and its ascidian prey Clavelina lepadiformis resulted in the isolation of one known and four new alkaloids. The novel metabolites lepadins B (73) and C (74) were found to possess decahydroquinoline carbon skeletons, while villatamines A (75) and B (76) were identified as pyrrolidine alkaloids. All five metabolites contain a conjugated diene and villatamine A (75) contains an unusual diene-yne functional group. All metabolites were found to be present in greater concentrations in the predator flatworm than in its prey, suggesting for the first time that the soft-bodied flatworms may sequester ecologically-useful metabolites in a manner similar to many gastropod molluscs. In summary, the results presented in this thesis add substantially to the body of knowledge on marine chemical ecology, particularly to the problem of the origin of secondary metabolites. Several new hypotheses have also been presented that need to be tested experimentally by chemists and ecologists. 172 5. Experimental section 5.1 Materials and Methods Solvents used were reagent or analytical grade, and were not treated unless in preparation for moisture sensitive reactions. Dichloromethane, THF, and methanol used in such reactions were dried by distillation, over calcium hydride, sodium with benzophenone, and magnesium, respectively. Moisture sensitive reactions were conducted under a positive pressure of argon dried by passage through sulfuric acid, sodium hydroxide, and Drierite®. Ozone was generated with a Welsbach ozonator. Chemicals were used as received from suppliers unless otherwise noted. Solvents were removed by evaporation using Biichi rotary evaporators, in vacuo. Normal phase silica gel column chromatography was performed using 230 - 400 mesh silica gel 60. Reversed phase silica gel was synthesized from the above silica and n-octadecyltrichlorosilane according to the method of Evans, Dale, and Little.141 Flash refers to the application of pressure from an aquarium air pump to column chromatography. Normal and reversed phase silica gel and alumina thin layer chromatography was carried out on commercial plates. Compounds were visualized by UV absorption at 254 nm and/or by staining the chromatograms with Dragendorff s reagent,142'143 a solution of vanillin in sulfuric acid and ethanol, or a solution of ninhydrin in ethanol143 followed by heating. Size-exclusion chromatography was performed using Sephadex LH-20 gel permation beads. High performance liquid chromatography was conducted on one of three systems: (1) Waters 501 HPLC pump with a Waters 440 UV absorbance detector or a Perkin Elmer LC-25 refractive index detector, (2) Waters 600E HPLC pump with a Waters 486 UV tunable absorbance detector, or (3) Waters 600E HPLC pump with a Waters 996 UV photodiode array 173 detector. Systems (1) and (2) were used with chart recorders while (3) was used in conjunction with a computer and Millenium™ 2021 chromatography software. HPLC columns employed were Waters Rad-Pak™ and Whatman magnum, both normal phase and reversed phase (octadecasilyl) 10 p. models. All HPLC solvents were filtered and degassed prior to use. Nuclear magnetic resonance spectroscopy was performed on Bruker AMX-500, WH-400, AM-400, and Varian XL-300 spectrometers. Spectra were processed using either Bruker Windows™ compatible WINNMR software, an Aspect 1000 workstation, or a Varian workstation. Spectra were referenced to the residual light solvent: CDCI3: 8 7.24 ('H), 8 77.0 (13C); C 6D 6: 8 7.15 ('H), 8 128.0 (13C); CD2C12: 8 5.32 ('H), 8 53.8 (13C); DMSO-d6: 8 2.49 (!H), 8 39.5 (13C); D 20: 8 4.65 (!H). Coupling constants (J) are reported in hertz. Post-acquisition modifications to NMR data may include magnitude calculation, symmetrization, zero-filling, and application of apodization functions to the free-induction decay. Ultraviolet measurements were performed on a Hewlett Packard 8452A diode array spectrophotometer. Maximum absorptions are reported as ^ m a x in nanometers with molar absorptivity coefficients e in units of L mol"'cm'1. Infrared spectra were measured on an ATI Mattson Genesis Series FTIR spectrophotometer, as thin films between sodium chloride plates. Optical rotation measurements were performed on either a Perkin Elmer 241 polarimeter or a JASCO J-700 spectropolarimeter, using a 1 cm quartz cell at the sodium D line (589 nm) and the values given in degrees. Gas chromatograms were run on Hewlett Packard 5890A gas chromatograph equipped with a HP3392A integrator. Radioactive decay counts were measured on a Beckman LS6000IC scintillation counter. Mass spectrometry was performed by the UBC Department of Chemistry Mass Spectrometry Centre. Measurements of compound cytotoxicity were performed by the laboratory of Professor Theresa M. Allen, at the University of Alberta. 5.2 Investigations of secondary metabolism of C. luteomarginata 174 5.2.1 New metabolite of C. luteomarginata Thirty-four specimens of C. luteomarginata (0.6 - 6.0 g each, wet weight) were collected from Barkley Sound, British Columbia by hand using SCUBA at depths of 3 - 20 m, and immersed in methanol, following which they were extracted exhaustively over five days with methanol (2 X) and methanol/dichloromethane 1:1 (4 X). The extracts were combined, reduced in vacuo, and partitioned between water (300 mL) and ethyl acetate (4 X 300 mL). The ethyl acetate extracts were dried over anhydrous magnesium sulfate, combined and reduced in vacuo to yield a fragrant yellow oil (~ 500 mg per 20 animals). This oil was fractionated on a silica gel flash column (15 cm long; 2 cm diameter) using a gradient eluent system (9:1 hexanes/ethyl acetate to ethyl acetate). Groups of compounds were pooled based on common TLC Rf's on silica gel using vanillin spray to visualize the spots. Each fraction was reduced in vacuo and passed through a reversed phase sep-pak™ with methanol as eluent to remove grease. The most polar fraction was further purified by reversed phase silica HPLC (methanol/water 9:1) using refractive index detection to yield pure seco-spongian 22 (3.3 mg) as a clear, colourless oil: [cc]D25 = -5.5°, (c = 0.2, CHC13); NMR data see Table 2.1; UV X m a x 248 sh (e = 434), 252 (e = 481); IR 2926, 1737, 1464, 1374, 1241 cm"1; HRDCIMS (m/z): M + calcd for C 2 4 H 3 8 O 6 , 422.26685; found, 422.26805. 175 5.2.2 Biosynthetic experiments with C. luteomarginata Between May and September 1996 six collections were made in Barkley Sound and Jervis Inlet, B.C., of between 20 and 44 specimens of C. luteomarginata each, as described above. Two collections were treated as control experiments and therefore not subjected to injections of putative biosynthetic precursors. These animals were immediately immersed in methanol and worked up according to the procedure outlined above. Collections targeted for injection experiments were treated as follows. Animals were kept alive in aquarium at 12 °C for 17 days following collection. On days 3, 5, 7, 9, 11, 13, and 15, each animal was subjected to injection of [l,2-13C2]sodium acetate, [l,2-13C2,18Oi]sodium acetate, or [2-13C]mevalonolactone. [1,2- *C?\ sodium acetate and [2-13C]mevalonolactone were purchased and used without further purification, while [l,2-13C2,18Oi]sodium acetate was prepared by synthesis as described below. The precursor used was dissolved in distilled water to a concentration of 0.55 mol/L. An injection of 100 - 200 p:L (proportional to the size of animal) was administered by 21G2 needle and syringe to the left dorsum (presumably into the hepatopancreas). On day 17, all specimens were immersed in methanol. Each collection of animals was processed separately following the same protocol, as described earlier. Yields of 18, 19, and 20 obtained from the various incorporation experiments are give in Table 2.2. Albicanyl acetate (18) was obtained as a clear, colourless oil matching previously published NMR and mass spectral data;55 'H NMR (CDC13) 8 0.71 (s, Me-15), 0.77 (s, Me-14), 0.84 (s, Me-13), 1.97 (s, Me-17), 2.36 (ddd, 7=13,4, 2, H-7') 4.14 (dd, 7=11, 9, H-12), 4.29 (dd, 7=11,4, H-12'), 4.47 (d, 7=1, H-ll), 4.81 (d, 7=1, H-ll ' ) ; 1 3C NMR data see Table 2.3; HREIMS (m/z): M + calcd for Ci 7 H 2 8 0 2 , 264.20892; found, 264.20887 (control experiment), 264.20970 (incorporation of [1,2- 13C2]acetate experiment), 264.20892 (incorporation of [1,2-176 13C2,18Oi]acetate experiment). Cadlinaldehyde (19) was obtained as a white, crystalline solid matching published NMR and mass spectral data;60 'H NMR (CDC13) 8 0.70 (s, Me-19), 0.77 (s, Me-18), 0.92 (s, Me-17), 0.98 (s, Me-21), 3.80 (dd, 7=17, 8, H-16), 3.88 (dt, 7=9, 3, H-16'), 10.07 (s, H-20); 1 3C NMR data see Table 2.4); HREIMS (m/z): M + calcd for C21H34O2, 318.25589; found, 318.25534 (control experiment), [M-Me]+ calcd for C 2 0 H 3 iO2 303.23239; found, 303.23243 (incorporation of [l,2-13C2]acetate experiment), 303.23181 (incorporation of [2-13C]mevalonate experiment). Luteone (20) was obtained as a white, crystalline solid matching published NMR and mass spectral data;53 *H NMR (CDC13) 8 0.57 (s, Me-21), 0.74 (s, Me-20), 0.91 (s, Me-19), 2.09 (s, Me-18), 2.30 (m, H-16), 2.57 (ddd, 7=18, 8, 5, H-16'), 4.40 (br s, H-23), 4.79 (d, 7=1, H-23'), 10.07 (br s, H-22); 1 3 C NMR data see Table 2.4; HREIMS (m/z): M + calcd for C23H36O2, 344.27155; found, 344.27170 (incorporation of [1,2- 13C2]acetate experiment), 344.27180 (incorporation of [2-13C]mevalonate experiment) (20 from control experiment decomposed before mass spectrum taken). Preparation of [l,2-13C2,18Oijsodium acetate To 1.00 g (12.4 mmol) of [l,2-13C2]acetyl chloride was added slowly by syringe [180]water (1.00 g; 50.0 mmol) while stirring at 0 °C. Dry THF (10 mL) was added to dilute the solution which was stirred for a further ten minutes. Powdered sodium bicarbonate (1.07 g; 12.7 mmol) was then added to neutralize hydrochloric acid produced by the reaction. After one hour the pH of the mixture was slightly acidic. At this time a second equivalent (1.07 g; 12.7 mmol) of sodium bicarbonate was added to yield a solid mixture of [l,2-13C2,l8Oi]sodium acetate and sodium chloride that upon lyophilization weighed 1.92 g (1.80 g expected; therefore some excess sodium bicarbonate present). This mixture was characterized and used for injection 177 into nudibranchs without further purification. [ 1,2-13C2,180i]sodium acetate: 'H NMR (D20) 8 1.19 (dd, 7=126.5, 5.6, H-2, 3 H); 1 3 C NMR (D20) 8 21.0 (d, 7=52.1, C-2), 179 (d, 7=52.1, C-l). 5.2.3 Investigations into distribution of metabolites in C. luteomarginata Collections of C. luteomarginata specimens were made as described above in Barkley Sound, B.C. and La Jolla, California, U.S.A. Three experiments were conducted: (1) variation between two populations, (2) variation between individuals of one population, and (3) variation within tissues of individuals. Eighteen specimens were collected in the waters off La Jolla, CA and divided into three pools: nine were immersed in methanol without further treatment, four were subjected to biosynthetic experiments using unlabelled mevalonolactone as precursor, and five were subjected to biosynthetic experiments using [ 1,2-13C2]sodium acetate as precursor. Similarly to specimens in the earlier described biosynthetic experiments, specimens in this study were injected with a 0.55 mol/L of one precursor on several occasions. In this case, the volume of injected precursor was 200 - 300 pL depending upon the size of individuals, and injections were performed once per day for four days, starting one week following collection. On the day following the last injection, the specimens were immersed in methanol. Thus, three extracts of La Jolla specimens of C. luteomarginata were generated. For comparison with B.C. specimens, 14 small specimens of C. luteomarginata were collected from Barkley Sound and immersed in methanol. In order to study variation amongst individuals of a single population, 19 specimens of C. luteomarginata were collected in Barkley Sound and each placed in a separate vial of methanol. Ten specimens of C. luteomarginata were collected from Barkley Sound and dissected 178 into the sections shown in Figure 2.20. The common tissues of the ten nudibranchs were combined in methanol. In all cases, animals or tissues were extracted with methanol for a period of one week. All extracts were treated identically. The whole animals or tissues were weighed and the volume of extract recorded. An aliquot of 2.0 mL of extract was removed from the original vessel and concentrated to 0.35 mL. The volume of this concentrated extract injected in each case into the gas chromatograph was 3 pL (80:1 split injection). Standards of the pure compounds albicanyl acetate (18), cadlinaldehyde (19) and luteone (20) were made from previously isolated material of known weight that was dissolved in methanol. Three dilutions of each standard solution was made, so that four different concentrations were available for calibration. Calibration curves were created with the integrated peak areas from multiple gas chromatograms of each standard. Each extract described above was subjected to GC analysis using a DB-5 column (120 °C for 2 min; increased to 275 °C at 35 °C per min; flow 1 mL He per min). The concentrations of metabolites 18,19, and 20 were calculated from the calibration data and the known weights of animals and volumes of extract and injected sample. Results of these experiments are shown in Tables 2.5, 2.6, and 2.7, and representative GC traces are shown in Figure 2.21. GC retention times were used to match standard peaks with those in the extracts. In certain cases where identification was uncertain, co-injection of a standard and extract sample was used to confirm a peak's identity. As well, GCMS performed with a similar GC system by the UBC Department of Chemistry Mass Spectrometry Centre was used to confirm the identity of many peaks as described below, including those of standards used. Albicanyl acetate (18) used as standard: GC peak at 7.17 min; HREIMS (m/z): M + calcd for CnHzgOz, 264.20892; found, 264.20887. Cadlinaldehyde (19) used as standard: GC peak at 11.38 min; HREIMS (m/z): M + calcd 179 for C 2 1 H 3 4 O 2 , 318.25589; found, 318.25534. Luteone (20) used as standard: GC peak at 13.42 min; HREIMS (m/z): M + calcd for C23H36O2, 344.27155; found, 344.27170. Extract from California specimens treated as control: GC peak at 7.13 min identified as albicanyl acetate (18): HREIMS (m/z): M + calcd for C17H28O2, 264.20892; found, 264.20859; GC peak at 11.39 min identified as cadlinaldehyde (19): HREIMS (m/z): [M-Me]+ calcd for C20H31O2, 303.23239; found, 303.23292. Extract from California specimens treated with [ 1,2- 13C2]acetate: GC peak at 7.14 min identified as albicanyl acetate (18), isolated by methods described earlier to yield a clear, colourless oil (5.6 mg) matching *H and l 3 C NMR data described in section above; GC peak at 11.40 min identified as cadlinaldehyde (19): HREIMS (m/z): [M-Me]+ calcd for C20H31O2, 303.23239; found, 303.23206; GC peak at 13.41 min identified as luteone (20) uncharacterized by MS. Extract of B.C. specimens treated as control: GC peaks at 6.78 min (albicanol), 7.13 min (18), 11.38 min (19), 13.39 min (20) in addition to 13 unidentified peaks exhibiting the following EIMS (m/z) (relative intensity): (i) 117 (27), 101 (100), 99 (39); (ii) 231 (7), 207 (8), 187 (17), 185 (52), 122 (100) 100 (8); (iii) 269 (7), 212 (6), 193 (10), 177 (100), 149 (11), 122 (33); (iv) 236 (6), 219 (21), 209 (68), 181 (62), 169 (12), 138 (44), 119 (54), 97 (67), 83 (100); (v) 231 (15), 177 (77), 162 (100); (vi) 281 (18), 269 (7), 205 (10), 191 (100); (vii) 222 (12), 177 (33), 149 (100), 119 (25); (viii) 292 (22), 253 (9), 230 (100), 215 (73), 202 (24), 187 (30); (ix) 288 (47), 289 (20), 288 (47), 219 (20), 207 (44), 188 (42), 168 (55), 148 (100), 119 (58); (x) 291 (100), 253 (13), 231 (12), 193 (20), 181 (25), 153 (33); (xi) 347 (100), 331 (7), 287 (6), 203 (15), 181 (10); (xii) 366 (100), 351 (7), 331 (7), 317 (6), 255 (8), 181 (30), 169 (32), 135 (40), 119 (39); (xiii) 378 (15), 316 (27), 253 (13), 219 (10), 207 (100), 181 (25). 180 Extract of MDFs from 10 dissected specimens of Barkley Sound C. luteomarginata: GC peaks at 6.76 min (albicanol), 7.15 min (18), 11.41 min (19), 13.45 min (20) in addition to six unidentified peaks exhibiting the following EIMS (m/z) (relative intensity): (i) 192 (54), 177 (100), 149 (18), 123 (30), 107 (26); (ii) 231 (17), 216 (47), 201 (35), 122 (100); (iii) 216 (34), 201 (100), 181 (18); (iv) 216 (28), 122 (100); (v) 362 (2), 347 (100), 287 (10), 207 (45); (vi) 326 (42), 283 (7), 189 (12), 159 (100), 131 (32). 5.3 Biosynthetic experiments with A. nanaimoensis and A. hudsoni Two experiments were conducted with A. hudsoni: a control experiment and a biosynthetic incorporation experiment with [l,2-l3C2,180i]acetate. Three experiments were conducted with A. nanaimoensis: all biosynthetic incorporation experiments, with [1,2-I3C2,180i]acetate, [l,2-13C2]farnesal, and [l,2-13C2]farnesyl pyrophosphate. All precursors were synthesized from commercially available 1 3C labelled starting materials. Preparation of (E,E)-[ 1,2- C2]farnesal and (E,E)-[ 1,2- C2]farnesyl pyrophosphate (£,£)-[l,2-13C2]Ethyl farnesoate: To a suspension of sodium hydride (0.106 g; 4.42 mmol) in dry THF under argon and at 0 °C, was added [l,2-13C2]ethyl 2-(diethylphosphono)acetate (1.00 g; 4.42 mmol). The mixture was stirred for one hour, following which geranylacetone (935 \iL; 4.20 mmol) was added dropwise and the temperature raised to 25 °C. After three days, the solvent was removed in vacuo and the residue was partitioned between saturated sodium bicarbonate (50 mL) and ethyl acetate (3 X 50 mL). The combined organic extracts were washed with brine, dried over anhydrous sodium sulfate and the solvent removed in vacuo, to yield 1.14 g of a yellow oil. This crude product was purified by flash silica 181 column chromatography (hexanes/ethyl acetate 49:1) to yield (E,E)-[\,2-13C2]ethyl farnesoate in 73.8 % yield as a clear, pale yellow liquid (0.824 g; 3.10 mmol): 'H NMR (CDC13) 6 1.23 (t, 7=7.4, 3H), 1.56 (s, 6H), 1.63 (s, 3H), 1.93 (m, 3H), 2.03 (t, 7=6.6, 2H), 2.11 (m, 6H), 4.09 (qd, 7=6.9, 2.7, 2H), 5.03 (m, 2H), 5.62 (d, 7=159.4); 1 3C NMR (CDC13) 8 14.2, 15.9, 17.6, 18.7, 25.6, 25.9, 26.6, 39.6, 40.8, 59.3, 115.6 (d, 100X intensity, 7=75.1, t, 7=72.4), 116.3 (d, 7=76.6), 122.8, 124.2, 131.3, 136.0, 166.8 (d, 100X intensity, 7=75.8); HRDCIMS (m/z): [M+H]+ calcd for Ci5H290213C2, 267.22348; found, 267.22227. (E,E)-[l,2-13C2]Farnesol: (£,£)-[l,2-13C2]Ethyl farnesoate 0.824 g (3.10 mmol) was dissolved in dry THF (50 mL) and cooled to -78 °C. DIBAL in hexanes (10.0 mL of 1.00 mol/L solution; 10.0 mmol) was added by syringe. The reaction mixture was allowed to warm to room temperature after two hours, and then was returned to -78 °C to be quenched with a saturated solution of ammonium chloride. Anhydrous magnesium sulfate was added, and the product mixture filtered through a sintered glass funnel. The solvent was removed in vacuo and the resulting clear, colourless oil was purified by flash silica column chromatography (hexanes/ethyl acetate 19:1), to give the desired product, (£'(£)-[l,2-13C2]farnesol (0.576 g; 2.60 mmol) in 83.9 % yield: 'H NMR (CDC13) 8 1.56 (s, 6H), 1.64 (s, 6H), 1.90-2.12 (m, 8H), 4.11 (ddd, 7=141.5, 6.5, 3.8, 2H), 5.06 (dd, 7=13.0, 6.5, 2H), 5.38 (dt, 7=153.0, 6.1, 1H); 1 3C NMR (CDC13) 8 15.9, 16.2, 17.6, 25.6, 26.2, 26.6, 39.4, 39.6, 59.3 (d, 100X intensity, 7=47.0), 123.3 (d, 100X intensity, 7=47.6), 123.7, 124.3, 131.2, 135.2, 139.6 (d, 7=72.8); HREIMS (m/z): M + calcd for Ci 3H 2 60 1 3C2, 224.20508; found, 224.20532. (E,E)-[\,2- 13C2]Farnesal: A dichloromethane solution of /V-methylmorpholine-/V-oxide (0.338 g; 2.54 mmol in 5 mL) was added to (E,E)-[l,2-13C2]farnesol (0.376 g; 1.70 mmol) in dichloromethane (5 mL) and finely ground dry 4 A molecular sieves (1.0 g). While stirring at room temperature, two crystals of TPAP were added and the reaction was monitored by TLC. 182 After three hours, the mixture was filtered through silica gel and the solvent evaporated to yield 0.309 g (1.39 mmol; 81.9 % yield) of (£,£)-[l,2-13C2]farnesal as a clear, colourless oil: 'H NMR (CDC13) 8 1.54 (s, 3H), 1.56 (s, 3H), 1.62 (s, 6H), 1.90-2.20 (m, 8H), 5.03 (m, 2H), 5.83 (dd, 7=157.9, 8.0, IH), 9.93 (ddd, 7=169.4, 24.4, 8.0, IH); 1 3C NMR (CDC13) 8 16.0, 17.5, 17.6, 25.6, 25.6, 26.5, 39.5, 40.5, 115.2 (d, 7=76.3), 122.4, 124.0, 127.3 (d, 100X intensity, 7=55.7), 131.3, 136.4, 191.2 (d, 100X intensity, 7=54.9); HREIMS (m/z): M + calcd for Ci 3 H 2 40 1 3 C 2 , 222.18942; found, 222.18961. (£,£)-[l,2-13C2]Farnesyl chloride: Dimethyl sulfide (172 |iL; 2.34 mmol) was added dropwise to /V-chlorosuccinimide (0.299 g; 2.24 mmol) in dry dichloromethane (60 mL) at -30 °C while stirring. The resulting white suspension was warmed briefly to 0 °C then cooled to -40 °C before the addition of (£,£)-[ l,2-I3C2]farnesol (0.432 g; 1.95 mmol) by syringe over a period of three minutes. The reaction mixture was warmed to 0 °C over 45 minutes, at which temperature it was maintained for one hour. The clear, pale yellow solution was then stirred at room temperature for 15 minutes. The reaction mixture was transferred to a separatory funnel in which it was washed with cold brine. The aqueous phase was extracted with two portions (20 mL) of pentane. The organic extracts were combined, increased in volume by a further 20 mL of pentane, and washed with cold brine (2X10 mL). Anhydrous magnesium sulfate was used to dry the pentane solution, and upon solvent removal, (£,£)-[l,2-13C2]farnesyl chloride was obtained as an unstable, clear, yellow oil (0.453 g; 1.87 mmol) in 95.9 % yield: *H NMR (CDCI3) 8 1.58 (s, 3H), 1.66 (s, 3H), 1.70 (s, 3H), 1.72 (s, 3H), 1.92-2.13 (m, 8H), 4.08 (ddd, 7=150.3, 7.6, 4.2, 2H), 5.07 (t, 7=6.5, 2H), 5.43 (dt, 7=158.3, 8.2, IH); 1 3C NMR (CDC13) 8 16.0, 16.1, 17.6, 25.7, 26.1, 26.7, 39.4, 39.7,41.1 (d, 100X intensity, 7=47.0), 120.3 (d, 100X intensity, 7=47.0), 120.5,123.4, 124.3, 131.3, 135.6. 183 (£,£)-[l,2-13C2]Farnesyl pyrophosphate: The phosphorylating agent tetrabutylammonium hydrogenpyrophosphate ((NBu4)3HP207) was prepared according to a previously reported method.82 Thus, disodium dihydrogen pyrophosphate (Na2H2P207) (3.34 g; 15.0 mmol) was passed through a DOWEX AG 50W-X8 (50 - 100 mesh) cation exchange column (hydrogen form) and then titrated to pH 7.30 with tetrabutylammonium hydroxide (30 mL of a 40 % w/w aqueous solution). The solution was then lyophilized and the tacky white solid was stored under vacuum at -15 °C until used, at which time 3.37 g (3.73 mmol) was dissolved in 3 mL of acetonitrile. (£,£)-[l,2-13C2]Farnesyl chloride (0.453 g; 1.87 mmol) in 2 mL of acetonitrile was added by cannula to the solution and the reaction mixture stirred at room temperature under argon for two hours. (E,E)-[l,2-13C2]Farnesyl pyrophosphate (as the ammonium salt) was produced from the tetrabutylammonium salt by cation exchange chromatography. Thus, elution from the column with a buffer of 25 mN NFLHCOa/iPrOH, followed by lyophilization, led to (E,E)-[l,2-13C2]farnesyl pyrophosphate (0.839 g mostly as the ammonium salt), a white solid. This material was immediately injected into specimens of A. nanaimoensis (see below). Collection of specimens, administration and incubation of precursors Specimens of A. nanaimoensis and A. hudsoni were collected for each experiment by hand using SCUBA in Barkley Sound, B.C., at depths of 2 to 20 m. For each biosynthetic experiment, nudibranchs were returned to UBC in refrigerated seawater and kept alive in aquarium for several days, after which time they were immersed whole in methanol and extracted. A. hudsoni control experiment: Fifty individuals were collected and placed directly in methanol. The isolation scheme above yielded 31 (0.5 mg) and 32 (1.7 mg). A. hudsoni [ 1,2-13C2,180i]acetate incubation experiment: Seventeen specimens were 184 collected and kept alive in seawater brought from Barkley Sound for 17 days. Each nudibranch was injected seven times into the hepatopancreas using a syringe equipped with a 21G2 needle, with 100 \iL of an approximately 0.55 mol/L (total) aqueous solution of [ 1,2-13C2,180i]sodium acetate/sodium chloride 1:1. Following extraction and isolation, this experiment yielded 31 (0.3 mg) and 32(1.5 mg). A. nanaimoensis [l,2-13C2,18Oi]acetate incubation experiment: Fifty specimens were collected and treated as A. hudsoni above, except that the 30 larger specimens were administered 200 pJL of precursor solution instead of 100 pL. The pure compounds isolated were 33 (36.2 mg) and 31 (6.1 mg). A. nanaimoensis [l,2-13C2]farnesal incorporation experiment: Thirty-two specimens were collected and injected once with 100 pX of 0.15 mol/L [l,2-13C2]farnesal in water with a few drops of DMSO to aid dissolution. Because the nudibranchs became slow and unresponsive to touch, the next injection was changed to 50 pL of 0.10 mol/L [l,2-13C2]farnesal in DMSO for each specimen. By the end of five days after collection, all specimens had died. 13 A. nanaimoensis [1,2- C2]farnesyl pyrophosphate incorporation experiment: Thirty-three specimens were collected and kept alive for seven days, during which they were injected 1 ^ twice with 50 pX each of 0.05 mol/L [1,2- C2]farnesyl pyrophosphate (ammonium salt) in DMSO. Isolation and purification of the extract yielded 33 (4.9 mg) and 31 (1.2 mg). In all experiments extracts of A. nanaimoensis and A. hudsoni were treated alike. Specimens were extracted twice with methanol and four times with 1:1 methanol/dichloromethane. The extracts were combined, the solvents removed, and the residue was partitioned between water (200 mL) and ethyl acetate (4 X 200 mL). The combined organic extracts were dried over anhydrous magnesium sulfate and the solvent removed by rotary evaporation. The resulting fragrant yellow oil was fractionated on silica gel (flash; 10 cm long; 185 2 cm diameter) using a gradient eluent system from hexanes/ethyl acetate 4:1 to ethyl acetate. After removal of solvents in vacuo, fractions exhibiting a fruity smell and staining with vanillin on TLC were combined, and grease removed by passing the oil through a reversed phase sep-pak™ with methanol. At this point, reduction with sodium borohydride was performed in order to convert the volatile aldehydes to their corresponding alcohols. Thus, the mixture of aldehydes (10 - 200 mg for each extract) was treated with an excess of NaBFL (60 mg) in 30 mL of isopropyl alcohol, and stirred at room temperature for 16 hours. After this time, 30 mL of water was added and the mixture was stirred for a further two hours, then partitioned between water and chloroform (4 X 25 mL). The combined chloroform layers were dried over anhydrous magnesium sulfate and the chloroform removed in vacuo to yield a mixture of alcohols. Finally, reversed phase HPLC (methanol/water 4:1) with refractive index detection was used to purify nanaimool (33), isoacanthodorol (31), and acanthodorol (32). Nanaimool (33): clear, colourless oil matching NMR and mass spectral data previously determined;77 ! H NMR (CDC13) 8 0.85 (s, 3H), 0.94 (s, 3H), 0.95 (s, 3H), 1.30-1.95 (m, 14H), 3.70 (br m, 2H); 1 3 C NMR (CDC13) 8 19.4, 21.3, 24.8, 27.8, 28.0, 30.7, 31.7, 33.5, 34.7, 39.8, 43.9, 44.0, 59.6, 125.5, 133.3; HRDCIMS (m/z): M + calcd for C i 5 H 2 6 0 , 222.19836; found, 222.19899. Isoacanthodorol (31): clear, colourless oil matching physical data previously determined;77 lH NMR (CDC13) 8 0.86 (s, 3H), 0.98 (s, 3H), 1.10 -1.50 (m, 6H), 1.59 (s, 3H), 1.50- 1.70 (m, 2H), 1.80- 1.95 (m, 3H), 2.03 (dt, 7=13.4, 7.3, IH), 3.67 (brm, 2H), 5.08 (s, IH); 1 3 C NMR (CDC13) 8 19.2, 19.9, 23.4, 26.7, 29.0, 32.1, 34.0, 37.4, 37.9,40.1,45.4,46.7, 60.1, 131.6, 134.1; HRDCIMS (m/z): M + calcd for C i 5 H 2 6 0 , 222.19836; found, 222.19800. Acanthodorol (32): clear, colourless oil; NMR data see Table 2.8; HRDCIMS (m/z): M + calcd for C i 5 H 2 6 0 , 222.19836; found, 222.19817. 186 Oxidation of 33, 31, and 32 to yield the corresponding aldehydes was accomplished using the NMO/TPAP methodology as described in the synthesis of farnesal. Thus, 4.3 mg of 33 was converted to 3.0 mg of nanaimoal (24); 31 (2.5 mg) was converted to isoacanthodoral (27) (1.5 mg), and 32 (1.7 mg) was converted to acanthodoral (28) (1.0 mg). Nanaimoal (24): clear, colourless, fragrant oil; NMR data see Table 2.9; HRDCIMS (m/z): [M+H]+ calcd for C i 5 H 2 5 0 , 221.19054; found, 221.19058. Isoacanthodoral (27): clear, colourless, fragrant oil; NMR data see Table 2.10; HRDCIMS (m/z): [M+NH4]+ calcd for C1 5H280N, 238.21709; found, 238.21789. Acanthodoral (28): clear, colourless, fragrant oil that decomposed prior to collection of NMR spectral data; HRDCIMS (m/z): [M+H]+ calcd for C i 5 H 2 5 0 , 221.19054; found, 221.19070. 5.4 Biosynthetic experiments with D. sandiegensis Five biosynthetic experiments were completed using D. sandiegensis during different months (April, August, September, and December). In each instance, the procedure used for isolation of the diaulusterols was the same. Three of the five experiments involved the use of [1,2-13C2]acetate as putative precursor, whereas the fourth and fifth involved [2-14C]mevalonate and [4-14C]cholesterol, respectively. 13 [1,2- C2] Acetate as precursor: On three occasions 18-37 specimens of D. sandiegensis were collected in Barkley Sound, B.C., by hand using SCUBA, at depths of 5 - 20 m. The animals were stored live in refrigerated seawater and transported to UBC. For 17 days they were maintained in aquarium with seawater brought from Barkley Sound. On the third, fifth, seventh, 187 ninth, 11th, 13th, and 15th days after collection, the nudibranchs were each injected into the hepatopancreas with 100 pL of a 0.55 mol/L solution of [ 1,2-13C2]sodium acetate in distilled water, by syringe equipped with a 21G2 needle. Nudibranchs that died during the experiments (< 5 %) were discarded. On the 17th day the nudibranchs were immersed whole in methanol. [2-14C]Mevalonate as precursor: Three specimens of D. sandiegensis were collected in September 1997 in Barkley Sound and transported to UBC as above. On the third, fifth, seventh and ninth days, each nudibranch was injected with 100 pL of [2-14C]mevalonate solution, which consisted of 1.85 MBq of [2-14C]mevalonate/Y,/Y,/y',/Y'-dibenzylethylenediamine salt, dissolved in 1.2 mL distilled water. On the 11th day the nudibranchs were immersed in methanol. Extracts of D. sandiegensis from this experiment were spiked with 0.8 mg of pure unlabelled diaulusterol A (34) prior to the silica column step of the isolation, in order to facilitate isolation of 34 from such a small quantity of extract. [4-14C]Cholesterol as precursor: Three specimens of D. sandiegensis were collected and transported as above. On the third, fifth, and seventh days, each nudibranch was injected with 50 pL of [4-14C]cholesterol solution, which consisted of 1.85 MBq of [4-14C]cholesterol, dissolved in 50 pL of DMSO and diluted with 400 pL of water. After this time, one nudibranch had died and was discarded, and another appeared sluggish so the experiment was terminated on the ninth day. The two remaining nudibranchs were immersed in methanol. Extracts of D. sandiegensis from this experiment were spiked with 0.8 mg of pure unlabelled diaulusterol A (34) prior to the silica column step of the isolation, in order to facilitate isolation of 34 from such a small quantity of extract. The isolation of diaulusterols was conducted as follows, as previously reported in the QQ literature: the specimens of D. sandiegensis were extracted with methanol (2 X) and with methanol/dichloromethane 1:1 (4 X) and the extracts were pooled and concentrated. The 188 resulting brown oil was partitioned between ethyl acetate (4 X 200 mL) and water (200 mL), and the ethyl acetate-soluble materials were combined, dried over anhydrous magnesium sulfate, and concentrated in vacuo. The residue was fractionated on a flash silica column (20 cm long; 2 cm in diameter) using acetone/dichloromethane (1:1) as eluent. Fractions containing a TLC spotting material (normal phase silica TLC; column eluent as developing solution; Rf ~ 0.35) that stained with vanillin were combined and subjected to HPLC purification. HPLC conditions used were: reversed phase silica magnum column; eluent methanol/water 4:1; detection by UV at 254 nm. Concentration and freeze-drying of the HPLC fraction corresponding to the largest UV absorption peak yielded pure diaulusterol A (34). Diaulusterol A (34) from [ 1,2- 13C2]acetate incorporation experiments: clear colourless oil (2 - 10 mg per experiment); *H and 1 3C NMR data see Table 2.11; 1 3C incorporation data see Table 2.12; HRDCIMS (m/z): M + calcd for C3iH 4 8 06, 516.34509; found, 516.34411. Diaulusterol A (34) from [2-14C]mevalonate incorporation experiment: clear colourless oil (1.2 mg); matched 34 by TLC and HPLC retention time; after one purification by HPLC, scintillation counting showed 36 DPM (aliquot tested 10 pL of 1 mL total; background count 32 DPM). Diaulusterol A (34) from [4-14C]cholesterol incorporation experiment: clear colourless oil (< 0.5 mg); matched 34 by TLC and HPLC retention time; after one purification by HPLC, scintillation counting showed 95 DPM (aliquot tested 10 pL of 1 mL total; background count 32 DPM). Geographic variation study: Seven specimens of D. sandiegensis were collected by hand using SCUBA off La Jolla, California, in June 1997, and immersed in methanol. Upon return to UBC, the nudibranchs were extracted according to the same procedure as used for the diaulusterols above, and diaulusterol B (35) was isolated. 189 Diaulusterol B (35) from La Jolla: clear, colourless oil (0.5 mg); ! H NMR data matches literature values89 (*H NMR spectrum, Figure 2.38); HRDCIMS (m/z): [M+H]+ calcd for C27H43O4, 431.31613; found, 431.31611. 5.5 Biosynthetic experiments with T. catalinae [l-13C]Sodium butyrate and L-[U-14C6]arginine monohydrochloride were purchased for use in biosynthetic incorporation experiments. All other precursors ([2,l'-13C2]disodium 2-ethylmalonate, [2,3-13C2]sodium butyrate, and [l,3-13C2]sodium isobutyrate) were synthesized from commercial 1 3C labelled materials. Synthesis of precursors [2,l'-13C2]Diethyl 2-ethylmalonate: To a 0 °C suspension of NaH (0.142 g; 5.90 mmol) in 5 mL of dry THF under argon, was added [2-13C]diethyl malonate (1.00 g; 6.21 mmol). The mixture was stirred at 0 °C for one hour, after which time [l-13C]iodoethane (0.926 g; 5.90 mmol) was added to the pale yellow solution by gas-tight syringe. The reaction mixture was heated at reflux for sixteen hours, yielding a white solid present in the colourless liquid. The mixture was then partitioned over water (50 mL) and extracted with diethyl ether (3 X 50 mL). The combined organic extracts were dried over anhydrous magnesium sulfate and the solvent evaporated to yield a clear, yellow oil. This product was further purified by flash silica column chromatography using petroleum ether/diethyl ether 19:1 to yield [2,l'-13C2]diethyl 2-ethylmalonate, a clear colourless oil in 81.6 % yield (0.915 g; 4.81 mmol): *H NMR (400 MHz, CDCI3) 5 0.90 (m , H-2', 3H), 1.20 (t, 7=7.2 , H-2", 6H), 1.85 (dpd, 7=103.6, 7.2, 4.3, H-l', 2H), 3.16 (dtd, 7=131.9, 7.2, 4.3 , H-2, IH), 4.12 (q, 7=6.8 , H-l", 4H); 1 3C NMR (100 MHz, CDC13) 190 8 11.7 (d, 7=34.3, C-2'), 13.9 (C-2"), 22.1 (d, 100X intensity, 7=33.6, C-l'), 53.5 (d, 100X intensity, 7=33.6, C-2), 61.0 (C-l"); 169.3 (d, 7=57.2, C-l); HRDCIMS (m/z): [M+H]+ calcd for C 7H 1 704 I 3C2, 191.11940; found 191.11982. [2,r-13C2]Disodium 2-ethylmalonate: Sodium hydroxide (0.403 g; 10.1 mmol) was dissolved in methanol and stirred while [2,l'-13C2]diethyl 2-ethylmalonate (0.915 g; 4.81 mmol) was added dropwise. This mixture was heated at reflux for 16 hours after which time the pH was slightly basic. The solvent was removed by rotary evaporation to yield a solid white mixture of residual sodium hydroxide and [2,1- C2]disodium 2-ethylmalonate, which was characterized without further purification: ! H NMR (400 MHz, D 20) 8 0.89 (m, H-2', 3H), 1.78 (dm, 7=129.9, H-1', 2H), 3.21 (dm, 7=131.8, H-2, 1H); 1 3C NMR (100 MHz, D 20) 8 14.4 (d, 7=34.2, C-2'), 25.3 (d, 100X intensity, 7=31.8, C-l'), 59.7 (d, 100X intensity, 7=34.1, C-2), 177.4 (d, 7=54.8, C-l); HRFABMS (m/z): [M+H]+ calcd for C3H70413C2Na2, 179.02074; found 179.02055. [2,3-13C2]Butyric acid: The mixture of [2,l'-13C2]disodium 2-ethylmalonate and residual sodium hydroxide was treated with 50 mL of 0.4 N HC1 at reflux for three days. The decarboxylation product was isolated by extraction with diethyl ether (3 X 100 mL) and the combined ethereal extracts dried over anhydrous sodium sulfate. Solvent removal afforded [2,3-13C2]butyric acid as a clear, colourless and pungent oil (0.360 g; 4.00 mmol) in 83.2 % yield over the last two steps. [2,3-13C2]Butyric acid: 'H NMR (400 MHz, CDC13) 8 0.96 (m, H-4, 3H), 1.65 (d-sextet-d, 7=128.9, 7.4,4.6, H-3, 2H), 2.32 (dtd, 7=127.8, 7.3,4.6, H-2, 2H); 1 3C NMR (100 MHz, CDC13) 8 13.6 (d, 7=34.3, C-4), 18.1 (d, 100X intensity, 7=34.3, C-3), 22.4 (d, 100X intensity, 7=32.1, C-2), 179.9 (d, 7=55.7, C-l); HRFABMS (m/z): [M-H]+ calcd for C 2H 702 1 3C2, 89.05132; found 89.05166. [2,3-13C2]Sodium butyrate: [2,3-13C2]Butyric acid (0.360 g; 4.00 mmol) was converted to [2,3-l3C2]sodium butyrate by titration with 1.0 mol/L sodium bicarbonate solution to pH 7 - 8. 191 The solution was stirred to allow for the loss of the carbon dioxide and then lyophilized to afford 0.242 g (2.16 mmol) of a solid white powder, [2,3- 13C2]sodium butyrate in 54.1 % yield: 'H NMR (400 MHz, CDC13) 8 0.77 (m, H-4, 3H), 1.42 (dm, 7=125.0, H-3, 2H), 2.00 (dm, 7=125.0, H-2, 2H); 1 3 C NMR (100 MHz, CDC13) 8 13.94 (d, 7=34.3, C-4), 20.05 (d, 100X intensity, 7=33.2, C-3), 40.34 (d, 100X intensity, 7=33.2, C-2), 185.0 (d, 7=55.0, C-l); HRFABMS (m/z): [M+Na]+ calcd for C 2H 702 1 3C2Na2, 135.03091; found 135.03085. [l,3-13C2]Sodium isobutyrate: To a 0 °C solution of diisopropylamine (4.02 mL; 30.6 mmol) in 100 mL of dry THF stirring under argon was added by syringe 19.2 mL of a 1.6 mol/L solution of n-butyllithium (30.6 mmol) in hexanes. HMPA (6.95 mL; 40.0 mmol) was then added and the solution was cooled to -78 °C. [l-13C]Propionic acid (1.00 g; 13.3 mmol) was added by syringe very slowly, and the reaction mixture was warmed to 0 °C over one hour. After that time the mixture was cooled back down to -78 °C and [13C]methyl iodide (0.92 mL; 14.7 mmol) was added by syringe, turning the solution colourless from pale yellow. The reaction vessel was warmed to 0 °C and stirred for 30 minutes. The initial work-up consisted of the addition of 2 mL of undistilled diethyl ether (water in the diethyl ether quenched any remaining LDA), followed by removal of solvent by rotary evaporation and removal of most of the HMPA by vacuum distillation. The remaining contents of the stillpot (including [l,3-13C2]lithium isobutyrate) were dissolved in 50 mL of water and titrated with IN HC1 (18 mL) to slightly acidic pH. The organic materials were extracted with diethyl ether (3 X 125 mL) and dried over anhydrous sodium sulfate. Removal of solvent in vacuo provided a slightly pungent, yellow oil, which was purified by vacuum Kugelrohr distillation (< 1 mm Hg; to 60 °C with -78 °C trap) yielding a clear, colourless, slightly pungent oil identified as [l,3-13C2]isobutyric acid (1.20 g) in 100 % yield: 'H NMR (400 MHz, CDC13) 81.16 (dm, 7=7.2, H-4, 3H), 1.16 (ddd, 7=128.3, 7.2, 5.4, H-3, 3H), 2.54 (septet-m, 7=7.2, H-2, 1H), 9.4 (br s, OH); 1 3C NMR (100 MHz, CDC13) 8 192 18.7 (d, 100X intensity, 7=1.5, C-3), 18.7 (C-4), 33.8 (dd, 7=54.9, 34.3, C-2), 183.5 (d, 100X intensity, 7=1.5, C-l); HREXMS (m/z): M + calcd for C 2 H 8 02 I 3 C 2 ) 90.05914; found 90.05953. [l,3-13C2]Isobutyric acid was dissolved in 5 mL of methanol, diluted with 20 mL of water, titrated to neutral pH with 1.0 N NaOH, and finally lyophilized to yield [l,3-13C2]sodium isobutyrate (1.00 g; 8.94 mmol; 67.1 % yield from [l-13C]propionic acid). Collection of specimens For each experiment, 30 specimens of Triopha catalinae were collected by hand using SCUBA from rocky reefs in Barkley Sound, British Columbia, Canada at -2 m to -15 m. Those intended for control experiments were submersed in methanol immediately upon their removal from the ocean water. Collections meant for incorporation studies were transferred to holding tanks onshore with circulating seawater. Feeding of 1 3C labelled precursors: Each specimen of Triopha catalinae was subjected to dorsal injections into the hepatopancreas of 0.55 mol/L precursor (100 - 200 pL per injection per nudibranch) in distilled water using a syringe equipped with a 21G2 needle, on two occasions (four and 20 hours after collection). An incubation period of seven days followed, during which time the nudibranchs were left undisturbed in the aquarium. After this time, the specimens were immersed in methanol with minimal handling to avoid shedding of metabolite. Three different feeding experiments were conducted, using [l-13C]sodium butyrate, [2,l'-13C2]disodium 2-ethylmalonate, and [2,3- C2]sodium butyrate. Feeding of L-[U-14C6]arginine as precursor: Each of 30 specimens of Triopha catalinae were injected twice (after four and 20 hours from collection) with 100 pL of a solution that contained 1.85 MBq L-[U-14Ce]arginine monohydrochloride in 6 mL of water (0.5 % ethanol) and also 0.55 mol/L unlabelled L-arginine and sodium acetate (0.275 mol/L in each). After an 193 incubation period of seven days during which the nudibranchs were left undisturbed in aquarium, they were immersed quickly in methanol. Isolation of triophamine (43): The same isolation procedure was followed for each separate experiment (one control, three 1 3C labelling, one 1 4C labelling). Specimens were extracted with methanol (2 X 250 mL) and methanol/dichloromethane (1:1) (4 X 250 mL). Organic extracts were combined, reduced in vacuo, and then partitioned between water (200 mL) and ethyl acetate (200 mL X 4). The ethyl acetate extracts were combined, dried over anhydrous magnesium sulfate, and the solvent removed to afford 0.5 - 0.7 g of a dark yellow oil. Flash silica gel column chromatography using hexanes/ethyl acetate 17:3 as eluent was used to separate components of the extract. Triophamine (43) was identified in column fractions by the presence of a UV-active spot on normal phase silica gel TLC. Combined fractions containing 43 were then subjected to reversed phase silica HPLC (methanol/water 4:1 eluent) to afford pure 43 as a white crystalline solid (8-12 mg per 30 specimens of Triopha catalinae): 'H NMR (CDCI3) 8 0.89 (t, 7=7.5, Me-10, 10', 6H), 0.94 (t, 7=7.5, Me-8, 8', 6H), 1.50 (m, H-9b, 9b', 2H), 1.54 (d, 7=6.8, Me-6, 6', 6H), 1.59 (m, H-9a, 9a', 2H), 2.00 (q, 7=7.5, H-7a, 7b, 7a', 7b', 4H), 2.11 (m, H-3b, 3b', 2H), 2.30 (m, H-2, 2', 3a, 3a', 4H), 5.19 (q, 7=6.8, H-5, 5', 2H), 8.0 - 10.0 (br s, NH); 1 3C NMR see Table 2.13; HREIMS (m/z): M + calcd for C21H37N3O2, 363.28857; found, 363.28830 (control experiment), 363.28857 ([l-13C]butyrate experiment), 363.28810 ([2,1'-13C2]ethylmalonate experiment, 363.28890 ([2,3-13C2]butyrate experiment). Isolation of triophamine (43) from L-[U-14C6]arginine labelling experiment: The isolation procedure was applied to the radioactive extract of Triopha catalinae as above, with the following deviation. Following the HPLC purification of 43 yielding 9.3 mg, an aliquot representing 2 % of the total amount was subjected to scintillation counting, which measured 238 DPM (decays per minute). Two further rounds of HPLC purification yielding 2.8 and 2.2 mg 43 respectively each resulted in scintillation counts of 39 and 33 DPM from 2 % aliquots. These counts were measured against a background of 29 DPM. Pure samples of 43 isolated in this experiment were identified by comparison of TLC Rf and HPLC retention times with authentic samples of triophamine (43). 5.6 New metabolites of P. villatus and C. lepadiformis Two hundred specimens of P. villatus flatworm and 13 g of C. lepadiformis colonial ascidian were collected by hand using SCUBA at -3 to -20 m off Storekinna Island near Bergen, Norway. The ascidians appeared as translucent sheaths through which yellow viscera were visible. Each member of a colony measured 1 - 2 cm in length, and was affixed to the rock face at a common point for the entire colony. The flatworms, 1 - 5 cm long and 0.5 - 3 cm wide, white with brown stripes running longitudinally, were found either on nearby rocks or feeding directly on the ascidians. Both were collected into mesh bags and immersed in methanol immediately upon removal from the sea. Jars were stored at -15 °C until isolation work began. Extracts of P. villatus and C. lepadiformis were treated according to the same isolation procedure. The contents of each jar of organisms were extracted with methanol (2 X), followed by methanol/dichloromethane 1:1 (3 X). The combined extracts were concentrated in vacuo and then partitioned between hexane (300 mL) and methanol/water 9:1 (300 mL). The methanol/water phase was further partitioned between carbon tetrachloride (300 mL) and methanol/water 4:1 (450 mL), and then between chloroform (2 X 300 mL) and methanol/water 3:2 (600 mL). The chloroform soluble extract was concentrated and separated according to molecular 195 weight using Sephadex LH-20 (1.5 m long, 2 cm diameter) with methanol as eluent. Fractions were pooled according to similar Rf by UV detection on reversed phase silica or alumina TLC. The earlier-eluting of two resulting fractions was further separated by LH-20 (1.5 m long, 3 cm diameter) using as eluent system EtOAc/methanol/water 20:5:2, resulting in four column fractions. Each of these was then separated by reversed phase silica flash column (10 cm long, 2 cm diameter) using a gradient eluent system of methanol/water 1:4 to methanol to EtOAc. Five compounds were isolated by these means and finally purified as their trifluoroacetate salts by reversed-phase silica HPLC, using the following eluent systems: acetonitrile/water(0.05 % TFA) 7:13 for lepadins A (66), C (74) and villatamine A (75); methanol/water(0.05 % TFA) 3:2 for lepadin B (73) and villatamine B (76). Lepadin A (66): clear colourless oil (0.2 mg per flatworm; 0.2 % dry weight); NMR spectroscopic data matched literature values129 (see Table 401); UV ^ m a x 234 nm (e = 16500); IR 1737 cm"1; HREIMS (m/z): M + calcd for C20H33NO3, 335.2460; found, 335.2453. Lepadin B (73): clear colourless oil (0.04 mg per flatworm; 0.05 % dry weight); [a]o25 = -96°, (c = 0.36, MeOH); NMR data see Table 402; UV Xm!a 232 (e = 17700); IR 3371 br; 1590 cm"1; HREIMS (m/z): M + calcd for C1gH3 .NO, 277.2406; found, 277.2407. Lepadin C (74): clear colourless oil (0.02 mg per flatworm; 0.02 % dry weight); [a]o25 = -25°, (c= 0.4, MeOH); NMR data see Table 403; UV k m a x 232 (e =16400); IR 3371 br, 1743, 1612 cm"1; HREIMS (m/z): M + calcd for C20H31NO4, 349.22531; found, 349.22566. Villatamine A (75): clear colourless oil (0.02 mg per flatworm; 0.02 % dry weight); [a]D 2 5 = 49°, (c = 0.1, MeOH); NMR data see Table 404; UV k m a x 266 (e =18700), 278 (8 = 14000); IR 1641, 1600 cm"1; HRDCIMS (m/z): [M+H]+ calcd for Ci 8 H 3 0 N, 260.23783; found, 260.23723. Villatamine B (76): clear colourless oil (0.09 mg per flatworm; 0.09 % dry weight); 196 [cc]D25 = 49°, (c = 2, MeOH); NMR data see Table 405; UV ? i m a x 230 (e = 19500); IR 3413 cm"1 (br); HRDCIMS (m/z): [M+H]+ calcd for Ci 8 H 3 4 N, 264.26913; found, 264.26879. In order to assist in the structure determination of the villatamines, reductive ozonolysis of 75 and 76 was required. Villatamine A (75) (0.3 mg) and villatamine B (76) (1.4 mg) were each dissolved in methanol and cooled to -78 °C. Ozone was bubbled through each solution for one minute, after which time a blue colour was observed. Oxygen was then bubbled through until each solution returned to a colourless state. Dimethyl sulfide (0.5 mL) was then added and the reaction mixture allowed to warm to room temperature while stirring. Water (15 mL) was added to the methanol solution, and the aldehyde products were extracted with diethyl ether (5 mL). A light stream of nitrogen gas was used to reduce the ether volume to 1 mL. These crude product solutions were then subjected to GC using a DB-1 column and their chromatograms compared with authentic samples of straight-chain aldehydes to determine the size of aldehydes cleaved from 75 and 76 by ozonolysis. Comparison of GC chromatograms and co-injection of products of 75 and 76 ozonolysis with pentanal, hexanal and heptanal, showed that the product of ozonolysis of 75 was pentanal, and that the product of ozonolysis of 76 was heptanal. These results were confirmed by GCMS: Pentanal from 75: EIMS (m/z) (relative intensity): 86.1 (8), 71.1 (5), 59 (72), 58 (100), 57 (55), 44 (97), 40(15), 29 (5). Heptanal from 76: EIMS (m/z) (relative intensity): 114 (5), 96 (18), 86 (20), 81 (24), 70 (100), 57 (40), 55 (45), 44 (42), 28 (4). 197 Figure A . l . HMQC spectrum of seco-spongian 22 in CDC13 (recorded at 500 MHz) Figure A.2. HMBC spectrum of seco-spongian 22 in CDC13 (recorded at 500 MHz) 199 Figure A.3. HMBC spectrum of seco-spongian 22 in CDC13 (recorded at 500 MHz) C l C2 C3 C4 C5 M M I I I M I 1 M M I I I M I I I I I I I M I I M M I I I M I I M I M I M I 38.9 19.0 41.8 33.3 55.0 C6 C7 C8 V l l l l l l l l l I I I I I I I I I I I M I M I M I I M I M I M I M I I I I I I I I 23.8 37.5 146.6 54.6 38.8 C l l C12 i M i I I M i i 61.4 C13 M i i i i i M i i i i i I I i i i 107.0 33.5 i i i i i i i i i i 21.6 C16 C17 I I I I I I I I I I I I I I I I I I M 171.0 20.9 Figure A.5. 13C NMR resonances from the spectrum of albicanyl acetate (18) (CDC1 125 MHz) incorporated with [l,2-13C2]acetate 203 Figure A.7. HMQC spectrum of lepadin B (73) in CDC13 (recorded at 500 MHz) 204 ,OH 1 NH 9 73 J U J U L 0 0 ? e 0 o o 6 0 0 9 o 6 1 0 ' 8 ' c 4 * 0 8 SO 120 160 (pom) 10 Figure A.8. HMBC spectrum of lepadin B (73) in CDC13 (recorded at 500 MHz) 1 205 Figure A. 10. HMQC spectrum of lepadin C (74) ih CDC13 (recorded at 500 MHz) Figure A . l 1. HMBC spectrum of lepadin C (74) in CDC13 (recorded at 500 MHz) Figure A. 12. HMQC spectrum of villatamine A (75) in CD2C12 (recorded at 500 MHz) Figure A. 13. COSY spectrum of villatamine A (75) in CD2C12 (recorded at 400 MHz) 210 Figure A.14. H M B C spectrum of villatamine A (75) in CD 2 C1 2 (recorded at 500 MHz) Figure A. 15. HMBC spectrum of villatamine A (75) in CD2C12 (recorded at 500 MHz) 76 Figure A. 16. COSY spectrum of villatamine B (76) in CDC13 (recorded at 400 MHz) 213 214 Figure A. 18. HMBC spectrum of villatamine B (76) in CDC13 (recorded at 500 MHz) References and Notes 215 Also termed secondary metabolites: metabolites not participatory in primary metabolism. 2 Faulkner, D.J. Nat. Prod. Rep. 1997,14, 259, and previous reviews cited therein. 3 e.g. Hamel, E. Med. Res. Rev. 1996,16, 207. 4 Hay, M.E. J. Exp. Mar. Biol. Ecol. 1996, 200, 103. 5 Krebs, C.J. Ecology; Third edition; Harper and Row: New York, 1985. 6 Pawlik, J.R. Chem. Rev. 1993, 93, 1911. 7 Bakus, G.J.; Schulte, B.; Jhu, S.; Wright, M.; Green, G.; Gomez, P. Antibiosis and Antifouling in Marine Sponges: Laboratory Versus Field Studies. In New Perspectives in Sponge Biology; Rutzler, K., Ed.; Smithsonian Institution Press: Washington, D.C., 1990; pp. 102-108. 8 Pawlik, J.R. Oceanogr. Mar. Biol. Annu. Rev. 1992, 30, 273. 9 Pawlik, J.R. Induction of Marine Invertebrate Larval Settlement: Evidence for Chemical Cues. In Ecological Roles of Marine Natural Products; Paul, V.J., Ed.; Comstock Publishing Associates: Ithaca, N.Y., 1992; pp. 189-236. 1 0 Hay, M.E.; Fenical, W. Oceanography 1996, 9, 10. 1 1 Hay, M.E.; Pawlik, J.R.; Duffy, J.E.; Fenical, W. Oecologia 1989, 81, 418. 1 2 Gil-Turnes, M.S.; Hay, M.E.; Fenical, W. Science 1989,245,116. 1 3 Hay, M.E.; Steinberg, P.D. The Chemical Ecology of Plant-Herbivore Interactions in Marine versus Terrestrial Communities. In Herbivores: Their Interactions with Secondary Plant Metabolites; Rosenthal, G.A.; Berenbaum, M.R., Eds.; Academic Press: San Diego, 1992; Volume 2, pp. 372-413. 1 4 Steinberg, P.D. Geographical Variation in the Interaction Between Marine Herbivores and Brown Algal Secondary Metabolites. In Ecological Roles of Marine Natural Products; Paul, V.J., Ed.; Comstock Publishing Associates: Ithaca, N.Y., 1992; pp. 51-92. 1 5 Faulkner, D.J. Nat. Prod. Rep. 1984, 7, 251. 1 6 Trench, R.K.; Greene, R.W.; Bystrom, B.G. J. Cell Biol. 1969,42,404. 1 7 Ireland, C; Scheuer, P.J. Science 1979, 205, 922. 216 Cimino, G.; Sodano, G. Transfer of Sponge Secondary Metabolites to Predators. In Sponges in Time and Space; van Soest, R.W.M.; van Kempen, Th.M.G.; Braekman, J.C, Eds.; A.A. Balkema Publishers: Rotterdam, Netherlands, 1994; pp. 459-472. 1 9 Schulte, G.; Scheuer, P.J.; McConnell, O.J. Helv. Chim. Acta 1980, 63, 2159. 2 0 Pawlik, J.R.; Kernan, M.R.; Molinski, T.F.; Harper, M.K.; Faulkner, D.J. J. Exp. Mar. Biol. Ecol. 1988,779,99. 2 1 Proksch, P. Toxicon. 1994, 32, 639. 2 2 Karuso, P. Chemical Ecology of the Nudibranchs. In Bio-organic Marine Chemistry; Scheuer, P.J., Ed.; Springer-Verlag: Berlin, 1987; Volume 1, pp. 31-60. 2 3 Thompson, J.E.; Walker, R.P.; Wratten, S.J.; Faulkner, D.J. Tetrahedron 1982, 38, 1865. 2 4 Faulkner, D.J.; Ghiselin, M.T. Mar. Ecol. Prog. Ser. 1983,13, 295. 2 5 Kobayashi, J.; Ishibashi, M. Chem. Rev. 1993, 93, 1753. 2 6 Oclarit, J.M.; Okada, H.; Ohta, S.; Kaminura, K.; Yamaoka, Y.; Iizuka, T.; Miyashiro, S.; Ikegami, S. Microbios 1994, 78,1. 2 7 Fenical, W.; Jensen, P.R. Marine Microorganisms: A New Biomedical Resource. In Marine Biotechnology; Attaway, D.H.; Zaborsky, O.R., Eds.; Plenum Press: New York, 1993; Volume 1, pp 419-457. 2 8 Unson, M.D.; Faulkner, D.J. Experientia 1993, 49, 349. 2 9 (a) Pettit, G.R.; Cichacz, Z.A.; Gao, F.; Herald, c.L.; Boyd, M.R.; Schmidt, J.M.; Hooper, J.N.A. J. Org. Chem. 1993, 58, 1302; (b) Pettit, G.R.; Cichacz, Z.A.; Gao, F.; Herald, C.L.; Boyd, M.R. Chem. Commun. 1993, 1166; (c) Pettit, G.R.; Herald, C.L.; Cichacz, Z.A.; Gao, F.; Schmidt, J.M.; Boyd, M.R.; Christie, N.D.; Boettner, F.E. Chem. Commun. 1993, 1805; (d) Pettit, G.R.; Herald, C.L.; Cichacz, Z.a.; Gao, F.; Boyd, M.R.; Christie, N.D.; Schmidt, J.M. Nat Prod. Lett. 1993, 3, 239; (e) Pettit, G.R.; Cichacz, Z.A.; Herald, C.L.; Gao, F.; Boyd, M.R.; Schmidt, J.M.; Hamel, E.; Bai, R. Chem. Commun. 1994, 1605; (f) Kobayashi, M.; Aoki, S.; Sakai, H.; Kawazoe, K.; Kihara, N.; Sasaki, T.; Kitagawa, I. Tetrahedron Lett. 1993, 34, 2795; (g) Fusetani, N.; Shinoda, K.; Matsunaga, S. J. Am. Chem. Soc. 1993,775,3977. 3 0 Hooper, J.N. A. In Guide to Sponge Collection and Identification; World Wide Web, 1996. 3 1 Simidu, U.; Kita-Tsukamoto, K.; Yasumoto, T.; Yotsu, M. Int. J. System. Bacteriol. 1990, 40, 331. 3 2 Garson, M.J. Nat. Prod. Rep. 1989, 6, 143. 3 3 Garson, M.J. Chem. Rev. 1993, 93, 1699. 217 3 4 (a) Rohmer, M.; Knani, M.; Simonin, P.; Sutter, B.; Sahm, H. Biochem. J. 1993, 295, 517; (b) Rohmer, M.; Seemann, M.; Horbach, S.; Bringer-Meyer, S.; Sahm, H. J. Am. Chem. Soc. 1996,118, 2564. 3 5 Eisenreich, W.; Menhard, B.; Hylands, P.J.; Zenk, M.H.; Bacher, A. Proc. Natl. Acad. Sci. USA 1996,93,6431. Cimino, G.; Sodano, G. Biosynthesis of Secondary Metabolites in Marine Molluscs. In Topics in Current Chemistry; Scheuer, P.J., Ed.; Springer-Verlag: Berlin, 1993; Volume 167, pp.77-115. 3 7 Faulkner, D.J. Chemical Defenses of Marine Molluscs. In Ecological Roles of Marine Natural Products; Paul, V.J., Ed.; Comstock Publishing Associates: Ithaca, N.Y., 1992; pp. 119-163. 38 Thompson, T.E. J. Mar. Biol. Ass. U.K. 1960, 39, 115. 3 9 Thompson, T.E. J. Mar. Biol. Ass. U.K. 1960, 39, 123. 4 0 Burreson, B.J.; Clardy, J.; Finer, J.; Scheuer, P.J. J. Am. Chem. Soc. 1975, 97, 4763. 4 1 Hagedone, M.R.; Burreson, B.J.; Scheuer, P.J.; Finer, J.S.; Clardy, J. Helv. Chim. Acta 1979, 62, 2484. 4 2 Johannes, R.E. Veliger 1963, 5, 104. 4 3 Cimino, G.; de Rosa, S.; de Stefano, S.; Sodano, G.; Villani, G. Science 1983, 279, 1237. 4 4 Cimino, G.; de Rosa, S.; de Stefano, S.; Morrone, R.; Sodano, G. Tetrahedron 1985, 41, 1093. 4 5 Cimino, G.; Sodano, G.; Spinella, A. J. Nat. Prod. 1988, 57, 1010. 4 6 Avila, C ; Cimino, G.; Crispino, A.; Spinella, A. Experientia 1991, 47, 306. 4 7 Gustafson, K.; Andersen, R.J. Tetrahedron 1985,47,1101. 48 Graziani, E.L; Andersen, R.J.; Krug, P.J.; Faulkner, D.J. Tetrahedron 1996, 52, 6869. 4 9 Graziani, E.L; Andersen, R.J. /. Am. Chem. Soc. 1996,118, 4701. 5 0 Graziani, E.L; Andersen, R.J. Chem. Commun. 1996, 2377. 5 1 Faulkner, D.J.; Molinski, T.F.; Andersen, R.J.; Dumdei, E.J.; de Silva, E.D. Comp. Biochem. Physiol. 1990, 97C, 233. Behrens, D.W. Pacific Coast Nudibranchs; Second edition; Sea Challengers: Los Osos, CA, 1991. 218 5 3 Hellou, J.; Andersen, R.J.; Rafii, S.; Arnold, E.; Clardy, J. Tetrahedron Lett. 1981,22, 4173. 5 4 Thompson, J.E.; Walker, R.P.; Wratten, S.J.; Faulkner, D.J. Tetrahedron 1982, 38, 1865. 5 5 Hellou, J.; Andersen, R.J.; Thompson, J.E. Tetrahedron 1982,35, 1875. 5 6 Gustafson, K.; Andersen, R.J.; Cun-heng, H.; Clardy, J. Tetrahedron Lett. 1985, 26, 2581. 5 7 Tischler, M.; Andersen, R.J.; Tetrahedron Lett. 1989, 30, 5717. 5 8 Tischler, M.; Andersen, R.J.; Choudhary, M.L; Clardy, J. J. Org. Chem. 1991, 56, 42. 5 9 Burgoygne, D.L.; Dumdei, E.J.; Andersen, R.J. Tetrahedron 1993, 49, 4503. 6 0 Dumdei, E.J.; Kubanek, J.; Coleman, J.E.; Pika, J.; Andersen, R.J.; Steiner, J.R.; Clardy, J. Can. J. Chem. 1997, 75, 773. 61 Schmitz, F.J.; Chang, J.S.; Hossain, M.B.; van der Helm, D. J. Org. Chem. 1985, 50, 2862. see Appendix Figures A . l , A.2, A.3, and A.4 for representative 2D NMR spectra 6 3 see Appendix Figure A.5 for all 1 3 C NMR resonances from incorporation experiment 6 4 this would be expected to occur with retention of the R-O bond, as has been previously demonstrated for certain alkaline phosphatases: e.g. Croteau, R. Biosynthesis and Catabolism of Monoterpenes. In Isopentenoids in Plants; Nes, D.W.; Fuller, G.; Tsai, L.S., Eds.; Marcel Dekker Inc.: New York, 1984; pp. 31-64. 6 5 Vederas, J. Nat. Prod. Rep. 1987, 4, 277. lalinowski, H.O.; New York, 1988 6 6 K . . Berger, S.; Braun, S. Carbon-13 NMR Spectroscopy; John Wiley & Sons: 6 7 Kubanek, J.; Graziani, E.I.; Andersen, R.J. J. Org. Chem. 1997, 62, 7239. 6 8 Riley, J.P.; Chester, R. Introduction to Marine Chemistry; Academic Press: London, 1971. 6 9 With assistance from SIO personnel, the specimens were collected before the author arrived to conduct the chemical studies, resulting in a seven day delay. sketch of Cadlina nudibranch taken from: Thompson, T.E. Biology of Opistobranch Molluscs; The Ray Society: London, 1984; Volume 2. 71 Avila, C. Oceanography and Marine Biology: an Annual Review 1995,53,487. 7 2 Avila, C ; Durfort, M. Veliger 1996,39, 148, and references cited therein. 7 3 Kubanek, J.; Andersen, R.J. Tetrahedron Lett. 1997, 38, 6327. 70 219 4 Ayer, S.W.; Hellou, J.; Tischler, M.; Andersen, R.J. Tetrahedron Lett. 1984, 25, 141. 5 Ayer, S.W.; Andersen, R.J.; Cun-heng, H.; Clardy, J. J. Org. Chem. 1984,49, 2653. 6 Engler, T.A.; Ali, M.H.; Takusagawa, F. J. Org. Chem. 1996, 61, 8456. 7 Graziani, E.I. Isolation, Structure Determination and Biosynthetic Studies of Secondary Metabolites from Dorid Nudibranchs, University of British Columbia, Vancouver, B.C. Canada, 1996. 8 Based upon observations of collecting teams in laboratory of R.J. Andersen, and personal communication with Dr. S. Millen, University of British Columbia. 9 Fenical, W.; Sleeper, H.L.; Paul, V.J.; Stallard, M.O.; Sun, H.H. Pure and Appl. Chem. 1979, 57, 1865. 0 Griffith, W.P.; Ley, S.V.; Whitcombe, G.P.; White, A.D. Chem. Commun. 1987, 1625. 1 Baggiolini, E.G.; Iacobelli, J.A.; Hennessy, B.M.; Batcho, A.D.; Sereno, J.F.; Uskokovic, M.R. J. Org. Chem. 1986, 57, 3098. 2 Davisson, V.J.; Woodside, A.B.; Neal, T.R.; Stremler, K.E.; Muehlbacher, M.; Poulter, CD. J. Org. Chem. 1986, 57,4768. 3 March, J. Advanced Organic Chemistry; Third Edition; John Wiley and Sons: New York, 1985. 14 Ruggeri, R.B.; Hansen, M.M.; Heathcock, CH. J. Am. Chem. Soc. 1988, 770, 8734. 5 Borthwick, E.B.; Connell, S.J.; Tudor, D.W.; Robins, D.J.; Schneier, A.; Abell, C; Coggins, J.R. Biochem. J. 1995, 305, 521. 16 Piersen, CE.; Prasad, R.; Wilson, S.H.; Lloyd, R.S. J. Biol. Chem. 1996, 277, 17811. 17 Fisher, H.F.; Viswanathan, T.S. Proc. Natl. Acad. Sci. USA 1984, 81; 2747. 8 Nakamura, N.; Moenne-Loccoz, P.; Tanizawa, K.; Mure, M.; Suzuki, S.; Klinman, J.; Sanders-Loehr, J. Biochemistry 1997,56,11479. 19 Williams, D.E.; Ayer, S.W.; Andersen, R.J. Can. J. Chem. 1986, 64, 1527. 0 (a) Fattorusso, E.; Lanzotti, V.; Magno, S.; Novellino, E. J. Org. Chem. 1985, 50, 2868. (b) Fattorusso, E.; Lanzotti, V.; Magno, S.; Novellino, E. J. Nat. Prod. 1985, 48, 784. (c) Aiello, A.; Fattorusso, E.; Magno, S. J. Nat. Prod. 1987, 50, 191. 11 Fukuzawa, A.; Kumagai, Y.; Masamune, T.; Furusaki, A.; Katayama, C ; Matsumoto, T. Tetrahedron Lett. 1981,22,4085. 2 Walker, R.P.; Faulkner, D.J. J. Org. Chem. 1981, 46, 1475. 220 9 3 Voogt, P.A. Int. J. Biochem. 1973, 4, 479. 9 4 Kerr, R.G.; Baker, B.J. Nat. Prod. Rep. 1991, 8 ,465, and references cited therein. 9 5 Goad, L.J. The Sterols of Marine Invertebrates: Composition, Biosynthesis, and Metabolites. In Marine Natural Products; Scheuer, P.J., Ed.; Academic Press: New York, 1978; Volume 2, pp. 75-172. 9 6 Castiello, D.; Cimino, G.; de Rosa, S.; de Stefano, S.; Izzo, G.; Sodano, G. Colloq. Int. C.N.R.S. 1979,297,413. 9 7 Andersen, R.J.; Ayer, S.W. Tetrahedron Lett. 1982,23, 1039. 9 8 Voogt, P.A. Lipids: Their Distribution and Metabolism. In The Mollusca; Hochachka, P.W., 99 Ed.; Academic Press: New York, 1983; Volume 1, pp. 329-370. Djerassi, C ; Silva, C.J. Acc. Chem. Res. 1991, 24, 371. 100 101 102 Silva, C.J.; Wunsche, L.; Djerassi, C. Comp. Biochem. Physiol. 1991, 99B, 763. Gustafson, K.; Andersen, R.J. J. Org. Chem. 1982, 47, 2167. Piers, E.; Chong, J.M.; Gustafson, K.; Andersen, R.J. Can. J. Chem. 1984, 62, 1. 1 0 3 (a) Westley, J.W.; Evans, R.H.; Harvey, G.; Pitcher, R.G.; Pruess, D.L.; Stempel, A.; Berger, J. J. Antibiotics 1974, 27, 288; (b) Hutchinson, C.R.; Sherman, M.M.; Vederas, J.C; Nakashima, T.T. J. Am. Chem. Soc. 1981,103, 5953. 1 0 4 Day, L.E.; Chamberlin, J.W.; Gordee, E.Z.; Chen, S.; Gorman, M.; Hamill, R.L.; Ness, T.; Weeks, R.E.; Stroshane, R. Antimicrobial Agents and Chemotherapy 1973,4, 410. 1 0 5 Higgs, M.D.; Faulkner, D.J. J. Org. Chem. 1978,43, 3454. 1 0 6 Stierle, D.B.; Faulkner, D.J. J. Org. Chem. 1980, 45, 3396. 1 0 7 Gunasekera, S.P.; Gunasekera, M.; Gunawardana, G.P.; McCarthy, P.; Burres, N. J. Nat. Prod. 1990, 53, 669. 1 0 8 D'Auria, M.V.; Minale, L.; Riccio, R.; Zampella, A.; Debitus, C. J. Nat. Prod. 1993, 56, 418. 1 0 9 Stierle, D.B.; Faulkner, D.J. /. Org. Chem. 1979,44, 964. 1 1 0 (a) Berlinck, R.G.S. Nat. Prod. Rep. 1996, 13, 377. (b) Berlinck, R.G.S. Prog. Chem. Org. Nat. Prod. 1995, 66, 119, and references cited therein. 1 1 1 Schantz, E.J.; Chazarossian, V.E.; Schnoes, H.K.; Strong, F.M.; Springer, J.P.; Pezzanite, J.O.; Clardy, J. J. Am. Chem. Soc. 1975, 97, 1238. 221 1 1 2 Goto, T.; Kishi, Y.; Takahashi, S.; Hirata, Y. Tetrahedron 1965,27, 2059. 1 1 3 Shimizu, Y. Pure and Appl. Chem. 1986, 58, 257. 1 1 4 Harbour, G.C.; Tymiak, A.A.; Rinehart, K.L.; Shaw, P.D.; Hughes, RG.; Mizsak, S.A.; Coats, J.H.; Zurenko, G.E., Li, L.H. Kuentzel, S.L. J. Am. Chem. Soc. 1981,103, 5604. 1 1 5 Moore, B.S.; Eisenberg, R.; Weber, C ; Bridges, A.; Nanz, D.; Robinson, J.A. J. Am. Chem. Soc. 1995,777, 11285. 1 1 6 Sherman, M.M.; Yue, S.; Hutchinson, CR. J. Antibiot. 1986, 39, 1135. 1 1 7 Pospisil, S.; Sedmera, P.; Havranek, M.; Krumphanzl, V.; Vanek, Z. J. Antibiot. 1983, 36, 617. 1 1 8 Omura, S.; Tsuzuki, K.; Tanaka, Y.; Sakakibara, H.; Aizawa, M.; Lukacs, G. J. Antibiot. 1983, 36, 614. 1 1 9 (a) Barnes, R.D. Invertebrate Zoology; Saunders College Publishing: Philadelphia, 1987. (b) Morris, R.H. Intertidal Invertebrates of California; Stanford Press: Stanford, 1980. 1 2 0 Kobayashi, J.; Ishibashi, M. Chem. Rev. 1993, 93, 1753. 1 2 1 Kobayashi, J.; Sato, M.; Ishibashi, M. J. Org. Chem. 1993, 58, 2645. 1 2 2 Davidson, B.S. Chem. Rev. 1993,93,1771. 1 2 3 Rinehart, K.L.; Gloer, J.B.; Cook, J.C; Mizsak, S.A.; Scahill, T.A. J. Am. Chem. Soc. 1981, 103, 1857. 1 2 4 (a) Wright, A.E.; Forleo, D.A.; Gunawardana, G.P.; Gunasekera, S.P.; Koehn, F.E.; McConnell, O.J. J. Org. Chem. 1990, 55, 4508. (b) Rinehart, K.L.; Holt, T.G.; Fregeau, N.L.; Stroh, J.G.; Keifer, P.A.; Sun, F.; Li, L.H.; Martin, D.G. J. Org. Chem. 1990, 55, 4512. (c) Rinehart, K.L.; Holt, T.G.; Fregeau, N.L.; Stroh, J.G.; Keifer, P.A.; Sun, F.; Li, L.H.; Martin, D.G. J. Org. Chem. 1991, 56, 1676. 1 2 5 Shen, G.Q.; Baker, B.J. Tetrahedron Lett. 1994,35,1141. 1 2 6 Smith, M.J.; Kim, D.; Horenstein, B.; Nakanishi, K. Acc. Chem. Res. 1991,24, 117. 1 2 7 He, X.; Kustin, K.; Parry, D.L.; Robinson, W.E.; Ruberto, G.; Nakanishi, K. Experientia 1992,48,367 128Steffan, B.;Brix, K.;Putz,W. Tetrahedron 1993,49,6223. 1 2 9 Steffan, B. Tetrahedron 1991,47,8729 1 3 0 Biard, J.F.; Guyot, S.; Roussakis, C ; Verbist, J.F.; Vercauteren, J.; Weber, J.F.; Boukef, K. Tetrahedron Lett. 1994, 35, 2691. 222 1 3 1 Copp, B.R.; Ireland, C M . J. Org. Chem. 1991, 56, 4596. 1 3 2 Kong, F.; Faulkner, D.J. Tetrahedron Lett. 1991, 32, 3667. 1 3 3 Raub, M.F.; Cardellina, J.H.; Choudhary, M.L; Ni, C ; Clardy, J.; Alley, M.C. J. Am. Chem. Soc. 1991,773,3178. 1 3 4 Raub, M.F.; Cardellina, J.H.; Spande, T.F. Tetrahedron Lett. 1992, 33, 2257. 1 3 5 Kupchan, S.M.; Britton, R.W.; Lacadie, J.A.; Ziegler, M.F.; Sigel, CW. J. Org. Chem. 1975, 40, 648. 1 3 6 see Appendix Figures A.6, A.7, and A.8 for representative 2D NMR spectra 1 3 7 see Appendix Figures A.9, A. 10, and A. l 1 for representative 2D NMR spectra 1 3 8 see Appendix Figures A.12, A.13, A.14, and A.15 for representative 2D NMR spectra 1 3 9 see Appendix Figures A.16, A.17, and A.18 for representative 2D NMR spectra 1 4 0 Kubanek, J.; Williams, D.E.; de Silva, E.D.; Allen, T.; Andersen, R.J. Tetrahedron Lett. 1995, 36, 6189. 1 4 1 Evans, M.B.; Dale, A.D.; Little, CJ. Chromatographia 1980, 73, 5. 1 4 2 Munier, R.; Macheboeuf, M. Bull. Soc. Chim. Biol. 1951, 33, 846. 1 4 3 Stahl, E., Ed. Thin Layer Chromatography; Springer-Verlag: Berlin and New York, 1969.