@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Science, Faculty of"@en, "Chemistry, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Northcote, Peter T."@en ; dcterms:issued "2010-10-18T16:39:11Z"@en, "1989"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """A chemical study of the marine petrosid sponge Xestospongia vanilla has led to the isolation of nine new isoprene derived secondary metabolites. Their proposed structures were elucidated by a combination of spectroscopic analysis and chemical degradation and interconversions. Xestodiol (94), a C₁₈ apocarotenoid, appears to be a degradation product of the abundant marine carotenoid fucoxanthin (103). The xestovanins (98-102) are triterpene glycosides; their isolation represents the second reported occurrence of this type of compound from sponges. Their triterpene carbon skeletons are unique, and are either monocyclic (secoxestovanane skeleton) or bicyclic (xestovanane skeleton). All the xestovanins contain the same disaccharide fragment composed of L-rhamnose alpha linked to the 4 position of D-fucose. The fucose residue is beta linked to the same position on the aglycone in all isolated xestovanins. Xestovanin D (102) contains an extra L-rhamnose residue attached to a different position on the aglycone. Xestovanin A (98) was found to be an inhibitor of fungal growth, while xestovanin C (101) inhibited the growth of bacteria. A series of three smaller apparently related terpenes was also isolated. Xestenone (95) and secoxestenone (97) both contained new C₁₉ carbon skeletons. Secoxestenone (97), a monocyclic compound, could be converted into the bicyclic xestenone (95) by an intramolecular aldol condensation. The C₂₀ xestolide (96), with a similar structure to both xestenone and secoxestenone, had an unique carbon skeleton that could not be derived readily from an unrearranged diterpene skeleton. It is suggested that these three smaller terpenes (95-97) are degraded triterpenes, derived from a secoxestovanane carbon skeleton. The secondary metabolite chemistry of the petrosid sponges is reviewed, and an overview of triterpenes of marine origin is presented."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/29253?expand=metadata"@en ; skos:note "N O V E L T E R P E N O I D M E T A B O L I T E S F R O M T H E M A R I N E S P O N G E XESTOSPONGIA VANILLA by P E T E R T. N O R T H C O T E B.Sc. University of British Columbia, 1982. A T H E S I S S U B M I T T E D I N PAJRTIAL F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E S T U D I E S Department of Chemistry We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A October 1989 ©Peter Northcote, 1989 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. The University of British Columbia Vancouver, Canada Department Date A3.Qc} \\Hrj DE-6 (2/88) - i i -A B S T R A C T A chemical study of the marine petrosid sponge Xestospongia vanilla has led to the isolation of nine new isoprene derived secondary metabolites. Their proposed structures were elucidated by a combination of spectroscopic analysis and chemical degradation and interconversions. Xestodiol (94), a C j g apocarotenoid, appears to be a degradation product of the abundant marine carotenoid fucoxanthin (103). The xestovanins (98-102) are triterpene glycosides; their isolation represents the second reported occurrence of this type of compound from sponges. Their triterpene carbon skeletons are unique, and are either monocyclic (secoxestovanane skeleton) or bicyclic (xestovanane skeleton). All the xestovanins contain the same disaccharide fragment composed of L-rhamnose alpha linked to the 4 position of D-fucose. The fucose residue is beta linked to the same position on the aglycone in all isolated xestovanins. Xestovanin D (102) contains an extra L-rhamnose residue attached to a different position on the aglycone. Xestovanin A (98) was found to be an inhibitor of fungal growth, while xestovanin C (101) inhibited the growth of bacteria. A series of three smaller apparently related terpenes was also isolated. Xestenone (95) and secoxestenone (97) both contained new Cjo carbon skeletons. Secoxestenone (97), a monocyclic compound, could be converted into the bicyclic xestenone (95) by an intramolecular aldol condensation. The C20 xestolide (96), with a similar structure to both xestenone and secoxestenone, had an unique carbon skeleton that could not be derived readily from an unrearranged diterpene skeleton. It is suggested that these three smaller terpenes (95-97) are degraded triterpenes, derived from a secoxestovanane carbon skeleton. - iii -The secondary metabolite chemistry of the petrosid sponges is reviewed, and an overview of triterpenes of marine origin is presented. - iv -T A B L E O F C O N T E N T S A B S T R A C T i i T A B L E O F C O N T E N T S iv LIST O F F I G U R E S vii LIST O F S C H E M E S xv LIST O F T A B L E S , xvi A C K N O W L E D G E M E N T S xviii A B B R E V I A T I O N S xix I N T R O D U C T I O N 1 SPONGES AND CHEMISTRY 1 TAXONOMY OF SPONGES 2 DESCRIPTION OFXESTOSPONGIA VANILLA 3 TAXONOMY OF XESTOSPONGIA VANILLA 5 NATURAL PRODUCTS FROMPETROSIDS 8 Steroids 8 Linear Acetylenic Compounds. 9 Alkaloids. 11 Compounds of Mixed Biogenesis 23 - v -Taxonomic Significance of the Secondary Metabolites 26 TRITERPENOIDS FROM MARINE SOURCES. 27 D I S C U S S I O N 41 ISOLATION OF SECONDARY METABOLITES FROM XESTOSPONGIA VANILLA 41 XESTODIOL (94) 44 XESTENONE (95) 54 XESTOLIDE (96) 71 SECOXESTENONE (97) 81 XESTOVANIN A (98) 88 Fragment A 92 Fragment B 94 Fragment C 97 Assembly of the Aglycone Fragments 99 Elimination Product (111) 101 Sugar Residues 107 SECOXESTOVANINA (99) 132 XESTOVANIN B (100) 149 - vi -XESTOVANIN C (101) 167 XESTOVANIN D (102) 180 C O N C L U S I O N S 195 PROPOSED BIOGENESIS OF THE TERPENES ISOLATED FROM XESTOSPONGIA VANILLA 195 BIOLOGICAL ACTIVITY. 196 E X P E R I M E N T A L 199 B I B L I O G R A P H Y 213 - vii -LIST O F F I G U R E S Figure 1. Map of the collection site of Xestospongia vanilla Figure 2. Xestospongia vanilla 6 Figure 3. Phylogenetic classification of Xestospongia vanilla 7 Figure 4. Steroids with extended side chains isolated from petrosid sponges 10 Figure 5. Brominated acetylenic fatty acids from petrosid sponges. 12 Figure 6. Long chain acetylenic compounds from petrosid sponges 13 Figure 7. Petrosins 14 Figure 8. Xestospongins 16 Figure 9. Sarains 17 Figure 10. Manzamines : 20 Figure 11. Miscellaneous nitrogen containing compounds 22 Figure 12. Compounds of mixed biogenesis from the genus Xestospongia 24 Figure 13. Compounds of mixed biogenesis from the genus Strongylophora 25 - viii -Figure 14. Triterpenoids isolated from marine algae 28 Figure 15. Further triterpenoids isolated from marine algae 30 Figure 16. Triterpenoids from holothurians 31 Figure 17. Further triterpenoids from holothurians 32 Figure 18. Triterpenoids from cnidarians 33 Figure 19. Terpenoids from the sponge Luffariella variabilis 35 Figure 20. Malabaricane triterpenoids 36 Figure 21. Triterpenoids from the sponge Siphonochalina siphonella 37 Figure 22. Triterpenoids from the sponge Siphonochalina siphonella 38 Figure 23. Glycosides isolated from the sponge Asteropus sp 40 Figure 24. Xestodiol (94) 45 Figure 25. Acyclic fragment of xestodiol (94) 46 Figure 26. Cyclic fragment of xestodiol (94) 49 Figure 27. Comparison of xestodiol (94) to fucoxanthin (103) 49 Figure 28. % N M R (300 M H z ) spectrum of xestodiol (94) in C D C 1 3 50 Figure 29. UC N M R (75 M H z ) spectrum of xestodiol (94) in C D C 1 3 52 Figure 30. The fragments of xestenone (95) 55 Figure 31. Fragment D of xestenone (95) 60 Figure 32. Assembly of the fragments of xestenone (95) 62 Figure 33. Comparison of xestenone (95) to model compounds 64 Figure 34. lH N M R (300 M H z ) spectrum of xestenone (95) in CDCI3 65 Figure 35. 2 H N M R (300 M H z ) spectrum of the acetate of xestenone (104) in CDCI3 67 Figure 36. * H N M R (300 M H z ) spectrum of the reduction product of xestenone 105 in CDCI3 68 Figure 37. 1 3 C N M R (75 M H z ) spectrum of xestenone (95) in CDCI3 70 Figure 38. Xestolide (96) 72 Figure 39. Acyclic fragment of xestolide (96) 72 Figure 40. E I mass spectrum of xestolide (96) 74 Figure 41. * H N M R (300 M H z ) spectrum of xestolide (96) in CDCI3 76 Figure 42. 1 3 C N M R (75 M H z ) spectrum of xestolide (96) in CDCI3 78 Figure 43. A H N M R (300 M H z ) spectrum of the keto methyl ester of xestolide 103 in CDC13 80 Figure 44. Secoxestenone (97) 82 Figure 45. * H N M R (300 M H z ) spectrum of secoxestenone (97)in CDCI3 85 Figure 46. 1 3 C N M R (75 M H z ) spectrum of secoxestenone (97) in CDCI3 87 Figure 47. Xestovanin A (98) 89 Figure 48. Aglycone of xestovanin A (98) 91 Figure 49. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) spin systems of fragment A 93 Figure 50. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) spin systems of fragment B 95 Figure 51. Model diterpene compound 96 Figure 52. C O S Y spectrum of the hexaacetate of xestovanin A 109 in CDCI3 correlations of fragment C 98 Figure 53. Structural proposals for the aglycone 01 xestovanin A 100 Figure 54. Elimination product of xestovanin A 111 102 Figure 55. 2 H N M R (400 M H z ) spectrum of the elimination product of xestovanin A 111 in CDCI3 104 - xi -Figure 56. 1 : > C N M R (75 M H z ) spectrum of the elimination product of xestovanin A 111 in CDCI3 106 Figure 57. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) fucose spin system 108 Figure 58. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) rhamnose spin system 109 Figure 59. Sugar linkages of xestovanin A (98) I l l Figure 60. * H N M R (400 M H z ) spectrum of xestovanin A (98) in acetone-d^ 113 Figure 61. C O S Y spectrum (400 M H z ) of xestovanin A (98) in acetone-d^ 114 Figure 62. 2 H N M R (400 M H z ) spectrum of xestovanin A (98) in Me 2SO-d£ 117 Figure 63. C O S Y spectrum (400 M H z ) of xestovanin A (98) in M e 2 S O - d 6 118 Figure 64. 1 3 C N M R (75 M H z ) spectrum of xestovanin A (98) in M e 2 S O - d ^ 121 Figure 65. 1 3 C N M R (75 M H z ) spectrum of the anomeric region of xestovanin A (98) in M e 2 S O - d 6 122 Figure 66. ! H N M R (400 M H z ) spectrum of xestovanin A hexaacetate 109 in CDCI3 123 Figure 67. C O S Y spectrum (400 M H z ) of xestovanin A hexaacetate 109 in CDCI3 124 Figure 68. Some nOe difference experiments (400MHz) on xestovanin A hexaacetate 109 in CDCI3 125 - xii -Figure 69. 1 3 C N M R (75 M H z ) spectrum of xestovanin A hexaacetate 109 in CDCI3 128 Figure 70. One bond H E T C O R spectrum of the downfield region of xestovanin A hexaacetate 109 in CDCI3 . 129 Figure 71. Secoxestovanin A (99) 133 Figure 72. Aglycone of secoxestovanin A 135 Figure 73. C O S Y spectrum (400 M H z ) of secoxestovanin A (99) in acetone-dg 137 Figure 74. lH N M R ( 4 0 0 M H z ) spectrum of secoxestovanin A (99)in Me2SO-d^ 1 4 0 Figure 75. C O S Y spectrum (400 M H z ) of secoxestovamn A (99) in Me2SO-dg 141 Figure 76. 1 3 C N M R (75 M H z ) spectrum of secoxestovanin A (99)m Me2SO-d^ 144 Figure 77. * H N M R ( 4 0 0 M H z ) spectrum of secoxestovanin A hexaacetate 112 in CDCI3 145 Figure 78. C O S Y spectrum (400 M H z ) of secoxestovanin A hexaacetate 112 in CDCI3 146 Figure 79. Xestovanin B (100) 150 Figure 80. Aglycone of xestovanin B (100) ....152 Figure 81. X H N M R (400 M H z ) spectrum of xestovanin B (100) in acetone-d^ 154 Figure 82. C O S Y spectrum (400 M H z ) of xestovanin B (100) in acetone-d^ 155 - xiii -Figure 83. * H N M R (400 M H z ) spectrum of xestovanin B (100) in M e 2 S O - d 6 158 Figure 84. C O S Y spectrum (400 M H z ) of xestovanin B (100) in M e 2 S O - d 6 159 Figure 85. 1 3 C N M R (75 M H z ) spectrum of xestovanin B (100) in acetone-d^ 162 Figure 86. * H N M R (400 M H z ) spectrum of xestovanin B hexaacetate 113 in CDCI3 163 Figure 87. C O S Y spectrum (400 M H z ) of xestovanin B hexaacetate 113 in CDCI3 164 Figure 88. Xestovanin C (101) 168 Figure 89. * H N M R of the aglycone of xestovanin C (101) in acetone-d^ 170 Figure 90. ! H N M R (400 M H z ) spectrum of xestovanin C (101) in acetone-dg 171 Figure 91. C O S Y spectrum (400 M H z ) of xestovanin C (101) in acetone-dg 172 Figure 92. ! H N M R (400 M H z ) spectrum of xestovanin C (101) in M e 2 S O - d 6 175 Figure 93. C O S Y spectrum (400 M H z ) of xestovanin C (101) in M e 2 S O - d 6 176 Figure 94. 1 3 C N M R (75 M H z ) spectrum of xestovanin C (101) in Me 2 SO-d 6 „ 179 Figure 95. Xestovanin D (102) 181 Figure 96. M e 2 S O - d 6 C O S Y spectrum of xestovanin D (102) additional rhamnose spin system 183 Figure 97. X H N M R (400 M H z ) spectrum of xestovanin D (102) in acetone-d^ Figure 98. C O S Y spectrum (400 M H z ) of xestovanin D (102) in acetone-d^ Figure 99. 1 H N M R (400 M H z ) spectrum of xestovanin D (102) in Me 2SO-d£ Figure 100. C O S Y spectrum (400 M H z ) of xestovanin D (102) in M e 2 S O - d ^ Figure 101. i3C N M R (75 M H z ) spectrum of xestovanin D (102) in Me 2 SO-dg Figure 102. 1 3 C N M R (75 M H z ) spectrum of the anomeric region of xestovanin D (102) in Me 2 SO-dg - XV -LIST O F S C H E M E S Scheme 1. Braekman's biogenetic correlation of the petrosins, xestospongins, and sarains from a common progenitor 18 Scheme 2. Isolation scheme 43 Scheme 3. Reduction of xestenone (95) 57 Scheme 4. Retro Diels-Alder reaction ofxestolide (96) 74 Scheme 5. Conversion of secoxestenone (97) to xestenone (95) 84 Scheme 6. Proposed biogenesis of the xestovanins 196 Scheme 7. Proposed biogenesis of the nor-triterpenes 198 - xvi -LIST O F T A B L E S Table 1. 2 H N M R Data for xestodiol (94) in CDCI3 51 Table 2. Comparison of proton chemical shifts of xestodiol (94) and fucoxanthin (103) 51 Table 3. Carbon assignments of xestodiol (94) compared to fucoxanthin (103) 53 Table 4. 1 H N M R data for xestenone (95) in CDCI3 66 Table 5. % N M R data for the reduction product of xestenone 105 in CDCI3 69 Table 6. lH N M R data for xestolide (96) in CDCI3 77 Table 7. Carbon assignments of xestenone (95) and xestolide (96) in CDCI3 79 Table 8. ! H N M R data for secoxestenone (97) in CDCI3 86 Table 9. % N M R data for elimination product 111 in CDCI3 105 Table 10. ^-H N M R data for xestovanin A (98) in acetone-dg 115 Table 11. % N M R data for xestovanin A (98) i n M e 2 S O - d 6 119 Table 12. 1 H N M R data for xestovanin A hexaacetate 109 in CDCI3 126 Table 13. Carbon assignments for xestovanin A (98) and its hexaacetate 109 130 Table 14. X H N M R data for secoxestovanin A (99) in acetone-d^ Table 15. N M R data for secoxestovanin A (99) in Me2SO-d^ Table 16. * H N M R data for secoxestovanin A hexaacetate 112 in CDCI3 Table 17. % N M R data for xestovanin B (100) in acetone-d^ Table 18. 1 H N M R data for xestovanin B (100) in Me2SO-d(3 Table 19. 1 H N M R data for xestovanin B hexaacetate 113 in CDCI3 Table 20. ! H N M R data for xestovanin C (101) in acetone-dg Table 21. lU N M R data for xestovanin C (101) in Me2SO-dg Table 22. ! H N M R data for xestovanin D (102) in acetone-dg Table 23. ^ N M R data for xestovanin D (102) in Me2SO-d(j A C K N O W L E D G E M E N T S I would like to extend my sincere thanks to my research supervisor, Dr. Raymond Andersen, for his advice, encouragement, and patience. I am grateful to Mike LeBlanc for performing the bioassays, his encouragement underwater, and his help in collecting the sponge. The assistance of the staff of the departmental N M R and mass spectroscopy laboratories and also the staff of the Bamfield Marine Station is greatly appreciated. My co-workers in the laboratory have been a great source of support and advice. I would also like to thank my wife Lisa for her great encouragement, help, and understanding during my years as a graduate student. A B B R E V I A T I O N S A P T = Attached Proton Test br = broad CDCI3 = Chloroform-dj C O S Y = Homonuclear Correlation Specroscopy d = doublet ED50 = 50% effective dose E I H R M S = Electron Impact High Resolution Mass Spectrum E I L R M S = Electron Impact Low Resolution Mass Spectrum E t O A c = Ethyl acetate E t 2 0 = Diethyl ether F A B M S = Fast Atom Bombardment Mass Spectrum G C = Gas Chromatography H E T C O R = Heteronuclear Correlation H P L C = High Performance Liquid Chromatography IR = Infrared J = scalar coupling constant - XX -M + = Parent ion M e O H = Methanol mult or m = proton resonance with unresolvable couplings m/z = mass to charge ratio nOe = nuclear Overhauser effect O D S = Octadecylsilane q = quartet quin = quintet rel. int. = relative intensity s = solvent resonance T L C = Thin Layer Chromatography U V = Ultra violet w = water resonance * H N M R = Proton nuclear magnetic resonance 1 3 C N M R = Carbon-13 nuclear magnetic resonance J O T R O D U C T I O N SPONGES AND CHEMISTRY Sponges (phylum Porifera) are considered to be the oldest and most primitive metazoans or multicellular animals. They lack many features that are normally thought to be ubiquitous in animals, such as a determinate body and a nervous system. As a consequence, their animal nature was not established beyond doubt until the early nineteenth century.-^ Sponges have a rich fossil record of over 500 million years extending to well before the early Cambrian, the period in which many other invertebrate animals first appear. In some periods, tropical reefs were made up of the skeletons of sponges much as modern reefs are made of coral skeletons. Sponges still comprise a significant proportion of the biomass of present marine ecosystems. * Being sessile organisms, sponges have been subjected to a strong selective pressure to evolve defences against predation, encrustation, and encroachment. Lacking the more complex bodies of other sessile invertebrates, perhaps they have had to rely more heavily on novel biochemical pathways to produce chemical defences. With their long evolutionary history, and apparent need for chemical defence, it is perhaps not surprising that sponges have provided a rich assortment of novel secondary metabolites for natural product chemists to study. Indeed, in the nine years between 1977 and 1985, of the 1009 new natural products reported from marine invertebrates, slightly under one half (447) were isolated from 2 sponges. TAXONOMY OF SPONGES - 2 -The primitive nature of sponges makes their identification and classification very difficult. The external morphology of sponges is notoriously variable because their indeterminate growth pattern and simple body plan allow their shapes and sizes to vary greatly in response to the environment. This external variability has led taxonomists to rely heavily on embryology, histology, skeletal characteristics, and biochemistry to classify sponges. The nature of the skeleton is the characteristic used almost exclusively by taxonomists to identify a particular specimen to the genus and species level. The phylum Porifera, containing about 5000 living species, is'divided into four classes based on the chemical nature of the inorganic skeletal material and how the various components of it are organized.* The Sclerospongiae have a tissue made of collagen and containing siliceous spicules (small discrete spines) that is restricted to a thin layer of living material surrounding a massive calcareous skeleton. The Calcarea have a skeleton of crystalline calcium carbonate. The Hexactinellida or glass sponges have a skeleton comprised entirely of siliceous spicules that frequently makes up over 9 0 % of the mass of the sponge. The Demospongiae, containing 95% of known sponges, have a skeleton comprised of siliceous spicules and fibres of collagen} There is little debate among sponge taxonomists about the class to which a particular sponge species belongs as the classes are quite distinct and well defined. This is not true for lower levels of classification. For almost all other types of multicellular organisms classification is well accepted down to the order and family level. The lack of such organization in the Porifera has made scientific study, both - 3 -biological and chemical, very difficult. Some of the taxonomic confusion has arisen because various workers have based their classification schemes on sponges collected largely in one region of the world. 3 It also must be realized that since classification into orders, families, and genera is based very much on the appearance of one feature, spicules, sponge taxonomy is largely artificial in that it may not reflect actual evolutionary relatedness. This again creates a real problem in trying to find trends in the occurrence of interesting secondary metabolites in sponges. DESCRIPTION OF XESTOSPONGIA VANILLA This study examines secondary metabolites isolated from the shallow water sponge Xestospongia vanilla De Laubenfels 1930. In his original description of the species, De Laubenfels describes this sponge as a stony hard, white or pale yellowish encrusting sponge up to 1cm thick. The type specimens were collected at Pacific Grove, California where it occurred in the intertidal zone on the undersides of boulders.^ De Laubenfels describes Xestospongia vanilla as one of the most abundant sponges in central California. Austin lists the range of this sponge as from Baja California to British Columbia.^ The specimens used in this study were collected around a group of small islands near the Bamfield Marine Station on the west coast of Vancouver Island. Although it is known from the intertidal zone in B.C.f* in the Bamfield area this sponge seems to be confined to a zone several meters below the low tide mark. In our collection area it was an uncommon sponge, never forming a dominant part of the sponge community. Almost all of the material collected for this study came from two small areas (see Figure 1), the only sites where significant amounts could be found. Generally it occurred on vertical solid rock walls at depths of 7 to 10 m, Figure 1. Map of the collection site of Xestospongia vanilla - 5 -below the zone of greatest biological abundance. It was rarely encrusted with other organisms, and was never seen to be fed upon by any species of nudibranch, as is commonly seen for most sponges growing in the same area. The specimens were often much larger than those described by De Laubenfels, sometimes reaching up to 10 cm off the substrate. Austin notes, that when alive, this sponge sometimes has a light pink colour associated with the tips of the osculi.^ Some of the specimens collected for this study had this pink colouration (see Figure 2). TAXONOMY OF XESTOSPONGIA VANILLA Xestospongia vanilla is a Demosponge (it has siliceous spicules and a collagen skeleton) that belongs to a group, the petrosids, that have been classified differently by various authors. I will follow Austin's classification,^ but it must be recognized that other opinions exist. Most authorities place the group in the sub-class Ceractinomorpha, characterized by monaxic megascleres (linear large spicules). Austin^ follows Hartman's^ placement of the group in its own order, the Petrosiida, equivalent to Bergquist's Nepheliospongidae (excluding fossil forms). V a n Soest lowers it to the family level and places it in the large order Haplosclerida along with the Haliclonidae, Niphatidae, and Callyspongiidae families. 3 The family Petrosiidae is characterized by oxeas or strongyles (linear large spicules with pointed or rounded ends) that dominate the collagen in the skeleton to such an extent that they obscure some of the normal internal features of the sponge and give the organism a hard or stony texture; hence the name petrosids (origin: petros; Greek for stone). The family is comprised of four genera: Petrosia with more than 20 species, Strongylophora with four species, Xestospongia with approximately 30 species, and an un-named genus reserved by Van Soest for the sponge previously described as Haliclona pellasarca (De Laubenfels 1934).^ Figure 2. Xestospongia vanilla (approx. life size) I i -7-Kingdom Metazoa (multi-cellular animals) Phylum Porifera (sponges) Class Hexactinellida Demospongiae Calcarea Subclass Tetractinomorpha Ceractinomorpha Order Haplosclerida Petrosiida Family Petrosiidae Genus Strongylophora Xestospongia Petrosia Species X. muta X. vanilla X.caycedoi Figure 3. Phylogenetic classification of Xestospongia vanilla - 8 -De Laubenfels, in his original description of the genus, separates Xestospongia from Petrosia by its lack of a specialized ectosomal skeleton and its single sized megescleres (oxeas). Petrosia is described as having a specialized ectosomal skeleton and two distinct size classes of megascleres. The genus Xestospongia occurs throughout the tropics and in the Antarctic and northern Pacific oceans. NA TURAL PRODUCTS FROM THE PETROS1DS Before describing the new natural products isolated in this study, 1 will review similar work that has been done on this group of sponges (the petrosids) and give an overview of triterpenes from marine sources. A wide variety of secondary metabolites have been isolated from petrosid sponges. The compounds fall into four distinct groups: steroids, linear acetylenic compounds, alkaloids, and compounds of mixed polyketide and terpenoid biogenesis. The steroids and alkaloids have attracted the most attention; the steroids, due to their biosynthetic novelty and their presumed taxonomic significance; the alkaloids, because of their potent biological activity. Steroids o Sponges have been known for some time to be sources of novel steroids, and they are the only marine source of steroids with alkyl extended side-chains. The petrosids have been found to contain steroids with as many as four extra carbons added. Biosynthetic studies have shown that the carbons are added sequentially with S-adenosyl-methionine the presumed biological source. It is beyond the scope of - 9 -this review to describe all the steroids that have been isolated from the petrosids. In many cases these novel steroids replace cholesterol as the principal steroid so that some of them must be considered primary metabolites of the sponge. A few of the steroids isolated first from petrosid sponges will be presented as representatives of the kinds of steroids found in this group of sponges. The first steroids described from petrosids were the principal steroids isolated from each of the three genera: xestosterol (1) from Xestospongia muta? strongylosterol (2) from Strongylophora durissima}®^ and petrosterol (3) from Petrosia ficijronnis-.12,13 Qf t h e s e compounds make up more than 5 0 % of the steroid content of the organisms from which they were isolated. Petrosterol is of particular interest due to the presence of a cyclopropane ring in the side chain. This compound has also been isolated as the principal steroid from a Halichondria species.^ Examination of trace steroids from these sponges has also yielded interesting compounds. Xestospongesterol (4), and isoxestospongesterol (5) have been found in Xestospongia sp. and Strongylophora durissima, respectively, ^ while 25-methyIxestosterol (6) was isolated from both of these sponges in trace -i /- -in amounts. From Xestospongia muta, mutasterol (7), was isolated. The genus Petrosia has also yielded interesting trace sterols: from Petrosia ficiformis, (24R)-24,26-dimethylcholesta-5,26-dien-3/3-oI (8), 1 8 from Petrosia hebes, hebesterol (9). 1 9 Linear Acetylenic Compounds Schmitz and Gopichand isolated the novel linear dibromo-C^-acetylenic acid 10 from Xestospongia muta in 1978.^ A series of similar acids have been isolated from other members of this genus: compound 11 from Xestospongia testudinaria^ and compounds 12-16 from an unidentified species.^ A Petrosia - 10-2 strongylosterol 4 xestospongesterol 5 isoxestospongesterol 6 25-methylxestosterol Side chain Figure 4. Steroids with extended side chains isolated from petrosid sponges - 1 1 -species has yielded a C3Q linear polyacetylene tetra-ol and its corresponding tetraketo analogue 17,18.^3'^ A series of very long linear polyacetylenes ( C ^ -C55) were isolated from Petrosia ficiformis (see Figure 6 ) . ^ ' ^ These compounds were only partially purified and the structures only partially solved, however, they appear to be closely related to the compounds previously described from Petrosia. Long chain acetylenic acid derived compounds are reasonably common in some terrestrial plant families (for example, the Arailiaceae). Marine examples are from members of the Petrosiida with a few notable exceptions from other sponge genera such as the siphonodiols isolated from the sponge Siphonochalina truncata^ and straight chain polyacetylenes from Reniera fulva?^ Alkaloids Perhaps the most unusual secondary metabolites isolated from the petrosids are the alkaloids. Many of these compounds are biologically active and most have unprecedented skeletons. The first to be isolated and characterized were the petrosins from Petrosia seriata?^ The structure of the first of these, petrosin (19), isolated as an optically inactive solid, was solved by X-ray diffraction. Two enantiomers co-crystalized indicating that the material was racemic. The C 2 symmetry of the molecule was reflected in the fact that only 15 resonances were observed in the N M R spectrum of this C3Q compound. Although it is not illustrated in the structural drawing of this study, the computer generated stereoscopic diagram presented indicates that the quinolizidine ring systems of both halves are trans-fused. Two minor metabolites, petrosin-A (20) and petrosin-B (21), were later isolated from the same sponge.^ Petrosin-A, like petrosin, contained an element of symmetry as evidenced by N M R . The structure has recently been - 12-12-14 from Xestospongia sp. Figure 5. Brominated acetylenic fatty acids from petrosid sponges - 13 -H C = C C H O H C H = < ; H C H 2 R 1 C H 2 C H = C H RJ + R 2 = CnHjn* where n = 25,28 H C ^ C C H O H C T ^ H C H ^ C H i C H ^ H ^ R i + R 2 = C n H ^ where n = 28,31,34 A series of partial) characterized fatty acid derived compounds isolated from the sponge Petrosia ficiformis 9 0 Figure 6. Long chain acetylenic compounds from petrosid sponges - 14-19-21 f r o m Petrosia seriata Figure 7. Petrosins -15-corrected by the original authors to the meso structure illustrated. u A related group of bis-l-oxaquinolizidine compounds, the xestospongins A to D (22, 23, 24, and 25), were discovered by Nakagawa and co-workers in Xestospongia exigua?^ The methylene chains connecting the two bicyclic systems contained one more carbon than in the petrosins, while in the bicyclic systems one carbon was replaced by oxygen. Single crystal X-ray diffraction analysis of xestospongin C (24) showed that one of the 1-oxaquinolizidine ring systems was cis-fused. No mention was made of the crystal structure being composed of a mixture of enantiomers as was the case in petrosin. Indeed, the authors record optical rotations for all of the xestospongins. Xestospongin A (22), however, is a diastereomer of xestospongin C (24). It is symmetrical and shows half the N M R resonances of the other three compounds. Only xestospongin B (23) has methyl substitution at position 3, like the petrosins, and only on one of the oxaquinolizidine systems. A series of alkaloids sharing many structural features with the petrosins and xestospongins has been isolated from the sponge Reniera sarai, a member of the order H a p l o s c l e r i d a ^ ' 3 3 . The sarains 1-3 (26, 27, and 28), and isosarain (29) contain one rro/ts-quinolizidine system with similar substitution and relative stereochemistry as petrosin. The sarain quinolizidine system has an extra substitution. The similarities have lead Braekman and co-workers to propose a common origin for these three series of compounds (petrosins, xestospongins, and sarains, see Scheme l ) . 3 3 The proposed progenitor is a tricyclic alkaloid comprised of two identical halves. Two piperidine rings are envisioned as being linked via linear alkyl chains at positions 1 and 3. These can further cyclize to form the quinolizidines of the petrosins and sarains, or the oxaquinolizidines of the petrosins. - 16-22-25 from Xestospongia exigua Figure 8. Xestospongins Figure 9. 26-29 isolated from the Haploscleridian sponge Reniera sarai Sarains Scheme 1. Braekman's biogenetic correlation of the petrosins, xestospongins, and sarains from a common progenitor -19-In the sarain series, one of the piperidine systems remains intact. In the petrosins and xestovanins the presumed progenitor is identical, with Cg alkyl chains, while in the sarain series the length of one of the alkyl chains is variable. A n even more remarkable series of alkaloids, the manzamines, have been isolated by various groups from species of Xestospongia, Haliclona, and Pellina. Manzamine-A (30) was isolated independently from a Haliclona species 3 4 , and from Pellina sp., a member of the Oceanapiidae.3^ This compound was reported to be cytotoxic, with an IC5Q of 0.07ug/mL against P388 mouse leukemia ce l l s 3 4 . In both studies, the structure was solved by single crystal X-ray analysis on the hydrochloride. Although the aromatic B carboline ring system is quite different from the quinolizidine systems of the previously described alkaloids, the remaining fragments do show some similarities. One of the remaining nitrogens is included in a piperidine ring alkylated by alkyl chains at positions 1 and 3. The remaining eight membered ring though, is again quite different. Later, manzamines B and C (31 and 32) were isolated from the same Haliclona species.3^ Manzamine B (31) could be seen to be clearly related to manzamine A (30). Manzamine C (32), although related through the presence of a common B carboline system, contained a completely unprecedented, nitrogen containing eleven membered ring. Both of these compounds were mildly cytotoxic towards mouse leukemia cell lines. More recently, two new manzamines (E and F (33 and 34)) have been isolated from a species of Xestospongia?^ Both of these compounds have the same carbon skeleton as manzamine A (30), differing only in oxidation of the B carboline and eight membered ring systems. The remaining alkaloids that have been reported from petrosid sponges bear no obvious relation to the other alkaloids. Recently, a pair of epimeric amino -20-32 manzamine C 33 manzamine E 34 manzamine E 30,31,32 from Haliclona sp. 33^34 from Xestospongia sp. Figure 10. Manzarmnes -21-alcohols (35 and 36) was reported from a Xestospongia species. 3 8 These compounds are described as being biosynthesized from fatty acids and serine. Two small alkaloids, renierol (37) and the previously known mimosamycin (38), were obtained from Xestospongia caycedoi?^ Renierol is closely related to l,6-dimethyl-7-methoxy-5,8-dihydroisoquinoline-5,8-dione (39), a compound previously reported from a Reniera species.4^ A novel alkaloidal pigment, petrosamine (40), has been reported from an unidentified Petrosia species.4* In solution, the colour of this quaternary ammonium salt depends on the solvent (green in acetonitrile, blue in methanol, and purple in aqueous solutions). It is suggested that this colour change reflects a shift in the dynamic equilibrium between keto and enol forms. Several authors have suggested that the alkaloids isolated from this group of sponges may be of microbial origin 3 9 ' 4 2' 3 5>41 Ichiba and co-workers state that the presence of the manzamines in three different sponge genera implies that they are microbial metabolites.4^ A review of all of the alkaloids isolated from the petrosids, however, leads to the conclusion that these compounds are probably produced by the sponges themselves. The structural similarities of many of these alkaloids suggest a common biological origin. If a microbe is the source, the same organism would have to be found in a select group of sponges distributed widely throughout the world's oceans. A n examination of the taxonomy of the non-petrosid sponges that also contain some of these alkaloids reveals that they may well be related to the petrosids. The petrosids have been classified as a family of the order Haploscleridae.^ Haliclona and Reniera, genera that contain similar or identical compounds to those isolated from petrosids, are members of the Haliclonidae, a family in the same order. Pellina, from which manzamines A (30) and E (34), both metabolites of Xestospongia, have been isolated, is a member of the Haploscleridian family Oceanapiidae. The simplest explanation for the presence of structurally 40 petrosamine 37 and 38 from Xestospongia caycedoi 39 from Reniera sp., 40 from Petrosia sp. Figure 11. Miscellaneous nitrogen containing compounds - 2 3 -related alkaloids occuring only in a group of related sponges is that they are produced by these organisms and not by commensal microbes. Compounds of Mixed Biogenesis Polyketides and terpenoids have not been obtained in any great number or variety from the petrosid sponges. Apart from the ubiquitous steroids, no purely terpene derived metabolites have been isolated until this study. A number of prenylated hydroquinone derived compounds have been isolated from petrosid sponges. The pentacyclic halenaquinone (41) has been isolated from Xestospongia exigua.^ Xestospongia sapra yielded the related compounds xestoquinone (42), the dihydroquinone of halenaquinone 43, and its sulfonate 4 4 . ^ ' ^ Strongylophora durissima is the source of a series of ichthyotoxic compounds, the strongylophorines 1-3 (45-47).^ Puupehenone (48), a cytotoxic and antineoplastic compound first isolated from a Heteronema species^ was reported from Strongylophora hartmani.^ This compound appears to be similar in some respects to the strongylophorines. -24-41 from Xestospongia exigua 41-43 from Xestospongia sapra Figure 12. Compounds of mixed biogenesis from the genus Xestospongia C O O H 45 strongylophorine-1 O 46 strongylophorine-2 45-47 f r o m Strongylophora durissima 48 f r o m Strongylophora hartmani Figure 13. Compounds of mixed biogenesis from the genus Strongylophora - 2 6 -Taxonomic Significance of the Secondary Metabolites Zoologists have expressed an interest in using the distribution of secondary metabolites in sponges to help with their classification.^ The steroids have received most attention along these lines. Bergquist has used sterol composition to reclassify orders of the Demospongiae.^ She notes that there are great differences in the sterol composition between members of the petrosids (Xestospongia testudinaria and Strongylophora durissima). Since this study was published (1980), many more members of this group have been examined and a large number of new steroids have been isolated. Many of the steroids are found in more than one genus of this group. Although common steroids support the grouping of these three genera (Xestospongia, Petrosia, and Strongylophora) in one family and order, it must be noted that they also share many of the same steroids with members of other orders (for example Halichondria and Japsis). It seems that the presence of side chain extensions and cyclopropane systems are widely distributed among the Demospongiae, and their occurrences are useful in determining the relatedness of subclasses and orders rather than families and genera. The significance of non-steroidal secondary metabolites in the classification of sponges has received little attention. Many compounds have been described since Bergquist's book was published in 1978. The number of similar compounds found in Xestospongia and Petrosia indicates that these two genera are quite closely related. Strongylophora does not show the same similarity in secondary metabolite composition, but this may be a reflection of the fewer published studies on this genus. A number of genera that do not belong to the petrosids seem to share similar, and in some cases identical, compounds with members of this order. This relationship is especially pronounced in the genera Reniera and Haliclona, members of the Haploscleridae. Similarities in secondary metabolite chemistry suggest that the petrosids are closely related to members of the Haploscleridae, in particular the family Haliclonidae. This does not conflict with the ideas of Bergquist and Austin, however, as the petrosids can be thought of as either a distinct family in the diverse order Haploscleridae, or as a closely related but separate order. It must be noted when comparing reported metabolites that most of this work is done by chemists who are primarily interested in biological activity or chemical novelty and not systematics. Little attempt is made to report the occurrence of compounds that have been previously described and this could lead to a distorted view of their distribution in sponge genera. TRITERPENOIDS FROM MARINE SOURCES A l l but one of the new compounds described in this thesis are either squalene based triterpenoids or nortriterpenoids. In order to better understand their significance, an overview of triterpene derived compounds of marine origin will be undertaken. This review is not intended to be comprehensive but instead to cover the types of carbon skeletons that have been isolated from the marine environment. Special attention will be payed to triterpenoids isolated from sponges. Squalene derived triterpenoids have been isolated from the genus Laurencia, from which many unique mono-, sesqui-, and diterpenes have also been isolated. (10R, l lR)-( + )-squalene-10,ll-epoxide (49) has been obtained f romL. okamurai.^ A tetracyclic polyether, thyrsiferol (50), was isolated from L. thryrsifera^ Thyrsiferol and three derivatives, thyrsiferyl 23-acetate (51), 15(28)-anhydrothyrsiferyl diacetate (52), and 15-anhydrothyrsiferyl diacetate (53) were isolated from L. obtusa along with the related cyclic polyethers, teurilene (54), and - 2 8 -49 (10R.1 lR)-(+)-squalene-10,l 1-epoxide J v . J 50 R=H, thyrsiferol B r ^ 51 R=Ac, thyrsiferyl -23-acetate 53 15-anhydrothyrsiferyl diacetate Figure 14. Triterpenoids isolated from marine algae - 2 9 -magireols A to C (55-57).^>53 Venustatriol (58), differing from thyrsiferol only in the configuration of two chiral centres, has been obtained from L. venusta^ Interestingly, all of the triterpenoids isolated from Laurencia species so far contain uncyclized carbon skeletons. With the exception of teurilene (54), all of the Laurencia triterpenoid metabolites described are very cytotoxic with ED^QS against murine P388 cells ranging from 100 to 0.3 ng/mL. Thyrsiferol (50), thyrsiferyl 23 acetate (51), and venustatriol (58) are reported to be potent antiviral compounds.^ A large number of triterpenoid glycosides containing the lanostane carbon skeleton have been isolated from holothurians. Holothurins A (59) and B (60) were isolated from Holothuria leucospilotar>->,->^ The holotoxins A (61) and B (62), isolated from Stichopus japonicus, differed from holothurin A mainly by the presence of an uncyclized side chain on the aglycone and the addition of two extra glucose residues.^ The echinosides A (63) and B (64), obtained from Actinopyga echnites, have the same carbohydrate structure as the holothurins, and a similar aglycone to the holotoxins, differing in the oxidation pattern of the lactone ring and by having a fully saturated side chain.^'-^ Very similar compounds have been isolated from other holothurian genera/^' 0 ^'^ The holotoxins and echinosides are reported to be antifungal while other related compounds are ichthyotoxic. Only two triterpenoids, echinolactones A (65) and B (66), have been reported from the phylum Cnidaria. These pentacarbocyclic lactones were isolated from the anthozoan Echinopora lamellosa.^ The authors of this work report that this organism also contains a number of other similar metabolites that are known to occur in terrestrial plants. No biological activity was reported for these compounds. The greatest diversity of triterpenoid carbon skeletons in the marine -30-58 venustatriol Figure 15. Further triterpenoids isolated from marine algae -31 -Figure 16. Triterpenoids from holothurians Figure 17. Further triterpenoids from holothurians 65 echinolactone A 66 echinolactone B Figure 18. Triterpenoids from cnidaria - 3 4 -environrnent is found among the sponges. Perhaps the most unusual are the mokupalides, C3Q terpenoid compounds isolated from the Indo-Pacific sponge Luffariella variabilis.64,65 m o k u p a l i d e s 1-3 (67-69) contain six isoprene units connected in a head to tail fashion, while squalene and all its many derivatives are composed of two C15 units linked tail to tail. A series of very similar compounds, the C25 sesterterpenoid manoalide (70) and its derivatives, were isolated from the same sponge.^ '^ The first series of squalene based triterpenoids reported from a sponge were the malabaricanes isolated from Japsis stellifera^ The malabaricanes 1-7 (71-77) all contain a five membered ring spanning the tail to tail linkage of squalene. In an independent study, a single crystal X ray diffraction structure was obtained for malabaricane 3 . ^ The stereochemistry of the 5-6 ring junction methine was shown to be incorrect in the original description. The later authors have proposed the name isomalabaricane for the corrected carbon skeleton. The Red Sea sponge Siphonochalina siphonella is the source of at least ten novel triterpenoids based on three different carbon skeletons. Sipholenols A - E (78-82) and sipholenones A - C (83-85) share the same tetracyclic carbon ske le ton ,^ '^ while siphonellinol (86) is tricyclic, lacking a five membered r i n g . ^ The third carbon skeleton is found in neviotine (87)7^ This pentacyclic compound, like the malabaricanes, contains a five membered ring spanning the tail to tail junction of the squalene precursor. 67 mokupalide 1 68 mokupalide 2 69 mokupalide 3 70 manoalide Figure 19. Terpenoids from the sponge Luffariella variabilis - 3 6 -Drawn with the Corrected Stereochemistry P OAc Figure 20. Malabaricane triterpenoids 7 8 sipholenol A 7 9 sipholenol B O 8 0 sipholenol C 8 1 sipholenol D 82 sipholenol E Figure 21. Triterpenoids from the sponge Siphonochalina siphonella - 3 8 -87 neviotine A (no stereochemistry reported) Figure 22. Triterpenoids from the sponge Siphonochalina siphonella - 3 9 -The most recent group of sponge triterpenoids to be described are the pouosides A to E (88-92) isolated from an Asteropus species.^3 These compounds contain a unique carotenoid-like triterpene skeleton. Along with a steroidal saponin sarainoside A j (93) isolated from the same spongej 4 the pouosides were the first glycosides reported from a sponge. Five of the compounds described in this thesis add to the growing number of triterpenoid compounds isolated from the marine environment. In contrast to terrestrial triterpenoids, many of these compounds have few if any carbocyclic rings. Apart from the holothurian glycosides and the echinolactones, the marine triterpenoids are characterized by cyclization modes very different from those observed in their terrestrial counterparts. - 4 0 -R i R 2 R 3 R4 8 8 Pouoside A OAc Ac H H 8 9 Pouoside B OAc H H H 9 0 Pouoside C H Ac H H 91 Pouoside D OAc Ac Ac H 92 Pouoside E OAc Ac H Ac OH Figure 23. Glycosides isolated from the sponge Asteropus sp. - 4 1 -DISCUSSION ISOLATION OF SECONDARY METABOLITES FROM XESTOSPONGIA VANILLA O n a routine collecting trip at the Bamfield Marine Station, a small colony of what was later identified by B i l l Austin as Xestospongia vanilla was discovered. Guided by our knowledge of the many interesting compounds that had been previously isolated from other members of the genus, and by antifungal and antibacterial activity of the crude extracts of the sponge, a more detailed chemical investigation of this sponge was initiated. G e l permeation chromatography on Sephadex LH-20 separated the crude organic soluble extract of this sponge into two major fractions on the basis of molecular size. The early eluting fraction, containing larger molecules, was not mobile on silica even though it was soluble in relatively non-polar solvents such as dichloromethane and carbon tetrachloride. The later eluting fraction was composed of a mixture of fats, steroids and a few smaller terpenoid compounds. The investigation was first directed towards the smaller terpenoids, as they were easily isolated by conventional silica chromatography. Four new terpenoids, xestodiol (94), xestenone (95), xestolide (96), and secoxestenone (97) were eventually isolated and structures were proposed on the basis of spectroscopy and chemical interconvertions. Biological testing revealed that most of the antifungal and antibacterial activity noted in the crude extracts of this sponge was associated with the larger molecules of the early eluting LH-20 fraction. Further fractionation by reverse - 4 2 -phase flash chromatography revealed a complex mixture of similar compounds from which the triterpene glycosides xestovanin A (98), secoxestovanin A (99), xestovanin B (100), xestovanin C (101), and xestovanin D (102) were isolated. Their structures were proposed on the basis of spectroscopy and chemical degradations. Xestovanin A (98) proved to be active in inhibiting the growth of the fungus Pythium ultimum, while xestovanin C (101) was active against the bacterium Bacillus subtilis. - 4 3 -Scheme 2. Isolation scheme Extraction of sponge with methanol l Liquid/l iquid extraction: CH2CI2 : H2O C H 2 C I 2 layer J G e l permeation chromatography (Sephadex LH-20) 9 : 1 M e O H : H 9 0 / \\ Early elutions Late elutions Xestovanins Fats, steroids smaller terpenes \\ Reverse phase flash chromatography A : 6 5 % M e O H : H 2 0 to 8 0 % M e O H : H 2 0 B : Further purification of fractions with: 5 5 % or 7 0 % acetone : H 2 0 i) Xestovanin A (98) ii) Secoxestovamn A (99) ii i) Xestovanin B (100) iv) Xestovanin C (101) v) Xestovanin D (102) XESTODIOL (94) - 4 4 -The molecular formula of xestodiol ( 9 4 ) was determined from the parent ion observed in the E I H R M S (m/z 308.1981 D a C 1 8 H 2 8 0 4 calc. 308.1989). A l l eighteen carbons were apparent in the * 3 C N M R spectrum and an A P T experiment revealed that 26 protons were attached to carbon. A strong band at 3369 cm\"* in the IR spectrum suggested that the remaining two hydrogens were attached to hydroxyl groups. Three of the five degrees of unsaturation required by the molecular formula of xestodiol were accounted for by a ketone and two double bonds indicated by five deshielded 1 3 C N M R resonances (6198.0 (C), 145.5 (C), 137.5 (CH), 136.0 (CH), and 124.3 (CH)). The remaining two degrees of unsaturation required that xestodiol had a bicyclic structure. A series of double resonance and nOe experiments (see Table 1) established a nine carbon acyclic fragment as illustrated in Figure 25. The protons of a methyl doublet (61.37,d,J=5.5Hz,3H) assigned to position 14 showed scalar coupling to a carbinol methine proton ( H ^ : 64.51,quintet,J=5.5Hz) which in turn was coupled to an olefinic proton ( H j 2 : 66.18,dd,J = 15.1,5.5Hz) with a coupling constant of 5.5 Hz. Further scalar couplings revealed that all three olefinic protons observed in the * H N M R spectrum were part of a contiguous spin system starting from the resonance at 67.04 ( H J Q ) continuing through a proton resonating at 66.64 ( H J J ) and connecting to the previously described olefinic proton resonance at 66.18 ( H 1 2 ) . The chemical shift of the carbonyl carbon (6198.0) and the frequency of the C = 0 stretching band in the IR spectrum (1664 cm\"*) suggested that the ketone was conjugated to a double bond, implying a dienone functionality. The observation of a U V absorption band at 276 nm was consistent with this assignment. The coupling pattern of the olefinic protons established the placement of an olefinic methyl Figure 24. Xestodiol (94) - 4 6 -CH 3 OH nOe correlations Figure 25. Acyclic fragment of xestodiol (94) - 4 7 -(61.92,s,3H) a to the ketone (position 15). A n isolated A B spin system 62.59,d,J = 18.3Hz and H-^: 63.65,d,J = 18.3Hz) was assigned adjacent to the carbonyl on the basis of the chemical shifts of the protons and nOes observed in the downfield olefinic proton (HJQ ) when either of these methylene protons were irradiated. NOe enhancements established both double bonds as E oriented. The * H N M R correlations that established the nature of the acyclic portion of xestodiol are summarized in Figure 25. Another isolated spin system, composed of a carbinol methine between two sets of methylene protons, was elucidated in a similar manner as discussed above (see Figure 26). The coupling constants observed for this system (see Table 1) suggested that it was part of a six membered ring in a chair conformation. The large coupling constant observed between the carbinol methine proton (H3) and an adjacent methylene proton ( H 4 a : 61.79,dd,J = 14.6,9.3Hz) indicated that the alcohol functionality was equatorial. What remained to be assigned were three aliphatic methyl singlets, a quaternary carbon, and two fully substituted ether carbons. Since all other oxygenated carbons had been accounted for, the ether carbons had to be linked through a single oxygen. The chemical shifts of the ether carbons suggested a tetrasubstituted epoxide. Biogenic reasoning lead to the assembly of the fragments into the bicyclic structure illustrated in Figure 24. Support for the proposed constitution of xestodiol and assignment of the relative stereochemistry around the cyclohexane ring came from the comparison of the *H and N M R data of xestodiol (94) with the values reported for fucoxanthin (103) (see Figure 27 and Table 2)?^^^ Fucoxanthin is one of the most abundant carotenoids of the marine environment. It is estimated that several million metric tons of this compound are produced by phytoplankton annually.^8 Since sponges feed primarily upon phytoplankton and bacteria, Xestospongia vanilla probably receives significant - 4 8 -amounts of this compound in its diet, raising the possibility that xestodiol is a degradation product of fucoxanthin. -49 -1.04 0.96 Figure 26. Cyclic fragment of xestodiol (94) 17 OAc Figure 27. Comparison of xestodiol (94) to fucoxanthin (103) Table 1. H N M R data for xestodiol (94) in CDC1 6(ppm) mult.(Hz) decoup.to nOe to nOe from — 1.68 m 3 1.52 ddd(12.6,3.2,1.5) 2a,3,4a 3 3,16, 3.82 m 2a,2b,4a,4b 2b,4a,16 2b, 16, 2.34 ddd( 14.6,4.6,1.5) 2b,3,4b 3,17, 1.79 dd( 14.6,9.3) 3,4a _ 17 3.65 d(18.3) 7b 7b, 10,15 7b,15,16 2.59 d(18.3) 7a 7a, 10,17 7a,17, 7.04 d(11.0) 11 • 7a,7b,12 6.64 ddd(15.1,11.0,1.3) 10,12,13 6.18 dd(15.1,5.6) 11,13 10,13 4.51 quin.(5.5) 11,12,14 12 1.37 d(5.5)3H 13 0.96 s,3H 2b,7a 7a 1.04 s,3H 2b,3,7a 7a 1.22 s,3H 4a,4b,7b 7b 1.92 s,3H 1 2a 2b 3 4a 4b 5 6 7a 7b 8 9 10 11 12 13 14 15 16 17 18 Table 2. Comparison of proton chemical shifts of xestodiol (94) and fucoxanthin (103) # xestodiol fucoxanthin 6(ppm) 6(ppm) 7a 3.65 3.55 7b 2.59 2.55 15 0.96 0.97 16 1.04 1.03 17 1.22 1.20 18 1.92 1.92 HO' OAc fucoxanthin (103) k i j j Ll - i iL .A. • hid L 4 - ^ t . M lil*|l».iAfcJlriirtvkl J j . . a i I . I . J J g . . l l l ju JllJl | | J J i h L u i ^ i u i M L J i j M i i U A i i t i d i i J ^ ^ i i | i i i i i i i i i i i i i i i i i i i i i I I i i i f i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i 200 1B0 160 I 1 ' ' 1 I 140 120 100 JJLIJiJJL i i I i i i i I i i i i I i i i i I i i i i | i i i i I i i i i | i i i i 80 60 40 20 PPM i | l i l l | l l i i | i l l l | l i l i Figure 29. 1 3 C N M R (75 M H z ) spectrum of xestodiol (94) in C D C 1 3 Table 3. Carbon assignments of xestodiol (94) compared to fucoxanthin (103) position xestodiol fucoxanthin e 6(ppm) 6(ppm) 1 35.3 (C) 35.2 2 40.9 ( C H 2 ) a 40.9 3 64.3 ( C H J 64.2 4 41.7 ( C H 7 ) a 41.8 5 66.2 (C)|f 66.2 6 67.0 ( C ) D 67.2 7 47.2 ( C H 2 ) a 47.3 8 198.0 (C) 197.9 9 136.0 (C) 134.5 10 145.5 ( C H ) C 145.1 11 124.3 ( C H ) C 123.4 12 137.5 ( C H ) C 139.2 13 68.3 ( C H ) . 14 23.5 ( C H ? ) ,d 15 21.2 ( C F K ) 5 21.5 16 25.1 ( C H o ) ° 25.0 17 . 2 8 . 2 ( C H o ) ° 28.2 18 1 1 . 8 ( C H 3 ) d 11.9 a-d: values can be exchanged e: The values listedior fucoxanthin are those that correspond to the xestodiol resonances as the C spectrum of fucoxanthin has not been assigned 17 94 xestodiol (94) - 5 4 -XESTENONE (95) The E I H R M S of xestenone (95) showed a parent ion at m/z 288.2024 D a demonstrating a molecular formula of C joF^gO^ (calc. 288.2030) requiring six sites of unsaturation. Six olefinic resonances (6172.3 (C), 144.3 (C), 137.3 (C), 134.4 (C), 119.9 (CH), and 115.6 ( C H ) ) and a carbonyl resonance at 6212.9 in the 1 3 C N M R spectrum (see Table 7) demonstrated the presence of three carbon-carbon double bonds and a ketone functionality, requiring the presence of two rings in the molecular structure of xestenone. Spectroscopic analysis revealed the presence of four fragments (A, B, C, and D Figure 30) that could be linked together to give the proposed structure of xestenone. A six carbon substructure containing a secondary alcohol functionality (fragment A , Figure 30) could be routinely identified. The facile loss of H 2 0 in the E I H R M S of xestenone (95) (m/z 270.2029 Da: C i o ^ g O calc. 270.1984 ) and a strong O H stretching band in the IR spectrum suggested the presence of an alcohol functionality. A * 3 C N M R resonance at 676.3 (CH), and a carbinol methine proton resonating at 64.17 ( H i 2 ) in the * H N M R spectrum that was shifted downfield to 65.22 in the acetate of xestenone (104), determined its secondary nature. A series of double resonance and C O S Y experiments (see Table 4) demonstrated that the carbinol methine ( H i 2 ) proton was coupled to a pair of allylic methylene protons resonating at 62.34 (position 13) which were in turn coupled to an olefinic proton resonating at 65.16 (H14). The olefinic and methylene resonances were both weakly coupled into a pair of olefinic methyl resonances at 61.73 (Mej^) and 61.65 (Me 17) completing this isolated proton spin system. A n nOe enhancement was observed in the methyl resonance at 61.73 when the olefinic resonance at 65.16 was irradiated, allowing the proton resonances of the two terminal methyls to be assigned. The observation of a strong fragment ion peak at m/z 219.1382 D a ( M + --55-xestenone (95) H 4.17 H 5.92 CH 3 1^3 Fragment B O CH 3 1.95 1713 212.9 Fragment C H1.93 H134 Fragment D Figure 30. The fragments of xestenone (95) - 5 6 -C5H9 calc. 219.1386) in the E I H R M S is consistent with the loss, by allylic cleavage, of the olefinic portion of this substructure. Another isolated proton spin system was assigned to a three carbon olefinic substructure (fragment B, Figure 30). A methyl resonance at 61.53 (Mejg) in the * H N M R spectrum showed allylic coupling into the downfield olefinic singlet resonance at 65.92 (H J Q ) . Since no further couplings were observed to either of these resonances, they were assigned to a trisubstituted double bond (fragment B). Strong nOe enhancements between the olefinic proton and the carbinol methine of fragment A established the attachment of the trisubstituted double bond to fragment A and revealed an E configuration for the trisubstituted olefin (see Table 4). The existence of a fully substituted enone system (fragment C, Figure 30) was suggested by evidence in the * 3 C N M R , * H N M R , IR, and U V spectra of xestenone (95). The IR stretching frequency of the ketone carbonyl (1698 cm\"*) along with the presence of a very deshielded olefinic resonance ( C 2 : 6172.6) in the 1 3 C N M R spectrum, the U V absorption 258 nm, and a deshielded olefinic methyl resonance at 61.92 (Mej) in the * H N M R spectrum of xestenone (95), indicated the existence of an a/3 unsaturated ketone with a, B methyl substituent. The N M R resonance of the ketone (Cg: 6212.9) was seemingly inconsistent with this assignment, so confirmation was sought from a sodium borohydride reduction of xestenone. The reduction reaction yielded a mixture of products. One of these, compound 105, was relatively easy to separate by H P L C chromatography. This compound failed to show a parent ion in the E I L R M S but yielded fragment ions of m/z 274 ( M + - H 2 0 ) and 223 ( M + - C5H9) Da, indicating a parent molecular -57-Scheme 3. Reduction of xestenone (95) O ' Xestenone (95) NaBH4 MeOH, 0*C,0.5h Reduction product 105 - 5 8 -formula of C19H32O2. The addition of four hydrogens indicated that this product was the result of the reduction of both the ketone and the double bond conjugated into it. A n expected new carbinol methine resonance (Hg) appeared at 63 .80 in the * H N M R spectrum run in CDCI3. The downfield olefinic methyl resonance of xestenone (95) was replaced by a methyl doublet (60.87,d,J = 7.2 Hz) . The olefinic singlet resonance at 65.92 (H J Q , fragment B) was shifted upfield to 65.58, and became a broad doublet (J = 10.5 Hz) . A series of double resonance experiments (see Table 5) revealed that the new carbinol methine resonance at 63 .80 ppm (Hg) was coupled to a methine resonance (Hg: 63.05,dt,J = 10.5, 7.0 Hz) which was in turn coupled to the down field olefinic doublet (HJQ) and another methine resonance (H2; 62.25,ddq,J = 7.0,8.4,7.2 Hz). The H 2 methine resonance showed coupling to the methyl doublet resonance (60.87,d,J=7.2Hz) assigned to carbon 1. Together with the hydroxyl hydrogen, these three new protons (H2.Hg.H9) accounted for the four sites of hydrogenation suggested by the mass spectrum of compound 105, and confirmed the presence of a fully substituted enone system with a methyl attached to the B carbon in xestenone (95). The 10.5 H z coupling observed between H9 and H J Q established a linkage between the trisubstituted double bond of fragment B, and the a position of the original enone system of fragment C of xestenone. The H g carbinol methine resonance of the reduction product (105) appeared as a triplet (J = 7.0 H Z ) in the * H N M R spectrum run in CDCI3. Proton N M R spectra of this compound run in benzene-dg, however, showed the same proton resonance as a doublet (Hg: 63.72,brd,J = 6.6 Hz) coupled only into the H9 methine resonance. This change in multiplicity suggests that the H g carbinol methine proton shows coupling to the hydroxyl proton in the * H N M R spectra run in CDCI3. The lack of any further vicinal coupling from this methine (other than to H9 and the hydroxyl proton) suggested that the only remaining non-protonated carbon (the quaternary carbon (C7: 654.7 ppm)) bearing the tertiary methyl singlet be placed adjacent to this centre. The H 2 methine proton (attached to what was originally the B position of the enone of xestenone (95)) showed coupling into another proton resonating at 61.97 ppm (H3) indicating a linkage into the methine of the remaining portion of the molecule. In summary, the reduction product 105 confirmed the presence of fragment C in the molecule. It revealed the linkage of the previously described acyclic part of xestenone (fragments A and B) to the a carbon of the enone system, and suggested the placement of the quaternary carbon adjacent to the carbonyl. A comparison of the molecular formula of xestenone (95) to the partially elucidated structures described indicated that a methine, three methylenes, and a quaternary carbon with its associated methyl, remained to be accounted for. These remaining atoms were assembled in a six carbon substructure containing a five membered ring (fragment D, Figures 30 and 31). A one bond H E T C O R experiment assigned a deshielded proton resonance (H3: 62.70,brd,J = 9.3 Hz) to the methine carbon (C3: 656.6). The chemical shift of the proton resonance allowed the assignment of this methine adjacent to the tetrasubstituted double bond (equivalent to the methine proton at 61.97 ppm in the reduction product 105). Evidence obtained from double resonance, C O S Y , * H 2DJ, and one bond H E T C O R experiments demonstrated that the remaining three methylenes formed a contiguous array with the allylic methine. The methine proton (H3) was coupled to pair of protons ( H ^ : 61.78,m and H ^ : 61.68,m) which in turn were coupled to two more protons ( H ^ : 61.58,m and H ^ : 1.24,m) (Table 4). The remaining two methylene protons at 61.93 ( H g a ) and 61.34 ( H ^ ) showed coupling only to the previously mentioned pair, indicating that they were at the end of the spin system, and attached to the non protonated carbon (C7). The remaining linkage between - 6 0 -Schematic diagram of the correlations of fragment D H 3 H 4 a m H 5 a Hga = scalar coupling = nOe correlation One bond HETCOR correlations of fragment D H 3 H 4 a H 5 a H^ 56.6 28.8 24.7 37.4 H 5b Hf5b = HETCOR correlation H168 Hl-58 Hl-24 H193 Hl34 Fragment D Figure 31. Fragment D of xestenone (95) - 6 1 -the allylic methine (G3) and the quaternary carbon (G7) was confirmed by a nOe observed in the H3 methine proton when the methyl protons at 61 .20 ( M e ^ ) were irradiated. This through-space correlation also established a cis relationship between M e 17 and H3. The assembly of the described fragments of xestenone (95) into the proposed structure was guided by a number of spectral features of the natural compound 95 and the reduction product 105. As previously mentioned, nOe enhancements observed between H J Q of fragment B and H ^ of fragment A established the linkage of these two substructures (Table 4). The linkages of fragments B and D to fragment C were best illustrated in the spectra of the reduction product 105 (Table 5). Vicinal coupling between H Q and H J Q of 105 demonstrated the connection of the acyclic substructure (fragments A and B) to the a carbon of the enone system in fragment C of xestenone (95). Similarly, vicinal coupling between the H 2 and H3 methines of 105 confirmed the linkage of the B olefinic carbon of fragment C of xestenone (95) (C2) to the allylic methine carbon, of fragment D (C3). The remaining connection between the ketone carbon (Cg) and the quaternary carbon (C7) was evidenced by an nOe enhancement observed between the H g carbinol methine resonance and the aliphatic methyl singlet resonance ( M e ^ ) of the reduction product (105). Validation of the proposed structure was obtained from S I N E P T correlations, nOes, and comparison to model compounds (see Figure 32). A S I N E P T experiment optimized for polarization transfer via a 7 H z 1 3 C - * H coupling showed a three bond coupling between the olefinic proton at 65.92 (H J Q ) and carbons at 676.3 (C^) and 6172.3 (C2) in agreement with the hypothesized structure. NOes from the deshielded olefinic methyl resonance ( M e i : 61.95) to the - 6 2 -nOe enhancements observed in Xestenone (95) Figure 32. Assembly of the fragments of xestenone (95) - 6 3 -olefinic singlet resonance (H J Q : 65.92), and from the ring junction methyl resonance (Mei7: 61.20) to the allylic methine resonance (H3: 62.70) are also consistent with the proposed ring system. Spectral data previously reported for two model bicyclo [3.3.0] octenones l O l 7 ^ and 102*^ is shown in Figure 33. The chemical shift of the carbonyl resonance of compound 101 is in good agreement with that of xestenone (95); the placement in a five membered ring seems to counteract the normal upfield shift of a carbonyl upon conjugation. Compound 102, an excellent model for the proposed structure of xestenone, showed very good agreement in the IR and *H N M R data. -64-106 IR: 1700 cm 1 1 J g 107 Figure 33. Comparison of xestenone (95) to model compounds Table 4. A H N M R data for xestenone (95) in CDCl # 6(ppm) mult.(Hz) decoup.to nOe to nOe from 1 1.95 s,3H 10 10 10 z 3 2.70 brd(9.3) 4a,4b 17 4a 1.78 m 3,4b,5a 4b 1.68 m 3,4a,5a,5b 5a 1.58 m 4a,4b,5b,6a,6b 5b 5b 1.24 m 4b,5a,6a,6b 5a 6a 1.93 dd(12,6) 5a,5b,6b 5b,6b 6b 6b 7 o 1.34 dt(6,12) 5a,5b,6a 6a 6a o 9 10 5.92 brs 1,3,12,18 12 1,12 11 - - - - -12 4.17 t(6.9) 13 10,13,14 10,14 13 2.34 t(6.9)2H 12,14,16,19 12 14 5.16 t(6.9) 13,16,19 12,13,16 16 15 - - - -16 1.73 s,3H 13,14 14 17 1.20 s,3H 3 18 1.53 s,3H 10 19 1.65 s,3H 13,14 - 6 7 -OH 1 — i — i — i — | — i — i — i — i — I — i — i — i — i — | — i i i i — | — i — i — i — i — j — i — i — i — i — I — i — i — i — i — | — i — i — i — i — 1 — i — i — i — i — 1 — i — i — i — \" j — i — i i i I i I ' Figure 36. J H N M R (300 M H z ) spectrum of the reduction product of xestenone 105 in CDCI3 Table 5. H N M R data for the reduction product of xestenone 105 in CDCI3 # 6(ppm) mult(Hz) scalar cor. nOe t o a nOe f r o m a 1 0.87 d(7.2)3H 2 2 2.25 ddq(7.0,8.4,7.2) 1,3,9 17 3 4 5 1.97 q(8.5) 2 6 7 8 3.80 t(7.0) b 9 17,9 17 9 3.05 dt( 10.5,7.0) 2,8,10 8 10 5.58 brd(10.5) 9,12,18 11 - - - - -12 4.08 brt(6.9) 13 13 2.4-2.2 m,2H 12,14 14 5.09 t(6.9) 13,16,19 15 - - - -16 1.70 s,3H 14 17 1.11 s,3H 8,3 18 1.50 s,3H 10 19 1.62 s,3H 14 a: recorded in benzene-dQ b: appears as a doublet in benzene-d^ Reduction product 105 -70 -XESTOLIDE (96) - 7 1 -The largest ion observed in the E I H R M S of xestolide (96) at m/z 334.2158 D a was assigned as the parent ion with a molecular formula of C2OH30O4 (calc. 334.2145). A l l twenty carbons were visible in the 1 3 C N M R spectrum and an A P T experiment established that 28 protons were attached to carbon (see Table 7). The presence of a strong O H stretch in the IR spectrum (3360 cm\" 1 ) and a M + - 18 fragment ion in the mass spectrum (316.2039 020*^03, c a ^ c - 316.2038) were consistent with the presence of two hydrogens attached to oxygen. O f seven deshielded carbon resonances, one was assigned to an a,B unsaturated ester functionality ( 1 3 C N M R : 6165.5 (C), IR: 1688 cm\" 1 ), while the remaining six were assigned to three double bonds. The remaining two degrees of unsaturation not accounted for by the carbonyl and olefinic functionalities required the structure of xestolide (96) to be bicyclic. A series of C O S Y and decoupling experiments established the presence of two independent spin systems that could be linked by nOes into an acyclic fragment similar to that of xestenone (95) (see Figure 39 and Table 6). The connecting nOes observed between protons H I Q and H i 2 also established the E configuration of the C i j - C12 double bond. This acyclic fragment differed from that of xestenone (95) only by the presence of an extra methylene (CQ) as evidenced by coupling between the downfield olefinic resonance (HJQ: 65.68,dd,J = 10.6,5.5 Hz) and a pair of proton resonances ( H 9 a : 62.66,ddJ = 14.5,5.3 H z and H 9 b : 62.51,dd,J = 14.5,10.6 Hz) assinged to Cg. Corroboratory evidence for this differing acyclic substructure came from the E I H R M S . Both compounds showed intense fragment ion peaks attributed to the loss of the terminal C5H9 fragment (in xestolide: 265.1431 calc. 265.1440). Xestolide (96) also showed an intense fragment ion peak at m/z 181.0871 D a (C1QH13O3 calc. 181.0865) attributed to the loss of the entire acyclic side chain. 'V' Mass spectrum fragments of acyclic fragment The remaining bicyclic portion of xestolide (96) had to contain a fully substituted double bond, ester and hydroxyl functionalities, a ketal carbon, an olefinic methyl, an aliphatic methyl with its associated quaternary carbon, and three methylenes. The relatively shielded carbonyl (6165.5 (C)) and substituted olefinic (6117.5 (C)) resonances, together with the deshielded olefinic resonance at 6164.7, suggested that the olefinic and ester functionalities were conjugated. The frequency of the C = 0 stretching band in the IR spectrum (1688 cm\" 1 ) and the presence of a chromophore in the U V spectrum of xestolide (96) (233nm, € = 7260) confirmed this assignment. The oxygen count of this substructure required that the hydroxyl and ester functionalities both be connected to the ketal carbon. In a manner similar to the structural elucidation of xestenone (95), a series of C O S Y correlations demonstrated that the three methylenes existed as an isolated, contiguous spin system. The obvious biogenic similarity between xestolide (96) and xestenone (95) led to the assembly of the identified components into the proposed structure illustrated in Figure 38. Evidence confirming the proposed structure was found in a pair of S I N E P T experiments optimized for polarization transfer through C - H couplings of 7 Hz. Irradiation of the downfield olefinic methyl proton resonance ( M e i j : 61.80) gave polarization through three bonds to the ester carbonyl resonance ( C j : 6165.5 (C)). This correlation confirmed the placement of the olefinic methyl in the a position. Irradiation of the protons on the aliphatic methyl (Me^g: 61.14) gave polarization transfer through three bonds to the ketal (Cg: 6105.3 (C)) and B olefinic (C3: 6164.7 (C)) carbons. This correlation confirmed the placement of the quaternary carbon adjacent to the ketal and B olefinic carbons, forming the six membered ring system and requiring the remaining methylenes to form a five membered ring completing the structure. Further support for this structure came from a prominent - 7 4 -90 80 71 60 50 40 30 20 10 • I 11 i i 240 I I I I I I I iZ8 4¥ T 280 I I I I I I I I I | I I I I I I 3*0 | I I I I I I 320 | I I I I 340 90 80 70 ( 0 50 40 30 10 0 • M ' I ' I ' I ' P . ' M ' I A»7 I T n ' i ' i ' I ' I ' i ' i I ' i ' i ' i 1 1 ' I * 111' I' t' T • I T 180 Figure 40. EI mass spectrum of xestolide (96) Scheme 4. Retro Diels-Alder reaction of xestolide (96) - 7 5 -ion in the E I H R M S . The peak at m/z 136.0885 D a ( C 9 H 1 2 0 , calc. 136.0889) could be attributed to a fragment formed by a retro Diels Alder fragmentation as shown in Figure 40 and Scheme 4. Irradiation of the aliphatic methyl protons assigned to Me induced nOes into both methylene protons attached to Cg, establishing the cis relative stereochemistry as illustrated. No evidence was obtained in any of the N M R spectra of xestolide (96) for the existence of the trans isomer that should be in equilibrium with the cis compound through the open chain form of the Cg hemiketal. That such an equilibrium does exist was demonstrated by the rapid conversion of xestolide (96) to the keto methyl-ester 103 (see Figure 43) upon treatment with diazomethane. The extra thermodynamic stability of the cis isomer may be due to an anomeric effect. Table 6. A H N M R data for xestolide (96) in CDC1 # 6 (ppm) mult.(Hz) decoup.to nOe to nOe from 1 2 3 4 2.51 m,2H 5,17 -5 1.97 m,2H 4,6a,6b 6a 2.11 m 5 6b 1.62 m 5 7 - - - -8 _ _ 9a 2.66 dd(14.5,5.3) 9b,10,19 18 9b 2.51 dd(14.5,10.6) 9a, 10,19 18 10 5.68 dd(10.6,5.3) 9a,9b,19 12 11 - - - -12 4.12 t(6.0) 13 10 13 2.2-2.4 m,2H 12,14,16,20 14 5.08 t(6.9) 13,16,20 15 - - • 16 1.71 s,3H 13,14 17 1.80 s,3H 4 18 1.14 s,3H 9a,9b 19 1.67 s,3H 9a,9b,10 20 1.64 s,3H 13,14 Xestolide (96) 1 1 1 1 1 1 1 1 1 1 1 1 [ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 200 180 160 140 120 100 80 60 40 20 PPM 0 Figure 42. \" C N M R (75 M H z ) spectrum of xestolide (96) in CDC1 Table 7. Carbon assignmentsaof xestenone (95) and xestolide (96) in CDCI3 position xestenone xestolide 6(ppm) <5(ppm) 1 16.7 ( C H , ) 165.5 (C) 2 172.3(C) 117.2(C) 3 56.6 (CH) 164.7 (C) 4 28.8 (CH?) 29.3 (CH?) 5 24.7 (CH?) 22.5 (CH?) 6 37.4 (CH?) 34.0 (CH?) 7 54.7 (C) 49.5 (C) 8 212.9 (C) 105.3 (C) 9 137.3(C) 34.8 (CH?) 10 115.6 (CH) 117.0 (CH) 11 144.3(C) 144.1(C) 12 76.3 (CH) 76.8 (CH) . 13 33.9 (CH?) 34.1 (CH?) 14 119.9 (CH) 119.6 (CH) 15 134.4 (C) 135.3 (C) 16 25.9 ( C H , ) 26.0 ( C H , ) 17 22.5 ( C H , ) 13.0 ( C H , ) 18 14.4 ( C H , ) 21.3 ( C H , ) 19 18.0 ( C H , ) 12.1 ( C H , ) 20 - 18.0 (CH3) a: protonated carbons assigned by H E T C O R b: may be exchanged 14 1 0 Xestolide (96) SECOXESTENONE (97) -81-The highest mass ion observed in the E I H R M S of secoxestenone (97) was at m/z 288.2081 D a ( C 1 9 H 2 g 0 2 calc. 288.2090). Examination of the 1 3 C N M R spectrum of this compound revealed 19 carbon atoms, in agreement with the mass spectrum. The presence of two carbonyl resonances (6212.6 and 210.6) and a deshielded carbon (677.2) clearly demonstrated that this molecule contained three oxygen atoms. The carbinol methine carbon, and the presence of an O H stretching band in the IR spectrum ( 3444 cm**) suggested that secoxestenone contained a secondary alcohol functionality. It was assumed that the highest observable ion in the mass spectrum was due to the loss of water from the parent ion and that the true molecular formula was CJ9H3QO3. The chemical shifts of the carbonyl carbons and the frequency of the C = 0 stretch in the IR spectrum (1705 cm\" 1 ) established the presence of two saturated ketone functionalities in the molecule. Four more deshielded carbon resonances (6139.9, 134.7, 119.9, and 118.2) were assigned to two double bonds. The remaining degree of unsaturation required by the molecular formula implied a monocyclic structure. The * H N M R spectra of secoxestenone (97) were very similar to those of xestolide (96). The same acyclic fragment was identified by a series of C O S Y , decoupling, and nOe experiments (see Table 8). Relative to xestolide (96), the C9 methylene proton resonances of secoxestenone (97) were shifted downfield, (63.34 and 3.26 ) leading to assignment of an adjacent ketone functionality at position 10, replacing the ketal of xestolide (96). A sharp methyl singlet at 62.16 in the * H 13 N M R spectrum and the second carbonyl resonance of the lJC N M R spectrum were assigned to a methyl ketone residue (Cj and C 2 ) . C O S Y correlations established the presence of an isolated spin system comprised of three contiguous methylenes and a downfield methine similar to fragment D of xestenone (95). The assembly of - 8 2 -Secoxestenone (97) 1.59 9H3 3.3.2H 1.63 CH 3 3.34,326 1.71 Chemical shifts of protons on the acyclic fragment 2.86 2.10,1.88 H H 1.79,1.64 H 224,1.% Chemical shifts of protons on the cyclic fragment Figure 44. Secoxestenone (97) - 8 3 -the fragments into the proposed structure illustrated in Figure 44 was guided by the presumed relatedness of this compound to xestenone (95) and xestolide (96). Upon irradiation of the aliphatic methyl singlet protons (Me-^) resonating at 61.29 nOes were observed into the methine proton resonance at 62.86 (H3) and the Cg methylene proton resonances (63.34 and 3.26). These nOe enhancements confirmed the placement of the quaternary carbon (C7) between the Cg ketone and the C3 methine as well as establishing the cis relationship between the C17 aliphatic methyl carbon and the methine proton on the five membered ring (H3). The proposed structure of secoxestenone (97) suggested that this compound was related to xestenone (95) through an intra-molecular Aldol condensation and subsequent dehydration. To confirm the proposed structure, secoxestenone (97) was treated with hot aqueous N a O H . T L C analysis and an * H N M R spectrum of the reaction mixture indicated that secoxestenone (97) had been converted to xestenone (95) clearly demonstrating the chemical relationship between these two compounds as illustrated in Scheme 5. Scheme 5. Conversion of secoxestenone (97) to xestenone (95) secoxestenone (97) 84-NaOH, H 2 0 xestenone (95) LA* . I—i—l—i—i—i—•—I—i—i—i—r-j|—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—|—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—]—r 1 r T PPM Figure 45. ! H N M R (300 M H z ) spectrum of secoxestenone (97)in C D C 1 3 00 Table 8. A H N M R data for secoxestenone (97) in CDCI3 # 6(ppm) mult.(Hz) decoup.to nOe to nOe from 1 2.16 s,3H 3 2 _ _ _ 3 2.86 dd(8.9,5.0) 4a,4b l,4a,17 17 4a 2.10 m 3,4b,5a 3 4b 1.88 m 3,4a,5a 5a 1.79 m 4a,4b,5b,6a 5b 1.64 m 5a,6a 6a 2.24 m 5a,5b,6b 6b 1.96 m 6a 7 - - - - -8 _ _ _ 9a 3.34 dd(18.4,6.7) 9b,10,18 17,10,18 17 9b 3.26 dd(18.4,6.7) 9a, 10,18 17,10,18 17 10 11 5.61 t(6.7) 9a,9b,18 12 9a,9b,12 11 12 4.05 t(6.7) 13 10 10 13 3.3 m,2H 12,14 14 5.09 t(7.3) 13,16,19 15 - - - -16 1.71 s,3H 14 17 1.29 s,3H 3,9a,9b 3 18 1.63 s,3H 9a,9b,10 9a,9b 19 1.59 s,3H 14 o Secoxestenone (97) I I 1 I I I I I I I I I I I 1 I I I I I I I I I I I I 1 I I I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I I I I I I I I I 220 200 180 160 140 120 100 80 60 40 20 PPM 0 Figure 46. 13C N M R (75 M H z ) spectrum of secoxestenone (97) in CDC13 - 8 8 -XESTOVANINA (98) The highest mass ion observed in the E I H R M S of xestovanin A (98) was m/z 746.4592 Da, consistent with a formula C 4 2 H 6 6 O n (calc. 746.4605). The F A B M S , however, showed a strong peak at m/z 787 D a ( M + + Na) indicating that the true molecular weight was eighteen mass units higher. It was assumed that the molecule lost water in the E I H R M S and, therefore, that its true molecular formula was ^42^68^12 requiring nine degrees of unsaturation. A l l 42 carbons were observed in the * 3 C N M R spectrum, and an A P T experiment revealed 61 protons attached to carbon (see Table 13). The * H N M R spectrum in M e 2 S O - d ^ showed seven resonances that were absent from the * H N M R spectrum in acetone-d^; six doublets (65.03,d,J = 4.0Hz, 64.92,d,J=4.3Hz, 64.67,d,J = 4.1Hz, 64.64,d,J = 5.8Hz, 64.57,d,J = 4.2Hz, 64.48,d,J = 5.7Hz), and a singlet (64.40,s) (see Table. 11). These resonances could be attributed to hydroxyl protons on six secondary alcohols and the hydroxyl of a tertiary alcohol. The IR spectrum of xestovanin A (98) showed a very strong hydroxyl O H stretch (3391 cm\"*) consistent with a polyhydroxylated compound. Formation of hexaacetate 109 upon treatment of xestovanin A (98) with acetic anhydride and pyridine ( 1 3 C N M R : 6170.8, 170.1, 170.1, 169.8, 169.6, 169.1; * H N M R (CDC1 3 ): 62.07,s,3H, 2.06,s,3H, 2.03,s,3H, 2.01,s,6H) provided further support for this assignment. The hexaacetate of xestovanin A 109 still showed a hydroxyl O H stretch in the IR spectrum (3385 cm\"*) which was attributed to the tertiary alcohol that was not acetylated under mild conditions. A n examination of the JC N M R spectrum and an A P T experiment on the hexaacetate of xestovanin A 109 confirmed that all but one of the protons in the molecule were attached to carbon (Table 13). One degree of unsaturation in xestovanin A (98) could be attributed to a saturated ketone which was indicated by an IR band at 1701 cm\"* and a carbonyl - 8 9 -Figure 47. Xestovanin A ( 9 8 ) - 9 0 -resonance at 6212.8 in the N M R spectrum. Eight olefinic resonances in the 1 3 C N M R spectrum (5139.4, 133.7, 132.2, 131.3, 126.3 (CH), 121.3 (CH), 121.2 (CH), 120.1 (CH)) indicated the presence of four double bonds, accounting for four more degrees of unsaturation. The large number of carbinol methine resonances (10) and the presence of two ketal methine resonances (6101.2, 98.6) in the 1 3 C N M R spectrum suggested the existence of two monosaccharides in the molecule. Acid hydrolysis of a few milligrams of xestovanin A (98) yielded a water soluble fraction that was identified tentatively by T L C analysis as a mixture of fucose and rhamnose. Examination of C O S Y spectra run in acetone-d^ (Figure 57) and Me2SO-dg (Figure 58) confirmed the existence of two separate spin systems, each composed of a methyl doublet resonance, four carbinol methine resonances, and an anomeric methine resonance respectively. These experiments firmly established the presence of two 6-deoxyaldohexose moieties in the molecule. The rings of the two sugars accounted for another two degrees of unsaturation. Subtraction of the twelve carbons and eight oxygens of the sugars from the molecular formula, indicated that the remaining part of the molecule (the aglycone) was a bicyclic C^Q structure with four oxygen atoms attached. Due to the large size of the molecule and the presence of very similar fragments, it was not possible to resolve all the proton resonances and their coupling details in one solvent system. By comparing proton spectra run in Me 2 SO-d6 and acetone-d^ on the natural product (98), and the proton spectra run in CDCI3 on the hexaacetate 109, it was possible to construct consistent fragments and assign all the proton resonances. The * H N M R spectra run in Me2SO-d^ were particularly useful because scalar couplings were observable to the otherwise exchangeable hydroxyl -91 -Proton chemical shifts of fragment A in acetone-d^ 1.61 1.54 Proton chemical shifts of fragment B in acetone-dg 1.08 CH 3 H H 1.47 139 Proton chemical shifts of fragment C in acetone-d* Figure 48. Aglycone of xestovanin A (98) -92 -protons. Due to the greater viscosity of dimethylsulfoxide, and size of xestovanin A (98), nOe enhancements were difficult to observe in this solvent. Acetone-d^ was used to observe nOe enhancements on the natural product, while CDCI3 was used for the same purpose with the hexaacetate 109. Unless otherwise noted in the text, the chemical shifts of the protons will be reported for the natural compound in acetone-d^, however, almost all of the protons could be assigned in all three solvents (see Tables 10,11, and 12). Analysis of the various * H N M R spectra led to the identification of several large fragments that could later be assembled into the complete structure of the aglycone of xestovanin A . Most of the atoms of the aglycone could be placed into one of three fragments: two very similar C I Q fragments representing the two uncyclized ends of the squalene precursor (fragments A and B, Figure 48), and a fragment forming part of the cyclic middle portion (fragment C, Figure 48). Fragment A The first fragment of the aglycone to be identified was composed of two isolated spin systems that were linked together across a double bond on the basis of nOe enhancements. The first spin system contained a carbinol methine proton (H5: 64.02,t,J=7.2Hz) coupled to two allylic methylene protons ( F L i a : 62.32,m and H ^ : 62.28,m) as evidenced by C O S Y correlations and double resonance decoupling (Table 10). In the Me2SO-d^ C O S Y spectrum, the carbinol methine proton (H5) did not show a correlation to a hydroxyl proton, indicating a linkage through oxygen to another part of the molecule at this position (see Figure 49). The allylic methylene protons ( H ^ , H ^ ) were coupled to each other, to an olefinic proton (H3: 65.10,t,J = 7.2Hz), and via homo-allylic coupling, to two olefinic methyls: Figure 49. Me2SO-d Q C O S Y spectrum of xestovanin A (98) spin systems of fragment A - 9 4 -(Me^: 61.65,s,3H and Me25: 61.60,s,3H). The olefinic triplet resonance (H3) also showed allylic coupling to the two olefinic methyl resonances. The second spin system of fragment A contained an olefinic proton (H7: 65.44,brd,J = 10.0Hz) that showed vicinal coupling to a pair of allylic methylene protons ( H g a : 62.43,dd,J = 15.1,9.5Hz, and H g D : 62.22,m) and allylic coupling to an olefinic methyl (Me26: 61.63,s,3H). The allylic methylene protons showed no further coupling indicating that the next carbon in the chain had no attached protons. NOes (negative in Me2SO-d Q , positive in acetone-dg) observed between the carbinol methine proton of the first spin system (H5) and the olefinic proton of the second (H7) revealed the linkage between the two spin systems across an E trisubstituted double bond. Fragment B Another very similar substructure (fragment B, Figure 48) could be identified by examining the C O S Y spectra and nOe experiments. In this fragment the first spin system differed from that of fragment A only at the carbinol methine resonance (H20) which showed vicinal coupling to a hydroxyl proton resonance (64.67,d,J=4.1Hz) in the C O S Y spectrum run in N ^ S O - d ^ (see Figure 50). The presence of a secondary alcohol at this position was confirmed by the nearly 1 ppm downfield shift of this carbinol methine resonance (from 64.20 to 65.17) in the hexa-acetate of xestovanin A (109). The pair of methylene resonances of the second spin system of fragment A were replaced in fragment B by an allylic methine resonance ( H 1 7 : 63.75,d,J = 9.9Hz). In the M e 2 S O - d 6 C O S Y spectrum, this methine resonance showed a weak coupling to the hydroxyl singlet of the tertiary alcohol. This correlation was attributed to a long range W coupling, indicating that the - 9 6 -1.65 H H 1.63 2.432.22 1.60 Fragment A of xestovanin A (98) *H chemical shifts (acetone dg) 5.05 Q H Fragment B of xestovanin A (98) H chemical shifts (acetone d^ ) Figure 51. Model diterpene compound -97-tertiary alcohol was vicinal to this methine proton and that they were held in an anti relationship to each other. As before, the two spin systems of fragment B were linked through a E trisubstituted double bond as evidenced by nOes observed between the carbinol methine proton of the first spin system (H20) a n d t n e olefinic proton of the second (Hjg). 12-(S)-hydroxygeranylgeraniol (110), an acyclic diterpene with an oxidation pattern similar to the two uncyclized fragments (A and B) of xestovanin A (98), has been isolated from a brown alga.^ The proton resonances from the methyl terminal end to the second olefinic proton were in good agreement with the two proposed substructures of xestovanin A (98) (see Figure 51). In particular, the downfield olefinic proton is coupled to the single upfield methyl, while the upfield olefinic proton is coupled to the two terminal methyls. Fragment C With twenty of thirty carbons of the aglycone accounted for, attention turned to the bicyclic substructure. In this portion of the aglycone an aliphatic methyl and the remaining two methine and six methylene protons could be assembled in a single spin system (fragment C, Figure 48). Although parts of this spin system could be resolved in the C O S Y spectra of xestovanin A (98) run in M ^ S O - d g and acetone-dg, the correlations were most distinct in the C O S Y spectrum of the hexaacetate 109 (see Figure 52). The upfield methyl doublet resonance (acetate 109: M e 2 7 : 61.08,d,J=7.0Hz) showed coupling to a methine, (HIQ: 62.27,dq,J=5.8,7.0Hz) which in turn showed coupling to another proton: (H11: 61.99,m). Examination of a one bond H E T C O R experiment run on the upfield region of Xestovanin A (98) in M ^ S O - d g showed that the H j j resonance in the - 9 8 -Figure 52. C O S Y spectrum of the hexaacetate of xestovanin A 109 in CDCI3 correlations of fragment C -99-proton spectrum was connected to a methine resonance at 655.0 in the carbon spectrum. Careful examination of the C O S Y spectrum of the hexaacetate (see Figure 52) revealed that the remaining three methylene proton pairs formed a contiguous linear array, coupling into the methine resonating at 61.99 ( H J J ) at one end, and being adjacent to a non protonated carbon ( C ^ ) at the other. In summary, fragment C is an isolated proton spin system containing six carbons (a methyl, two methines, and three methylenes) arranged as shown in Figures 48 and 52. Assembly of the Aglycone Fragments Fragments A , B, and C contained 26 of the 30 carbons, and 43 of the 47 protons attributable to the aglycone. What remained were the ketone, the tertiary alcohol, and the tertiary methyl with its associated quaternary carbon. The final assembly of the aglycone fragments was guided by biogenic reasoning and by the obvious similarity of some of the fragments to parts of the previously isolated terpenoids from this sponge. It was possible to construct two candidate structures (I and II Figure 53) possessing five membered rings similar to secoxestenone (99) with no rearrangements of a triterpene skeleton. Both structures were consistent with the data discussed so far. A number of nOe enhancements and two SINEPT experiments revealed that structure II was correct. The downfield 1 3 C N M R resonance (5212.8) and the ketone stretching frequency of 1701 cm\"* in the IR spectrum were consistent with a cyclohexanone. The long range coupling betweeri the methine proton of fragment B and the tertiary hydroxyl proton allowed these two functionalities to be placed 1,2 trans diaxial on a chair-like six membered ring. Irradiation of the protons of the methyl doublet resonance (Me2SO-d^: Me27: Proposed Aglycone Structure I O H Proposed Aglycone Structure JJ nOe correlations around six membered ring Figure 53. Structural proposals for the aglycone of xestovanin A - 1 0 1 -60.95) in a SINEPT experiment optimized for a J Q H °* ^rlz led to the enhancement of the 1 3 C N M R resonance (682.1) assigned to the tertiary alcohol. A second SINEPT experiment ( J ^ - H = 7Hz) showed polarization transfer from the methine of fragment B (Hj^) to both the tertiary alcohol and the ketone carbon resonances. These results clearly placed the tertiary alcohol between the methine carbon of fragment B (C17) and the methyl carbon at end of fragment C (Me27). Observation of an nOe enhancement in the methine at 61.99 ( H l l ) in the hexaacetate 109, when the methyl doublet resonance (Me 2 7) was irradiated, confirmed the cis ring fusion. Similar nOe enhancements to the remaining two methine protons in the aglycone (H10 and H17) were assigned to 1,3-diaxial interactions between these protons and the tertiary methyl, establishing the relative stereochemistry around the six membered ring. Elimination Product (111) Further confirmation of the proposed structure of the aglycone came from the spectroscopic analysis of an elimination product obtained by treatment of xestovanin A (98) with aqueous potassium hydroxide. The presence of a /3-hydroxy ketone in the structure suggested that the six membered ring was the result of an Aldol condensation. The treatment with base in an excess of water was an attempt to perform a retro Aldol reaction to produce a monocyclic product (secoxestovanin A (99)). Analysis of the reaction products by T L C , however, revealed a product that migrated on silica gel in non-polar solvents clearly demonstrating that the aglycone portion of the molecule had separated from the sugar residues. Spectroscopic analysis of the non-polar fraction revealed a double elimination product 110 (Figure 54). The E I H R M S of compound 110 showed a parent ion at m/z 436.3336 Da, corresponding to a formula of C 3 0 H 4 4 O 2 . Examination of the 1 3 C N M R and IR - 102 -4 . 1 8 O H Elimination product (111) formed on treatment of xoiovanin A (98) with aqueous base Figure 54. Elimination product of xestovanin A 111 spectra revealed the presence of a conjugated ketone ( N M R : 6204.4; IR: 1651 cm\" 1 ) and a secondary alcohol ( 1 3 C N M R : 676.8(CH); IR: 3410 cm\" 1 ). The ketone, along with six double bonds indicated by the twelve olefinic carbons (6C, 6CH) displayed in the 1 3 C N M R spectrum, accounted for seven of the nine degrees of unsaturation indicated by the molecular formula. The remaining unsaturation required the elimination product 110 to be bicyclic. The presence of a conjugated ketone, two extra double bonds, and only two oxygen atoms suggested that compound 110 was formed from xestovanin A (98) by the 1-2 elimination of the elements of water from the /3-hydroxy ketone and 1-4 elimination of the sugar moiety from C5 of the side chain (fragment A ) . A n olefinic proton resonance, assigned to H^g on the basis of a C O S Y correlation into a single olefinic methyl resonance (Me 2Q: 61.46,s), and a nOe correlation to the carbinol methine proton (H 2Q: 64.18,t,J = 6.4Hz) of fragment B, appeared as a broad singlet indicating that the Hy] methine proton of xestovanin A (98) had been eliminated. The removal of the tertiary alcohol functionality at Cg was evidenced by a downfield shift of the H J Q methine (xestovanin A-acetate 109: 62.27, compound 111: 63.01) and a homoallylic coupling from H J Q to H i g across a double bond. A pair of downfield olefinic doublet resonances which were coupled to each other (66.58,d,J = 16.4Hz, 66.29,d,J = 16.4Hz) (see Table 9) were assigned to H7 and H g respectively of an E-disubstituted double bond on the basis of their chemical shifts and nOe enhancements to neighboring protons (from H7 to H5 and H j n ; from H g to M e 2 ^ and M e 2 9 ) . A methylene resonance (62.86,t,J = 7.2Hz,2H) was assigned to C4 as it coupled to two olefinic protons (H3 and H5). The remaining spectroscopic features of the elimination product were consistent with the proposed structure. Compound 111 confirmed the presence of a /3-hydroxy ketone in the six membered ring of xestovanin A and firmly established the attachment of the side chain bearing the sugar residues adjacent to the methyl doublet (C 2 7). i oLuJ i i i i i I —1 i 1 1 1 1 1 1 1 i 1 i 1 1 1 1 1 1 1 1 1 1 1— 6.8 6.6 6.« 6.2 6.0 5.B 5.6 5.4 5.2 5.0 4.8 4.6 «.« 4.2 4. 0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.« 2.2 2.0 l .B 1.6 1 1 1 2 1 0 PPH Figure 55. % N M R (400 M H z ) spectrum of the elimination product of xestovanin A 111 in CDCI3 - 105 -Table 9. A H N M R data for elimination product 111 in CDCI3 # 6(ppm) mult.(Hz) C O S Y corr. nOe to nOe from 1 1.70 s,3H 3,4 3 z 3 5.12 t(6.5) 1,4,25 1 5 4 2.86 t(7.2)2H 1,3,4,25 5 5 f. 5.55 t(7.4) 4,26 3,4,7,10(-) 7,10(-) U 7 6.58 d(16.4) 8 5,10 5,10 8 Q 6.29 d(16.4) 7 26,29 10 3.01 qnt(6.7) 11,27 5(-),7,11,27 5(-),7,ll 11 2.18 dd(13.8,6.9) 10 10,12a,28 10,28 12a 1.95 m 11,12b 11 12b 1.60 m 12a 13a 1.66 m 14a, 14b 13b 14a 2.23 m 13a,14b 14b 14b 1.56 m 13a,14a 14a 15 - - - - -16 - - - - -17 - - - - -18 6.13 brs 29 20 19 - - - - -20 4.18 t(6.4) 21 18,21,22,29 22,18 21 2.35 m,2H • 20,22 20 22 5.22 t(6.3) 21,24,30 20,24 20 23 - - - -24 1.74 s,3H 22 25 1.65 s,3H 3,4 26 1.79 s,3H 4,5 8 27 1.15 d(6.8)3H 10 10 10 28 1.16 s,3H 11 11 29 1.46 s,3H 18 8 30 1.65 s,3H 22 25 26 27 Elimination Product of Xestovanin A 111 i i | i i i i | i i i i | i i i i | i i i i | i i i i | i i 60 40 20 PPM O - 107 -Sugar Residues Close examination of the C O S Y spectra obtained on Xestovanin A (98) and its acetate (109) revealed two further spin systems attributable to the two sugar residues. Starting at the upfield anomeric resonance of one of the sugar residues, it was possible to follow correlation peaks around the ring to the 6-deoxy methyl resonance: (acetone-d6: H r : 64.17, H 2>: 63.56, H3.: 63.58, H 4>: 63.82, H 5>: 63.61, Me^>: 61.29) (see Table 10). A n examination of the C O S Y spectrum run in M ^ S O - d ^ revealed, in addition to a similar coupling pattern, correlations to hydroxyl protons for protons H2> and H3 ' (2 'OH: 64.57, 3 ' O H : 63.46) (see Figure 57). The lack of hydroxyl couplings to Hy and indicated that these were linkage positions in this sugar residue. Correspondingly, in the hexaacetate 109, which also displayed a similar correlation series, only and H3 ' were shifted downfield by the acetylation ( H 2 ' : 65.20, Hy: 64.93) (see Table 12). A series of double resonance decoupling experiments carried out on both the natural product in acetone-d^ and on the hexaacetate in CDCI3 revealed most of the coupling constants between the carbinol methine protons and helped establish the relative stereochemistry around the ring. A large coupling constant between Hy and H2> (8Hz) indicated a rram-diaxial relation between these two protons. A large coupling to H3 ' (10Hz) and a small coupling from H3 ' to placed these two protons as axial and equatorial respectively. These stereochemical relationships are consistent with a fucose system. The proposed stereochemical relationships within this ring system were confirmed by nOe experiments on the hexaacetate 109. The 1-3 diaxial relation between protons H ^ , H3' , and H^- was indicated by nOe enhancements between them (see Table 11). - 108 -Fucose system Figure 57. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) fucose spin system - 109-Rhamnose system Figure 58. M e 2 S O - d 6 C O S Y spectrum of xestovanin A (98) rhamnose spin system The connectivities, relative stereochemistry, and linkages for the second sugar residue, rhamnose, were worked out in a similar manner. Hydroxyl couplings and acetate shifts demonstrated that this sugar was linked only through the anomeric center (see Figure 58). Large coupling constants from H ^ to H y and H3» implied that all three were axial, while small coupling constants from H2\" to H3\" and H i \" reflected H^-'s equatorial orientation. These relative configurations, which were confirmed by nOe enhancements (see Tables 11 and 12), were consistent with a rhamnose ring system. At this stage it was necessary to establish how the two sugar residues and aglygone were linked and determine the stereochemistry of the anomeric centres. A series of nOe experiments provided the linkage information. NOes observed between the anomeric proton of the fucose ring and H5 (carbinol methine on fragment A ) established that the sugars were linked to the aglycone at position 5 through the anomeric position of the fucose, leaving rhamnose to be linked via its anomeric center to position 4 of fucose as was demonstrated by nOes (Hi<> to Hy). The large coupling between Hi> and H2 ' of the fucose system dictated axial assignments to both of these protons. Placement of Hj» axial established the linkage from the fucose to the aglycone as a /3 linkage. The small coupling constant (2 Hz) displayed by the anomeric proton of the rhamnose ring did not establish an a configuration for the remaining linkage, however, as the equatorial orientation of H2\" would cause a small coupling to Hj« regardless of the configuration at the anomeric centre. By measuring the magnitude of the one-bond carbon to hydrogen coupling constant it is possible to distinguish between a and B l inkages.^ To this end, a fully coupled carbon spectrum was obtained for xestovanin A (98) and the one-bond J ^ - H coupling constants for the anomeric carbons were measured. Using Figure 59. Sugar linkages of xestovanin A (98) Fucose - 112 -H E T C O R results to assign the resonances, it was possible to confirm that the fucose was B linked, and determine that the glycosidic linkage between the two sugars was a (rhamnose C I : 6101.2, 1 J C . H 171 Hz, fucose Cl:698.6, 1 J C . H 159 Hz). To confirm the assignment of the sugar portion of the molecule, and to determine the absolute configurations of these sugars, glycosides of the free sugars were formed with a chiral alcohol. After acetylation the resulting compounds have G C retention times that depend on the absolute stereochemistry of the sugar. The water soluble material formed when xestovanin A (98) was treated with aqueous potassium hydroxide was hydrolyzed with aqueous trifluoroacetic acid. The resulting monosaccharides were reacted with ( + ) or (-) 2-octanol in the presence of a catalytic amount of trifluoroacetic acid and then treated with acetic anhydride/pyridine to produce acetylated-2-octyl glycosides. Similarly, acetylated-( + )-2-octyl and acetylated-(-)-2-octyl glycosides of D- and L-fucose and L-rhamnose standards were prepared. Comparison by capillary G C analysis led to the identification of L-rhamnose and D-fucose in the mixture obtained from xestovanin A (98). 8 3 These arguments established the structure of xestovanin A (98) to be that shown in Figure 47 with the absolute stereochemistry of the sugar residues as shown. The relative configurations of the chiral centers in the bicyclic portion of the molecule are as depicted, however, since their absolute configurations have not been determined, the set of configurations shown for these centers is arbitrary. It was not possible to determine the configurations at C5 or C ^ Q -- 114 -- 115 -Table 10. 1 H N M R data for xestovanin A (98) in acetone-d, # 6 (ppm) mult.(Hz) C O S Y corr. nOe to nOe from 1 1.65 s,3H 3 2 _ _ 3 5.10 t(7.2) l,4a,4b,25 4a 2.32 m 3,5 4b 2.28 m 3,5 5 f, 4.02 t(7.3) 4a,4b o 7 5.44 brd(10.0) 8a,8b,26 8a 2.43 dd(15.1,9.5) 7,8b,26 8b Q 2.22 m 7,8a,26 y 10 2.36 m 11,27 28 11 2.06 m 10,12a, 12b 28 12a 2.01 m 12b, 13a, 13b 12b 1.62 m 12a,13a,13b 13a 1.47 m 12a,12b,13b,14a,14b 13b 1.39 m 12a,12b,13a,14a,14b 14a 2.58 ddd(12.9,6.1,5.1) 13a,13b,14b 14b 1.17 m 13a,13b,14a 15 - - - - -16 - - - - -17 3.75 d(9.9) 18 28 18 5.63 d(9.9) 17,29 19 - - - - -20 4.20 dd(8.3,6.4) 21a,21b 21a 2.26 m 20,21b,22 21b 2.21 m 20,21a,22 22 5.05 t(6.9) 21a,21b,24,30 23 - - - -24 1.66 s,3H 22 25 1.60 s,3H 3 26 1.63 s,3H 7,8a,8b 27 1.08 d(6.9) 10 28 1.38 s,3H 17,10,11 29 1.54 s,3H 18 30 1.61 s,3H 22 Table 10. (continued) 1 H N M R data for xestovanin A (98) in acetone-d 6 # 6 (ppm) mult.(Hz) C O S Y corr. nOe to nOe from r 4.17 d(7.3) 2' 2' 3.56 m r 3' 3.58 m 4' 4' 3.82 br d(3.0) 3\\5' 5' 3.61 m 4\\6' 6' 1.29 d(6.4) 5' 1\" 5.27 d(1.2) 2\" 2\" 4.02 br s 1\",3\" 3\" 3.71 dd 2',4' 4\" 3.41 t(9.1) 3\",5\" 5\" 3.80 m 4\",6\" 6\" 1.21 d(6.2) 5\" - 119-Table 11. A H N M R data for xestovanin A (98) in M e 2 S O - d ( # 6 (ppm) mult.(Hz) C O S Y corr. nOe to nOe from 1 1.60 s,3H 3,4a,4b L 3 4.93 t(7.0) l,4a,4b,25 4a 2.25 mult l,3,4b,5,25 4b 2.20 mult l,3,4a,5,25 5 f. 4.04 dd(8.7,5.9) 4a,4b 7(-) 7(-) u 7 5.22 br d(8.1) 8a,8b,26 5(0 5(-) 8a 2.24 mult 7,8b 8b o, 1.97 mult 7,8a 9 0 H 4.40 s 17 10 2.07 dq(5.5,6.2) 11,27 11 1.86 mult 10 12a 1.83 mult 13a 12b 13a 1.32 mult 12a, 14a, 14b 13b 14a 2.43 ddd(13.0,5.5,8.0) 13a, 14b 14b 1.10 mult 13a,14a 15 - - - -16 - - - - -17 3.59 d(9.9) 18,90H 18 5.46 br d(9.6) 17,29 20(-) 20(-) 19 - - - -20 3.90 dt(4.1,7.5) 21a,21b,20OH 18(-),20OH(-) 18(-) 20OH 4.67 d(4.1) 20 20 21a 2.17 mult 20,21b,22,24 21b 2.05 mult 20,21a,22,24 22 4.96 t 21a,21b,24,30 23 - - - - -24 1.58 s,3H 21a,21b,22 25 1.55 s,3H 3,4a,4b 26 1.53 s,3H 7 27 0.95 d(6.2)3H 10 28 1.35 s,3H 29 1.41 s,3H 18 30 1.53 s,3H 21a,21b,22 - 120 -Table 11. (continued) X H N M R data for xestovanin A (98) in M e 2 S O - d 6 # 6 (ppm) mult.(Hz) C O S Y corr. nOe to nOe from 1' 3.97 d(7.3) 2' 2' 3.28 ddd(9.6,7.3,4.3) 1',2'OH 2 ' O H 4.92 d(4.3) 2' 3' 3.31 ddd(9.6,4.0,1.8) 4',3'OH 3 ' O H 5.03 d(4.0) 3' 4' 3.61 br d(1.8) 3',5' 5' 3.39 br q(6.3) 4',6' 6' 1.15 d(6.3)3H 5' 1\" 5.10 d(2.0) 2\" 2\" 3.76 ddd(4.2,3.1,2.0) l\",3\",2\"OH 2\"OH 4.57 d(4.2) 2\" 3\" 3.46 ddd(9.3,5.7,3.1) 2\",4\",3\"OH 3\"OH 4.48 d(5.7) 3\" 4\" 3.18 dt(5.8,9.3) 3\",5\",4\"OH 4\"OH 4.64 d(5.8) 4\" 5\" 3.56 dq(9.2,6.2) 4\",6\" 6\" 1.10 d(6.2)3H 5\" 6\" OH s r OH 1 1 1 1 1 1 1 j 1 1 1 1 1 1 1 1 1 j 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 200 ISO 160 140 120 100 80 60 40 2( ' ' I | I 0 P P M Figure 64. 1 3 C N M R (75 M H z ) spectrum of xestovanin A (98) in M e 2 S O - d 6 to Fully coupled 1 3 C N M R (75 M H z ) Spectrum of Xestovanin A (98) to determine the UQ.H Coupling Constants of the Anomeric Centres ( i n M e 2 S O d 6 ) i I i i i i I i i i i i i i i i I i i i i I i i i i I i i i i i i i i i I i i i i i i i i i I i i i i i i i i i I i i i i i i i i i I i i i i i i i i 106 104 102 100 98 96 94 PPM Figure 65. 1 3 C N M R (75 M H z ) spectrum of the anomeric region of xestovanin A (98) in Me 2 SO-d6 - 1 2 4 -- 125 -PPM Offset Irrad. Figure 68. Some nOe difference experiments (400MHz) on xestovanin A hexaacetate 109 in CDCI3 - 126-Table 12. A H N M R data for xestovanin A hexaacetate 109 in CDCI3 # $ (ppm) mult.(Hz) C O S Y corr. nOe to nOe from 1 1.67 s,3H 3,4a L 3 5.02 br t(7.2) l,4a,4b,25 4a 2.42 mult l,3,4b,5,25 4b 2.23 mult 3,4a,5 5 f. 4.09 dd(9.5,5.0) 4a,4b 7,l',4a 7,1\\3' 0 7 5.27 br t(6.7) 8a,8b,26 5,17 5 8a 2.28 mult 7,8b 8b Q 2.15 mult 7,8a y 10 2.27 dq(5.8,7.0) 11,27 11 1.99 mult 10,12a,12b 10,12b,27,28 27,28 12a 1.83 mult ll,12b,13a,13b 12b 1.69 mult ll,12a,13a 13a 1.55 mult 12a, 12b, 13b, 14a, 14b 13b 1.41 mult 12a,13a,14a,14b 14a 2.64 ddd(13.1,8.7,5.3) 13a,13b,14b 14b 14b 14b 1.20 mult 13a,13b,14a 14a 14a 15 - - - - -16 - - - - -17 3.60 d(9.8) 18 7,10,28,29 7,28 . 18 5.63 br d(9.8) 17,201r,929 20 19 - - - - -20 5.17 br t(6.7) 21a,21b 21a 2.42 mult 20,21b,22 21b 2.28 mult 20,21a,22 22 4.92 t 21a,21b,24,30 30 23 - - - - -24 1.65 s,3H 22 25 1.61 s,3H 3 26 1.50 s,3H 7 27 1.08 d(7.0)3H 10 28 1.35 s,3H 10,11,17 29 1.52 s,3H 18 30 1.60 s,3H 22 - 127 -Table 12. (continued) 1 H N M R data for xestovanin A hexaacetate 109 in C D C 1 3 # 6 (ppm) mult(Hz) C O S Y corr. nOe to nOe from r 4.32 d(8.0) 2' 5,3',5' 5,3',5' 2' 5.20 dd(10.4,8.0) l',3' 3' 4.93 dd(10.4,3.1) 2',4' 5,1',4\\5' l',4',5' 4' 3.82 br d(3.1) 3' 3',5M\" 3\\4\\6' 5' 3.56 br q(6.4) 6' l',3',4' l',3',6' 6' 1.34 d(6.4)3H 5' 4\\5\\5\" 1\" 4.84 d(1.6) 2\" 4*,2\" 4',2\" 2\" 5.42 mult 1\",3\" 1\" 1\" 3\" 5.40 mult 2\",4\" 4\",5\" 5\" 4\" 5.08 t(9.5) 3\",5\" 3\",6\" 5\" 4.08 dq(6.0,9.5) 4\",6\" 3\",4\",6\" 3\",6\" 6\" 1.20 d(6.0)3H 5\" 4\",5\" 5\" 6\" OAc Xestovanin A Hexaacetate 109 s | I M I | M M j M I I | M I M I I I I | l l l l j l l l l | l l l l l | M I | l l l l j l M I | l l l l | l l l l | I I I I | l l l [ | l l l l | l l l l | l l l l | l l l l | l l l l | l l l l | l l l l | 220 200 180 160 140 120 100 80 60 40 20 PPM 0 Figure 69. N M R (75 M H z ) spectrum of xestovanin A hexaacetate 109 in CDC1, >1 F l (PPM) - 129 -F2 (PPM) Figure 70. One bond H E T C O R spectrum of the downfield region of xestovanin A hexaacetate 109 in CDCI3 Table 13. Carbon assignments for xestovanin A (98) and its hexaacetate 109 xestovanin A (98) hexaacetate 109 # inMe 2 SO-5 E J L R M S m/z (rel. int.) 746 (0.5), 728 (1.0), 582 (31.8), 436 (17.7), 367 (23.2) Determination of the absolute stereochemistry of the sugars: 10 mg of xestovanin D was hydrolyzed with 3 M aqueous trifluoroacetic acid at 60 °C for 4 hr. The peracetylated (-) octylglycosides were prepared in the usual manner. The mixture and standards were dissolved in C H C 1 3 and analyzed by capillary G C (DB17 column; temperature program: 180°C for 2 min, increase at 5 ° C per min, final temperature 220°C). The retention times and relative peak intensities of the mixtures were compared to those observed for the standards. The observed retention times were as follows: peracetylated ( + )-octyl-L-fucoside: 9.81, 10.56 min; peracetylated (-)-octyl-L-rhamnoside: 8.91, 9.68 min; mixture of sugars from secoxestovanin A : 8.90, 9.66, 9.80,10.55. B I B L I O G R A P H Y Bergquist, P. R. Sponges 1978, University of California Press, Berkeley and Los Angeles Ireland, C. M . ; Rol l , D. M . ; Molinski, T. R ; McKee, T. C.;Zabriskie, T. M ; swersey, J . C. in Biomedical Importance of Marine Organisms E d . Fautin, D. G. 1988, California Acad, of Sciences Memoirs 13,41-57 Van Soest, R. W. M . Marine Sponges from Curacao and other Caribbean Localities (Part II Haplosclerida) in Studies on the Fauna of Curacao and other Caribbean Islands, 191 De Laubenfels, M . W. Proc. of the Nat. Museum 1930,81 57-171 Austin, W. C.An Annotated Checklist of Marine Invertabrates in the Cold Temperate Northeast Pacific, 1985 Inst, of Ocean Sciences, Sidney, B. C. Austin, W. C. Personal communication, 1989 Hartman, W. D, In \"Synopsis and Classification of Living Organisms\" Parker, S. P., Ed. ; McGraw-Hil l : New York, 1982, Vol.1, 640-666 Stoilov, I. L.; Thompson, J . E.; Cho, J . ; Djerassi, C.I. Am. Chem. Soc. 1986, 108, 8235-8241 Kokke, W. C. M . C ; Tarchini, C ; Stierle, D. B.; Djerassi, C.7. Org. Chem. 1979,44, 3385-3388 Goad, L. J . in Marine Natural Products Vo l . II Scheuer, P. J . Academic Press, 1978, New York Theobald, N . ; Djerassi, C. Tetrahedron Lett. 1978, 45, 4369-4372 -214-12 Sica, D.; Zollo, F. Tetrahedron Lett. 1978, 9, 837-838 13 Mattia, C. A . ; Mazzarella, L.; Puliti, R.; Sica, D. Zollo, F. Tetrahedron Lett. 1978, 41, 3953-3954 14 Ravi, B. N . ; Kokke, W. C. M . C ; Delesh, C ; Djerassi, C. Tetrahedron Lett. 1978,45, 4379-4380 15 L i , L. N . ; Djerassi, C. /. Am. Chem. Soc. 1981,103, 3606-3608 16 L i , L. N . ; Sjostrand, U.; Djerassi, C. /. Org. Chem. 1981,46, 3867-3870 17 L i , L. N.; Sjostrand, U.; Djerassi, C.J. Am. Chem. Soc. 1981,103, 15-19 18 Khalil , M . W.; Djerassi, C. Steroids, 1980,35, 707-719 19 Cho, J.; Djerassi, C. /. Chem. Soc, Perkin Trans. 11987, 1307-1318 20 Schmitz, F. J . ; Gopichand, Y . Tetrahedron Lett. 1978, 39, 3637-3640 21 Quinn, R. J . ; Tucker, D. J . Tetrahedron Lett. 1985,26, 1671-1672 22 Hirsh, S.; Carmely, S.; Kashman, Y . Tetrahedron 1987,43, 3257-3261 23 Fusetani, N . ; Kato, Y . ; Matsunaga, S.; Hashimoto, K. Tetrahedron Lett. 1983, 24, 2111-211A 24 Fusetani, N . ; Sugano, M . ; Matsunaga, S.; Hashimoto, K. Tetrahedron Lett. 1987,25,4313-4314 -215 -25 Castiello, D.; Cimono, G.; De Rosa, S.; De Stefano S.; Sodano, G . Tetrahedron Lett. 1980,21, 5047-5050 26 Fusetani, N . ; Sugano, M . ; Matsunaga, S.; Hashimoto, K. Tetrahedron Lett. 1987,25, 4311-4312 27 Fusetani, N . ; Yasukawa, K.; Matsunaga, S.; Hashimoto, K. Tetrahedron Lett. 1987,25, 1187-1190 28 Braekman, J . C ; Daloze, D.; Macedo de Abreu, P.; Piccinni-Leopardi, C ; Germain, G.; Van Meerssche, M . Tetrahedron Lett. 1982,23, 4277-4280 29 Braekman, J . C ; Daloze, D.; Defay, N . ; Zimmermann, D. Bull. Soc. Chim. Belg. 1984,95, 941-944 30 Braekman, J.C.; Daloze, D.; Cimino, G.; Trivellone, E . Bull. Soc. Chim. Belg. 1988, 97, 519-524 31 Nakagawa, M . ; Endo, M . Tetrahedron Lett. 1984,25, 3227-3230 32 Cimino, G.; Spinella, A . ; Trivellone, E . Tetrahedron Lett. 1989,30, 133-136 33 Cimino, G. ; De Stefano, S.; Scognamiglio, G.; Sodano, G.; Trivellone, E . Bull Soc. Chim. Belg. 1986, 95, 783-800 34 Sakai, R.; Higa, T.J. Am. Chem. Soc. 1986,108, 6404-6405 35 Nakamura, H . ; Deng, S.; Kobayashi, J . ; Ohizumi, Y . Tetrahedron Lett. 1987, 28, 621-624 36 Sakai, R.; Kohmoto, S.; Higa, T. Tetrahedron Lett. 1987,25(45), 5493-5496 37 Ichiba, T.; Sakai, R.; Kohmoto, S.; Saucy, G . Tetrahedron Lett. 1988,29, 3083-3086 -216-38 Gulavita, N . K.; Scheuer, P. J . /. Org. Chem. 1989,54, 366-369 39 McKee, T. C ; Ireland, C. M . /. Nat. Prod. 1987,50, 754-756 40 Fusetani, N . ; Yasukawa, K.; Matsunaga, S.; Hashimoto, K. Tetrahedron Lett. 1987 25, 1187-1190 41 Molinski, T. F.; Fahy, E.; Faulkner, D. J.J. Org. Chem. 1988,53, 1340-1341 42 Ichiba, T.; Sakai, R.; Kohmoto, S.; Saucy, G . Tetrahedron Lett. 1988,29, 3083-3086 43 Rol l , D. M . ; Scheuer, P. J . ; Matsumoto, G. K.; Clardy, J . J. Am. Chem. Soc. 1983,105, 6177-6178 44 Nakamura, H ; Hirata, Y . Chem. Lett. 1985, 713-715 45 Schmitz, F. J . ; Bloor, S. J . /. Org. Chem. 1988,53, 3922-3925 46 Braekman, J . C.; Daloze, D.; Hulot, G.; Tursch, B.; Declercq,J. P.; Germain, G.; Van Meerssche, U.Bull. Soc. Chim. Belg. 1978,87, 917-926 47 Ravi, B. N . ; Perzanowski, H . P.; Ross, R. A . ; Erdman, T. R.; Scheuer, P. J . Pure and Appl. Chem. 1919,51, 1893-1900 48 Kohomoto, S.; McConnell, O. J.; Wright, A,;Koehn, F.jThompson, W.; Lui , M . ; Snader, K. M . /. Nat. Prod. 1987 50,336 49 Bergquist, P. R.; Hofheinz, W.; Oesterhelt, G . Biochemical Systematics and Ecology 1980,8, 423-435 50 Kigoshi, H . ; Ojika, M . ; Shizuri, Y. ; Niwa, H . ; Yamada, K. Tetrahedron Lett. 1982,23,5413-5414 -217-51 Blunt, J . W.; Hartshorn, M . P.; McLennan, T.J.; Munro, M . H . G . Tetrahedron Lett. 1978, 69-72 52 Suzuki, T.; Suzuki, M . ; Furusaki, A . ; Matsumoto, T.; Kato, A . ; Imanaka, Y . ; Kurosawa, E . Tetrahedron Lett. 1985,26,1329-1332 53 Suzuki, T.; Takeda, S.; Suzuki, M . ; Kurosawa, E.; Kato, A . ; Imanaka, Y . Chem. Lett. 1987, 361-364 54 Shinichi, S.; Higa, T. Tetrahedron Lett. 1986, 27, 4287-4290 55 Kitagawa, I.; Nishino, T.; Kyogoku, Y . Tetrafiedron Lett. 1979,16, 1419-1422 56 Kitagawa, I.; Nishino, T.; Matsuno, T.; Akutsu, H . ; Kyogoku, Y . Tetrahedron Lett. 1978,11 985-988 57 Kitagawa, I.; Yamanaka, H . ; Kobayashi, M . ; Nishino, T.; Yosioka, I.; Sugawara, T. Chem. Pharm. Bull. 1978,26, 3722-3731 58 Kitagawa, I.; Inamoto, T.; Fuchida, M . Okada, S.; Kobayashi, M . ; Nishino, T.; Kyogoku, Y . Chem. Pharm. Bull. 1980,28, 1651-1653 59 Kitagawa, I.; Kobayashi, M . ; Inamoto, T.; Fuchida, M . ; Kyogoku, Y . Chem. Pharm. Bull 1985,33,5214-5224 60 Kitagawa, I.; Kobayashi, M . ; H r o i , M . ; Kyogoku, Y . Chem. Pharm. Bull. 1981, 29,282-285 61 Garneau, F.; Simard, J . ; Harvey, O.; ApSimon, J . W.; Girard, M . Can. J. Chem. 1983, 61,1465-1471 218 -62 Zurita, M . B.; Ahond, A . ; Poupat, C ; Potier, P. J. Nat. Prod. 1986, 49, 809-813 63 Sanduja, R.; Weinheimer, A . J . ; Alam, M . ; Hossain, M . B.; van der Helm, D. 7. Chem. Soc., Chem. Commun. 1984, 1091-1093 64 Yunker, M . B.; Scheuer, P. J . J. Am. Chem. Soc. 1978,100, 307-309 65 de Silva, E . D.; Scheuer, P. J . Tetrahedron Lett. 1980,21, 1611-1614 66 de Silva, E. D.; Scheuer, P. J . Tetrahedron Lett. 1981, 22, 3147-3150 67 Ravi, B. N . ; Wells, R. J . Aust. J. Chem. 1982, 35, 39-50 68 McCabe, T.; Clardy, J . ; Minale, L.; Pizza, G ; Zollo, F.; Riccio, R. Tetrahedron Lett. 1982, 23, 3307-3310 69 Shmueli, U . ; Carmely, S.; Groweiss, A . ; Kashman, Y . Tetrahedron Lett. 1981, 22, 709-712 70 Carmely, S.; Kashman, Y . /. Org. Chem. 1987,45,3517-3525 71 Carmely, S.; Loya, Y . ; Kashman, Y . Tetrahedron Lett. 1983,24, 3673-3676 72 Carmely, S.; Kashmam, Y . /. Org. Chem. 1986,51, 784-788 73 Ksebati, M . B.; Schmitz, F J . , Gunasekera, S. P. / Org. Chem. 1988,53, 3917-3921 74 Schmitz, F. J . ; Ksebati, M . B.; Gunasekera, S. P.; Agarwal, S.J. Org. Chem. 1988,53, 5941-5947 - 2 1 9 -75 Bonnett, R.; Mallams, A . K.; Spark, A . A . ; Tee, J . L.; Weedon, B. C. L.; McKormick, A . / . Chem. Soc.(C) 1969, 429-454 76 Galasko, G.; Hora, J . ; Toube, T. P.; Weedon, B. C. L.; Andre, D.; Barbier, M . ; Lederer, E.; Villanueva, V . R.J. Chem. Soc. (C) 1969, 1264-1265 77 De Vi l le , T. E.; Hursthouse, M . B.; Russell, S. W.; Weedon, B. C. L, Chem. Comm. 1969,1311-1312 78 Liaaen-Jensen, S. In \"Marine Natural Products\" Scheuer, P. J . , Ed . ; Academic Press: New York 1978 V o l . II, 1-73 79 Whitesell, J . K.; Mathews, R. S. J. Org. Chem. 1977,42, 3878 80 Manzardo, G. G.; Karpf, M . ; Drieding, A . S. Helv. Chim. Acta 1986, 69, 659 81 Vails, R.; Banaigs, B.; Francisco, L.; Codomier, L.; Cave, A . Phytochemistry 1986 ,25, 751-752 82 Liptak, A . ; Nanasi, P.; Neszmeli, A . ; Wagner, H Tetrahedron 1980,36, 1261-1268 83 Leontein, K; Lindberg, B.; Lonngren,/. Carbohydr. Res. 1978, 62, 359-362 84 Kuhler, T. C ; Lindsten, G . R. / Org. Chem. 1983, 48, 3589 PUBLICATIONS Peter T. Northcote and Raymond J . Andersen, X e s t o l i d e and Secoxestenone, Degraded Triterpenoids from the Sponge Xestospongia vanilla, Can. J. Chem. 6 7 , 1359 (1989). Peter T. Northcote and Raymond J. Andersen, Xestovanin A and Secoxestovanin A, Triterpenoid Glycosides with New Carbon Skeletons from the Sponge Xestospongia vanilla, Journal of the American Chemical Society, 111, 6276 (1989). Peter T. Northcote and Raymond J. Andersen, Xestenone, a New B i c y c l i c C Terpenoid from the Marine Sponge Xestospongia vanilla, Tetrahedron L e t t e r s , V o l. 29, No. 35, 4357-4360 (1988). Peter T. Northcote and Raymond J. Andersen, X e s t o d i o l , a New Apocarotenoid from the Sponge Xestospongia vanilla, Journal of Natural Products, V o l . 50, No. 6, 1174-1177 (1987). "@en ; edm:hasType "Thesis/Dissertation"@en ; edm:isShownAt "10.14288/1.0060261"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Chemistry"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Novel terpenoid metabolites from the marine sponge xestopongia vanilla"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/29253"@en .