AN ABSTRACT OF THE THESIS OF
Namthip Sitachitta for the degree ofDoctor ofPhilosophy in Pharmacy presented on March 9. 2000. Title: Natural Products Studies ofthe Marine
Cyanobacterium Lyngbva majuscula Redacted for Privacy Abstract approved: --~--~------.c-7------William H. Gerwick
The marine cyanobacterium Lyngbya majuscula has proven to be extraordinarily rich in bioactive secondary metabolites. This dissertation describes the chemistry ofL. majuscula collected from Grenada, Fiji and Papua New Guinea, and the biosyntheses of two L. majuscula metabolites, curacin A and barbamide.
The chemical studies with a Grenada collection ofL. majuscula revealed three new metabolites, grenadadiene, debromogrenadadiene and grenadamide. These three compounds are the only reported cyclopropyl-containing fatty acids from Lyngbya species.
The chemistry of a mixed assemblage ofL. majuscula/Schizothrix species from
Fiji was investigated and shown to contain two novel depsipeptides, yanucamides A and
B. Both compounds possessed the unique 2,2-dimethyl-3-hydroxy-7-octynoic acid, a unit that has only been described in the structures ofkulolide-1 and kulokainalide-1, metabolites from the marine mollusk Phi/inopsis speciosa.
Chemical investigation ofthe highly brine shrimp toxic extract ofPapua New
Guinea collection ofL. majuscula led to the isolation ofthe previously described cytotoxins, curacins A and D. Upon further investigation ofthe same extract, two new depsipeptides, clairamide and carliamide, were discovered. Clairamide contains 3-amino
2-methyl-pentanoic acid, a component unique to cyanobacterial metabolites, while carliamide possesses the 3-amino-2-methyl-7-octynoic acid, a unit that has only been found in the structure of onchidin A, a metabolite from the marine mollusk Onchidium sp.
Biosynthetic investigations ofbarbamide, a unique trichloromethyl-containing metabolite, were carried with the cultured L. majuscula. Results from the feeding experiments have established the biosynthetic units ofthe compound. In addition, isotope-incorporation studies also revealed that barbamide biosynthesis involves chlorination that exclusively occurs at the unactivated pro-S methyl group of leucine.
The biosynthesis ofcuracin A was also examined. Stable isotope feeding experiments have illustrated the cysteine-initiated (or a thiazoline acyl CoA-initiated) polyketide chain assembly of curacin A with C 17 and the OCH3 arising from methionine.
Moreover, the labeling pattern ofacetate at C 18-C22 of curacin A is consistent with the five-carbon unit deriving from a branched triketide-derived precursor or isopentyl diphosphate (IPP)/dimethylallyl diphosphate (DMAPP). ©Copyright by Namthip Sitachitta March 9, 2000 All Rights Reserved Natural Products Studies ofthe Marine Cyanobacterium Lyngbya majuscula
by
Namthip Sitachitta
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the degree of
Doctor ofPhilosophy
Presented March 9, 2000 Commencement June 2000 Doctor ofPhilosophy thesis ofNamthip Sitachitta presented on March 9, 2000
APPROVED:
1\/1 II Redacted for Privacy Major Professor, Pharmacy
Redacted for Privacy Dean ofthe cpllege ofPharmacy 1
Redacted for Privacy
Dean of Gr~te'School -
I understand that my thesis will become part ofthe permanent collection ofOregon State
University libraries. My signature below authorizes release ofmy thesis to any reader upon request.
Redacted for Privacy
~ Namthip Sitachitta, Author ACKNOWLEDGEMENTS
First and foremost I would like to express my gratitude to my major advisor, Dr.
Bill Gerwick, for his guidance and support throughout my studies at Oregon State
University. He did a superb job ofmentoring me and made my research experience both enjoyable and productive. I would also thank my graduate committee members, Drs.
Christopher Mathews, Philip Proteau, Mark Zabriskie and Bruce Coblentz for their advice and support.
I would like to acknowledge Rodger Kohnert for his assistance with NMR experiments, and Brian Arbogast for providing mass spectral data. I am grateful to Dr. T.
Murray (U. ofGeorgia) for performing the cannabinoid binding assay and Dr. V. Di
Marzo (the National Institute for the Chemistry ofBiological Systems, C.N.R., Italy) for conducting the FAAH assays. I would like to acknowledge the National Cancer Institute for performing the cytotoxicity assays. I thank Dr. Christine Willis (University of Bristol,
UK) for providing the synthetic chirally 13C-labeled precursors. My very special appreciation goes to Mary Ann Roberts for her talents in cultivating of marine cyanobacteria, which have been essential to my research.
My graduate career would not be complete without the support, assistance and friendship from my fellow labmates. I especially thank R. Thomas Williamson for his helpful insights into my research and his assistance with NMR experiments. My thanks also go to Lisa Nogle, Drs. George Constantine, Tatsufurni Okino and Inder Pal Sigh for critically reading this manuscript.
I would like to thank the people in my private life who always been there for me and deserve to share my accomplishment. I am forever indebted to my parents for their love, guidance and support. I am grateful to my husband Wesley Yoshida for his loving support, patience and encouragement over the past years. CONTRIBUTION OF AUTHORS
Dr. William H. Gerwick was involved in the design, analysis, and writing ofeach manuscript. R. Thomas Williamson assisted in data collection for Chapters III and V.
Claire Dickie, Carly Me Elligot and Lik Tan Tong assisted in obtaining data for Chapter
IV. Mary Ann Roberts, James V. Rossi, Dr. Matthew D. Fletcher (U. ofBristol, UK) and
Dr. Christine L. Willis (U. ofBristol, UK) were involved in conduction the experimental work for Chapters V and VI. The insecticidal assays against tobacco bud worm (Heliothis virescens) that appear in Appendix A were performed by Dow Agro Science. The sea urchin assays that appear in Appendix C were conducted in Dr. Robert Jacobs' laboratory
(UCSB). TABLE OF CONTENTS
CHAPTER I: GENERAL INTRODUCTION 1
Marine Natural Products 2
Natural Products from Cyanobacteria 5
Biosynthetic Studies ofNatural Products from Cyanobacteria 11
Natural Products Studies ofthe Marine Cyanobacterium 16 Lyngbya majuscula
CHAPTER II: GRENADADIENE AND GRENADAMIDE, CYCLOPROPYL CONTAINING FATTY ACID METABOLITES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
Abstract 17
Introduction 18
Results and Discussion 20
Experimental 34
CHAPTER III: YANUCAMIDES A AND B, TWO NOVEL DEPSIPEPTIDES AN ASSEMBLAGE OF THE MARINE CYANOBACTERIA SCHIZOTHRIX SPECIES ANDLYNGBYA MAJUSCULA
Abstract 37
Introduction 38
Results and Discussion 40
Experimental 50
CHAPTER IV: CLAIRAMIDE AND CARLIAMIDE, NOVEL DEPSIPEPTIDES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
Abstract 53 TABLE OF CONTENTS (Continued)
Introduction 54
Results and Discussion 56
Experimental 68
CHAPTER V: THE BIOSYNTHESIS OF BARBAMIDE
Abstract 71
Introduction 72
Results and Discussion 75
Experimental 95
CHAPTER VI: THE BIOSYNTHESIS OF CURACIN A
Abstract 104
Introduction 105
Results and Discussion 109
Experimental 137
CHAPTER VII: CONCLUSION 144
BIBLIOGRAPHY 148
APPENDICES
APPENDIX A: THE ISOLATION OF MICROCOLINS A, B 157 AND C FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
APPENDIX B: THE ISOLATION OF MALYNGAMIDE F 163 ACETATE, MALYNGAMIDES H, I, K AND INDOLE DERIVATIVES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA TABLE OF CONTENTS (Continued)
APPENDIX C: THE ISOLATION OF KALKITOXIN AND 170 CYMOPOL FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA LIST OF FIGURES
Figure Page
I.1 The Biosynthetic Building Blocks ofMicrocystin-LR (5). 12
I.2 A Proposed Biosynthetic Pathway of(2R, 3S)-3-Methylaspartic 12 Acid (L-Masp).
I.3 A Proposed Biosynthetic Pathway of Saxitoxin Derivatives 14
I.4 Incorporation Patterns of Labeled Precursors into Borophycin (24), 15 Boromycin (25) and Aplasmomycin (26).
I.5 Incorporation Patterns of Labeled Precursors into Tolytoxin (27). 16
11.1 Structures of Fatty Acids and Their Derivatives Produced by 19 L. majuscula.
11.2 EI-Mass Spectrum of Grenadadiene (34). 22
11.3 HMQC Spectrum of Grenadadiene(34). 23
11.4 Partial Structures of Grenadadiene(34). 24
11.5 1H NMR Spectra of (A) Grenadadiene (34) and (B) 27 Debromogrenadadiene (35).
11.6 1H-1H COSY Spectrum ofGrenadamide (36). 28
11.7 Partial Structures of Grenadamide (36). 29
11.8 NCI 60 Cell Line Tumor Growth Inhibition Response Curves 32 for Grenadadiene (34).
11.9 NCI 60 Cell Line Growth Inhibitor Mean Graph for Grenadadiene 33 (34) (GI Growth Inhibitor, TG =Total Growth Inhibition).
111.1 Partial structures ofYanucamide A (39). 41 LIST OF FIGURES (Continued)
Figure Page
111.2 1H_IH COSY spectrum ofYanucamide A (39). 42
111.3 HMBC spectrum ofYanucamide A (39). 43
111.4 1H NMR ofYanucamides A (39) and B (40) in Methanol-d4 48
IV.1 Natural Products Isolated from a Single Extract ofL. majuscula 55 Collected from Cura~ao.
IV.2 Partial Structures ofClairamide (51). 57
IV.3 TOCSY Spectrum ofClairamide (51). 58
IV.4 HMBC Spectrum ofClairamide (51). 59
IV.5 Partial Structures ofCarliamide (52). 63
IV.6 TOCSY Spectrum ofCarliamide (52). 64
IV.7 HMBC Spectrum ofCarliamide (52). 65
V.l Biogenesis ofBarbamide (47) from Primary Precursors. 74
V.2 Comparison of Selected Regions ofthe 13C NMR Spectra of(A) 77 Natural Abundance Barbamide (47), and (B) Barbamide (47) Produced During Supplementations with [1-13C]-L-Leucine and [2-13C]-L-Leucine.
V.3 Comparison ofSelected Regions of (A) Natural Abundance 78 Barbamide (47), (B) Barbamide \47) Produced During Supplementation with (4S)-L-[5- 3C]Leucine, (C) Barbamide (47) Produced During Supplementation with (4R)-L-[5-13C]Leucine.
V.4 Proposed Formation of4,4,4-Trichloroisovalerate (58) via Leucine 79 Catabolic Pathway.
1 V.5 Comparison ofSelected Regions ofthe (A) H NMR (toluene-d8) 80 with the 2H NMR (toluene-Hs) ofBarbamide (47) Produced During Supplementation ofCultures with eHIO]-L-Leucine. LIST OF FIGURES (Continued)
Figure Page
V.6 Summary ofResults from the 13C-Labeled Leucine and 3(R)-[4- 82 13C]Methylbutanoic Derivatives Feeding Experiments to Study the Biosynthesis ofBarbamide (47).
V.7 Comparison of Selected Regions ofthe 13C NMR Spectra of 83 Barbamide ( 47) Produced During Supplementation with [l,2-13C2]Acetate (A and C), and Natural Abundance Barbamide (47) (Band D).
V.8 Comparison of Selected Regions ofthe 13C NMR Spectra of(A) 84 Barbamide ( 47) Produced During Supplementation with 13 18 [1- C, 0 2]Acetate, and (B) Natural Abundance Barbamide (47).
V.9 Comparison of Selected Regions ofthe 13C NMR Spectra of(A) 86 Barbamide ( 47) Produced During Supplementation with L-[3-13C] Phenylalanine, and (B) Natural Abundance Barbamide (47).
V.lO Metabolic Relationship of(A) Glycine and Serine/Cysteine, 87 (B) Glycine and Methionine, and the Expected Labeling Pattern in Barbamide ( 47) Produced During Supplementation of [2-13C,15N]Glycine.
V.ll (A) Schematic Representation ofthe Expected Coupling Patterns 90 from the Modified GHNMBC Experiment, and (B) GHNMBC Spectrum Acquired for Barbamide (47) Produced During Supplementation with [2-13C, 1 ~]glycine.
V.12 Comparison ofthe 13C NMR Spectra of(A) Barbamide (47) 92 Produced During Supplementation with [2-13C, 15N]Glycine, and (B) Natural Abundance Barbamide (47).
V.13 Summary of 13C-Labeled Precursors Incorporated into 93 Barbamide ( 47).
V.14 Revised Biogenesis ofBarbamide ( 47) from Primary 94 Precursors.
VI.l Three Alternatives for the Polyketide Chain Assembly Leading 106 to Curacin A (15). LIST OF FIGURES (Continued)
Figure
VI.2 Possible Biosynthetic Precursors to the C18-C22 Five-Carbon 108 Fragment of Curacin A (15).
VI.3 The Be NMR Spectrum ofCuracin A (15) Produced During 112 Supplementation with Sodium [ 1,2-BC 2] Acetate.
VI.4 The Be NMR Spectrum ofCuracin A (15) Produced During 114 18 Supplementation with Sodium [ 1-BC, 0 2]Acetate.
VI.5 Incorporation of [1-13C]Acetate into Isoprenoids: (A) Glyoxylate and 115 Tricarboxylic Acid Cycles. (B) Glyceraldehyde-3-phosphate/Pyruvate Pathway. (C) Classical Mevalonate Pathway.
VI.6 Tricarboxylic Acid Cycle (TCA Cycle). 116
VI.7 Schematic Representation ofthe Interrupted TCA Cycle in Some 117 Cyanobacteria.
VI.8 The Be NMR Spectrum ofCuracin A (15)Produced During 118 Supplementation of Sodium [2-BC]Acetate.
VI.9 Two Hypotheses Concerning the Decarboxylation ofthe Polyketide 121 Intermediate in the Biosynthetic Pathway ofCuracin A (15).
VI.10 The Be NMR Spectrum ofCuracin A (15) Produced During 126 Supplementation with Sodium eH3, 1-BC]Acetate.
VI.ll The 1H- and 2H-Decoupled Be NMR Spectrum ofCuracin A (15) 127 Produced During Supplementation with Sodium eH3,2-BC]Acetate.
VI.12 Comparison ofthe Be NMR Spectra ofCuracin A (15) Produced 128 During Supplementation with [2-BC]-DL-Mevalonolactone, (B) Natural Abundance Curacin A (15)
VI.13 Metabolic Relationship of(A) Glycine, Serine/Cysteine, (B) 129 Glycine and Methionine, and the Expected Labeling Pattern of Curacin A (15) Produced During the Supplementation with [ 2-Bc, 1 ~]Glycine LIST OF FIGURES (Continued)
Figure
VI.14 Comparison of Selected Regions ofthe Be NMR Spectra of 131 (A and C) Natural Abundance Curacin A (15), (Band D) Curacin A (15) Produced During Supplementation with [2-BC,15N]Glycine.
VI.15 Comparison of Selected Regions ofthe Be NMR Spectra of 132 A) Natural Abundance Curacin A (15), (B) Curacin A (15) Produced During Supplementation with [1-BC]Glycine.
VI.16 Comparison of Selected Regions ofthe Be NMR Spectra of 133 (A) Natural Abundance Curacin A (15), (B) Curacin A (15) Produced During Supplementation with [methyl-BC]-L Methionine.
VI.12 The Proposed Modification Steps ofthe Post-Assembly 136 Polyketide Chain ofCuracin A (15), based on the Information Obtained from Acetate Feeding Experiments. LIST OF TABLES
Table Page
11.1 NMR Data for Grenadadiene (34) and Debromogrenadadiene 25 (35) in CDCh.
11.2 NMR Data for Grenadamide (36) in CDCh. 30
111.1 NMR Data for Y anucamide A (39) in Methanol-d4. 45
111.2 NMR Data for Yanucamide B ( 40) in Methanol-d4. 49
IV.l NMR Data for Clairamide (50) in DMSO-d6. 61
IV.2 NMR Data for Carliamide (51) in CDCh. 66
V.l NMR Data for the Major Conformer ofBarbamide (47) in 81 DMSO-d6 and Toluene-ds.
V.2 Calculation of4(S)-[5-13C]-L-Leucine Incorporation into 100 Barbamide (47).
VI.2 13C NMR Data ofCuracin A (15/ Derived from Sodium 119 13 18 13 3 [1- C, 0 2]-, [2- C]-, and [1,2- C2]Acetate
VI.3 2H Isotope-Induced Shifts Observed in the 13C NMR 122 Spectra ofCuracin A (15) Derived from Sodium eH3, 1-13C] and eH3,2-13C]Acetate.
VI.4 13C NMR Data ofCuracin A (15) Derived from [l-13C]- and 130 [2-13C, 15N]Glycine and [methy/- 13C]-L-Methionine.
13 18 VI.5 Calculation of [l- C, 0 2]Acetate Incorporation into 140 Curacin A (15).
VI.6 Calculation of [1,2-13C2]Acetate Incorporation into Curacin A (15). 141 LIST OF ABBREVIATIONS
ACP Acyl Carrier Protein
Adda 3-Amino-9-methoxy-2,6,8-trimethyl-1 0-phenyldeca-4,6-dienoic acid Amoya 3-Amino-2-methyl-7-octynoic acid
Ampa 3-Amino-2-methylpentanoic acid
{3-Aia {3-Alanine br Broad
COSY 1H-1H chemical shift correlation spectroscopy d Doublet
DEPT Distortionless enhancement by polarization transfer
Dhoya 2,2-Dimethyl-3-hydroxy-7-octynoic acid
DMAPP Dimethylallyl diphosphate
DMSO Dimethylsulfoxide
ElMS Electron-impact mass spectrometry
EtOAc Ethyl acetate
FAAH Fatty acid amide hydrolase
FABMS Fast atom bombardment mass spectrometry
FDAA 1-Fluoro-2,5-dinitrophenyl-5-L-alanine amide
FTIR Fourier transform infrared spectroscopy
GC/MS Gas chromatography
GI Growth inhibition
Hiv 2-hydroxyisovaleric acid
HMBC Heteronuclear multiple-bond coherence spectroscopy
HPLC High performance liquid chromatography
HR-EIMS High resolution electron-impact mass spectrometry HR-FABMS High resolution fast atom bombardment mass spectrometry
HSQC Heteronuclear single quantum coherence
HMG-CoA B-Hydroxy-B-methylglutaryl-CoA
Hmp 2-hydroxy-3-methylpentanoic acid
GHNMBC Gradient 1H-15N heteronuclear multiple bond correlation
IPP Isopentenyl diphosphate
IR Infrared spectroscopy
Lac Lactic acid
LD Lethal dose
LR-EIMS Low resolution electron-impact mass spectrometry
LR-FAB Low resolution fast atom bombardment mass spectrometry m Multiplet mm Minute
MeOH Methanol
MS Mass Spectrometry
NAC N-Acetylcysteamine
NCI National Cancer Institute
N-MeAla N-Methylalanine
N,O-Me2Try N-Methyl-0-methyltyrosine
N-MePhe N-Methylphenylalanine
N-MeVal N-Methylvaline
NMR Nuclear magnetic resonance
NP-HPLC Normal phase high performance liquid chromatography
NOE Nuclear Overhauser effect
PKS Polyketide synthase rei. int. Relative intensity
RP-HPLC Reversed phase high performance liquid chromatography SAM S-adenosyl-L-methionine t Triplet tR Retention time TFA Trifluoroacetic acid
THF Tetrahydrofolate
TGI Total growth inhibition
TLC Thin layer chromatography
TMS Trimethylsilyl
TOCSY Total correlated spectroscopy uv Ultraviolet spectroscopy Val Valine
VLC Vacuum liquid chromatography This thesis is dedicated to my family. NATURAL PRODUCTS STUDIES OF THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
CHAPTER I
GENERAL INTRODUCTION
Since prehistoric times man has been using natural products for a variety of purposes, ranging from medicines to chemical warfare agents. A recent review has estimated that approximately 60% ofthe antitumor and antiinfective agents that are commercially available or currently in late stages ofclinical trials are ofnatural products origin.l Historically, a large number ofnatural products, particularly plant-derived drugs, were discovered on the basis oftraditional or empirical local medical practices. Today, novel bioactive metabolites are being discovered by bioassay-guided isolation using various molecular targets. With advances in molecular and biochemical pharmacology, new cellular receptors are rapidly being identified. As a result, an increasing number of natural products with useful biological properties are being uncovered.
Despite the structural diversity ofsecondary metabolites, each molecule is manufactured from only a limited set ofprimary metabolites including amino acids, acetate, mevalonic acid, sugars, and intermediates ofthe shikimic acid pathway. Hence, another aspect ofnatural product studies is to understand the detailed processes involved in the biosynthesis ofeach secondary metabolite. Based on this knowledge, biosynthetically engineered secondary metabolites with useful biological activities can potentially be produced.
In this introductory chapter, several examples ofmarine natural products will be highlighted. These examples are intended to illustrate the exceptional chemical diversity provided by sources in the marine environment. Then the focus will be narrowed down 2
and divided into two categories; the chemistry ofcyanobacteria with an emphasis on those ofmarine origin, and biosynthetic studies ofmarine cyanobacterial secondary metabolites.
Marine Natural Products
The marine environment was nearly devoid ofethnomedical history until the
1960's when SCUBA diving permitted scientists to explore the ocean. Since then, studies ofmarine natural products have attested to the richness and the uniqueness of secondary metabolites from marine sources. Over the last several years, the increasing number ofnovel marine natural products reported in the literature has been comprehensively reviewed annually by Faulkner. 2-15
One ofthe triumphs ofmarine natural products chemistry was the structure elucidation ofan extremely potent non-proteinous toxin, palytoxin (1), a metabolite ofthe zoanthid Palythoa toxica and other Palythoa species.16 Its intravenous lethality (LD50) ranges from 0.025 Jlg/kg in the rabbit to 0.45 Jlg/kg in the mouse. Tetrodotoxin (2), a potent neurotoxin, was initially known as a metabolite produced by pufferfish ofthe family Tetraodontidae but subsequent studies have shown that 2 is in fact a product of many strains ofmarine bacteria.17 Tetrodotoxin is a specific sodium channel blocker that has served as an important probe for the study ofion channels.l8
Okadaic acid (3) is the most important selective protein phosphatase inhibitor from marine organisms.19 Compound 3 was originally isolated from two sponges ofthe genus Halichondria but it was later discovered that the marine dinoflagellate
Prorocentrum lima is the true producer ofthis metabolite.20 Today, okadaic acid is a valuable molecular probe for understanding regulatory phenomena and signal transduction pathways in eukaryotic cells.21 Bryostatin 1 (4), a metabolite from the bryozoan Bugula neritina, is an exceptionally potent anticancer agent with remarkable 3 selectivity against human leukemia, renal cancer, melanoma, and non-small cell lung cancer cell lines. 22 The mode of action of4 involves modulation ofsignal transduction of a protein kinase C (PKC). Bryostatin I is currently in phase II human clinical trials for treatment ofmelanoma. 23 In addition to the examples mentioned above, an increasing number ofmarine natural products are being selected for in-depth investigations for their biological properties. Hence, marine natural products chemistry will continue to be a promising scientific discipline in the future.
OH Palytoxin (1) 4
OH OH Tetrodotoxin (2)
Okadaic Acid (3)
0
0
... ~0 OCf-la
Bryostatin (4) 5
Natural Products from Cyanobacteria
Cyanobacteria (blue-green algae) are photosynthetic prokaryotic microorganisms that are extraordinarily rich in chemically unique bioactive metabolites. Prominent among these are toxins, anticancer agents, enzyme inhibitors, and algicides.
Interestingly, more than two-third ofthe bioactive secondary metabolites from cyanobacteria are depsipeptides and lipopeptides.24
Toxic cyanobacterial waterblooms are found worldwide in lakes, ponds, and coastal waters, where they cause poisoning to animals and threaten human health. Many genera ofcyanobacteria are found to produce cyanotoxins, with Microcystis being the most common. Over 50% of Microcystis waterblooms showed hepatotoxicity to mammals?5 The microcystins, such as microcystin-LR (5), are the most well known hepatotoxins produced by Microcystis. Today, more than 50 structural variations in the microcystins have been identified not only from Microcystis but also from Anabaena,
Nostoc, and Oscillatoria. Although the microcystins have mainly been obtained from terrestrial cyanobacteria, microcystins have also been found in the marine environment.
For example, microcystin-LR was found to be a major toxin found in mussels in Gillam
Island, British Columbia?6 Another class ofhepatotoxins that are structurally related to the microcystins are the nodularins. Nodularin ( 6) was first identified from the brackish water-dwelling cyanobacterium Nodularia spumigena. 21 A metabolite with a similar structural identity to nodularins, motuporin (7), was obtained from a marine sponge
Theone/la swinhoei, which is known to harbor symbiotic cyanobacteria.Z8
Other groups oftoxins associated with poisonous cyanobacterial blooms are the neurotoxins, anatoxin-a (8) and anatoxin-a(s) (9). These deadly cyanobacterial toxins are produced by several strains ofcyanobacteria including Anabaena, Oscillatoria, and
Microcystis.29 Anatoxins mimic the neurotransmitter acetylcholine and cannot be degraded by acetylcholinesterase or by any other enzyme present in eukaryotic cells. 30 6
As a result, the muscles remain overstimulated, leading to twitching, cramping, and subsequently fatigue and paralysis.
Microcystin-LR (5)
Nodularin (6)
Motuporin (7) 7
H~N(CHs)2 }-N,o 4cH, HN T ·o-r-ocHs 0 Anatoxin-a (8) Anatoxin-a(s) (9)
Lyngbya majuscula is one ofthe most toxic cyanobacteria ofthe Oscillatoriaceae family. Lyngbyatoxin A (10), an alkaloid from Lyngbya majuscula collected in Hawaii, was found to be responsible for severe oral and gastrointestinal inflammation suffered by 31 32 people who accidentally ingest the alga. ' Debromoaplysiatoxin (11) was identified to be an active component in a toxic strain ofL. majuscula that causes acute dermatitis in ocean swimmers in Hawaii (USA) and Okinawa (Japan).33 Both 10 and 11 are potent activators ofprotein kinase C and are currently being used as research tools for probing certam. b.to 1 ogtca . I processes. 34
~: 0 OH OH
Lyngbyatoxin A (10) Debromoaplysiatoxin (11)
Thus far, the most promising anticancer agents ofcyanobacterial origin are the 35 cryptophycins. The parent compound, cryptophycin-1 (12) was originally isolated from
Nostoc sp. by researchers at Merck as an antifungal agent but the compound appeared to be too toxic for any practical uses. Subsequently, cryptophycin-1 was discovered to be a potent cytotoxic agent during a collaborative effort between researchers at the University 8
ofHawaii and Wayne State University. Cryptophycin-1 inhibits cell proliferation by blocking cell cycle progression at the prometaphase/metaphase stage ofmitosis at picomolar concentrations by an apparent action on microtubules. Cryptophycin-52 (13), a synthetic member ofthe cryptophycins, was found to be the most potent mitotic 36 inhibitor yet characterized. It acts by suppression ofspindle microtubule dynamics, and is currently undergoing clinical evaluation. Although the cryptophycins have only been isolated from terrestrial cyanobacteria, their structures are very similar to that of arenastatin A (14), a metabolite from the marine sponge Dysidea arenaria, another sponge that is known to harbor symbiotic cyanobacteria. 37
OCHa
Curacin A (15)
Cryptophycin-1 (12) R1 =CH3, R2 =H, R3 =Cl Cryptophycin-52 (13) R1 =CH3, R2 =CH3, R3 =Cl Arenastatin A (14) R1 =H, R2 =H, R3 =H
Curacin A (15), an exceptionally potent antiproliferative agent with selectivity for colon, renal, and breast cancer-derived cells, was isolated from the organic extract ofa
Curayao collection ofL. majuscuta?8 The mechanism ofaction of curacin A involves 9
the inhibition oftubulin polymerization by interacting with the colchicine-binding site.
All other clinically useful tubulin depolymerizing anticancer drugs are known to bind at a different region ofthe tubulin monomer called the vinca alkaloid domain. Therefore, a new class oftherapeutically useful anticancer drugs with this binding property can potentially be developed based on the structure ofcuracin A.
Extracts ofa mixture ofa L. majuscula/Schizothrix calcicola assemblage and L. majuscula collected near Guam yielded two potent cytotoxins, lyngbyastatin 1(16) and dolastatin 12 (17).39 Compounds 16 and 17, at concentrations 2 and 0.2 J.Lg/mL, respectively, caused the complete loss offilamentous (F)-actin coincident with dramatic changes in cell morphology when tested against A -10 cells and smooth muscle cells.
Dolastatin 12 was originally discovered as a constituent ofthe sea hare Do/abel/a auricularia, which is known to be a generalist algal herbivore.40 Thus, the isolation of dolastatin 12 from an assemblage ofL. majuscula and S. calcicola suggests that 17 is of cyanobacterial origin.
R
Lyngbyastatin-1 (16) R = OCH3 Dolastatin-12 (17) R = H
Proteolytic enzymes play important roles in many biological processes. For example, plasmin, a serine protienase that regulates blood coagulation, has been implicated in cardiovascular diseases such as stroke and coronary artery occlusion. 10
Micropeptins 478A (18) and B (19) are two potent plasmin inhibitors (ICso 0.11 and 0.45
J.tM, respectively) isolated from M aeruginosa NIES 299. 41 The unique Ahp (3-amino
6-hydroxy-2-piperidone) residue in these peptides is found only in cyanobacterial metabolites25 with the exception ofdolastatin 13.42 This observation further substantiates, the hypothesis that cyanobacteria are the true producers ofthe dolastatins.
The most intriguing algicide from a cyanobacterium is cyanobacterin (20), isolated from a culture ofthe freshwater cyanobacterium Scytonema hofmanni UTEX
2349.43 S. hofmanni was shown to inhibit the growth ofother species ofalgae when grown on Petri dishes. Cyanobacterin was shown to inhibit electron transport of electrons in the oxygen-evolving system ofphotosynthesis (photosystem II), killing vanous. spectes . o f a1 gae. 44
Micropeptin 478A (18) R = H Micropeptin 4788 (19) R = SO4H
OCH3
Cyanobacterin (20) 11
Biosynthetic Studies ofNatural Products from Cyanobacteria
Although cyanobacteria are producers ofmany structurally intriguing natural 45 products, biosynthetic studies have been quite limited. A majority ofthe reported biosynthetic investigations to date involve toxins, since an understanding ofthese biosynthetic pathways can aid the development ofDNA-based protocols for the detection oftoxic algal strains. In addition to toxins, a few structurally unique cyanobacterial metabolites have also been targets for biosynthetic studies.
Moore and coworkers studied the biosynthesis ofthe potent hepatotoxin, 46 microcystin-LR ( 5) in M. aeruginosa Kiitzing. Incorporation ofsodium [ 1,2
13C2]acetate into 5 revealed that C-1 through C-8 ofthe (28, 38, 88, 98)-3 amino-9 methoxy-2,6, 8-trimethyl-10-phenyldeca-4,6-dienoic acid (Adda) are derived from acetate units. L-[methy/-13C]Methionine labeled the 2-, 6-, and 8-methyl and 9-methoxy carbons ofthe Adda unit. Interestingly, experimental results showed that the C2 methyl group also derived from a second source involving propionate. The remaining carbons of the Adda unit were shown to be phenylalanine-derived from the [U-13C]-L-phenylalanine feeding experiment. The (2R, 38)-3-methylaspartic acid (Masp) residue in 5 was shown to arise from a condensation ofacetate and pyruvate. The latter two precursors likely have undergone a condensation to form citramalic acid (21) followed by rearrangements to form the Masp unit (Figure 1.2).
B6emer and coworkers isolated a 2982 bp DNA fragment, mapepl, which encodes a complete peptide synthetase module that specifically hybridized to DNA from 41 hepatotoxic strains ofM. aeruginosa. When the nucleotides on either side ofthis gene fragment were sequenced, additional peptide synthetase gene clusters ofthe microcystins were revealed. Insertional replacement ofmapepl with the chloramphenicol resistance gene cassette by homologous recombination resulted in a mutant that was unable to produce microcystins. 12
C2, C3 and Ph of L-Phenylalanine
acetate r pyruvate T methyl of methionine 5
Figure 1.1. The Biosynthetic Building Blocks ofMicrocystin-LR (5).46
Figure 1.2. Proposed Biosynthetic Pathway of(2R, 3S)-3-Methylaspartic Acid (L-
Masp).45 13
Saxitoxin (22) and neosaxitoxin (23), metabolites ofseveral strains of dinoflagellates and freshwater cyanobacteria, are known as the toxic principles of paralytic shellfish poisoning. 48 The biosynthesis ofthese toxins has been extensively studied in cyanobacterium Aphanizomenon flos-aquae and dinoflagellate Alexandrium tamarense. Results from feeding experiments with isotopically labeled arginine, acetate, and methionine showed that the perhydropurine skeleton ofthe toxins is formed by the
Claisen-type condensation ofacetate with the a.-carbon ofarginine followed by the loss ofthe carboxyl group ofarginine, followed by ami dation and cyclization. The carbamate group was derived from the methyl group ofmethionine via S-adenosylmethionine
(SAM). The proposed biosynthetic pathway ofsaxitoxin and its derivatives is summarized in Figure 1.3.
Saxitoxin (22) R = H Neosaxitoxin (23) R =OH
Moore and coworkers had isolated and conducted a biosynthetic investigation of borophycin (24), a potent boron-containing cytotoxin from a culture ofa marine strain of the cyanobacterium Nostoc linkia.49 The ionophore 24 is structurally related to boromycin (25) and aplasmomycin (26), antibiotics from terrestrial and marine strains of 50 Streptomyces sp., respectively. All three metabolites are acetate-derived polyketides that utilize a C3 precursor for the starter unit and methionine for the methyl groups on the 14
polyketide chain. However, the C3 starter unit for the biosynthesis of24 is derived from acetate and methionine, whereas phosphoglycerate or phosphoenolpyruvate has been suggested to be the C3 starter unit in the biosynthesis of25 and 26.
Figure 1.3. A Proposed Biosynthetic Pathway ofSaxitoxin Derivatives. 48
Tolytoxin (27), a potent fungicide and cytotoxin, was first isolated from a terrestrial cyanobacterium Tolypothrix conglutinata and subsequently found as a major metabolite associated with several strains ofScytonema sp.51 Tolytoxin is structurally related to many marine natural products such as the swinholide, kibiramides, and halichondramides. In the biosynthesis oftolytoxin inS. mirabile, glycine is utilized as a starter unit and extended by fifteen acetate units to form the polyketide chain. The one 2 carbon branches were found to arise from the tetrahydrofolate C1pool.5 15
C-3 Structure Unit
C-3 Structure Unit Borophycin (24)
Glycerol
).OH =[1,2· 13CiJActate
.a. =[methyt-13c]methionine
-Glycerol
Aplasmomycin (26)
Figure 1.4. Incorporation Patterns ofLabeled Precursors into Borophycin (24), 49 50 Boromycin (25) and Aplasmomycin (26). • 16
H
JOH =[1 ,2- 13~]Actate Tolytoxin (27) A = [methyP:t]methionine
Figure 1.5. Incorporation Patterns ofLabeled Precursors into Tolytoxin (27).51
Natural Products Studies of the Marine Cyanobacterium Lyngbva majuscula
Our research group has specialized on the natural products ofmarine algae and cyanobacteria. My doctoral research projects have focussed on the marine cyanobacterium (blue-green alga) Lyngbya majuscula Gomont (Oscillatoriaceae). This dissertation describes research projects that focussed on two aspects oftheir natural products chemistry; 1) the isolation and structure determination ofstructurally unique secondary metabolites, and 2) the biosynthetic investigation oftwo classes ofL. majuscula metabolites. The following chapters are divided into two categories; Chapters
II, III, and IV will describe the isolation and structural characterization ofthree classes of novel metabolites from L. majuscula collected from the Caribbean Sea and the South
Pacific Ocean, respectively. Chapter V and VI will show detailed results ofthe biosynthetic investigation oftwo unique L. majuscula metabolites, barbamide and curacin
A, respectively. Both ofthese latter compounds were previously isolated and characterize. d m. our Ia b oratory. 38,53 17
CHAPTER II
GRENADADIENE AND GRENADAMIDE, CYCLOPROPYL-CONTAINING FATTY ACID METABOLITES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
Abstract
A Grenada collection ofLyngbya majuscula was examined for the presence of novel bioactive metabolites. The organic extract ofthe collection yielded three new structurally unique cyclopropyl-containing metabolites, grenadadiene (34), debromogrenadadiene (35), grenadamide (36), and a previously described fatty acid, 7 methoxytetradec-4(E)-enoic acid (28). The structures and the relative stereochemistries of(34-36) were determined using spectroscopic methods. Grenadadiene is an inhibitor of fatty acid amide hydrolase (FAAH), an enzyme that catalyzes the hydrolysis ofbioactive fatty acid amides with an estimated ICso of22.5 JlM. Grenadadiene also exhibited an interesting profile ofcytotoxicity in the NCI 60 cell line assay, while grenadamide showed strong brine shrimp toxicity (LDso = 5 ppm) and cannabinoid receptor binding activity (Ki = 4. 7 f.l.M). 18
Introduction
The marine cyanobacterium Lyngbya majuscula is recognized as a rich producer ofbiologically active and structurally diverse metabolites. Nearly one half ofthe natural products isolated from L. majuscula have fatty acid/polyketide-derived biogenetic subunits. One ofthe fatty acids that is found in many varieties ofL. majuscula is 7 methoxytetradec-4(E)-enoic acid (28),54 which is also a subunit found in most members ofthe malyngamide family, such as malyngamide A (29).55 Other fatty acids that are less commonly encountered in L. majuscula extracts include 7-methoxy-9-methylhexadeca
4(£),8(£)-dienoic acid (30),56 malyngic acid (31),57 and 2,5-dimethyldodecanoic acid
(32).58 Fatty acid 32 was found as a subunit ofcompound 33 isolated from an Australian 9 collection ofLyngbya species.5 Our chemical investigation ofL. majuscula collected from Grand Anse Beach, Grenada, led to the isolation ofthree novel cyclopropyl containing fatty acid-derived metabolites , grenadadiene (34), debromogrenadadiene (35), grenadamide (36) and compound 28. 19
QCHs OH 28 riMe N~:Y CHs I OCH Cl 3 29
OH 30
OH
32
0~srh> CHs l._O)l...CH 3 33
Figure ll.l. Structures ofFatty Acids and Their Derivatives Produced by L. majuscula. 20
Results and Discussion
The cyanobacterium L. majuscula was collected from Grand Anse Beach,
Grenada, in July 1995 and kept cold in isopropanol until extracted. A portion ofthe lipid extract (6 g) was subjected to silica gel vacuum liquid chromatography (VLC) using a mixture ofhexanes and EtOAc as eluent. A non-polar fraction (2% EtOAc/hexanes) showed a UV-active, brown-charring spot on TLC. This fraction was further purified by
NP-HPLC to yield two novel metabolites, grenadadiene (34) and debromogrenadadiene
(35). A more polar fraction (50% EtOAc/hexanes) from the VLC fractionation ofthe crude sample was found to be brine shrimp toxic. This latter fraction was sequentially fractionated by Sephadex LH-20 and NP-HPLC to yield a previously described 7 methoxytetradeca-4(E)-enoic acid (28) and a new metabolite, grenadamide (36).
The ElMS ofgrenadadiene (34) displayed equal intensities of [Mt and [M+2t ions at m/z 414 and 416 (Figure 11.2), respectively, indicating that 34 contained one bromine atom. This observation was confirmed by HREIMS measurement, which analyzed for a molecular composition ofC2oHJt79Br04 (m/z at 414.1405). Offive degrees ofunsaturation, implied by the molecular formula, two were defined as ester
1 carbonyls from Be NMR (o 170.17, 169.63) and IR (uc--a 1759 cm- ) data. The observed four sp2hybridized carbons at o108.70, 121.35, 124.29, and 137.56 in the Be NMR spectrum indicated the presence oftwo double bonds. Hence, compound 34 contained onenng.
Interpretation ofthe 1H and Be NMR (Table 11.1), 1H-1H COSY, and 1H-BC
COSY data (Figure 11.3), three partial structures, 34a-34c, were generated (Figure 11.4).
Partial structure 34a composed ofan ester ofa fatty acid with a cyclopropane ring substituted at C-4 (o 17.88) and C-6 (o 18.88). Characteristically shielded methylene protons (H2-5, o0.19) and two methine protons (H-4, o0.42 and H-6, o0.42) were diagnostic for a cyclopropyl ring. HREIMS ofa fragment ion at mlz 195.1748, analyzing 21
for C13H230, further confirmed partial structure 34a. Partial structure 34b was defined by sequential 1H-1H COSY correlations between olefinic protons of8 5.75 (H-2') and 8 7.27
(H-1 '),and 8 6.96 (H-3'), indicating that the two double bonds in 34 formed a conjugated diene (Amax =254 nm, MeOH). Two additional desheilded methylene protons at 8 4. 79
(H2-5 ') displayed a weak correlation to H-3 ', concluding partial structure 34b. Partial structure 34c was readily generated as an acetate group consisting ofan ester carbonyl (8
170.17) and an acyl methyl group eH 8 2.08 and 13C 8 20.75).
The HMBC spectral data (Table 11.1) were used to assemble partial structures
34a-34c as well as to confirm the above structural assignments. The correlation observed from C-1 to H-1' allowed the assembly ofpartial structures 34a and 34b. The three bond correlation between the acetate carbonyl and H2-5' enabled connection ofunit 34c to partial structure 34b. The remaining bromine atom could only be attached to the molecule at the olefinic carbon (C-4') thereby completing the structure ofa novel fatty acid-derived metabolite, grenadadiene (34).
The almost identical chemical shifts observed for H2-5 (8 0.19) indicated that the methylene protons ofthe cyclopropane ring are in a similar chemical environment, implying the relative stereochemistry ofthe ring to be trans. This proposed relative stereochemistry was confirmed by comparison of 1H NMR chemical shifts ofH2-5 to those ofsynthetic reference compounds. 60 The double bond geometries in 34 were determined by nOe difference spectroscopy. Irradiation ofH-1' resulted in the enhancement ofH-2', indicating a Z geometry for the double bond. This was further supported by a distinctive 6.5 Hz coupling value between these two protons, the decrease in magnitude (typically~ 10 Hz) due to the presence ofan electron withdrawing group on
C-1 '. Similarly, a Z geometry for the C3 '-C4' double bond was indicated by the observed enhancement ofH-3' when H2-5' was irradiated. l5 twi!HWI.IS R1= 12:51 iCI SJf 14-111"1"56 15:28 TIC: ~ te;e: 2941& El lmiS 180 141
121 177 135
Figure ll.2. EI-Mass Spectrum of Grenadadiene (34). 23
13 5' g ~OACHs H 1' H Br 34
Solvent OAc H5' H8-H12 H13 H1' Hal H4H5 H3' H2' ""' I I 1' \.v~l1l ppm t ~ I ~ f-- 20 ~ I ~ 1 40
f- 60
I ' I f- 80
f-- 100 • f-- 120 I
I f- 140 'T 'I 'I , I 'T 'I 'I 7 6 5 4 3 2 1 ppm
Figure 0.3. HMQC Spectrum ofGrenadadiene (34). 24
13 ~ ~ ~o-~ :0,~ 3'8Brr' 1 • ~ ~~- o: H 1' H m/z195~ 34a 34b 34c
Figure ll.4. Partial Structures for Grenadadiene (34).
13 5' g ~O)l__CH3 1 H 1' H R
Grenadadiene (34) R=Br Debromogrenadadiene (35) R =H 25
Table ll.l. NMR Data for Grenadadiene (34) and Debromogrenadadiene (35) in CDCb.
Grenadadiene (34)A Debromogrenadadiene (35l C# 13C DEPT 1H mult (J =Hz) 13C DEPT 1H mult (J =Hz) 1 169.63 c 170.51 c 2 34.02 CH2 2.49 t (7.3) 34.14 CHz 2.52 t (7.3) 3 29.25 CH2 1.55 m 29.67 CH2 1.60 m 4 17.88 CH 0.42m 18.94 CH 0.50 m 5 11.71 CH2 0.19m 11.78 CH2 0.22 m 6 18.88 CH 0.42m 18.01 CH 0.50 m 7 33.95 CHz 1.20m 34.04 CH2 1.22 m 8-10 29.07 CHz 1.23m 29.44 CH2 1.29 m 11 31.80 CH2 1.23 m 31.87 CH2 1.29 m 12 22.58 CH2 1.23 m 29.04 CH2 1.29 m 13 14.02 CH3 0.90 t (7.5) 14.07 CH3 0.88 t (6.7) 1' 137.56 CH 7.27 d (6.2) 134.98 CH 7.12 d (6.2) 2' 108.70 CH 5.75 dd (10.5, 6.2) 111.37 CH 5.49 dd (10.5, 6.2) 3' 124.29 CH 6.96 d (1 0.5) 126.34 CH 6.67 dd (15.5, 10.5) 4' 121.35 c 126.75 CH 5.78 td (15.5, 6.7) 5' 68.93 CH2 4.79 br s 64.74 CH2 4.62 d (6.7) OAc 20.75 CH3 2.08 s 22.65 CH3 2.10 s 170.17 c 171.08 c
Aspectral data recorded on a Broker AC300 spectrometer, operating at 300.13MHz for 1H and 75.46 MHz for Be. BSpectral data recorded on Bruker DRX600 spectrometer, operating at 600.01 MHz for 1H and 150.90 MHz for Be. All spectra were referenced to the solvent at 7.26 ppm for 1H and at 77.0 ppm for Be.
A minor compound, debromogrenadadiene (35), analyzed by HREIMS to be a debromo analogue ofgrenadadiene (C2oH3204). Both compounds 34 and 35 possessed a
very similar 1H and 13C NMR spectra, differing primarily in the presence ofan additional olefinic proton at o5.78 (H-4') in the 1H NMR spectrum of35 (Figure IT.S). By 1H_IH COSY, H-4' was placed adjacent to H-3' and H-5'. TheE geometry ofthe C3'-C4' disubstituted olefin was established by measurement ofa 15.5 Hz coupling between H-3' and H -4'. All other features ofthe structure and stereochemistry ofthis minor metabolite, debromogrenadadiene (35) were essentially identical to those of grendadiene, except for the sign ofthe optical rotation ([a.]o =+5° for 35 and [a.]o = -8° for 34). 26
High resolution ElMS measurement ofgrenadamide (36) defined a molecular composition ofC21 H33NO, requiring six degrees ofunsaturation. The Be NMR spectrum of36 indicated the presence ofa monosubstituted phenyl moiety [o 139.01 (s), 128.61
(2C, d), 128.73 (2C, d), 126.48 (1C, d)] and an amide carbonyl (o 171.58), which accounted for five ofthe six required degrees ofunsaturation. The shielded signals at o 0.38 and o0.16 in the 1H NMR spectrum indicated that grenadamide (36) contained a cyclopropyl ring, which fulfilled the last degree ofunsaturation. Two partial structures of grenadamide were generated from interpretation ofthe 1H and Be NMR (Table 11.2), 1H
1H COSY (Figure 11.6), and 1H-BC COSY spectral data. Partial structure 36a (Figure
II.7) is a cyclopropyl-containing fatty acid chain similar to those that found in 34 and 35.
The second spin system (36b) consisted ofa (3-phenylethylamine moiety. HMBC correlations showed from the Hz-1 ' protons to the amide carbonyl connected the two spin systems, completing the structure ofgrenadamide (36). OAc
A
HS-12 H5' H13
H1' H7 HS H3' H3 '\ H2' "' 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm
HS-12
B.
OAc H13
H2 H1' H5' H3' H4' H2'
7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 ppm
Figure ll.5. 1H NMR Spectra of (A) Grenadadiene (34) and {B) Debromogrenadadiene (35). 28
H N 3' 1 o ~ . v6' 36
H4'-H8' H8-H12 H13
Hl' NH ppm 0
1
2
• 3 , . 4
5
6 • 7 7 6 5 4 3 2 1 0 ppm
Figure ll.6. 1H-1H COSY Spectrum of Grenadamide (36). 29
36a 36b
Figure II.7. Partial structures ofGrenadamide {36).
Grenadamide {36) 30
Table 11.2. NMR Data for Grenadamide (36) in CDCh.
C# De DEPT IH 1 171.88 c 2 36.89 CH2 2.18 (t, 7.4) 3 30.33 CH2 1.50 (br q, 8.5) 4 18.88 CH 0.35(m) 5 11.73 CH2 0.16 (m) 6 18.88 CH 0.35 (m) 7 34.10 CH2 1.09 (m) 8-10 29.44 CH2 1.25 (m) 11 31.87 CH2 1.25 (m) 12 29.60 CH2 1.25 13 14.02 CH3 0.87 (t, 6.5) 1' 40.46 CH2 3.53 (t, 6.5) 2' 35.67 CH2 2.81 (t, 6.6) 3' 139.01 c 4' (8') 128.61 CH 7.20 (m) 5' (7') 128.73 CH 7.31 (m) 6' 126.48 CH 7.25 (m) NH 5.43 (br s)
Spectral data recorded on a Bruker AC300 spectrometer; operating at 300.13 MHz for 1H and at 75.46 MHz for 13C. All spectra were recorded referenced to the solvent at 7.26 ppm for 1H and at 77.0 ppm for 13C.
Besides grenadadiene (34), the only other examples ofbromine-containing metabolites from Lyngbya species are the aplysiatoxins61 and 2-bromopropenyl2,5 dimethyldodecanoate (33).59 From a biosynthetic point ofview, compound 33 is similar to 34 in that it is derived from a fatty acid that is esterified with a bromoalkene subunit.
However, metabolites 34-36 represent the only examples ofmetabolites deriving from cyclopropane-containing fatty acid isolated from Lyngbya species. It is perhaps noteworthy that a C2o cyclopropane-containing fatty acid was isolated from the digestive 62 gland ofthe sea hare Bursatella leachii. However, sea hares are known to sequester secondary metabolites from their algal diet, including mat-forming cyanobacteria.63
Thus our finding ofstructurally similar fatty acids in L. majuscula suggests that the B. leachii metabolite is ofcyanobacterial origin. 31
Grenadadiene (34) is an inhibitor offatty acid amide hydrolase (F AAH), an enzyme that catalyzes the hydrolysis ofbioactive fatty acid amides and esters such as the endogenous cannabinoid receptor ligands, anandamide and 2-arachidonoylglycerol, with
64 an estimated IC5oof22.5 J.!M. In addition, grenadadiene also showed an interesting profile ofcytotoxicity in the NCI 60 cell lines assay and had been selected for in vivo evaluation (Figures 11.8 and 11.9). Grenadamide (36) exhibited strong brine shrimp 65 toxicity (LDso = 5 ppm) and cannabinoid receptor binding activity (Ki = 4.7 J.!M). • .. • A lt •1------ r j
··.~----~~----~"------~.~----~------~ _, .. ·-. ··"'·:.------+------!.,~-----+.------+----,._~ '-t~o ut1UI!Jikat~(!lal.., .4.. flfl&s)tl en...raaia. .,.,.., LcatOfi(.C..~~-(llalu) COJI.(D( __..,_ Rl.AI(JB) __.._ r:1f __ _.,. __ C.fCl .... _..... MOl.f-4 ...... r...... -a:-fQ.ml!M _ .... __ m\"1l --...- UC..a ·--4·-· BIJII.IJ ···•·D··•·• OCI'·IIS-+-- ICC'J-C' .-... 12'Joa.CD6 --·-- 0 ··-··- ,.,,Q) --·-
I l J' ~\b r '\-···\ JDI------+1':--1:\ ...J-----~"'----1
~.,~----~~-----+,------~.~----~~~----~ ·'"'.:------+------!.,~-----t.. ------!4~------'. .,.,>!.------t------~~r------l.~-----t------:! t..a,olfffl~aJU~•(t.lll.a) lAiiiiJIII..... D:II•ud•~ l..-j.aiii~C411~..(tttaJ.) "'" rl rl \.
·*~'~----~~----~.------~.~----~~~~~ .~.,~------!~----~~~-----..~----~~--~--'. ~.,:------!~------!,~-----+.------+------~ •••10 ...... ~-Dl.otv) t..,.. ~,_plt~a(YD1c) I..Jttiai~C.c.Miilab()ko!ltt ,.tCJI7 ____.,...... ,_.zn; ·-+-- ~o:sncr ...... JAI>A-MJUU Mf'U.," --·-
Figure 11.8. NCI 60 Cell Line Tumor Growth Inhibition Dose Response Curves for Grenadadiene (34).
w N PalltiiCtll Litle GISO ..... Till L
MO-~ ·3.31 -IJi 154 I.~"L -It 2.30 l.TS . ~:\
Figure 11.9. NCI 60 Cell Line Growth Inhibitor Mean Graph for Grenadadiene (34) (GI= Growth Inhibition, TG Total Growth Inhibition. 34
Experimental
GeneraL Nuclear magnetic resonance (NMR) spectra ofgrenadadiene (34) and grenadamide (36) were recorded on a Bruker AC300 instrument at a proton frequency of
300.13 MHz and a carbon frequency of75.46 MHz. NMR spectra of debromogrenadadiene (35) were recorded on a Bruker DRX600 instrument at a proton frequency of600.01 MHz and a carbon frequency of 150.90 MHz. Proton spectra were referenced to the residal signal ofCHCh at 7.26 ppm. Carbon spectra were referenced to centerline ofthe CDCh signal at 77.0 ppm. LR- and HR-EIMS were recorded on a
Kratos MS50TC mass spectrometer. UV and IR spectra were recorded on Hewlett
Packard 8452A UV-vis and Nicolet 510 spectrometer, respectively. Optical rotations were measured on a Perkin-Elmer Model141 polarimeter. High-performance liquid chromatography (HPLC) utilized a Waters M6000A pump, Rhoedyne 7125 injector, and a Waters Lambda-Max 480 LC spectrometer or a R 401 differential refractometer.
Merck aluminum-backed thin-layer chromatography (TLC) sheets (silica gel 60 F24s) were used for TLC. Vacuum liquid chromatography (VLC) was performed with Merck
Silica gel G for TLC. Compounds were detected by UV illumination or by heating plates sprayed with a 50% H2S04 solution. All solvents were distilled from glass prior to use.
Collection. The marine cyanobacterium L. majuscula was collected at a depth of
4-6 mat Grand Anse Beach, Grenada, in July 1995 and immediately preserved in isopropanol for transport, then stored at low temperature ( -20°C) until work-up. The voucher specimen is available as GGA-29 Jul-95-02.
Isolation and Purification. Approximately 3 L ofthe preserved alga was defrosted and extracted with CH2Ch/MeOH (2: 1) twice and dried in vacuo to give 11.3 g ofthe crude organic extract and 741 g dry weight ofextracted algal material. A portion ofthe crude extract ( 6 g) was fractionated by using vacuum liquid chromatography
(VLC, 8.5 em x 5 em, Merck Silica Gel G for TLC) using increasingly polar mixtures of 35
hexanes, EtOAc, and methanol. Eluted material was collected in 14 x 250 mL fractions and monitored by TLC. Similar fractions were combined to give eight fractions.
Fraction 2 (1.2 g, eluted with 2% EtOAc/hexanes) showed UV active, brown charring reactions on TLC. This fraction was further fractionated twice on flash column chromatography (Merck Kieselgel60, 230-400 mesh) using stepwise gradient from 2%
EtOAc/hexanes to 50% EtOAc/methanol. Fractions eluted with 2%-4% EtOAc/hexanes contained the UV-active and brown charring components. These fractions were combined and further purified by NP-HPLC [Phenomenex Maxsil Silica, 500 mm x 10 mm, 10 J..~.m) using 3% EtOAc/hexanes as eluent, detection at 254 nm] to give grenadadiene (34, tR = 8-9 min, 58.5 mg) and debromogrenadadiene (35, tR 12-13 min,
3.2 mg). Fraction 5 ofthe initial VLC fractionation showed brine shrimp toxicity (LDso
=50 Jlg/mL). This fraction was fractionated by Sephadex LH-20 [2.5 J.D. x 60 em, using
EtOAc/MeOH ( 1: 1) as eluent] and followed by a final purification on NP-HPLC [ 500 x
10 mm Maxsi110 J.A.m) using 30% EtOAc/hexanes as eluent and detected by a differential refractive index detector] to give grenadamide (36, 13.2 mg).
Grenadadiene (34). Pure grenadadiene is a colorless oil and had the following properties: [a]o -8° (CHCh, c = 0.1); UV(MeOH) Amax (E) 252 nm (12 200); FTIR (neat)
1 1 13 Vmax2924,2855, 1759,1219,1115, 1026cm- ; Hand CseeTablell.l;LR-EIMS(70 eV) mlz (rei. intensity) 416 ~. 16), 414 ~. 18), 222 (18), 220 (18), 195 (90), 177
(34), 141 (100), 135 (24), 121 (38), 99 (58), 97 (50), 95 (45), 83 (56), 69 (60); HR-EIMS [Mi m/z 414.1405 {calcd for C2oHJt0479Br, -0.1 mmu dev), 416.1385 (calcd for C2oH3t0481 Br, -0.1 mmu dev).
Debromogrenadadiene (35). Pure debromogrenadadiene is a colorless oil and had the following properties: [a]o +5° (CHCh, c = 0.1); UV(MeOH) Amax {E) 232 nm (8
1 1 13 000); FTIR (neat) Vmax 2954,2850, 1746, 1221, 1020 cm- ; H and C see Table 11.1;
LR-EIMS (70 eV) mlz (rei. intensity) 336 ~. 8), 330 (13), 315 (28), 195 (100), 177 36
(38), 142 (40), 135 (24), 121 (35), 107 (22); HR-EIMS [Ml mlz 336.2301 (calcd for C2oH3204, +0.1 mmu dev.).
Grenadamide (36). Pure grenadamide is a colorless oil and had the following properties: [a]o -11 o (CHCh, c =0.1); UV (MeOH) Ama,. (E) 206 nm (2 600);FTIR (neat)
1 1 13 Vmax 3300, 2924, 1645, 1552, 1456, 700 cm- ; H and C NMR see Table 11.2; LR-EIMS
(70 eV) mlz 315 (38), 230 (45), 163 (40), 105 (60), 104 (100), 91 (10), 72 (25); HR
EIMS [M+] mlz 315.2563 (calcd for C21H33NO, +0.1 mmu dev.).
Brine Shrimp Toxicity Assay. In a method slightly modified from the original 66 description, about 15 newly hatched brine shrimp (Artemia salina) in ca. 0.5 mL artificial seawater were added to each well in a 140-well plate containing different concentrations ofthe sample in 50 f.lL EtOH and 4.5 mL artificial seawater to make a total volume ofca. 5 mL. Samples and controls were run in duplicate. After 24 h at 28°C, the brine shrimp were observed and the numbers dead and alive were counted with a dissecting light microscope to generate LDso values. 37
CHAPTER III
YANUCAMIDES A AND B, TWO NOVEL DEPSIPEPTIDES FROM AN ASSEMBLAGE OF THE MARINE CYANOBACTERIA LYNGBYA MAJUSCULA AND SCHIZOTHRIXSPECIES
Abstract
Yanucamides A (39) and B (40) were isolated from the lipid extract ofaLyngbya majuscu/a and Schizothrix sp. assemblage collected at Yanuca Island, Fiji. The structures ofcompounds 39 and 40 were determined by spectroscopic methods. Both compounds contain a unique 2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya) which has previously been described only as a component ofkulolide-1 (41) and kulokainalide-1 (42), metabolites from the marine mollusk Phi/inopsis speciosa. Thus, the isolation ofthe yanucamides from this cyanobacterial assemblage supports the hypothesis that the kulolides and related metabolites are ofcyanobacterial origin. 38
Introduction
Marine invertebrates have proven to be an exceptionally productive source of new bioactive metabolites. Examples include the dolastatins, a family ofcytotoxic agents that 40 42 61 80 were isolated from the Indian Ocean sea hare Do/abel/a auricularia. ' • - However, these invertebrates are generally known to consume cyanobacteria or prey on organisms which consume cyanobacteria, and many compounds isolated from marine invertebrates contain structural motifs that are similar to those ofcyanobacterial metabolites. Thus, questions remain in the exact origin ofthese natural products. In a recent literature report the isolation ofsymplostatin I (37) from the marine cyanobacterium Symploca hydnoides, 81 an analog ofa potent antineoplastic agent dolastatin I 0 (38), 70 supports the hypothesis that the dolastatins and perhaps other biologically active metabolites from marine invertebrates are ofcyanobacterial origins.
H~~~~~~ CHs 0 A CHs OCH:P CHsO 0 ~
37 R=CH3 v 38 R=H
A chemical investigation of a lipid extract of a L majuscula and Schizothrix sp. assemblage collected from the north of Y anuca Island, Fiji led to the isolation of two new depsipeptides, yanucamides A (39) and B (40). Both 39 and 40 contained a unique substructure, 2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya), which has been reported only as a subunit of the marine mollusk Philinopsis speciosa metabolites, kulolide-1 (41) 82 83 and kulokainalide-1 ( 42). • Since P. speciosa are known to prey on the herbivorous 39
sea hare Stylocheilus longicaudus which feeds on mat-forming blue-green algae, cyanobacteria are likely the true producers of the kulolides and their related metabolites. 40
Results and Discussion
The assemblage ofthe cyanobacteria was collected from the north ofthe Yanuca
Island, Fiji, and kept cold in isopropanol until extracted. A portion ofthe organic extract
(800 mg) was subjected to silica gel vacuum liquid chromatography (VLC) using an increasing gradient ofEtOAc in hexanes. Purification ofa fraction containing the yanucamides utilized Cts vacuum liquid chromatography, using a stepwise gradient elution from 60% MeOH in H20 to 100% MeOH. A yanucamide-containing fraction was subjected to a final purification over reversed phase HPLC (ODS) to afford yanucamides A (39) and B ( 40).
Yanucamide A (39) was assigned a molecular composition ofC33R.sN307 by
HRFABMS data. The IR spectrum of39 displayed strong absorption bands at 1732 em-t
1 and 1661 cm· , indicating the presence ofester and amide functionalities. Ofthe twelve degrees ofunsaturation implied by the molecular formula, nine were defined from 13C
NMR and DEPT spectral data as five amide/ester carbonyls and a monosubstituted phenyl group.
Partial structures 39a-39d (Figure IlL 1) were constructed as valine (Val), N methylphenylalanine (N-MePhe), 13-alanine (13-Ala), and 2-hydroxyisovaleric acid (Hiv), respectively, using lD and 2D NMR data (Table III.1). The H3-H6 spin system ofpartial structure 39e was deduced from the COSY correlation spectrum (Figure III.2). The connection between H6 and H8 was equivocal due to the overlapping nature oftheir proton signals (S 2.22). Although DEPT 135 spectral data indicated that C8 (S 70.21) was a methine carbon, it showed no correlation to its corresponding proton in the HMQC
1 spectrum. Instead, an intense one bond C-H satellite to C8 with a large J cH coupling of
243 Hz was observed in the HMBC spectrum, characteristic ofan alkyne functionality.
The combination ofthis large C-H coupling value taken together with the 13C chemical shift ofC-7 (o 84.58), a triple bond was placed next to the C-6 methylene carbon. 41
Geminal dimethyl protons (H39 and H3-1 0) showed HMBC correlations to a carbonyl carbon at 8 179.72 (C1), a quaternary carbon at 8 47.0 (C2), and an oxygenated tertiary carbon at 8 78.44 (C3), supportive ofpartial structure 39e as 2,2-dimethyl-3-hydroxy-7 octynoic acid (Dhoya). This residue was reported as part ofkulolide-1 (41) and kulokainalide-1 (42) structures, and fulfilled two ofthe three degrees ofunsaturation.
f~ ux~~ H 0 ~/~~ 6=~ ~ 0 39a 39b 39c
9 c~
1 ~~ ~:!x:l 8 0 ~~ 39d 39e
Figure 111.1. Partial Structures ofYanucamide A (39). 42
H10H14 Solvent ll23a H9\ H15 Solvent NCH3 H7 juts \ r~ ppm
I
Q • 1 a • ' • ' + 0 ' 2 .0 • 3 .,t • *I ..+· +- . ~ .. ... u 4 Q +·. to \ .~f· • ... 5 t
6 ' 0 +• • l!i' 7
7 6 5 4 3 2 1 ppm
Figure m.:z. 1H)H COSY Spectrum ofYanucamide A (39). 43
H
100
120 AI. • • • • .. 140
160 .. It • • ...... 180 7 6 5 4 3 2 1 ppm
Figure ll.3. HMBC Spectrum ofYanucamide A (39). 44
Correlations observed in the HMBC spectrum of39 (Table 111.1 and Figure 111.3) from the a.-CH's to the neighboring carbonyl carbons were used to deduce a sequence of
Dhoya!Val!Hiv. The correlation shown fromN-CH3of39b to the carbonyl carbon of39d linked partial structure 39b to the Hiv end ofthe depsipeptide sequence. Three-bond coupling from H3 of Dhoya to the carbonyl carbon of(3-Ala completed the planar and cyclic structure ofyanucamide A (39), thus accounting for the remaining unassigned degree ofunsaturation. Acid hydrolysis of39 followed by Marfey's analysis revealed the
L-configuration ofvaline and D-configuration ofN-methylphenylalanine.84 The absolute configuration ofthe L-Hiv residue was determined by analysis ofthe corresponding methyl ester derivative ofL-Hiv following hydrolysis of39 using chiral
GC-MS under optimized conditions and in comparison with standards. Due to the limited supply of39, the absolute configuration ofthe Dhoya unit was not determined.
2 -:7 D- N-MePhe ::::::.... H 30 ~-A~a33 21 N,.....CHi 20 O 6 •'' L-Hiv ~•' 19 31 s~ Dhoya ? ~ y. ,~~x1'H ··-.. L-Val 15 9 10 / 14 Yanucamide A (39) 45
Table 111.1. NMR Data for Yanucamide A (39) in Methanol-d4.
Unit C# I3c DEPT 1H mult (J=Hz) HMBCto Dhoya 1 179.72 c 2 47.00 c 3 78.44 CH 5.25 dd (10.7, 2.2) C: 1, 2, 9, 10,31 4a 30.21 CHz 1.56 m C: 3, 5, 6 4b 1.76m C: 5, 6 5 26.31 CHz 1.43 m C: 3, 4, 6, 7 6 18.75 CHz 2.22m C: 4, 5, 7, 8 7 84.58 c 8 70.21 CH 2.22m C: 6, 7 9 17.02 CH3 1.32 s C: 1, 2, 3, 10 10 24.77 CH3 l.lls C: 1, 3, 9
L-Val 11 174.31 c 12 60.59 CH 3.88 d (6.8) C: 1, 11, 13, 14, 15 13 30.48 CH 2.14m C: 11, 12, 14, 15 14 19.46 CH3 1.05 d (6.8) C: 12, 13, 14 15 19.97 CH3 1.06 d (6.9) C: 12, 13, 15 L-Hiv 16 172.14 c 17 76.64 CH 4.94 d (3.5) C: 11, 16, 18, 19, 20 18 30.30 CH 0.71 m C: 19, 20 19 17.02 CH3 0.66 d (6.2) C: 17, 18, 20 20 19.84 CH3 0.77 d (6.5) C: 17, 18, 19 D-NMePhe 21 171.63 c 22 64.10 CH 4.74 dd (9.7, 4.7) C: 21, 23, 24, 30 23a 36.23 CHz 3.30 dd (14.0, 4.7) C: 21, 24, 25, 29 23b 3.03 dd (14.0, 9.7) C: 21, 24, 25, 29 24 139.22 c 25,29 130.65 CH 7.22m C: 23,27 26,28 129.96 CH 7.30m C: 24 27 128.16 CH 7.25 m 30 30.60 CH3 2.90 s C: 16,22 (3-Ala 31 173.25 c 32a 33.06 CHz 2.59 ddd (18.8, 11.3, 3.7)C: 31, 33 32b 2.79 ddd (18.8, 3.9, 2.3) C: 31 33 36.50 CHz 3.32m C: 21. 31.32
Spectral data recorded on a Broker DRX600 spectrometer, 1H and 13C spectra referenced to solvent signals at 4.87 ppm and 48.15 ppm, respectively. HMBC optimized for 8Hz coupling. 46
The HRFABMS of yanucamide B ( 40) established its molecular formula as
C34H5oN30 7, one more methylene unit than that of yanucamide A (39). Further confirmation of this relationship between 39 and 40 was provided by the 1H NMR spectrum of 40 (Figure 111.4), which was similar to that of 40 except for two new signals at 8 1.40 (H14a) and 8 1.64 (H14b). These two new resonances were assigned to a methylene carbon (8 27.14) based on HSQC data. Analysis of 1D and 2D NMR data of
40 indicated that the additional methylene carbon belonged to an isoleucine (lie) residue that replaced L-Val in 39. The absolute stereoconfigurations of L-alla-isoleucine and D
N-MePhe in 2 were determined by Marley's analysis whereas the L-configuration of the
Hiv residue was again established based on GC-MS analysis of the corresponding methderivative. The Dhoya stereochemistry was not determined.
D-N-MePhe
Yanucamide B (40)
Both yanucamides A (39) and B ( 40) exhibited strong brine shrimp toxicity (LD5o 5 ppm). Interestingly, the Dhoya unit has previously been found only in kulolide-1 (41) and kulokainalide-1 ( 42), metabolites isolated from the marine mollusk Philinapsis speciosa. A study by Scheuer and co-workers has shown that P. speciosa preys on the herbivorous sea hare Stylocheilus longicaudus, an organism well known to sequester 47
63 secondary metabolites from their diet ofmat-forming cyanobacteria. Thus, the discovery ofthe yanucamides from a field collected L. majuscula and Schizothrix sp. substantiates the hypothesis that marine cyanobacteria are the probable source ofthe kulolides and their related metabolites. Moreover, based on this reasoning, we speculate that the unassigned Dhoya stereocenter in 39 and 40 is the same (3S) as in k:ulolide-1
(41).
Kulolide-1 (41} Kulokainalide-1 (42) Solvent Solvent NCB3
Yanucamide A HlO H9 H14Ht8 H15 Hl H20 H6 H23a H8 H25-H29 H5 Hl2
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 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm
Solvent Yanucamide B
NCH3 H9HlO
Solvent H25-H29
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 ' 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm
Figure ID.4. 1H NMR. spectra of Yanucamides A (39) and B ( 40) in Methanol-d4. 49
Table 111.2. NMR Data for Yanucamide B (40) in Methanol-d4.
Unit e# Be DEPT 1H mult (Hz) HMBeto Dhoya 1 179.62 e 2 47.19 e 3 79.41 eH 5.28 brd d (10.7) e: 1, 2, 9, 10,31 4a 29.29 eH2 1.64 m 4b 1.79 m 5 26.29 eH2 1.46 m e: 6, 7 6 18.75 eH2 2.22m e: 4, 5, 7, 8 7 84.56 e 8 70.20 eH 2.22m e: 6, 7 9 17.13 eHJ 1.33 s e: 1, 2, 3, 10 10 24.73 eHJ 1.12 s e: 1, 2, 3, 9
L-allo-Ile 11 174.09 e 12 58.76 eH 4.04 d (6.0) e: 1, 11, 14, 15 13 36.88 eH 1.91 m e: 11, 14 14a 27.14 eH2 1.40 m e: 13, 14', 15 14b 1.64 m e: 13, 14', 15 14' 11.62 eHJ 0.95 t (7.3) e: 13, 14 15 15.94 eHJ 1.08 d (6.8) e: 12, 14, 15
L-Hiv 16 172.09 e 17 76.58 eH 4.95 d (2.0) e: 11, 16, 19, 20 18 30.29 eH 0.67m 19 17.01 eHJ 0.64 d (4.5) e: 17, 18,20 20 19.89 eHJ 0.78 d (5.8) e: 17, 18, 19
D-NMePhe 21 171.63 e 22 63.81 eH 4.75 dd (9.5, 4.6) e: 21,23 23a 35.86 eH2 3.03 dd (13.8, 9.5) e: 21, 22, 24, 25, 23b 3.44 dd (13.8, 4.6) 29 e: 21, 22, 24, 25, 29 24 139.23 e 25,29 130.64 eH 7.24m e: 23 26,28 129.95 eH 7.32 m e: 24 27 128.15 eH 7.27m 30 30.16 eHJ 2.97 s e: 16,22
(3-Ala 31 173.21 e 32a 33.02 eH2 2.59 brd dd (18.5, 3.9) e: 31 32b 2.81 ddd (18.5, 12.2, e: 31 2.8) 33 36.19 eH2 3.32 m e: 31 32
Spectral data recorded on a Broker DRX600 spectrometer, 1H and Be spectra referenced to solvent signals (MeOH-d4) at 4.87 ppm and 48.15 ppm, respectively. HMBe optimized for 8 Hz coupling. 50
Experimental
General. Nuclear magnetic resonance (NMR) spectra were recorded on a Broker
DRX600 spectrometer operating at a proton frequency of600.01 MHz and a carbon frequency of 150.90 MHz with the solvent used as an internal standard (Me0H-d4 at S 4.87 for 1H and S 48.15 for 13C). Mass spectra were recorded on a Kratos MSSOTC mass spectrometer. GC-MS data were obtained on a Hewlett-Packard 5890 Series II GC connected to a Hewlett-Packard 5971 mass spectrometer. Ultraviolet (UV) spectra were recorded on Hewlett-Packard 8452A UV-vis spectrometer. Infrared (IR) spectra were run as neat films with a Nicolet 510 Fourier transform IR spectrophotometer. Optical rotations were measured with a Perkin-Elmer Modell41 polarimeter. HPLC separation was performed with a Waters M-6000A pump, a Rheodyne 7010 injector, and a Waters
Lambda-Max 480 UV detector. Merck aluminum-backed thin layer chromatography
(TLC) sheets (silica gel Fzs4) were used for TLC. Vacuum liquid chromatography (VLC) was performed with Merck Silica gel G for TLC or with Baker Bonded Phase-octadecyl
(Cts). All solvents were distilled from glass prior to use.
Collection. The mixed assemblage ofmarine cyanobacteria Schizothrix and
Lyngbya majuscula (voucher specimen available as collection number VYI-5 Feb 97-1) was collected from shallow water (2-3m) at Yanuca Island, Fiji, in February 1997 and stored in isopropanol for transport, then kept at reduced temperature (-20°C) until extraction.
Isolation and Purification. Approximately 1 L ofpreserved alga was extracted with CHzCh/MeOH (2: 1) twice to give 1.29 g ofcrude organic extract and 67.8 g of dried alga. A portion ofthe organic extract (800 mg) was subjected to silica gel vacuum liquid chromatography (VLC) with a stepped gradient elution from 100% hexanes to
100% EtOAc to 50% MeOH in EtOAc, giving five distinctive fractions. Fraction 4 was purified by reversed phase (Cts) vacuum liquid chromatography followed by ODS HPLC 51
[Phenomenex Spherisorb ODS (2), MeOH/H20 (3: 1), flow rate 3 mVmin, detection at
215 nm, tR = 23 min for 39 and 30 min for 40] to yield yanucamide A (39, 7.5 mg) and yanucamide B (40, 5.5 mg).
Yanucamide A (39) was obtained as a colorless amorphous solid. It properties 20 were: [a] o = -33° (c 0.1, MeOH); UV (MeOH) Amax (E) 204 nm (13 000); IR (K.Br)
1 1 13 3400(br),3315,2963,1732, 1661,1379,1174cm" ; Hand CNMRseeTablell.1;
FABMS m/z 599 [M+1t (64), 507 (8), 460 (9), 356 (19), 329 (12), 307 (18), 244 (64),
154 (100), 135 (76), 120 (15), 107 (25), 89 (24), 77 (27), 69 (22), 55 (30); HR-FABMS m/z 598.3491 (calcd for C33f4sN307, 598.3489).
Yanucamide B (40) was obtained as a colorless amorphous solid. Its properties 20 were: [a] o = -31° (c = 0.1, MeOH); UV (MeOH) Amax (E) 204 nm (13 000); IR (K.Br)
1 1 3356 (br), 3307, 2964, 2926, 2875, 1732, 1660, 1467, 1453, 1412, 1380, 1196 cm" ; H and 13C NMR see Table 111.3; FABMS m/z 613 [M+It (16), 612 (25), 585 (19), 401
(14), 369 (20), 244 (100), 134 (47), 91 (16), 86 (26), 73 (9), 69 (22), 55 (19); HR
FABMS mlz 612.3651 (calcd for CJ.JisoNJ07, 612.3653). Acid Hydrolysis and Marfey Analysis ofYanucamides A and B. Y anucamides
A (39, 1 mg) and B (40, 1 mg) were separately hydrolyzed with 6 N HCl, 105 °C for 12 h. A portion ofthe hydrolysate was extracted with EtOAc, and the aqueous layer was added to 50 J!L of0.1% 1-fluoro-2,4-dinitrophenyl-5-L-alanine amide (FDAA) solution in acetone and 100 J!L of0.1 N NaHC03, followed by heating at 80 oc for 5 min. After cooling to room temperature, the reaction mixture was neutralized with 50 J!L of0.2 N
HCl and diluted with CH3CN:HzO:TF A (50:50:0.05). The solution was analyzed by RP
HPLC [MICROSORB-MV (Rainin), Cts, UV detection at 340 nm] using a CH3CN in
HzO linear gradient (20-50% CH3CN over 30 min and 50% CH3CN for 10 additional min). The retention time (tR. min) ofthe derivatized amino acids in the hydrolysate of39 matched those of£-Val (17.4; D-Val, 21.7) and D-N-MePhe (23.3; L-N-MePhe, 22.1). 52
The retention time of derivatized amino acids in the hydrolysate of40 matched those of
L-allo-Ile (20.9; D-allo-Ile, 26.8; L-Ile, 23.7; D-Ile, 27.0) and D-N-MePhe (23.3, L-N
MePhe, 22.1 ).
Absolute Stereochemistry ofthe 2-Hydroxyisovaleric Acid by Chiral GC-MS
Excess HCl from a portion ofthe 6 N HCI hydrolysate of39 was removed under a stream ofdry Nz. The hydrolysate was diluted in O.S mL of diethyl ether and treated with diazomethane for 10 min. Excess CHzNz was removed with a stream ofNz. Capillary
GC-MS analyses were carried out using a Chiralsil-Val column (Alltech, 2S m x 0.2S mm). The following conditions were used for GC: a 11 psi initial head pressure and a column temperature held at 40 °C for 10 min after injection ofthe sample, then increased from 40 °C to 100 oc at a rate of3 °C/min, then from 100 oc to ISO oc at a rate of IS °C/min, and fmally held at ISO oc for S min. The retention time found for the yanucamide A derived 2-Hiv was found at 7.7 min. Standards ofD- and L-Hiv were also converted to the corresponding methyl derivatives by the same procedure (D-Hiv, 8.4 min; L-Hiv, 7.7 min).
Brine Shrimp Toxicity Assay. In a method slightly modified from the original description,66 about IS newly hatched brine shrimp (Artemia salina) in ca. O.S mL artificial seawater were added to each well containing different concentrations of the sample in 50 JlL EtOH and 4.5 mL artificial seawater to make a total volume ofca. 5 mL. Samples and controls were run in duplicate. After 24 h at 28°C, the brine shrimp were observed and counted with a dissecting light microscope. The percentage of live brine shrimp versus total brine shrimp was used to determine LDso values. 53
CHAPTER IV
CLAIRAMIDE AND CARLIAMIDE, TWO NOVEL DEPSIPEPTIDES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
Abstract
Two new depsipeptides have been isolated along with two known metabolites, curacins A (15) and D (SO), from the lipid extract ofthe marine cyanobacterium Lyngbya majuscula collected at Wewak Bay, Papua New Guinea. The structures were elucidated by using spectroscopic methods. These compounds have been assigned the trivial names clairamide (51) and carliamide (52). Clairamide (51) contains a residue of3-amino-2 methylpentanoic acid (Ampa), a component that is unique to metabolites isolated from L. majuscu/a and the marine mollusk Dol/abel/a auricularia, such as lyngbyastatin 1 (16) and dolastatin 12 (17). Carliamide (52) possesses 3-amino-2-methyl-7-octynoic acid
(Amoya), a unit that has been only described in onchidin A (53), a metabolite from the marine mollusk Onchidium sp. Hence, the isolation of51 and 52 from L. majuscula supports the hypothesis that cyanobacteria are the true producers ofthe dolastatins and the onchidins. 54
Introduction
The marine cyanobacterium Lyngbya majuscula Gomont (Oscillatoriaceae) has proven to be a very rich source ofchemically diverse classes ofbiologically active metabolites. For example, from a single extract ofL. majuscula, collected in Curayao, we have isolated an antimitotic metabolite curacin A (15), 38 the ichthyotoxins antillatoxin (43) 85 and malyngamide H (44),86 two quinoline alkaloids (45 and 46),87 a molluscicidal agent barbamide (47),53 and the structurally unique linear peptides carmabins A (48) and B (49)88 (Figure IV.l).
In a continuing search for novel biologically active natural products, we have made marine cyanobacterial collections from various tropical and sub-tropical locations.
From a preliminary screening ofthese cyanobacterial extracts using the brine shrimp toxicity assay, we identified a lipid extract ofL. majuscula collected from Papua New
Guinea to be highly brine shrimp toxic (LDso 10 ppb). Processing ofthis extract led to the isolation ofthe previously described brine shrimp toxic lipids, curacin A (15) and D
38 89 (50). • Upon further examination ofa polar fraction ofthe same extract led to the isolation oftwo novel depsipeptides, clairamide (51) and carliamide (52). Described in this chapter are the isolations and structural elucidations ofclairamide (51) and carliamide (52). 55
OCH3
15
49
Figure IV.l. Natural Products Isolated from a Single Extract ofL. majuscula Collected from Curayao. 56
Results and Discussion
The marine cyanobacterium Lyngbya majuscula was collected at Wewak Bay,
Papua New Guinea in September 1998, and kept cold until extracted. Both lipid and aqueous extracts ofthis collection were found to be highly toxic to brine shrimp (LD50 = 10 ppb). A portion ofthe lipid extract (1.72 g) was subjected to silica gel vacuum liquid chromatography (NP-VLC) using an increasing gradient ofEtOAc in hexanes, giving eight distinct fractions. Fraction 4 showed the highest brine shrimp toxicity (LD50 = 1 ppm). Further purification ofthis fraction utilizing reversed-phase (C 18) vacuum liquid chromatography (RP-VLC) followed by normal phase HPLC provided the brine shrimp 8 89 toxic components, curacins A (15) and D (50)? •
OCHa
Curacin D (50)
In addition to fraction 4 from VLC, fraction 6 also exhibited strong brine shrimp toxicity (LDso = 5 ppm). Further purification ofthis fraction by RP-VLC and RP-HPLC
(ODS) led to the discovery oftwo new depsipeptides, clairamide (51) and carliamide
(52). Interestingly, after purification, neither compounds 51 nor 52 possessed brine shrimp toxicity at 10 ppm.
The structure ofclairamide was constructed using a combination of spectroscopic data. The high resolution FABMS ofclairamide (51) gave an [M+ It peak at 614. 3022 analyzing for a molecular composition of CJtH4~s 06S (+ 1.1 mmu deviation) which required thirteen degrees ofunsaturation. The IR spectrum of51 showed absorption bands 1732 cm"1 and 1669 em·• indicating the presence ofester and amide functionalities, 57
respectively. The 13e NMR spectrum of51 exhibited the presence offive ester/amide carbonyls and a monosubstituted phenyl ring [o 136.89 (s), 128.84 (d, 2C), 128.27 (d,
2e) and 126.57 (d)] (Table IV. I), accounting for ten ofthirteen units ofunsaturation.
Interpretation of 1D and 2D NMR spectra of51 allowed construction ofsix partial structures (Figure IV.2). Partial structures 5la-51d were assigned as N methylphenylalanine (N-MePhe), N-methylvaline (N-MeVal), 3-amino-2 methylpentanoic acid (Ampa) and lactic acid (Lac). Spin system 51e was established from the 1H_IH COSY correlations showing from a methine proton (H11) to methyl protons (H3-12) and amide proton (NH-2'). Partial structure 51fwas defined as a 2,4 disubstituted thiazole ring based on the observed characteristic 13e chemical shifts at o 170.01 (e10), 147.52 (e8) and 124.03 (e9) and HMBe correlations (Figure IV.4) detected from H9 to e8 and e 10. With the four out offive remaining degrees of unsaturation assigned to partial structure 51f, clairamide (51) was deduced to be a monocyclic depsipeptide.
22 28 yH3 ~?Ha H /~ ~N...~ ~ N,~ 5 ~+:L~Ha 19 0 l H3 CHs 6CH3
51 a 51b 51c
H2' ~~~ ~YN'~ 31CH3 I2CH3 l~k H 9 S ~ 51d 51e 51f
Figure IV.2. Partial Structures ofelairamide (51). 58
12 15 10 I ~ 1 N2' . 9
~ /2~6NH 5 27 JyN29 3 0 1 6 31 ° Clairamide (51}
WI WI otf11 ots.
817-Zl 830 824 Bt4I on/
• I 1 t I
I I 2 • • • 3
4
•0' ' ' 5 .... 4 I •• 6
7
8
9
9 8 7 6 5 4 3 2 1 0 ppm
Figure IV.3. TOCSY Spectrum ofClairamide {51). 59
1122, HZII
"iI.H1Sa
H9 H17-21 ppm •• • • • • t \I~··· .... 20 4 I I • 0 .. ~ • • I I .. • •. • ••• •• 40 ·> .. •.... ·'· • • • 60 ' I
I • 80
100
120 • • .... • • I II • 140
160 t •• • • •.• II • 9 8 7 6 5 4 3 2 1 ppm
Figure IV.4. HMBC Spectrum ofClairamide (51). 60
The partial structures of 51 were connected through interpretation ofthe HMBC spectral data (Table IV.1 and Figure IV.4). The connectivities ofpartial structures 51a,
51e, and 51fwere achieved through correlations from the amide proton of(NH-2') to
C10 and C13. A heteronuclear coupling from the amide proton (NH-1 ')to C7linked the
Ampa unit to the carbonyl of 51f. The coupling between H 14 and C23 of 51b connected the N-MeVal to the N-MePhe. Three bond couplings from H24 to C29 and from H30 to
C1 placed lactic acid (51 d) between the N-MeVal and the Ampa unit, completing the planar structure ofclairamide (51) as shown below.
The absolute configurations ofC11, C14 and C24 were determined as follows.
Clairamide (51) was first treated with 03 to destroy the aromaticity ofthe thiazoline ring, a structural feature that may facilitate the racemization ofC11 when nitrogen is protonated during the acid hydrolysis. 90 After the ozonolysis was completed, the depsipeptide was subjected to acid hydrolysis and followed by derivatization with
Marley's reagent. 84 The results from Marley's analysis revealed the L-configurations of
N-MePhe, N-MeVal and Ala. The L-configuration ofAla corresponded to an S configuration at C11 in 51. The absolute configurations at C2, C3 and C30 are currently under investigation. 61
Table IV. 1. NMR data for Clairamide (51) in DMSO-d6
C# 'jc DEPT 1H mult (Hz) HMBCto 1 171.22 c 2 44.42 CH 2.72 m C: 3 3 52.35 CH 4.08 m 4a 25.59 CH2 1.52 m C: 3, 5 4b 1.60 m C: 3, 5 5 10.95 CH3 0.94 t (7.3) C: 3, 4 6 14.53 CH3 1.11 d (6.9) C: 1, 3 NH-1' 9.01 d (10.6) C: 7 7 159.96 c 8 147.52 c 9 124.03 CH 8.21 s C: 8, 10 10 170.01 c 11 47.45 CH 5.11 m C: 10, 12 12 23.64 CH3 1.37 d (6.7) C: 10, 11 NH-2' 7.76 d (6.3) C: 10, 13 13 167.65 c 14 60.00 CH 5.39 t (7.5) C: 13, 16, 22, 23 15a 36.19 CH2 2.83 dd (14.0, 7.0) C: 13, 14, 16, 17 15b 3.19 dd (14.0, 7.1) C: 13, 14, 16, 17 16 136.89 c 17 128.84 CH 7.17 t(7.0) C: 16, 18 18 128.27 CH 7.14 t (7.0) C: 17, 19 19 126.57 CH 7.07 t (7.0) C: 17,21 20 128.27 CH 7.14 t (7.0) C: 19,21 21 128.84 CH 7.17 t(7.0) C: 16,20 22 28.67 CH3 2.99 s C: 14 23 168.75 c 24 57.08 CH 4.89 d (10.5) C: 23, 25, 26, 27, 28,29 25 26.46 CH 2.15 m C: 24, 26,27 26 17.97 CH3 0.27 d (6.5) C: 25 27 18.08 CH3 0.77 d (6.5) C: 25 28 29.70 CH3 3.00 s C: 24 29 172.53 c 30 67.54 CH 5.28 q (6.8) C: 1, 31 31 15.59 CH3 1.29 d (6.8) C: 29.30
All spectra were recorded on a Broker DRX600 spectrometer eH spectra referenced to the solvent signal at 2.49 ppm; 13C spectra referenced to the carbon signal ofthesolvent at 39.51ppm). 1H-13C connectivities assigned by HSQC; HMBC optimized for 8Hz coupling. 62
Carliamide (52) was obtained as a colorless oil from the same fraction that yielded 51. Pure 52 analyzed for a molecular composition ofC3sH53N4 0 7 (+2.7 mmu deviation) by high resolution FABMS. TheIR spectrum of 52 indicated the presence of
1 1 13 ester (1712 cm- ) and amide (1652 cm- ) carbonyl functionalities. The C NMR spectrum of52 possessed five ester/amide carbonyl signals together with signals at
6135.04 (s), 129.38 (2C, d), 114.06 (2C, d) and 159.03 (s), resonances characteristic ofa para-substituted phenyl ring (Table IV.2). This accounted for nine ofthe twelve degrees ofunsaturation implied by the molecular formula of52.
Interpretation of lD and 2D NMR spectral data of 52 generated five partial structures (52a-52e) (Figure IV.5). Partial structures 52a-52d were easily deduced as 2 hydroxy-3-methylpentanoic acid (Hmp), N,O-dimethyltyrosine (N, O-Me2 Tyr), valine
(Val), and N-methylalanine (N-MeAla), respectively. For partial structure 52e, H2-H6,
H3-9 and NH-1 spin systems were generated from the COSY spectral data. The correlation observed from the H2-6 to H8 resonance in the TOCSY spectrum (Figure
IV.6), together with the HMBC correlation (Figure N.7) observed from H2-6 to C7 and
C8, defined partial structure 52e as 3-amino-2-methyl-7-octynoic acid (Amoya). The presence ofthe terminal acetylene functionality in 52e accounted for two ofthe three remaining degrees ofunsaturation. Hence, carliamide (52) was a monocyclic.
The connectivities ofpartial structures 52a-52e were accomplished using HMBC spectral data (Figure N.7). The three-bond correlation showing from the a.-proton (H33) ofN-MeAla to C27 ofVallinked the two partial structures (52c and 52d). A two-bond heteronuclear coupling exhibited from NH-2 to C16 connectedN,O-Me2Tyr residue to
Val. The Hmp unit was placed adjacent to N,O-M~Tyr based on the HMBC correlations observed from H17 and H3-26 to ClO. Finally, the observed two-bond correlation from the amide proton (NH-1) ofAmoya (52e) to C32 ofN-MeAla and from H11 ofHmp to
Cl ofAmoya completed the cyclic structure assignment ofcarliamide (52). 63
1 _,~2' ~~0 ~ ~~~~ '~ HsC H3C CH3 15 14 CH3 30 31 52a 52b 52c
52e
Figure IV.S. Partial Structures Carliamide (52a-52e)
Carliamide (52) 64
Carliamide (52)
810,824 , NB-2 I /B21,Bll
I I ~ • 0 x~ • I :g IO I 0 •.iG 9 0 1 ;di tu eo • • I
I • •
c . ' l e' • 0
7 6 5 4 3 2 1 ppm
Figure IV.6. TOCSY Spectrum ofCarliamide (52). 65
Carliamide (52)
Solvml H2S \ HlO,II24 1134 ~H4bm• m ms NH2,~.1123 I 11211 Hl8b m ~HS~ H14 NH-1 Hll I \ , •• H3 8 Hd \~ \I 1!i J J,3ab .t r J.1 .• 'J 1.0. -- __,...... _ ppm
0 .. 10 . ••• • • •. • •...... :r...... 20 . . . t.: • : • • 30 .. • • • • •• ...• 40 . .. • 50 •.. • ... . t . . . •. • 60 .. 70 • • ...... 80 90
100 . .. 110 120 " . It ... . • 130
140
150 • • • 160 • •• • I . • • •• " 170 ' • • I 180 7.5 7.0 6.5 6.0 5.5 6.0 4.5 4.0 8.5 8.0 2.5 2.0 1.5 1.0 0.5 ppm
Figure IV.7. HMBC Spectrum ofCarliamide (52). 66
Table IV.2. NMR data for earliamide (52) in eDeh
Unit e# Be DEPT 1H mult (Hz) HMBeto Amoya 1 174.21 eH 2 44.98 eH2 2.80 m e: 4, 9 3 51.06 eH 4.18 br t 4a 26.64 eH2 1.53 m e: 5, 6 4b 1.89 m e: 5, 6 Sa 25.53 eH2 1.60 m e: 3, 4, 6, 7 5b 1.80 m e: 3, 4, 6, 7 6 17.81 eH2 2.20 m e: 5, 7, 8 7 84.01 e 8 68.64 eH 1.92 t (2.4) 9 13.81 eH3 1.20 d (7.2) e: 1,2,3 NH-1' 5.43 e: 32 Hmp 10 171.83 e 11 75.90 eH 4.52 d (8.3) e: 1, 10 12 36.45 eH 1.52 m 13a 23.38 eH2 0.55 m e: 11, 12, 14 13b 0.58 m e: 11, 12, 14 14 11.29 eH3 0.60 t (7.07) e: 12, 13 15 14.81 eH3 0.85 d (6.63) e:11,12,13 N,O-Me2 16 167.24 e Tyr 17 58.76 e 5.62 dd (13.0, 4.5) e: 10, 16,26 18a 32.91 eH2 2.82 dd (15.5, 13.0) e: 17, 20,24 18b 3.80 dd (15.5, 4.5) e: 17, 20,24 19 135.04 e 20,24 129.38 eH 7.11 d (8.5) e: 22 21,23 114.06 eH 6.81 d (8.5) e:22 22 159.03 e 25 55.16 eH3 3.76 s e: 22 26 31.33 eH3 3.01 s e: 10 Val 27 167.31 e 28 55.16 eH 4.48 dd (9.6, 9.6) e: 29, 30,31 29 29.69 eH 2.28 m 30 18.445 eH3 0.97 d (6.9) e: 28 31 20.18 eH3 0.96 d (6.6) e: 28 NH-2' 7.09 d (9.6) e: 16 N-MeAla 32 168.20 e 33 61.29 eH 3.43 q (6.8) e: 27, 32, 34, 35 34 12.48 eH3 1.48 d (6.8) e: 32 35 37.70 eH3 3.10 s e: 33
All spectra were recorded on a Broker DRX600 spectrometer ctH spectra referenced to the solvent signal at 2.49 ppm; 13e spectra referenced to the carbon signal ofthe solvent at 39.51ppm). 1H- 13e connectivities assigned by HSQe; HMBe optimized for 8Hz coupling. 67
From a biosynthetic perspective, clairamide (51) is unusual because it contains 3 amino-2-methylpentanoic acid (Ampa), a unit which has only previously found in the 39 structure ofmetabolites isolated from L. majuscula, such as lyngbyastatin 1 (16), and from the marine mollusk Do/abel/a auricularia such as dolastatin 12 (17).40 This latter metabolite was recently found in an extract ofan assemblage ofL. majuscula and
Schizothrix calcicola, thereby suggesting that the real source ofthis compound is a cyanobacterium. Cyanobacteria are thus far the only organisms known to produce the unique Ampa unit. Similarly, the unique 3-amino-2-methyl-7-octynoic acid (Amoya) of carliamide (52) has been reported only as a substructure ofa depsipeptide, onchidin A
(53), from the marine mollusk Onchidium sp., suggesting that cyanobacteria are possibly
91 92 the true producers ofthe onchidins. •
Lyngbyastatin-1 (16) R = OCH3 Dolastatin 12 (17) R = H
Onchidin A (53) 68
Experimental
GeneraL Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker
DRX600 instrument operating at 600.01 MHz for 1H NMR and at 150.90 MHz for 13C
NMR. Proton spectra were referenced to the solvent signal 8 2.49 for DMSO-d6 and 8
7.26 for CDCh. Carbon spectra were referenced to the carbon signal ofDMSO-d6 at
39.51 ppm and CDCb at 77.0 ppm. Infrared (IR) spectra were obtained using a Nicolet
510 Fourier transform IR (FTIR) spectrometer. Ultraviolet (UV) spectra were run on a
Hewlett-Packard 8452a spectrometer. Low and high resolution mass spectra were recorded on a Kratos MS 50 TC. High performance liquid chromatography (HPLC) utilized a Waters pump, Rheodyne 7125 injector and a Waters Lambda-Max 480 LC spectrophotometer. Optical rotations were measured on a Perkin-Elmer 141 polarimeter.
Merck aluminum-backed thin layer chromatography (TLC) sheets (silica gel F2s4) were used for TLC. Vacuum liquid chromatography (VLC) was performed with Merck Silica gel G for TLC or with Baker Bonded Phase-octadecyl (C1s). All solvents were distilled from glass prior to use.
Collection. The marine cyanobacterium Lyngbya majuscula (voucher specimen available as PNSB-4 Sept 98-01) was collected from shallow water (2-3m) at Wewak
Island, Papua New Guinea and immediately stored in isopropanol for transport, then kept at reduced temperature (-20°C) until extraction.
Isolation and Purification. Approximately 0.5 L ofpreserved alga was extracted with CH2Ch/MeOH (2: 1) twice to give 2.31 g ofcrude organic extract and 170.1 g dry weight ofextracted algal material. A portion ofthe organic extract ( 1.72 g) was subjected to silica gel vacuum liquid chromatography (NP-VLC) with a stepped gradient elution from 100% hexanes to 100% EtOAc, giving eight distinctive fractions. Fraction
4, a fraction with strong brine shrimp toxicity (LDso = 1 ppm), was sequentially purified by Cts Sep-Pak and NP-HPLC [Verapack Silica 10 Jl, 4.1 mm x 30 mm, 4% 69
EtOAc/hexanes, flow rate 9 ml/min, detection at 254 nm, tR = 24 min for 15 and 25.5 min for 50] to yield curacin A (15, 16 mg) and curacin D (50, 14 mg).
In addition to fraction 4 from VLC, fraction 6 also exhibited strong brine shrimp toxicity (LDso = 5 ppm). Processing this moderately polar fraction by reversed phase
(Cis) vacuum liquid chromatography followed by ODS HPLC [Phenomenex Spherisorb
ODS (2), 10 J..l, 250 x 10 mm, MeOHIH20 (7:3), flow rate 4 ml/min, detection at 215 nm, tR = 18 min for 51 and 20 min for 52] yielded clairamide (51, 3.2 mg) and carliamide (52,
1.1 mg). 20 Clairamide (51). A colorless oil having the following properties; [a.] 0 = -8° (c 0.1, MeOH); UV (MeOH) Amax (E) 214 nm (14 400); IR (K.Br) 3309,2964,2930, 1732,
1669, 1620, 1551, 1519, 1496, 1465, 1274, 1221, 1078, 1047, 789 cm-1; FABMS m/z 614
[M+1t (100), 522 (6), 437 (7), 329 (5), 293 (5), 273 (9), 244 (6), 217 (10), 196 (8), 168
(56), 134 (21), 97 (19), 86 (27), 70 (4), 58 (10); HR-FABMS mlz 614.3023 (calcd. for
C31Ht4Ns 06 S, (+ 1.1 mmu deviation); for 1H and 13C NMR data see Table IV.l.
AcidHydrolysis andMatfey's Analysis ofClairamide. Clairamide (2 mg) was hydrolyzed with 6 N HCl, 105 °C for 12 hrs. The hydrolysate was concentrated to dryness. The residue was dissolved in H20 (50 J..lL), and to the resulting mixture was added a 1% (w/v) solution of 1-fluoro-2,4-dinitrophenyl-5-L-alanine amide (FDAA) in acetone (100 J..lL) and 1 M NaHC03(20 J..lL). After being heated at 37 oc for 1 hr, the reaction mixture was cooled, acidified with 2 N HCl (20 J..lL), and evaporated to dryness.
The resulting product was then resuspended in 2 mL ofDMSO/H20 (1:1). The solution was analyzed by RP-HPLC [Waters Nova-Pak® Cts, 3.9 x 150 mm ], UV detection at 340 nm using a CH3CN in H20 linear gradient (10-50% CH3CN over 60 minutes with 0.05% TFA, 1 mL!min) for Ala andN-MeVal, and a CH3CN in 50 mM triethylammonium phosphate (pH 3.0) linear gradient (10%-50% CH3CN over 60 min, 1 mL/min) for N
MePhe. The retention time (tR, min) ofthe derivatized amino acids in the hydrolysate of 70
51 matched those L-Ala (25.1; D-Ala, 29.1), L-N-MeVal (38.0; D-N-MeVal, 41.0 ), and
L-N-MePhe (38.6; D-N-MePhe, 39.1).
20 Carliamide (52). A colorless oil having the following properties; [a] 0 =-76° (c 0.1, MeOH); UV (MeOH) Amax (c) 222 nm (16 400); IR (K.Br) 3361, 2961,2930,2872,
1 1712, 1652, 1513, 1458, 1300, 1248, 1226, 1095, 1036,829,782 cm" ; FABMS mlz 641
[M+1t (100), 625 (6), 376 (5), 345 (5), 288 (21), 260 (7), 249 (4), 164 (28), 135 (8), 121
(18), 135 (9), 121 (19), 96 (6), 77 (5), 58 (25); HR-FABMS mlz 641.3941 (calcd. for
1 13 C35H53N407, (+2.7 mmu deviation); for H and C NMR data see Table IV.2. Brine Shrimp Toxicity Assay. In a method slightly modified from the original description,66 about 15 newly hatched brine shrimp (Artemia salina) in ca. 0.5 mL artificial seawater were added to each well containing different concentrations of the sample in 50 JA.L EtOH and 4.5 mL artificial seawater to make a total volume ofca. 5 mL.
Samples and controls were run in duplicate. After 24 h at 28°C, the brine shrimp were observed and counted with a dissecting light microscope. The percentage of live brine shrimp versus total brine shrimp was used to determine LD5ovalues. 71
CHAPTERV
THE BIOSYNTHESIS OF BARBAMIDE
Abstract
Stable isotope incorporation studies have been used to investigate the biosynthesis ofbarbamide (47), a molluscicidal metabolite isolated from the marine cyanobacterium
Lyngbya majuscula. Feeding experiments conducted with cultured L. majuscula have provided clear evidences that barbamide biosynthesis involved chlorination ofthe unactivated pro-S methyl group ofleucine. Incorporations of [ 1,2- 13C]acetate and [3
13C]- L-phenylalanine confirmed the origins of C5-C6 and the phenyl group of 47 (C7,
C8 and C10-C16), respectively. The thiazole ring (C17-C18) of47 was shown to arise
1 13 from cysteine through a [2-13C, ~]glycine feeding experiment. Detection ofintact C
15N bond was observed by application ofa new GHNMBC NMR experiment. Results from the latter feeding experiment also indicated that the NCH3 and OCH3 groups of47 originated from the cl pool. 72
Introduction
Marine organisms are prolific sources ofstructurally unique halogenated 1 15 secondary metabolites. " A majority ofthese halogen atoms are incorporated into positions which are suggestive oftheir reaction as x+ species, although there are a few examples wherein passive incorporation ofX" species appears rational.93 Haloperoxidase enzymes responsible for the formation ofthe x+-halogenating species have been found in 94 virtually all classes ofmarine organisms and have been an area ofintense interest. In contrast, a number of sponge-cyanobacterial and cyanobacterial metabolites possess halogenated functional groups where the electronic nature ofthe halogenating species is uncertain. Such an example is the unique trichloromethyl group ofbarbamide (47), a molluscicidal metabolite we isolated from the marine cyanobacterium Lyngbya maJUSCU. Ia. 53
Discovery of47 is exceptional in several regards. First, it helps to clarify the origin ofother related metabolites from the sponge-cyanobacterial-bacterial complex.95
Second, unlike the sponge symbiotic cyanobacterium ( Oscillatoria spongeliae), the barbamide-producing cyanobacterium is a free-living organism and is potentially culturable. Therefore, it provides an experimental system in which to examine the chlorination reactions ofthe trichloromethyl moiety of47 as well as to identify the other biosynthetic units ofthe metabolite. To this end, we have brought back the barbamide producing strain ofL. majuscu/a from Curayao and cultured it in our laboratory. In culture, the organism (strain 19L) retains its capacity to produce barbamide in good yield
(ca. 2.4% ofextractable lipid).
For the biosynthesis ofbarbamide (47), we initially postulated that Cl-C4 and C9 ofthe molecule are derived from 5,5,5-trichloroleucine (57) that was biosynthesized by a pathway similar to that ofleucine via a proposed intermediate, trichloropyruvate (56)
(Figure V.l). The existence of57 is supported by the isolation ofdiketopiperazine 73
96 91 derivatives of57, such as 60 and 61 from marine sponge Dysidea herbacea. •
Transamination and decarboxylation of 57 could give rise to 58, which could be a ketide chain extended by malonyl CoA to provide intermediate 59 (Figure V.l). Our original proposal, presented in Figure V.l, also includes theN-methyl phenylalanine and cysteine as precursors to the phenyl and thiazole rings, respectively. A heterocyclization ofthe side chain ofcysteine with the carbonyl carbon ofphenylalanine followed by a two electron oxidation complete the formation ofthe thiazole ring. AnN-acylation of intermediate 59 with 55 is conceivably followed by a methylation at the olefinic oxygen by S-adenosylmethionine (SAM) to complete the biogenesis of47. This chapter describes feeding experiments conducted with the cultured marine cyanobacterium L. majuscula to test the hypotheses presented in Figure V.I.
9 yH3 _ ?CH3fCl3 ~J~'••l N~-3 ~H3 13v . 16 o D 17 18
Barbamide (47)
CI~~CCI, CH3
60 61 CCX()()£H+ -~-~,~·--~-0- .:~ J ·ooc ~ """""'
~Is a~ J~u 9 yCia (:> _?HyCia HOOC CHs ~ EnzSr~CHa Enz~CHa 5,5,5-Trichloroleucine (57) NADP+ NADPH 58 0 59 1l EnzSH C02 U Qc~SEnz SAM 0
Trichloropyruvate (56) t scr
HJcoo· Barbamide (47) Pyruvate
Figure V.l. Biogenesis ofBarbamide (47) from Primary Precursors. 75
Results and Discussion
Isolation of Barbamide ( 47). The marine cyanobacterium L. majuscula was cultured in our laboratory as described previously. 98 For the precursor-incorporation feeding experiments, a 50-75 mL ofwet-pack cell ofL. majuscula (strain 19L) was inoculated into fresh medium (SWBG 11 ). After three days ofequilibration, isotope labeled precursors were administered to the cultures, which were harvested on day 9 or
10. The crude organic extract ofthe cultures was subjected to normal phase vacuum liquid chromatography (NP-VLC}, Cts SepPak, and ODS-HPLC (Cts, 80% MeOH/FhO), respectively, to provide the labeled barbamide ( 47).
Leucine Feeding Experiments. The biogenetic proposal in Figure V.l predicts that the 5,5,5-trichloroleucine (57) is biosynthesized via trichloropyruvate (56).
Therefore, ifisotopically labeled leucine is supplied to the barbamide-producing culture, no direct incorporation should be observed. To test this hypothesis, L-[2-13C]leucine was fed to the cultures ofL. majuscula and the isolated 47 was analyzed by the 13C NMR spectroscopy.99 To our surprise, we observed a 452% increase in signal intensity at o 166.8 when normalized to unlabeled carbons. This carbon was originally assigned to the amide carbonyl (C6).53 However, upon the reexamination ofthe complex region in the original HMBC spectrum of47, correlations were observed from the OCH3 protons at o
3.60 too 166.80 and from NCH3 protons at o 2.88 too 167.04. From these new analyses, a correction was made in the published data, the resonance at o166.80 and o 167.04 were assigned to C4 and C6, respectively (Table V.l). Thus, the high incorporation level implies that the chlorination step follows the biosynthesis ofleucine from pyruvate.
Having established that C2 to C6 ofleucine contributed to C1-C4 plus C9 of47, it remained in question whether the carbonyl carbon ofleucine contributes to C5 of47 or if it is lost during the biosynthetic pathway. To probe the fate ofCl ofleucine during the biosynthesis ofbarbamide and to be able to draw firm conclusions even from a negative 76
incorporation result, similar amounts ofboth L-[l-13C] leucine and L-[2- 13C]leucine were provided to the cyanobacterium. Analysis ofthe 13C NMR spectrum ofbarbamide produced under these conditions (Figure V.2), in comparison with a natural abundance sample, showed an expected enhancement at C4 of47 (203% when normalized to the unenriched signals) whereas C5 of 47 showed no enhancement (1 02% ). This result clearly suggests that Cl of leucine is lost in the biosynthesis of47.
To examine the chirality ofthe chlorination ofthe prochiral methyl group of leucine (the C2 chirality of47), 4(S)-L-[5)3C]leucine and 4(R)-L-[S)3C]leucine were 100 101 synthetically prepared by C. Willis eta/., and separately provided to cultures. •
Analyses ofthe 13C NMR spectra ofbarbamide from these two experiments as well as a natural abundance spectrum (Figure V.3) demonstrated that 4(S)-L-[5-13C]leucine selectively enhanced the signal for the C9 trichloromethyl group of47 463% whereas
4(R)-L-[ 5-13C]leucine selectively enhanced the signal for the C 1 methyl group 593%. A slight enrichment was also observed for C9 signal from the latter feeding experiment
(177% increase in signal intensity) which was likely due to the 70% de at C4 in this synthetic leucine preparation. Results from the incorporation ofthe two chirally 13C labeled leucines into 47 demonstrated that chlorination reaction occurred at the pro-S methyl group ofleucine and established the 2(S) stereochemistry of47. Trichloromethyl moieties possessing the same stereochemistry at comparable centers have been observed 96 102 108 in other "sponge-cyanobacterial" metabolites. ' " Figure V.6 summarizes results from the 13C-labeled leucine feeding experiments. A.
C11-C15 1 C-4 ~ C-9 C-5 I C-10 ~;17 .l C-18 J I .Jll . Jl J IL. ..,.. ·' ' 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 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 ppm B.
C-4 C11-C15 ~ . I C-10 II C-9 C-5 I J . . C-17 j I C-18 I .J ....,._.~._.•..,.,_,..,,.,..J.,....,..,Witlll!llllll..- ..lrlu~•••-'*'''*"'•'illlWMtu"'-•Mtt~••...,_A.r., 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 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 ppm
Figure V.l. Comparison of Selected Regions ofthe 13C NMR Spectra of(A) Natural Abundance Barbamide (47), (B) Barbamide (47) Produced During Supplementations with [1)3C]-L-Leucine and [2}3C]-L-Leucine. ::j 78
A.
C-1
C-9
I IJ l .I
II I I I II ''I I I II I I I I' I I I II I j '''I' I I iii'' 'I'' I ••••• I I' ''''''''I'',,,''''1'''''''''1'''''' '''I IIIII Iii I 100 90 80 70 60 50 40 30 ppm
B.
C-9
C-1
100 90 80 70 60 50 40 30 ppm c. C-1
iii Iii '''I''' I II II 'I''' iii II 'I''',,,,, I, •••••••• 'I'' II I I '''I''' •••••. , ••••••••• ,. I,,,,, I., ••••••••• 100 90 80 70 60 50 40 30 ppm
Figure V.3. Comparison ofSelected Regions of(A) Natural Abundance Barbamide (47), (B) Barbamide (47) Produced During Supplementation with 4(S)-L-[5-13C]Leucine, (C) Barbamide (47) Produced During Supplementation with 4(R)-L-[5-13C]Leucine. 79
Although diketopiperazine derivatives of5,5,5-trichloroleucine (57) have been isolated, the possibility ofthe chlorination reaction arising through the leucine catabolic pathway could not be excluded. During the degradation ofleucine, a carboxylation of C4 ofintermediate 62 (or C5 ofleucine) would provide intermediate 63, which can be channeled through a successive electrophillic addition by Ct and a decarboxylation to yield 4,4,4-trichloroisovaleroyl CoA (58) (Figure V.4). To probe such a possibility, L eHIO]leucine was fed to the cultures ofL. majuscula and the labeled 47 isolated from this feeding experiment was analyzed by 2H NMR. The 2H NMR spectrum ofthe enriched 47 (Figure V.5) showed two bands, one centered at o3.13 (for H-2 and/or H-3) and the other 1 at o1.22 (for H3-1) [the assignments ofthese H NMR signals were confirmed by analysis of2D NMR data of47 in toluene-ds]. Integration ofthese two peaks showed the
2 2 2 ratio of H3-1 to H-2 + H2-3 to be 3.00:2.77, suggesting that no losses ofdeuteriums from C3 or C4 ofleucine occurred during its incorporation into 47. The retention ofsix deuterium atoms was confirmed by the observed 125% enrichment ofmlz 467 ([M+ 1t (natural abundance 47: FABMS m/z461 [M+lJl. In light ofthese results, the C9 methyl group of47 is not activated to electrophilic chlorine additions via a leucine catabolic pathway. Because chlorination occurs at a stage after leucine biosynthesis with no detectable methyl group activation, novel chlorination mechanisms may be involved.
yHs fjJH2 yHs ~ yHs ~ HsC~COO~ H 3~SC;:- H~SCoA- Leucine 62
")J_ CCI3 0 HaC ~ SCoA H3~SCoA 63 58
Figure V.4. Proposed Formation of4,4,4-Trichloroisovalerate (58) via Leucine Catabolic Pathway. H-1 A. B. Solvent
H-1 H2 H3a H8a
Solvent
H8b\-~ H2+H3ab
I ' ' ' ' I ' ' ' ' I ' ' ' ' I ' I I I I I I ' I I ' ' I I 3.5 3.0 2.5 2.0 ppm 3.5 3.0 2.5 2.0 ppm
1 2 Figure V.S. Comparison ofSelected Regions of(A} H NMR (Toluene-d8) with (B) the H NMR (Toluene) ofBarbamide (47) Produced During Supplementation with L-eHto]Leucine. gg 81
Table V.l. NMR data for Major Conformer ofBarbamide ( 47) in DMSO-d6and Toluene-ds.
C# H (mult, JHz) c H (mutt, JHz) c DMSO-d6 DMSO-d6 toluene-ds toluene-ds 1 1.10 (d, 6.3) 15.22 1.30 15.87 2 2.84 (obscured) 52.02 3.15 o scured) 53.34 3a 3.02 (obscured) 33.98 3.16 (obscured) 35.32 3b 3.02 (dd, 13.1, 12.4) 3.34 (dd 13.1, 12.4) 4 166.80 166.08 5 5.37 (br s) 94.00 4.78 (br s) 95.09 6 167.04 168.72 7 6.31 (dd, 10.3,5.5) 54.28 6.69 (dd 10.8, 3.8) 53.28 8a 3.25 (dd, 14.3, 10.3) 35.82 3.23 (obscured) 37.15 8b 3.56 (obscured) 3.39 (dd 14.3, 6.2) 9 105.67 106.66 10 137.72 obscured obscured 11 7.28 (m) 128.82 obscured obscured 12 7.28 (m) 128.13 obscured obscured 13 7.28 (m) 126.32 obscured obscured 14 7.28 (m) 128.13 obscured obscured 15 7.28 (m) 128.82 obscured obscured 16 169.42 169.91 17 7.79 (d, 3.2) 141.99 7.51 (d, 3.2) 142.23 18 7.71 (d, 3.2) 120.70 6.61 (d, 3.2) 119.62 NCH3 2.88 (s) 28.26 2.61 (s) 30.52 OCH3 3.60 s 55.48 2.88 s 54.33
13 C NMR data recorded at 100.61 MHz; referenced to 39.51 ppm (for DMSO-d6) and 1 20.4 ppm (for toluene-d8). H NMR data recorded at 400.13 MHz; referenced to 2.50 ppm (for DMSO-d6) and 2.09 ppm (for toluene-ds).
Sodium 3(R)-[4-13C]Methylbutanoate and 3(R)-[4-13C]Methylbutanoyl N
Acetyl-cysteamine (NAC) Feeding Experiments. As shown in Figure V.1, we hypothesized that the chlorination reaction takes place prior to the decarboxylation ofC 1 ofleucine. To examine this proposal, we separately provided the synthetically prepared
3(R)-[4- 13C]methylbutanoate and its N-acetylcysteamine (NAC) thioester derivative to cultures ofL. majuscula. The latter intermediate was provided to the cultures because it may undergo transesterification with a thiol group ofa biosynthetic enzyme without undergoing a metabolic transformation prior to the incorpomtion ofthe feasibility ofthis approach. Careful analysis ofthe 13C NMR spectra of47 from both feeding experiments 82
revealed no incorporation ofthe precursors. These results suggest that the chlorination reaction is likely to occur prior to the decarboxylation ofel of leucine. However, we have not excluded the possibility ofthe precursors not being taken up by the cyanobacterial cells.
13 [1·13c]Jeucine [2·13C]Ieucine 4(S)-L-[5.1 3C]Ieucine 4(R)·L-[5- C]Ieucine ~ ~,..----!-----,.!
yHs ~H3 fC1s• ~COOR )( ··r·· CHs N~-0 • R =H 3(R)-[4-13C]methylbutanoic acid R =NAC 3(R)-[4-13C]methylbutanoyl N-acetyl cysteamine 0 Barbamide (47)
Figure V.6. Summary ofResults from the Be-Labeled Leucine and 3(R)-[4 Be]Methylbutanoic Derivatives Feeding Experiments to Study the Biosynthesis of Barbamide ( 47).
Acetate Feeding Experiments. To determine ifes and e6 of47 originated from an intact acetate unit, a [1,2-Be2]acetate feeding experiment was conducted. The Be NMR spectrum (Figure V.7) of47 produced under these conditions exhibited a coupling oftwo spin-coupled doublets between es and e6 eJcc = 71.8 Hz), suggesting that es and e6 of47 are derived from an acetate unit. Because barbamide existed in two NeH3 rotamers, therefore Be signals ofthe minor conformations are labeled as e4', eS' and e6'. 83
C5 (J =71.8 Hz)
A. B. cs
C5'
I ' ' I ' I I I I I I I I I I ' I ' ' I I I I I I I 94 ppm 94 ppm
c. D. C6 (J =71.8 Hz) C4
C6
C6' C4'
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 167 ppm 167 ppm
Figure V.7. Comparison ofSelected Regions of 13C NMR Spectra ofBarbamide (47) Produced During Supplementation with [l,2-13C2]Acetate (A and C), and Natural Abundance Barbamide ( 47) (B and D). 84
To determine the origin of the oxygen atom on C6, we administered [1
13C, 1802]acetate to the cultures ofL. majuscula. The 13C NMR spectrum (Figure V.8) of 47 produced from this feeding experiment revealed an observed 0.03 ppm upfield isotope shift at C6 resonance (6 167.04), which is in the reported range of0.03-0.55 ppm of 180 labeled carbonyl carbons. 109 The results from this feeding experiment convincingly showed that C6 originates from the carbonyl of an acetate unit and the oxygen atom was not lost during the biosynthesis of 47.
A. C-6 B. L\() =0.03
C-6' A<>= 0.03 C-4 C-6 C-4
C-6'
167.0 ppm 167.0 ppm
Figure V.8. Comparison of Selected Regions of 13C NMR Spectra of (A) Barbamide 13 18 (47) Produced During Supplementation with [1- C, 0 2]Acetate, and (B) Natural Abundance Barbamide ( 47). 85
L-[3-13C)Phenylalanine Feeding Experiment. To probe the biosynthetic origin ofthe phenyl moiety of47 (C7-C8 and CIO-Cl6), L-[3)3C]phenylalanine was provided to the cultures. Analysis ofthe 13C NMR spectrum of47 (Figure V.9) isolated from this feeding experiment showed a 219% increase in signal intensity for C8. This incorporation result suggests that C7-C8 and CIO-Cl6 of47 arise from L-phenylalanine.
[2-13C,15N]Glycine Feeding Experiment. The biosynthetic proposal presented in
Scheme 1 requires cysteine to be a direct precursor to the thiazole ring of47 (except for
C16). However, a direct examination ofthis hypothesis is problematic, as isotopically labeled cysteine is very expensive, metabolically labile, and toxic to the organism (when more than 20 mg/L ofDL-cysteine was supplied to the cultures). These concerns prompted the exploration ofother approaches to provide information on the origin ofthe thiazole moiety in 47.
Because cysteine metabolically derives from serine, a feeding experiment utilizing this latter amino acid was attempted. Unfortunately, serine was also toxic to the organism (when more than 25 mg/L ofL-serine was supplied to the cultures). Conceived as an alternative to feeding cysteine or serine, it was thought that glycine, the metabolic precursor ofserine and cysteine, might be better tolerated by the cyanobacterium. 52 As shown in Figure V.lO A, in the conversion ofglycine into cysteine, the C2 carbon and nitrogen atom remain intact. Hence, ifan intact incorporation ofthe 13C-1~ doubly labeled glycine fragment is observed in 47 from a [2-13C,15N]-glycine feeding experiment, this would provide supporting evidence that the thiazole ring of47 arises from cysteine. C-8
OMe C-2 C-8' C-3 OMe'
jli
OMe C-2 C-3 C-8 OMe' C-8'
1"'"""1"111"''I I iiiI I II'I"" I Ill I I' Ill I I I I '1 '""""I""'"' I I"'"'"'I"Ill' I "1"'"""1" II I I I I I I"' "'"'I' I I"' I i 'I I II II'i"I"""''I III II II '"I'' Ill iii 'I' I II I ""I'" I Iii I I I' I I I'""l"iii'" I I'' I' I I I "I I I 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 ppm
Figure V.9. Comparison ofSelected Regions ofthe 13C NMR Spectra of (A) Barbamide (47) Produced During Supplementation with L-[3)3C]Phenylalanine, and (B) Natural Abundance Barbamide {47). A. B. ~ a a H2~...... COOH H~COOH \. C02 +NH4+ '"""".. C02 + NH4+ Glycine ( Glycine ~ N',N''-met>yle.,..THF THF Ni,N°-metlylene-THF ThtF l a"Y T u··THF x:H H2N...... COOH -~ J _,CsH H~-y COOH ;:s-CH' Glycine H2 OOH H2N COOH Serine Methionine Homocysteine I ~a, ~X CH3 ~ ~H - T H2 T COOH 0 ...~1· Cysteine T T Barbamide (47) C(XN 0 CHs u Barbamide (47)
Figure V.lO. Metabolic Relationship of(A) Glycine and Serine/Cysteine, (B) Glycine and Methionine, and the Expected Labeling Pattern in Barbamide (47) Produced During Supplementation of [2-13C,1'N]Glycine. 88
In the analysis of47 from the [2·13e, 1 ~]glycine feeding experiment, we encountered a number ofchallenges in the detection ofan intact 13e·15N bond incorporation. First, e2 ofglycine to serine and in so doing enriches e3 of serine/cysteine (Figure V.10 A). As a result, 13e.l3e coupling between e2 and e3 of cysteine (e17 and e18 in 47) greatly reduces the sensitivity ofdetection in the one
13 dimensional e NMR analysis. Second, barbamide existed in two different N·eH3 amide conformations in DMSO-d6, which further complicated the analysis.
Improvement in the detection of 13 e· 1 ~ intact incorporation in barbamide was obtained in two ways. First, theN-methyl amide rotamer population in 47 was
1 1 minimized in toluene-d8 at 305 K. Second, we developed a new gradient H· ~ heteronuclear multiple bond correlation (GHNMBe) method for observing the coupling within intact 13e·15N units based on a modification ofthe HMBe experiment proposed
13 13 110 111 by Seto, Watanabe, and Furihata for the detection of long range e- e couplings. •
As in Seto's experiment, our method relies on an HMBe pulse sequence without a low
1 1 112 pass J filter. In our experiment, the indirectly observed nucleus is ~• Protons 13 1 2 attached to e are split by the JcH coupling of 160-210 Hz for an sp hybridized carbon.
In the case ofthe incorporation of [2- 13 e- 1 ~]glycine (Figure V.1 0 A), the proton that resides on the 13e-Iabeled glycine-derived carbon will appear as a doublet ofca. 180 Hz.
A two-bond coupling from this proton to the thiazole 1 ~ will result in ca. 180Hz doublet in the GHNMBe spectrum at the nitrogen chemical shift. In contrast, ife 17 of the thiazole is unlabeled (12e), the correlation ofH·17 to nitrogen will give rise to an apparent singlet at the H-17 chemical shift. Both ofthese long·range correlations were observed in the spectrum of47 from the [2-13e, 1 ~]glycine feeding experiment (Figure V.11), confirming the intact 13e·15N bond incorporation. In addition, a similar coupling pattern was also observed between NeH3 protons and the nitrogen atom that they are 89
attached to. This two-bond coupling indicates the enrichment ofNCH3 that results from a contribution ofC2 ofglycine to the Ct pool (Figure V.l 0 B), a carbon source for both
NCH3 and OCH3 of47.
13 L-[Methyl- CJ-Methionine Feeding Experiment. To establish that the N-CH3 and O-CH3 of47 derive from the Ct pool via SAM, L-[methyl-13C]methionine was administered to cultures ofL. majuscula. Analysis ofthe isolated barbamide showed only slight enrichment when compared to natural abundance (124% for NCH3 and 136%
13 for OCH3). Unfortunately, L-[methyl- C]-methionine was found to be toxic to the organism when administered at a higher level(> 45 mg!L). The 13C NMR spectrum of
47 from the [2- 13C,15N]glycine feeding experiment (Figure V.l2), however, revealed increases in signal intensities ofNCH3 (214%) and OCH3 (207%). The observation resulted from the contribution ofC2 ofglycine to the C1 pool (Figure V.lO B), a carbon source for both NCH3 and OCH3 groups. Thus, results from this latter experiment suggested that the methyl groups oftheNCH3 and OCH3 of47 originated from the Ct pool.
Conclusion. The incorporation studies described above have provided insight into the biosynthetic origins ofall carbon atoms in 47, which resulted in a revision ofour original biosynthetic proposal (Figure V.l). The leucine feeding experiments have shown that Cl-C4 plus C9 of47 originate from L-leucine. Results from incorporation of the two chirally labeled leucines established the 2(S) stereochemistry ofbarbamide and that chlorination reaction occurs at the pro-S methyl of leucine. An incorporation experiment using L-eHto]leucine showed that the leucine pro-S methyl group is not activated via the leucine catabolic pathway. Since the chlorination reaction occurs at a stage after leucine biosynthesis with no detectable methyl group activation, novel chlorination reactions, perhaps involving radicals, are implicated. 90
A.
• • • 180Hz
NCH3 B.
114
Wi
118
120
122
124
126 185Hz
128
130 132 ~ ,, 134 t I
Figure V.ll. {A) Schematic Representation ofthe Expected Coupling Patterns from the Modified GHNMBC Experiment. {B) GHNMBC S8ectrum Acquired for Barbamide (47) Produced During Supplementation with [2-13C, ~Glycine. 91
As depicted in Figure V.l3, incorporations ofL-[3)3C]phenylalanine and [1,2
13C2]acetate into 47 provided insights into the origins ofthe phenyl moiety and CS-C6 of
13 1 the molecule, respectively. The intact incorporation of [2- C, ~glycine into the thiazole ring of47 using a new modified GHNMBC NMR experiment strongly supports cysteine as a direct precursor. Analysis ofthe 13C NMR spectrum of47 from this latter feeding experiment also provided convincing evidence that the NCH3 and OCH3 groups both derive from the C1 pool. Figure V.l4 presents our revised biosynthetic proposal of barbamide (47). A. OCH3
NCH3 C-2 C-3 C-8 C-7 C-8'
56 54 52 50 48 46 44 42 40 38 36 34 32 ppm
56 54 52 50 48 46 44 42 40 38 36 34 32 ppm
Figure V.12. Comparison of 13C NMR Spectra of(A) Barbamide (47) Produced During Supplementation with [2- 13 C, 1 ~]Glycine, and (B) Natural Abundance Barbamide (47). 93
0 ..-Jlo [1 ,2-13c2]Acetate
... I ... y~ ?CHafC1a 0 0 ··r··N~CHa __+Ha~oo- 0g 0 • 13 [2-13C,15N]Giycine [3- C]-L-Phenylalanine "- "- Barbamide (47)
Figure V.13. Summary of 13C-Labeled Precursors Incorporated into Barbamide (47). a.-KG Glu ~13 \ J. ul3 HOOC. C~3 ~ Enzs[ CH3 5,5,5-Tnchloroleuc~ne (57)NAD+ NADH 58 EnzSH C02 -oocn ~SEnz SAM Novel Chlorination 11 ~ Mechanism? 0 ~3 HOOC CH3 L-Leucine
Barbamide (47)
Figure V.14. Revised Biogenesis ofthe Biosynthesis ofBarbamide (47) from Primary Precursors. 95
Experimental
GeneraL Nuclear magnetic resonance (NMR) spectra were recorded on a Bruker
AM400 instrument operating at 400.13 MHz for 1H NMR and at 100.61 MHz for 13C
NMR. The GHNMBC experiment on 47 was operated on a Bruker DRX600 spectrometer operating at a proton frequency of600.01 MHz. Proton spectra were referenced to 2.50 ppm and 2.09 ppm for DMSO-d6 and toluene-ds, respectively. Carbon spectra were referenced to 39.51 ppm for DMSO-d6 and 20.4 ppm for toluene-ds. High performance liquid chromatography (HPLC) utilized Waters M6000A or Waters 505 pumps, a Rheodyne 7125 injector, and a Waters Lambda-Max 480 LC spectrophotometer or Photodiode Array Detector model 996. Merck aluminum-backed thin layer chromatography (TLC) sheets (silica gel60 F2s4) were used for TLC. Vacuum liquid chromatography (VLC) was performed with Merck Silica Gel G for TLC or with Baker
Bonded Phase-octadecyl (Cts). All solvents were distilled from glass prior to use. The synthetically prepared 4(S)-L-[5-13C]leucine, 4(R)-L-[ 5-13C]leucine, 3(R)-[4- 13C]butanoic acid, and 3(R)-[4- 13C]methylbutanoyl N-acetylcysteamine (NAC) were generously provided by Professor Christine Willis (U. ofBristol). All ofthe other stable isotope precursors were purchased from Cambridge Isotope Laboratories.
General Culture Conditions and Isolation Procedure. 3 g ofL. majuscula strain
19L was inoculated into a 2.8-L Fembach flask containing 1 L ofSWBG11 medium.
The culture was grown at 28 °C under uniform illumination (4.67 J.Ullol photon s-1 m"2), aerated, and equilibrated for 3 days prior to addition ofisotopically labeled precursors.
Cultures ofL. majuscula were harvested 10 days after inoculation, blotted dry, weighed, and repetitively extracted with 2: 1 CH2Ch/MeOH. The filtered organic extracts were dried in vacuo, weighed, and applied to silica gel columns (1.5 em I.D. x 15 em) in 5% EtOAc/hexanes, and eluted with a stepped gradient elution of5% EtOAc to 100%
EtOAc. Fractions containing barbamide (eluted with 50% EtOAc/hexanes) were further 96
fractionated by RP-VLC, using a stepped gradient elution from 60% MeOHIH20 to
100% MeOH. The fractions eluting with 80% MeOH (barbamide-containing fraction) were subjected to a final purification by ODS-HPLC [Phenomenex Spherisorb ODS (2),
4:1 MeOHIH20, flow rate 3 mL/min, detection at 254 nm] to give pure barbamide (47,
3.86 mg/L).
Calculation ofthe Results ofJJC-Labeled Precursor Feeding Experiments on
Barbamide. The followings explains how the calculations for 13C incorporations were performed. The method used with barbamide (47) produced from the 4(S)-L-[5
13C]leucine feeding experiment is presented as an example (Table V.2.). Column A shows all 20 carbons of47. The 13C NMR spectral data for both major and minor conformers of47 are shown in column B and C, respectively. Column D and E show the integration ofeach Be signal (both major and minor conformers) ofthe natural abundance 47 and the enriched 47, respectively. Next, the normalization factors in comparison to C 1 were calculated by individually dividing the integration values ofC1
Cl8, NCH3 and OCH3 (cells D3 to D22) by the integration value ofCl (cell D3). Those values appear in column F. Columns G-Y show the normalization factors used for comparson with other carbons. Multiplication ofthe normalization factors (column F) with the integration value ofCl signal ofthe enriched 47 (cell E3) will provide the normalized integration values for the 13C signals ofthe enriched 47 (column Z). The percentage for the Be enhancements ofenriched 47 in comparison to Cl (column AA) were calculated by dividing the integration values ofthe enriched 47 (column E) by the normalized integration values (column Z), and multiplying by 100. The percentages of enhancement compared to other carbons are shown in columns AC, AE, AG, AI, AK,
AM, AO, AQ, AS, AU, A W, A Y, BA, BC, BE, BG, BI, BK and BM. Finally, the average values of% relative enhancement(s) ofthe carbon signal(s) ofthe enriched 47 were calculated and reported in column BN. 97
Feeding Experiments. A. L-{J-13C]Leucine and L-{2-13C]Leucine. L-[1
13C]Leucine (180 mg) andL-[2-13e]leucine (120 mg), were provided to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of4.60 mg of labeled 1 was isolated from the crude organic extract. The 13e NMR spectrum of47
(DMSO-d6) from this feeding experiment showed 203% increase in signal intensity for e4 whereas no enrichment was observed for es.
B. 4(S)-L-{5J3C]Leucine. 4(S)-L-[5-13C]Leucine (180 mg) was added to 3 x 1-L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of
10.9 mg oflabeled 47 was isolated from the crude organic extract. The Be NMR spectrum of47 (in DMSO-d6) from this feeding experiment showed 463 % increase in signal intensity for e9.
C. 4(R)-L-{5J3C]Leucine. 4(R)-L-[5-Be]Leucine (180 mg) was added to 3 x 1
L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of
12.3 mg oflabeled 47 was isolated from the crude organic extract. The Be NMR spectrum of47 (in DMSO-~) from this feeding experiment showed 593% increase in signal intensity for e 1.
D. L-lHJo]Leucine. L-eHto]Leucine (160 mg) was added to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of6.70 mg of labeled 47 was isolated from the crude organic extract. The 2H NMR spectrum of47 (in toluene) showed two 2H bands, one centered at B3.13 (for H-2 and H2-3) and the other at 2 2 2 B1.22 (for H3-1). Integration ofthese two peaks showed the ratio of H3-1 to H-2 + H2 3 to be 3.00:2.77.
E. Sodium 3(R)-{4J3C]Methylbutanoate. Sodium 3(R)-[4-Be]methylbutanoate
(82 mg) was supplied to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of6.80 mg of47 was isolated from the crude organic 98
13 extract. The C NMR spectrum of47 (in DMSO-d6) from this feeding experiment showed no significant level ofincorporation ofany carbon signal.
F. 3(R)-[4-13C]Methylbutanoyl N-Acetylcysteamine (NAC). The NAC thioester of3(R)-[4-13C]methylbutanoate (145 mg) was supplied to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of4.9 mg of47 was isolated from the crude organic extract. Again, the 13C NMR spectrum of47 from this feeding experiment revealed no significant level ofincorporation ofany carbon signal.
13 3 G. Sodium [1,2- CiJAcetate. Sodium [1,2_1 C2]acetate (208 mg) was mixed with unlabelled sodium acetate (415 mg) and supplied to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of 15.2 mg oflabeled 47 was isolated from the crude organic extract. The 13C NMR spectrum of47 (DMSO-d6) from this feeding experiment showed a pair ofdoublets signals at C5 ctJcc = 71.8 Hz) and C6 eJcc 71.8 Hz). H. Sodium {IJ3C/80]Acetate. Sodium [1-13C, 180]acetate (375 mg) was mixed with unlabeled acetate (622 mg) and supplied to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of 10.5 mg of labeled 47 was isolated
13 from the crude organic extract. The C NMR spectrum of47 (in DMSO-d6) revealed an isotopic shift (AS = 0.03 ppm) at the C6 resonance (S 167 .04). Integration ofthis signal revealed that 0.62% ofthe 13C at C6 carries an 180 atom.
L L-[3-13C]Phenylalanine. L-[3-13C]Phenylalanine (250 mg) was supplied to 3 x
1-L cultures on days 3 and 6, and all three cultures were harvested on day 9. A total of
15.1 mg oflabeled 47 was isolated from the crude organic extract. The 13C NMR spectrum of47 (DMSO-d6) from this feeding experiment showed 219% increase in signal intensity for C8.
J. [2-13C/5N]Glycine. [2- 13 C, 1 ~Glycine (225 mg) was supplied to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of 18 99
mg oflabeled 47 was isolated from the crude organic extract. The 13e NMR spectrum of
47 showed 214% and 207% increase in signal intensities ofNeH3 and OCH3, respectively. The incorporation of intact 13 e- 1 ~ bond was detected by a modified
GHNMBe experiment (in toluene).
K. [ 13C-Methylj-L-methionine. [13e-Methyl]-L-methionine (60 mg) was supplied to 2 x 1-L cultures on days 3, 6, and 8, and all both cultures were harvested on day 10. A total of 17.9 mg of labeled 47 was isolated from the crude organic extract.
The 13e NMR spectrum of47 (in DMSO-d6) showed 124% and 136% increase in signal
13 intensities ofNeHJ and OCH3, respectively. However, a slightly higher amount of[ e methy/]-L-methionine (90 mg to 2 x 1 L) was found to be toxic to the organism. Table V.2. Calculation of4(S)-L-[5-13C]-Leucine Incorporation into Barbamide ( 47).
A B c D E F G H I J K L M N 0 R s T 1 Barbemldo from 4$)-L-[S"C)Ieuclno
Total Total Normell:zotlor Normellatlon Normellzatlon Normalization Normalization Normalization Normellatlon Normalization Normalization Normal izatlon Normalization Normell:zotlon Normoll211tlor 613C NMR 613CNMR Integral of Integral of Factor Factor Factor Factor Factor Factor Factor Factor Factor Factor Factor Factor FIICtor Major Minor Natural Enriched Compared to Compared to C Compared to C Compared to C Compared to C Compared to C Compared to C Compared to C Compared to C Compared to c Compared to C Compared to C Compared to C z Carbon tt Conformer Conformer Abundance Babemldo C-1 2 3 4 5 6 7 8 9 10 13 14 1& _I_ 1 15.22 3.5 4.2 1.0 0.9 0.9 2.7 1.0 2.0 1.5 0.9 2.3 1.8 0.9 0.9 1.0 ~ 2 52.02 4.0 3.5 1.1 1.0 1.1 3.1 1.2 2.3 1.7 1.0 2.6 2.1 1.1 1.1 1.1 & 3 33.98 33.73 3.7 4.1 1.1 0.9 1.0 2.9 1.1 2.2 1.6 0.9 2.5 1.9 1.0 1.0 1.1 ...L 4 166.80 166.31 1.3 1.8 0.4 0.3 0.3 1.0 0.4 0.8 0.6 0.3 0.9 0.7 0.3 0.3 0.4 .L 5 94.00 93.59 3.4 3.1 1.0 0.9 0.9 2.7 1.0 2.0 1.5 0.8 2.3 1.8 0.9 0.9 1.0 ...L 6 167.04 167.11 1.7 1.6 0.& 0.4 0.5 1.3 0.5 1.0 0.7 0.4 1.1 0.9 0.5 0.5 o.s II 7 64.28 2.3 2.2 0.7 0.6 0.6 1.8 0.7 1.4 1.0 0.6 1.5 1.2 0.6 0.6 0.7 l.i. 8 3&.82 36.47 4.1 3.4 1.2 1.0 1.1 3.2 1.2 2.4 1.8 1.0 2.7 2.1 1.1 1.1 1.2 [1_ 9 105.67 1.5 7.7 0.4 0.4 0.4 1.2 0.4 0.9 0.7 0.4 1.0 0.8 0.4 0.4 0.4 12 10 137.72 1.9 1.8 0.6 0.5 0.5 1.5 0.6 1.1 0.8 0.5 1.3 1.0 0.5 0.5 0.5 13 11 128.82 129.12 3.5 3.1 1.0 0.9 0.9 2.7 1.0 2.1 1.5 0.9 2.3 1.8 0.9 0.9 1.0 M 12 128.13 128.23 3.7 3.3 1.1 0.9 1.0 2.9 1.1 2.2 1.6 0.9 2.5 1.9 1.0 1.0 1.1 _n 13 126.32 126.58 3.7 3.5 1.1 0.9 1.0 2.9 1.1 2.2 1.6 0.9 2.5 1.9 1.0 1.0 1.1 HI 14 128.13 3.7 3.3 1.1 1.1 1.0 2.9 1.1 2.2 1.6 0.9 2.5 1.9 1.0 1.0 1.1 IZ. 15 128.82 3.5 3.1 1.0 1.0 0.9 2.7 1.0 2.1 1.5 0.9 2.3 1.8 0.9 0.9 1.0 ,. 16 169.42 170.02 1.5 1.5 0.4 0.4 0.4 1.1 0.4 0.9 0.6 0.4 1.0 0.8 0.4 0.4 0.4 111 17 141.99 142.31 3.2 2.8 0.9 0.8 0.9 2.5 0.9 1.9 1.4 0.8 2.1 1.7 0.9 0.9 0.9 ..!9 18 120.70 121.43 3.3 3.2 0.9 0.8 0.9 2.6 1.0 1.9 1.4 0.8 2.2 1.7 0.9 0.9 0.9 _2.]_ NCH3 28.26 2.8 2.3 0.8 0.7 0.7 2.2 0.8 1.6 1.2 0.7 1.9 1.4 0.8 0.7 0.8 22 DCH3 55.48 55.37 3.7 s.s 1.1 0.9 1.0 2.9 1.1 2.2 1.6 0.9 2.5 1.9 1.0 1.0 1.1 23
8- Table V.2. Calculation of 4(S)-L-[5-13C]-Leucine Incorporation into Barbamide (47) (continued).
u v w X y z M All AC AD AE AF AG AH AI AJ N{ I Normal!ad Normalim Normollad Normal lad Normallad Normelim:t lnteanotlon lntegnotlon Integration Integration lnt..gration lntegr.tion V.lueof V.luo of V.lueof V.lue ol Value of Vtllue af Normoll:ratlon Normall:zatlon Normall:ratlon Normall:ratlon Normellzatl"' Enriched Enriched Enriched Enriched Enriched Enrlehtd Factor Foetor Foetor Factor Foetor Bobamlde "Enrlchman Bebomldt "Enrichment Bobamldt " Enrlchmen Bebomlda "Enrichment . Bobamldt "Enrlohment Babomldt "Enrlchmon ~redtoC Compared to C ~redtoC ~red to Compor 2.6 1.2 1.1 1.3 1.0 4.5 90.4 3.3 126.3 4.1 100.0 5.2 78.4 3.4 .. 121.5 3.4 120.0 153.2 0.9 0.4 0.4 0.5 0.3 1.6 115.3 1.1 161.1 1.4 =127.6 l.& 100.0 1.2 155.0 1.2 2.3 1.1 1.0 1.2 0.9 4.2 74.4 3.0 103.9 3.8 82.3 4.8 64.5 3.1 100.0 3.1 9!.8 1.2 0,5 0.5 0.6 0.5 2.1 75.3 u 105.2 1.9 83.3 2.4 55,3 1.5 101.2 1.6 100.0 1.6 0.7 0.8 0.6 2.8 77.8 i 2.0 108.6 2.5 86.0 3.2 67.4 2.1 104.5 2. 103.3 2.8 1.3 1.2 1.S 1.1 s.o 69.4 i 3.5 97.0 4.5 76.8 5.7 50.2 93.3 3.7 92.2 1.0 o.s 0.5 0.5 0.4 1.8 421.7 1.3 589.2 1.7 465.6 2.1 365.6 1.4 555.9 1.4 550.1 1.3 0.6 0.5 0.7 0.5 2.3 1.7 105.4 2.1 84.3 2.7 56.0 1.7 102.4 1.8 101.2 2.4 1.1 1.1 1.3 0.9 4.3 3.1 102.0 3.9 80.8 5.0 53.3 3.2 98.1 3.2 97.0 2.5 1.2 1.1 1.3 1.0 ! 4.5 3.3 101.9 4.1 80.7 5.2 63.3 3.4 98.1 3.4 9&.9 2.& 1.2 1.1 1.3 1.0 i 4.5 s 3.3 108.2 4.1 5.2 67.1 ~r3.4 104.1 3.4 102.9 2.& 1.2 1.1 1.3 1.0 4.5 it73.0 3.7. 89.2 4.1 80.7 5.2 96.9 2.4 1.1 1.1 1.3 0.9 4.3 73.0 3.5 89.3 3.9 80.8 5.0 63.3 98.1 3.2 97.0 1.0 o.s 0.4 o.s 0.4 1.8 85.1 1.5 104.1 1.6 94.2 2.1 73.8 1.3 113.0 ll• 2.2 1.0 1.0 1.1 0.9 3.9 72.7 2.8 101.6 3.5 80.4 4.5 63.0 2.9 97.7 2.9 9&.5 0 2.3 1.0 1.0 1.2 0.9 4.0 80.5 2.9 89.0 4.6 69.8 3.0 1 08.1 3.0 106.9 11 Z..l =1==f5 1.9 0.9 0.8 1.0 0.7 3.4 67.6 2.4 94.4 1 74.8 3.9 58.6 2.5 90.8 2.6 u.a 2.5 1.2 1.1 1.3 1.0 -4.5 122.1 3.2 170.5 1 135.1 5.2 1-8"'B1f U Ml U 162.1 - -0 Table V.2. Calculation of4(S)-L-[5- 13C]-Leucine Incorporation into Barbamide (47) (continued). AI. AM AN AO AP AQ AR AS AT AU AV AW ,;x AY Ill BA BB 1 Normalized Normalized Normalized Normalized Normalized Normalized Normalized Normalized Norrrwllad Integration Integration Integration Integration Integration Integration Integration Integration Integration Value at Value Of Value of Value Of Value Of Value of Value Of Value of Value of Enriched Enriched Enriched Enriched Enriched Enriched Enriched Enriched Enriched Bebamlde "Enrlchmen Babamlde "Enrichmen Bebamlde " Enr lchmanl Bobomlde "Enrlchmen Bobamlde " Enr lchment Babamide "Enrlchmen Babamide "Enrlchmen Bobamlde "Enrlchmen Babamldl Compared to Compared to Compared to Compared to Compared to Compared to Compared to Compared to Comperedto Compared to Compared to Compared to Compared to Compared to Compared to Compared to Compared to 2 C-7 C-7 C-8 C-8 C-9 C-9 C-10 C-10 C-11 C-11 C-12 C-12 C-13 C-13 C-14 C-14 C-16 3 3.3 128.6 2.9 144.0 17.9 23.7 3.2 131.3 3.1 137.0 3.1 137.0 3.3 129.1 3.1 137.0 3.1 4 3.8 92.1 3.4 103.1 20.4 17.0 3.7 94.0 3.6 98.0 3.5 98.1 3.8 92.4 3.5 98.1 3.6 6 3.6 116.2 3.2 130.2 19.2 21.4 3.6 118.6 3.3 123.8 3.3 123.9 3.5 116.7 3.3 123.9 3.3 6 1.2 148.3 1.1 166.1 6.6 27.3 1.2 161.4 1.1 168.0 1.1 158.1 1.2 148.9 1.1 168.1 1.1 7 3.2 96.7 2.9 107.1 17.6 17.6 3.2 97.7 3.0 101.9 3.0 102.0 3.2 96.1 3.0 102.0 3.0 _a 1.6 96.8 1.4 108.4 8.8 17.9 1.6 98.8 1.5 103.1 1.6 103.2 1.6 97.2 1.5 103.2 1.6 9 2.2 100.0 2.0 112.0 11.9 18.4 2.1 102.1 2.1 106.6 2.1 106.6 2.2 100.4 2.1 106.6 2.1 10 3.9 89.3 3.4 100.0 20.9 16.6 3.8 91.2 3.6 96.1 3.6 95.2 3.8 89.7 3.6 96.2 3.6 Jj 1.4 542.4 1.3 607.3 7.7 100.0 1.4 653.6 1.3 577.6 1.3 578.0 1.4 544.5 1.3 578.0 1.3 12 1.8 98.0 1.6 109.7 9.9 18.1 1.8 100.0 1.7 104.3 1.7 104.4 1.8 98.4 1.7 104.4 1.7 1.1 3.3 93.9 3.0 105.2 18.1 17.3 3.3 96.9 3.1 100.0 3.1 100.1 3.3 94.3 3.1 100.1 3.1 4 3.5 93.8 3.2 106.1 19.2 17.3 3.5 95.8 3.3 99.9 3.3 100.0 3.5 94.2 3.3 100.0 3.3 15 3.5 99.6 3.2 111.6 19.2 18.4 3.5 101.7 3.3 106.1 3.3 106.1 3.5 100.0 3.3 106.1 3.3 16 3.6 93.8 3.2 106.1 19.2 17.3 3.6 96.8 3.3 99.9 3.3 100.0 3.6 94.2 3.3 100.0 3.3 3.3 93.9 3.0 106.2 18.1 17.3 3.3 96.9 3.1 100.0 3.1 100.1 3.3 94.3 3.1 100.1 3.1 18 1.4 109.6 1.2 122.6 7.6 20.2 1.4 111.7 1.3 116.6 1.3 116.7 1.4 109.9 1.3 116.7 1.3 19 3.0 93.6 2.7 104.7 16.4 17.2 3.0 96.4 2.8 99.6 2.8 99.6 3.0 93.9 2.8 99.6 2.8 20 3.1 103.6 2.8 116.9 17.0 19.1 3.1 105.6 2.9 110.2 2.9 110.3 3.1 103.9 2.9 110.3 2.9 21 2.6 86.9 2.4 97.3 14.3 16.0 2.6 88.7 2.5 92.6 2.4 94.7 2.8 82.6 2.6 92.6 2.6 22 3.5 157.0 3.2 175.8 19.1 28.9 3.6 160.2 3.3 167.2 3.3 167.3 3.5 157.6 3.3 167.3 3.3 23 - s Table V.2. Calculation of4(S)-L-[5-13C]-Leucine Incorporation into Barbamide ( 47) (continued). BC BD BE BF BG BH Bl BJ BK BL BM BN 1 Nomwll""' Nor,.ll""' Normell""' Normoll""' Normoll""' Integration Integration Integration Integration Integration Value of Value of Value of Value of Value of Enriched Enrlchocl Enrichocl Enriched Enriched "Enrlch...n Babamldo "Enrlchmont Babamldo "Enrloh..- Bobomldo "Enrlchmont Bobotnldo "Enrlohmon Babamlde "Enrlchmeni Awrago" Compared to Compared to Compared to Compared to Compared to Compared to Compared to Compared to Compared to C-redto C-redto Relative _2_ C-15 C-16 C-16 C-17 C-17 C-18 C-18 NCH, NCH, OCH, OCH, Enrichment 3 137.0 3.6 117.5 3.1 137.6 3.4 124.3 2.9 148.0 5.2 81.9 126.0 4 98.0 4.1 84.1 3.5 98.5 3.9 89.0 3.3 105.9 5.9 58.6 90.2 s 123.8 3.9 106.2 3.3 124.3 3.7 112.3 3.1 133.7 5.5 74.0 113.9 6 158.0 1.3 135.5 1.1 158.7 1.3 143.4 1.1 170.7 1.9 94.5 145.4 7 101.9 3.5 87.4 3.0 102.3 3.4 92.5 2.8 110.1 5.1 61.0 93.8 a 103.1 1.8 88.5 1.5 103.6 1.7 93.6 1.4 111.4 2.5 61.7 94.9 g 106.5 2.4 91.4 2.0 107.0 2.3 96.7 1.9 3.4 63.7 98.0 10 95.1 4.2 81.6 3.6 95.5 4.0 $6.3 3.3 7 6.0 56.9 87.5 1.1_ 577.6 1.6 495.5 1.3 580.2 1.S 524.2 1.2 ~11.0 2.2 345.5 531.5 12 104.3 2.0 89.5 1.7 104.8 1.9 94.7 1.6 112.7 2.9 62.4 96.0 3 100.0 3.7 35.8 3.1 100.4 3.5 90.8 2.9 108.0 5.3 59.8 92.0 14 99.9 3.9 85.7 3.3 100.4 3.7 90.7 3.1 108.0 s.s 59.8 92.0 15 106.1 3.9 91.0 3.3 106.5 3.7 96.3 3.1 114.6 s.s 63.5 97.6 16 99.9 3.9 85.7 3.3 100.4 3.7 90.7 3.1 108.0 5.5 59.8 91.3 17 100.0 3.7 $5.8 3.1 100.4 3.5 90.8 2.9 108.0 5.3 59.8~ 13 118.6 1.5 100.0 1.3 117.1 1.4 105.$ 1.2 125.9 2.2 69.7 106.5 19 99.6 3.3 85.4 2.8 100.0 3.1 90.4 2.6 107.6 4.7 69.6 91.6 20 110.2 3.4 94.5 2.9 110.7 3.2 100.0 2.7 119.0 4.9 65.9 101.4 21 92.6 2.9 79.4 2.5 93.0 2.7 84.0 2.3 100.0 4.1 55.4 85.0 22 167.2 3,9 143.4 3.3 167.9 3.7 151.7 3.1 180,6 s.s 100.0 1S3.S 23 -@ 104 CHAPTER VI THE BIOSYNTHESIS OF CURACIN A Abstract The biosynthesis ofcuracin A (15) was studied with the cultured marine cyanobacterium Lyngbya majuscula using stable isotope tracer techniques and high-field NMR spectroscopy. The results have demonstrated that curacin A has a polyketide origin with cysteine (or thiazoline-acyl CoA serving) as the starter unit. The thiazoline ring (except for C 18) and C3 of15 was shown to arise from cysteine through [I)3C]glycine and [2-13C,15N]glycine feeding experiments. Acetate feeding experiments showed that C4 through C 16 arise from seven contiguous acetate units followed by a proposed decarboxylation ofthe terminal carboxyl group. Incorporations oflabeled acetate into C 18 through C22 revealed a labeling pattern that is consistent with the five-carbon unit resulting from a mevalonic acid or a branched triketide-derived precursor. The methoxy carbon and C 17 were shown to originate from the methyl group ofmethionine. The post assembly polyketide chain processing is proposed based on information obtained from the en3, 1-13C]- and en3, 2-13C]acetate feeding experiments. 105 Introduction In 1994, our research group reported the discovery ofa potent cytotoxic agent, curacin A (15), from a Cura~ao collection ofthe marine cyanobacterium Lyngbya majuscula. The planar structure of 15 was elucidated using spectroscopic data,38 while its absolute configuration was determined by chemical degradation.113 A collaborative study with the National Cancer Institute (NCI) has shown that pure 15 is an antimitotic agent (ICso 6.8 ng/mL in the Chinese hamster Aux B 1 cell line) in that inhibits microtubule assembly by binding to the colchicine binding site. 17 H3 3 9 Curacin A (15) Examination ofthe structure of 15 immediately suggested a polyketide-derived alkene chain. However, we proposed three alternatives for assembling this polyketide chain (Figure VI.1 ). The first possibility (Figure VI.1 A) involves a cysteine (or a thiazoline-acyl CoA)-initiated linear heptaketide that exclusively derived from acetate units with a branched methyl group ( C 17) arising from S-adenosylmethionine (SAM). The second alternative (Figure VI.l B) also involves a cysteine (or a thiazoline acyl CoA)-initiated heptaketide but with a branched methyl group (C17) originating from C3 ofa propionate unit via methylmalonyl CoA. The third possibility (Figure VI.1 C) involves a preformed heptaketide chain that undergoes a condensation with cysteine (or a thiazoline-acyl CoA), the branched methyl group deriving from SAM. A. Direction of Polyketide Chain Assembly A-rA JJ..~H3SAM n ~ ~Hl """"' B. Direction of Polyketide Chain Assembly _Jl. =intact acetate =intact propionate ~ -~ = methyl carbon from Direction of Polyketide Chain Assembly JJ.. c. /,,. [methyi-13C]methionine SAM JJ..CHa n ~ .,.,.~H_i Figure VI.l. Three Alternatives for the Polyketide Chain Assembly Leading to Curacin A (15). 107 Formation ofthe thiazoline ring of 15 was proposed to involve the heterocyclization ofcysteine with a carbon from a hydrophobic amino acid such as leucine or valine, which also undergoes intramolecular cyclization to form the cyclopropyl ring (Figure VI.2 A). Alternatively, the methylene carbon (C20) ofthe cyclopropyl ring could originate from the electrophilic addition of a methyl group to an olefin of a diketide mediated by SAM (Figure VI.2 B). A third possibility exists whereby the cyclopropyl ring is the result of a five-carbon unit that fo.rms via an aldol condensation between an acyl SEnz and a carbonyl at C3 ofa diketide chain to form HMG CoA, and followed by decarboxylation and loss ofthe hydroxyl group (Figure VI.2 C). This type ofmechanism has also been proposed for the formation ofa methyl 114 115 116 branch m. the b.tosyn th ests . o f vrrgmtamycm, . . . . myxovrrecm,• • and oncor h yncoI'd 1 e. Finally, the cyclopropyl ring could result from the cyclization ofdimethylallyl diphosphate (DMAPP)/isopentyl diphosphate (IPP) which has been biosynthesized either through the classical mevalonate pathway or the non-mevalonate pathway (Figure VI.2 D). Lastly, the methoxy carbon of15 is likely to arise from methionine via SAM. In order to test the viability ofthese hypotheses, cultures ofL. majuscula (strain 19L) were fed with various 2H- and 13C-labeled precursors. This chapter describes the results from these feeding experiments. 108 A. ~ 1 H2 CHa / ~CHa ~~l18 Valine 20 'HOO~Ha NH2 CHa Leucine B. (+/ ~1-\ CHa-s, N~218 <=Enz~ 0 20 c. ~1-\ ·y-y <== CoA~" 18 N~ 64 OH~~o- 20 HMG-CoA D. ~1-\ N~18 pp~ <==H~ 20 lsopentyl diphosphate Mevalonic acid ~ ppn ~H JH ~YaH 2-C-methyi-D-erythritol-4-phosphate Figure VI.2. Possible Biosynthetic Precursors to the C18-C22 Five-Carbon Fragment of Curacin A (15). 109 Results and Discussion Isolation of Curacin A (15). The feeding experiments were conducted on the same strain ofL. majuscu/a (19L) that produces barbamide (47). The cultivation conditions and strain selection ofthe cyanobacterium were previously described. 98 For the precursor-incorporation feeding experiments, 50-75 mL ofwet-packed cells ofL. majuscu/a were inoculated into fresh medium (SWBG 11 ). After three days of equilibration, isotope-labeled precursors were administered to the 19L cultures and harvested on day 9 or 10. The cyanobacterium was blotted dry and repetitively extracted with 2:1 CH2Ch. The crude organic extract ofthe cultures was subjected to normal phase vacuum liquid chromatography (VLC) and NP-HPLC (4% EtOAclhexanes) to provide the labeled curacin A (15). Radioisotope Feeding Experiments. To explore the possible precursors to 15 and to establish a feeding protocol, James V. Rossi performed a series ofradioisotope feeding experiments.99 Table VI. I lists the experimental results from his radioisotope feedings. The high level of [1-14C]acetate incorporation is supportive ofthe hypothesis that the lipid chain of15 is polyketide in origin. Both e5S]-L-cysteine and [methy/- 14C] L-methionine were also taken up by the cultures at low yet significant levels. Table VI.1. Incorporation of 3H-, 14C-Labeled Precursors into Curacin A (15).99 Precursors Amount fed J.t.Ci % Incorporation in 15 Sodium [1-14C]acetate 66.20 4.05 [U-14C]-L-Valine 14.39 0.95 [U-14C]-L-Leucine 17.42 1.86 [U-14C]-L-Leucine + 5.30 2.20 *[3,4,5-3H]-L-Leucine 6.70 0.31 *[methyl-14C]-L-Methionine 14.95 0.11 35 [ S]-L-cysteine 16.80 0.13 *Values are average of each in which both 3H-labeled and 14C-labeled leucines were fed to the 19L cultures. 110 As illustrated in Figure VI.2 A, a cyclization ofeither valine or leucine was proposed in the formation ofthe cyclopropyl group of15. When [U-14e]-L-valine and [U- 14e]-L-leucine were supplied to the cultures, both amino acids were incorporated but a higher incorporation level was observed from the leucine feeding experiment. To determine whether leucine had undergone a transformation prior to its incorporation into 15, [3,4,5_3H]-L-leucine and [U-14e]-L-leucine were co-administered to the cultures. The radio labeled 15 resulting from this feeding experiment showed a high degree of discrimination favoring 14e, suggesting that more than one transformation is taking place before leucine is incorporated into 15. Leucine Feeding Experiments. The significant incorporation level of [U-14e]-L leucine into 15 suggested that the amino acid might serve as a precursor to the e l8-e22 carbon fragment. To prove this hypothesis, a [2- 13e]-L-leucine feeding experiment was performed.99 A careful analysis ofthe Be NMR spectrum of 15 produced during this feeding experiment revealed no Be enrichment at any carbon while a 452% increase in Be signal intensity for e4 ofbarbamide (47) was observed. Likewise, the Be NMR spectra produced during supplementation with 4(S)-[ 5-13e]-L-leucine and 4(R)-[ 5-13e]-L leucine also showed no incorporation ofleucine into 15. This finding suggested that although leucine is taken up by the organism, it is not a biosynthetic precursor to the cyclopropyl moiety (el8-e22) of 15, and that the incorporations from the radioisotope studies are likely a result ofthe metabolic breakdown ofthe amino acid. Sodium 3(R)-[4-13C]Methylbutanoate and 3(R)-[4-13C]-Methylbutanoate N Acetyl-Cysteamine (NAC) Feeding Experiments. Because we had demonstrated that leucine is not a biosynthetic precursor to the five-carbon unit (e18-e22) of 15, it became unlikely that its catabolic intermediate, 3(R)-[ 4-13e]methylbutanoate or its NAe-thioester would be incorporated into the molecule. However, as described in chapter V, these two 13e-labeled precursors were fed to the cultures to probe whether they are substrates for 111 the chlorination reaction in the barbamide biosynthetic pathway. As expected, analyses of the Be NMR spectra of 15 isolated from these two feeding experiments showed no enrichment ofany Be resonance. Acetate Feeding Experiments. As suggested by its structure, the lipid chain of 15 is likely to derive from a heptaketide and results from the radioisotope studies provided a preliminary support for this hypothesis. In order to clearly establish the labeling pattern ofthe acetate units and to distinguish among alternatives A, B and C shown in Figure VI.1, a [1,2-BC2]acetate feeding experiment was conducted.99 The Be NMR spectrum oflabeled 15 isolated from this initial feeding experiment showed the incorporation of acetate units into e4-e16 and C 18-C22. Unfortunately, a high level of inter-unit couplings of [ 1,2-Be2]acetate was observed. To minimize inter-unit coupling, sodium [1,2-Be2]acetate was co-administered with unlabeled sodium acetate to the 19L cultures. The Be NMR spectrum of 15 (Figure VI.3) produced under these conditions showed a lower level ofinter-unit coupling of [ 1,2-BC 2]acetate. Analysis ofthe coupling constants determined that intact acetate units were incorporated into carbon pairs C4-C5 (Jcc == 42.5 Hz}, C6-C7 (J == 44.0 Hz}, C8-C9 (J = 64.6 Hz}, C10-C11 (J= 42.9 Hz}, e12 Cl3 (J= 39.3 Hz}, C14-C15 (J= 42.6 Hz). Interestingly, intact acetate units were also incorporated into carbon pairs C18-C19 (J= 67.2 Hz) and C2l-C22 (J = 43.9 Hz) (Table VI.2). These results confirmed the polyketide origin ofthe lipid chain of15 and established the labeling pattern in the polyketide chain, which consistent with alternative A in Figure VI. I. In addition, the observed incorporation into e l8-C22 suggested that this five-carbon unit either arises from a branched triketide-derived precursor (64, Figure Vl.2 C) or DMAPPIIPP (Figure VI.2 D) To prove that CS, C7, C9, Cl1, Cl3 and CIS of 15 originated from Cl of acetate 13 18 and to probe the origin ofthe oxygen atom attached to C13, sodium [I- C, 0 2]acetate was fed to the 19L cultures. The 13C NMR spectrum oflabeled 15 showed the 112 C-20 C-22 C-17 C-6 C-1 C-14 C-11 C-5 \ C-21 I C-12 C-19 .I It ll ill ll I I I I I I 40 35 30 25 20 15 ppm OMe C-13 C-2 I I I I I I I I I I I I I I I I I 80 78 76 74 72 70 68 66 64 62 60 ppm C-16 C-9 C-15 C-18 .1. ll 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 165 160 155 150 145 140 135 130 125 ppm Figure VI.3. The 13C NMR Spectrum ofCuracin A (15) Produced During Supplementation with Sodium [ 1,2- 13C2]Acetate. 113 enrichments in the expected resonances (Table VI.2) including the isotope-induced upfield shift ofC13 from natural abundance by 0.02 ppm (Figure VI.4). 109 These results established the orientation ofthe acetate units as in possibility A in Figure VI. I, and that the oxygen attached to C 13 derives directly from acetate. In addition, the observed enhancements ofboth C18 and C21 signals suggested that ifDMAPPIIPP is the precursor to C 18-C22 of15, it is likely to formed via the classical mevalonate pathway since the enrichment is expected only either at C20 or C22 ifDMAPPIIPP is formed via non 117 mevalonate pathway (Figure VI.5). It is also worth mentioning that there is a significant level of 13C-13C couplings resulting from the random Be incorporation into all ofthe resonances (except for OCH3) 13 18 of 15 isolated from the [ l- C, 0 2]acetate feeding experiment. This observation can be 13 18 explained as follows. When [ 1- C, 0 2]acetate is supplied to the cultures , a portion of the labeled precursor will enter the tricarboxylic cycle (TCA cycle, Figure VI.6) where its Cl will be decarboxylated. The released 13C02 will in turn be incorporated back in the form ofsugars in the photosynthetic process, which are in turn metabolized and converted into various primary precursors. Although it is recognized that the interrupted TCA cycle, which lacks both succinyl CoA synthase and a.-ketoglutarate dehydrogenase, exist in some cyanobacteria (Figure VI. 7),118 no studies have shown that this also applies to L. majuscula. 13 18 To complement the ( l- C, 0 2]acetate feeding experiment, sodium [2 13C]acetate was supplied to the 19L cultures. Analysis ofthe Be NMR spectrum of15 produced under these conditions showed enrichments at the expected carbons [C4, C6, ClO, Cl2, C14, C16, C19, C20 and C22 (Figure VI.8, TableVI.2). The observed 13C)3C coupling ctJcc =12Hz) between C19 and C20 further substantiates the hypothesis that C 18-C22 five-carbon unit was derived from either precursor 64 (Figure VI.2 C) or DMAPIIPP (Figure VI.2 D). ~a=o.o2 C-4 80.0 ppm C-7 C-21 C-3 C-t C-17 \ C-11 C-14 C-5 C-15 C-6 C-1\ C-:!0 C-13 OMe C-1 C-16 19\ C-22: C-2 C-12 C-10 C-18 \ .I .f~/----~--~~~- ''1'''''''''1'''''''''1'''''''''1'''''''''1'''''''''1''''''//•····1''''1''''1''''1''''1''''1''''1''''1''''1''''1''''1''''1''''1''''1''''1 160 150 140 130 ppm 80 75 70 65 60 55 50 45 40 35 30 25 20 ppm Figure VI.4. The 13C NMR Spectrum of Curacin A (15) Produced During Supplementation with Sodium [ 1-13C, 1802]Acetate. -+:>. HJ-OH c ~OPP ~A HJ.ncHa 0 / B H~H H).ycH, ""~H3C OP e~OPP 0 0 OP OH ~OP/ ' OH Figure VI.5. Incorporation of[l-13C]Acetate into Isoprenoid: (A) Glyoxylate and Tricarboxylic Acid Cycles. (B) Glyceraldehyde-3 Phosphate/Pyruvate Pathway. (C) Classical Mevalonate Pathway. 117 -VI H~SCoA Acetyl CoA - 112• 1/2. \. 00 • 112e Y'coo· ~ ·ooc~coo· 0 HO coo- \ 1/a Oxaloacetate ( Citrate ~ 1/a • 112e •00~112• ·ooc coo· l coo- 112•coo· OH L·malate lsocitrate I 112e t -C~ • g 112e ,~J?o· ·ooc~coo· ·ooc Fumarate a-Ketoglutarate • !;> ~12. ~ ·ooc·~-...... '-..../'"'~ - --- ·ooc~SCoA -C02 Succinate Succinyl CoA Figure VI.6. Tricarboxylic Acid Cycle (TCA Cycle). ...----- H~coo· Pyruvate co l 0 HgCASCoA Acetyl CoA 00 ·ooycoo· ---- Ycoo· --~--=---- ·ooc~coo· NH2 0 HO coo· \ Aspartic acid Oxaloacetate Citrate "' ?H ·ooc~coo· ! coo· lsocitrate -ooc~coo· !-C02 Succinate 0 ·oo~coo· a-Ketoglutarate Figure VI.7. Schematic Representation ofthe Interrupted TCA Cycle in Some Cyanobacteria.118 1 1Jcc =12Hz Jcc= 12Hz C-19 C-20 C-7 C-3 . C-15\ C-1& C-13 OMe C-10 '\ C-2 \ Figure VI.8. The 13C NMR Spectrum ofCuracin A (15) Produced During Supplementation of Sodium (2- 13CJAcetate. -00 Table VI.2. 13C NMR Data ofCuracin A (15) Derived from Sodium [1-13C,1802]-, [2-13C]-, and [1,2-13C2]Acetate. r1,2-13C2]Acetate [1-13C, 11102]Acetate [2- 13C]Acetate #C Be Jcc(Hz) %Relative Isotopic %Relative %Relative Incorporationa Shift(ppm) Incorporationc Incorporationc 1 40.35 107.7 104.4 2 74.73 93.6 93.2 3 131.71 b 114.5 100.3 4 131.25 42.0 b 97.4 187.5 5 28.54 42.0 129.9 163.6 99.7 6 33.52 43.7 129.6 99.3 175.3 7 131.75 43.7 b 146.0 103.0 8 128.28 b b b 9 125.92 57.0 126.7 174.2 108.3 10 136.83 43.0 125.6 83.0 214.7 11 36.18 43.0 126.9 164.9 104.9 12 32.56 39.5 131.0 108.7 183.0 13 80.33 39.5 127.5 0.02 176.0 111.2 14 38.43 42.4 132.4 110.6 182.7 15 135.73 42.4 127.1 157.5 102.3 16 117.18 94.4 190.4 17 16.97 92.7 101.9 18 168.80 67.0 117.7 163.3 108.9 19 20.51 67.0 124.3 101.9 179.8 20 14.61 14.4 141.8 111.9 188.0 21 16.37 44.2 127.2 181.6 104.7 22 12.72 44.2 134.5 96.9 199.6 OMe 57.70 95.2 100.9 All u,C NMR spectra were recorded on a Broker AM400 spectrometer operatmg at 100.61 MHz., referenced to the centerline of 8 the solvent (CJ)6) at 128.39 ppm. % Relative incorporation= AlB, where A= total integrated NMR signal at that center and B =integrated NMR signal ofthe uncoupled resonance at that center. ~ot determined due to the overlapping ofNMR signals. ~elative incorporation = CID, where C = total integrated NMR signal at that center and D = normalized integrated NMR signal at that center when compared to C1, C2, C3, C17 and OCH3. -\0 120 The results from the acetate feeding experiments described above clearly support alternative A (Figure VI. I), which requires a decarboxylation ofthe terminal carbonyl group ofthe polyketide intermediate. This processing step, however, could take place on one ofthe two proposed polyketide intermediates, which differed in the degrees of reduction at C13 and C15 (Figure VI.9) and could be distinguished by observing the number ofacetate proton(s) retained at C14. Based on this rationale, sodium eH3, l 13C]acetate was fed to the 19L cultures. The 13C NMR spectrum oflabeled 15 produced 119 120 from this feeding experiment showed deuterium induced ~-isotope shift ' at C15 with a measured value (Ao = 0.078), which is consistent with C14 bearing two deuterium atoms. Therefore, this result supports alternative Bin Figure VI.9. In addition, the~ isotope shifts were also observed at C5, C9, C13 and C21 (Figure VI.10, Table VI.3), which are consistent with the conclusion that C4, C8 and C 12 each bears one deuterium atom and C22 bears two deuterium atoms. It is noteworthy that a value ofthe observed ~-isotope shift ofC5 (Ao = 0.1 07) was relatively high for the retention ofone deuterium atom at·C4. Although the possibility is unlikely, the large isotope shift value may result from the additions of ~-isotope effects from C4 and C6. Because C 16 and C20 of 15 derived from C2 ofacetate units from which the corresponding C 1 carbons were decarboxylated during the biosynthetic process, therefore the C 1 carbons cannot be used as reporter nuclei to determine the number ofdeuterium atom(s) retained at these centers. Moreover, the ~-isotope shift at C7 resonance could not be unambiguously measured and the number ofdeuterium atom(s) retained at C6 could not be established from the eH3, 1-13C]acetate feeding experiment because the chemical shift ofC7 is almost identical to that ofC3 . To examine the origin ofprotons attached to C6, C16 and C20 as well as to confirm the results from the eH3,1-13C]acetate feeding experiment, sodium eH3, 2-13C]acetate was administered to the 19L cultures. The proton and deuterium decoupled 13C NMR spectrum (Figure VI.ll) oflabeled 15 121 possessed deuterium induced a-isotopic shifts at C4, C6, Cl2, CI4, CI6, C20 and C22 (Table VI.3). 121 The measured a-isotope shifts ofthese 13C resonances indicated that C4, C6, Cl2 and C20 each retained single deuterium atom while C14, C16 and C22 each retained two deuterium atoms. In addition, this result also supports the alternative B of the decarboxylation process proposed in Figure VI.9. A. 0 0 }qo¥r~ 0 OCH3 B. Q)0S(,,__!q ~1)6R ~ ~I OCHa Figure VI.9. Two Hypotheses Concerning the Decarboxylation ofthe Polyketide Intermediate in the Biosynthetic Pathway of Curacin A (15). 122 Table VI.3. 2H Isotope-Induced Shifts Observed in the 13e NMR Spectra ofeuracin A (15) Derived from Sodium eH3, 1-13e]- and eH3,2-13e]Acetate. #e Sc rtH3, 1- 13e]Acetatea eH3, 2-13C]Acetateb as as (~-Isotope Shifts) (a.-Isotope Shifts) 1 40.35 2 74.73 3 131.71 4 131.25 0.351 5 28.54 0.107 6 33.52 0.368 7 131.75 8 128.28 9 125.92 0.055 10 136.83 11 36.18 12 32.56 0.385 13 80.33 0.055 14 38.43 0.736 15 135.73 0.078 16 117.18 0.546 17 16.97 18 168.80 19 20.51 20 14.61 0.295 21 16.37 0.138 22 12.72 0.558 OeH3 57.70 aThe 13e NMR spectrum was recorded on a Bruk.er DRX600 spectrometer operating at 150.90 MHz. bThe 2H-decoupled 13e NMR was recorded on a General Electric GN Omega500 spectrometer operating at 125.76 MHz. Both 13e NMR spectra were referenced to the centerline ofthe solvent at 128.39 ppm (e6D6). 123 [2-13C]-DL-Mevalonolactone Feeding Experiment. The results from the acetate feeding experiment suggested that C18-C22 could arise either from DMAPPIIPP or a triketide-derived precursor (64). To examine one ofthese possibilities, we supplied [2-13C]-DL-mevalonolactone to the cultures. Under careful analysis ofthe 13C NMR spectrum ofthe isolated 15 (Figure Vl.12), no enrichment ofany resonance was observed. This negative incorporation evidence did not by itself prove that mevalonic acid is not a precursor to 15. However, ifC18 through C22 are derived from acetate units via DMAPPIIPP then it might be reasonable to expect lower levels of incorporation of labeled acetates at these sites as compared to sites deriving directly from acetate (C4 Cl6). This results from the increased number ofbiosynthetic steps in transforming acetates into mevalonic acid prior its incorporation into 15. This speculation is based on an assumption that exogenously supplied acetate is not preferentially used by the cyanobacterium. The incorporation levels listed in Table VI.2 show that C 18-C22 are labeled with acetate to the same extent as the carbons on the polyketide chain (C4-C16). Taken together, these results suggest that 64 is likely to serve as a precursor to the Cl8 C22 five-carbon fragment rather than DMAPPIIPP. Glycine Feeding Experiments. The most probable biosynthetic precursor to the thiazoline ring of 15 (except for C18) is cysteine. Unfortunately, 13C-labeled cysteine is very expensive, metabolically labile and was found to be toxic to the cyanobacterium (when more than 20 mg/L ofDL-cysteine was supplied to the 19L cultures). Conceived as an alternative to feeding cysteine, serine, a metabolic precursor to cysteine, was fed to the cultures and was also found to be toxic to the organism (when more than 25 mg/L of L-serine was supplied to the 19L cultures). The next approach was to use glycine since it is a metabolic precursor to serine/cysteine and is likely to be better tolerated by the organism.121 As presented in 124 Figure VI.13, the conversion ofglycine to cysteine involves the loss ofCI to the C 1 pool while the bond between C2 and the nitrogen atom remains intact. Therefore, the observation ofintact incorporation ofthe 13C-15N bond would strongly support the cysteine origin ofthe thiazoline ring of 15. Based on this rationale, [2-13C,15N]glycine was administered to the 19L cultures. The 13C NMR spectrum of 15 (Figure VI.l4) produced from this feeding experiment revealed the expected upfield isotope-induced doublet eJCN = 2.6 Hz) flanking the C2 natural abundance signal. In addition to the intact 13C- 15N bond incorporation, a doublet ofdoublets was also observed flanking the C2 natural abundance signal of 15. The large 13C)3C coupling value eJcc = 29.9 Hz) ofthis doublet ofdoublet 1 resulted from the coupling ofC2 to Cl while the smaller coupling constant ( JCN = 2.6 Hz) resulted from the intact coupling ofC2 to 1 ~. The significant enrichment level at C I of15 (Table VI.4) is likely to result from the contribution ofC2 ofglycine to tetrahydrofolate (THF) to form~,N10-methylene-THF. This latter intermediate would shuttle its Ct carbon back to glycine in the conversion ofglycine to serine/cysteine. Similarly, N5methyl-THF also supplies the Ct carbon to homocysteine for its conversion to methionine. The donation ofthe methyl group ofmethionine enriched in 13C from glycine to C 17 and the OCH3 of15 via SAM were evident from the enhancements of their 13C resonances (Table VI.4) in the 13C NMR spectrum oflabeled 15. Therefore, this feeding experiment provides evidence that C I and C2 of 15 derive from cysteine while Cl7 and the OCH3 ofthe molecule originate from the Ct pool. To obtain a supporting evidence that C3 of15 arises from Cl ofcysteine, [1 13C]glycine was provided to the cultures. Examination ofthe 13C NMR spectrum of labeled 15 (Figure VI.15) produced from this feeding experiment showed the expected enhancement in the C3 resonance. Although the enrichment level (Table VI.4) did not 125 reach a desirable level (200%), these data are still sufficient to show that C3 of15 is likely to originate from Cl ofcysteine. [Methy/-13C]-L-Methionine Feeding Experiment. To demonstrate that C 17 and the OCH3 derived from the methyl group ofmethionine, the cyanobacterium was supplied with [methyl- 13C]-L-methionine. Analysis ofthe 13C NMR spectrum of15 (Figure VI.l6) produced from this supplementation showed significant enhancements of the C17 and OCH3 signals (Table V1.4), confirming that both carbons originated from methionine via SAM. 126 C-22 C-19 C-20 C-21 C-17 '''''''I'''''''''I'''''''''I''''''''' I''''''''' I'''''''''I'''''''''I'''''''''I'''''' 20 19 18 17 16 15 14 ppm C-1 C-14 C-~11 C-5 j~~-"'------U~'~..____l '"'"1'''''"''1"'''''''1'''''''"1''"''"'1'''''""1"'"''''1'"'""'1'''""''1"'""''1'''''"''1'''''''''1''''''' 40 39 38 37 36 35 34 33 32 31 30 ppm C-13 C-2 OMe 80 78 76 74 72 70 68 66 64 62 60 ppm C-3 C-7 C-4 C-9 ~, C-16 C-15 C-10 C-18 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 1~ 100 1~ 1~ 1~ 1® 1~ 1~ 1~ wm Figure VI.lO. The 13C NMR Sgectrum ofCuracin A (15) Produced During 13 Supplementation with Sodium [ H3,1- C]Acetate. 127 C-1 C-14 C-11 C-6 C-12 C-5 C-21 C-20 C-19 C-22 C-17 I I I 2S ppm C-2 C-13 OMe I I I 80 75 70 65 60 ppm C-15 C-10 I 135 125 120 ppm Figure VI.ll. The 1H- and 2H-Decoupled 13C NMR Spectrum ofCuracin A (15) Produced During Supplementation with Sodium eH3,2-13C]Acetate. A. II I. 'I".. II" I I I • ''I" I 'I''"II I I 'I I II II. I"I'I"I'. I I I I I II I .. I I I'I' I' I I I 'I' II .. ' H 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 II '' I 'I'I'''' I ' I I I I' I I I I I I I I II"II'' I' 160 150 140 130 120 110 100 90 80 70 60 50 40 30 ppm B. ... ~., ••• ,,,,''IIt' i IS I I,,. I ;'j I. j' I I' i I' t#i. I j I I''iII I 'J if,,' •• i 'II'';,,'''I I I I I.' II 'I''' I i it ill'';; i I'i''I ili¥1 IiI iIi i. it I I til II II t I I' i I'' 11 iiI I 'IiiIiI I I I 'IiI I I' I I I I 160 150 140 130 120 110 100 90 80 70 60 50 40 30 ppm Figure VI.12. Comparison ofthe 13C NMR Spectra of(A) Curacin A (15) Produced During Supplementation with [2- 13C]-DL Mevatonolactone, (B) Natural Abundance Curacin A (15). A. Cl B. Cl .... HY.cooH C02+NH4+ H2N"'-cooH Glycine ( \' Glycine THF tJI,N10-meflylene-THF .... ~5.-eH. J:H ~ j_ HY.cooH H2CI T COOH H2 COOH Serine Glycine Methionine Homocysteine j ~ JSH Hs H~ T COOH Cysteine ~ Hs OCH3 15 Figure VI.13. Metabolic Relationship of (A) Glycine and Serine/Cysteine, (B) Glycine and Methionine, and the Expected Labeling Pattern in Curacin A (15) Produced During Supplementation of [2-13C, 1 ~]Glycine. 130 Table VI.4. 13C NMR Data of Curacin A (15) Derived from [1- 13C]- and [2-13C, 1 ~Glycine and [methyl- 13C]-L-Methionine. 3 13 1 13 #C 5c [1_I C]Glycine [2- C, 'N]Glycine [Methyl- C] L-methionine % Relativea %Relativeb %Relativec Incorporation Incorporation Incorporation 1 40.35 100.0 253.4 98.2 2 74.73 100.7 199.9 98.9 3 131.71 132.7 100.0 103.1 4 131.25 101.5 132.0 106.0 5 28.54 111.3 122.5 101.9 6 33.52 107.9 115.9 102.6 7 131.75 104.3 119.7 109.4 8 128.28 d d d 9 125.92 118.9 130.4 104.1 10 136.83 103.0 193.2 113.5 11 36.18 111.9 117.8 117.1 12 32.56 110.6 121.9 104.7 13 80.33 109.8 130.5 105.1 14 38.43 109.6 128.8 106.9 15 135.73 104.3 146.4 106.0 16 117.18 110.7 130.9 112.8 17 16.97 101.8 413.9 201.3 18 168.80 88.5 109.9 95.3 19 20.51 106.6 120.1 92.4 20 14.61 109.2 121.0 98.4 21 16.37 106.8 109.7 100.1 22 12.72 123.5 135.9 103.2 OCH3 57.70 97.6 309.7 187.8 ~elative incorporation = AlB, where A = total integrated NMR signal at that center and B =normalized integrated NMR signal at that center when compared to C1, C2, C17 and OCH3 of15. ~elative incorporation = A/C, where A= total integrated NMR signal at that center and C normalized integrated NMR signal at that center when compared to C3 of 15. Normalization to multiple carbon centers was not applicable to this feeding experiment because C2 glycine will be converted into C2 of pyruvate and subsequently into C1 of acetate via its catabolic pathway. ~elative incorporation= AlB, where A= total integrated NMR signal at that center and B = normalized integrated NMR signal at that center when compared to C 1, C2 and C3. ~ot determined due to the overlapping ofthe NMR signal with the solvent signal. All 13C NMR s8ectra were recorded on a Bruker AM400 spectrometer operating at 100.61 MHz. C NMR spectra were referenced to the centerline ofthe solvent (C6D6) at 128.39 ppm. 131 T 3 Hooc-""~~ [2- 13C,15NJGiycine A. B. 'Jcc =29.9 Hz C-2 C-2 75.0 74.8 74.6 ppm 75.0 74.8 74.6 ppm 1 Jcc =29.9 Hz c. D. C-1 C-1 40.6 40.4 40.2 ppm 40.6 40.4 40.2 ppm Figure VI.l4. Comparison ofSelected Regions ofthe 13C NMR Spectra of(A and C) Natural Abundance Curacin A (15), (B and D) Curacin A (15) Produced During Supplementation with [2-13C,1'N]Glycine. 132 17 Hs HOO~NH2 9 [1- 13C]Giycine C-7 C-7 C-3 \ I 132.0 ppm 132.0 ppm Figure VI.15. Comparison ofSelected Regions ofthe 13C NMR Spectra of(A) Natural Abundance Curacin A (15), (B) Curacin A (15) Produced During Supplementation with [ 1-13C]Glycine. C-21 C-20 1 I OMe "1 • C-17I I ~ J lTrlr 1 J I 55 50 45 40 35 30 25 20 15 ppm C-17 OMe C-11 C-14 C-12 C-5 C-21 C-6 C-1 C-20 C-19 C-22 I I I I 1 T I I I 55 50 45 40 35 30 25 20 15 ppm Figure VI.16. Comparison ofSelected Regions ofthe 13C NMR Spectra of(A) Natural Abundance Curacin A (15), (B) Curacin A (15) Produced During Supplementation with [methyl- 13C]-L-Methionine. . 134 Conclusion. The feeding experiments described above indicated that the lipid chain of 15 derives from a heptaketide chain that consists of seven acetate units (alternative A in Figure VI.l). The starter unit ofthe polyketide appears to be cysteine as indicated by the labeling pattern of [1-13C]-glycine and [2-13C, 15N]glycine into 15. The methoxy carbon and C 17 of 15 were clearly demonstrated to arise from the methyl group ofmethionine from the [methyl- 13C]-L-methionine feeding experiment. The information obtained from these feeding experiments together with the acetate incorporation experiments, particularly with the doubly labeled eH3, 1-13C]- and eH3,2 13C]acetate, provide insights into the structure ofthe putative PKS product (65, Figure VI.17). The first eight processing steps (enolization, reduction and dehydration) ofthe PKS product are likely to occur while the polyketide chain is bound to an acyl carrier protein (ACP). The decarboxylation ofthe carboxyl group is proposed to occur after the polyketide chain is released from the ACP. The biosynthetic precursor ofthe C18-C22 section of15 is revealed by the labeling pattern ofacetates into this five-carbon fragment and the negative incorporation of [2-13C]-DL-mevalonolactone. These results suggest that this five-carbon unit is likely to result from a branched triketide-derived precursor (64). The formation of64 is envisioned to involve a condensation between an acyl-SEnz group and the carbonyl at C3 ofa diketide followed by a decarboxylation and dehydration (Figure VI.2 C). This type ofmechanism has been previously encountered in the biosyntheses ofvirginiamycin, 114 116 13 myxovirecin and oncorhyncolide. - However, the negative incorporation of [2- C] DL-mevalonolactone could also result from the inability ofthe cyanobacterial cell to take up the precursor. Ifthis latter postulation is true and DMAPPIIPP is the biosynthetic 13 18 precursor to the C 18-C22 section, then the labeling pattern from the [ 1- C, 0 2]acetate feeding experiment (enrichments were observed at C 18 and C21 instead ofonly at C20 or C22) would suggest that the DMAPPIIPP is formed via the classical mevalonate 135 117 pathway . Although there is no evidence for use ofthe mevalonate pathway in cyanobacteria, no studies has yet been done in L. majuscula. Enolization EnzS SH EnzS SH 65 NADp+ NADPH + W ) EnzS SH .\. EnzS SH NH2 '( CH3 68 H~ 67 lEnolization Methylation EnzS SH Enz ?' SH NH2 69 70 rNADPH+H' NADPH+W NADP+ NADP+ SH x .., SH EnzSH H~ 72 ~O,+H,O 71 SH SH NH2 NH2 74 Figure VI.17. The Proposed Modification Steps of the Post-Assembly Polyketide Intermediate of Curacin A (15) based on the Information Obtained from Acetate Feeding Experiment. 137 Emeriment GeneraL Most ofthe NMR spectra were recorded on a Bruker AM400 spectrometer operating at 400.13 MHz for 1HNMR and at 100.61 MHz for 13C NMR. The NMR spectra of 15 isolated from the eH3, 1-13C]acetate feeding experiment were recorded on a Bruker DRX600 spectrometer operating at a proton frequency of600.01 MHz and a carbon frequency of 150.90 MHz. The NMR spectra of15 isolated from the eH3,2- 13C]acetate feeding experiment were recorded on a General Electric ON Omega500 spectrometer operating at 500.11 for the 1H NMR and at 125.76 MHz for 13C NMR. Chemical shifts are adjusted to the centerline ofsolvent at 128.39 ppm (C6D6). High-performance liquid chromatography (HPLC) utilized a Waters M6000A pump, Rhoedyne 7125 injector, and a Waters Lambda-Max 480 LC spectrophotometer. Merck aluminum-backed thin layer chromatography (TLC) sheets (silica gel60 F254) were used for TLC. Vacuum liquid chromatography (VLC) was performed with Merck Silica Gel G for TLC. The synthetically prepared 4(S)-L-[5- 13C]leucine, 4(R)-L-[5-13C]leucine, 3(R) [4-13C]butanoic acid, and 3(R)-[4-13C]methylbutanoyl N-acetylcysteamine (NAC) were generously provided by Professor Christine Willis (U. ofBristol). [2- 13C]-DL mevalonolactone was purchased from Isotec Inc. and all ofthe other stable isotope precursors were purchased from Cambridge Isotope Laboratories. General Culture Conditions and Isolation Procedure. Approximately 3 g ofL. majuscula strain 19L were inoculated into a 2.8-L Fembach flask containing 1 L of SWBG11 medium. The culture was grown at 28 ·c under uniform illumination (4.67 J.Lmol photon s·tm-2), aerated, and equilibrated for 3 days prior to addition of isotope labeled precursors. Cultures ofL. majuscula were harvested on day 9 or 10 after the inoculation, blotted dry, and extracted with 2:1 CH2Ch/MeOH. The extract was gravity filtered through glass wool. The filtered organic extract was dried in vacuo, weighed, and applied to a silica gel column (1.5 em I.D. x 15 em) in 5% EtOAc/hexanes, and 138 eluted with a stepped gradient elution of5% EtOAc to 100% EtOAc. A fraction containing curacin A (eluted with 5% EtOAclhexanes) was subjected to a final purification by NP-HPLC [Verapack Silica 10 Jl,X 4.1 mm x 30 mm, 4% EtOAc/hexanes, and UV detection at 254 nm, tR = 22 min] to give pure curacin A (15, 11% ofthe extractable lipid). Calculation ofthe Results of13C-Labeled Precursor Feeding Experiments on Curacin A. The following explains how the calculations for 13C incorporations were 13 18 performed. The method used with curacin A (47) produced from the [1- C, 0 2]acetate feeding experiment is presented as an example {Table VI.5). Column A shows all 23 carbons of15, while column B displays their 13C chemical shifts. Columns C and I show the integration values ofthe13C signals ofnatural abundance 15 and enriched 15, respectively. Next, the normalization factors in comparison to C1 (column D) were calculated by individually dividing the integration values ofC 1-C22, and OCH3 (cells C3-C24) by the integration value ofC1 (cell C3). The normalization factors compared to C2, C3, C17 and OCH3 are shown in columns E, F, G and H, respectively. Multiplication ofthe normalization factors (column D) with the integration value for the C1 resonance ofenriched 15 (cell 13) will provide the normalized integration values compared to C1 (column J). The normalization values compared to C2, C3, C17 and OCH3are shown in columns L, N, P and R, respectively. The percent ofthe 13C enhancements ofeach carbon ofenriched 15 in comparison to C 1 (column K) was calculated by dividing the integration values ofenriched 15 (column I) by the normalized integration values (column K), and multiplying by 100. The percent ofthe 13C enhancements ofeach carbon in comparison to C2, C3, C17 and OCH3 are shown in columns M, 0, Qand S, respectively. The average values of% relative enhancement are displayed in column T. 13 13 The C relative enrichments of15 from the [1,2- C2]acetate feeding experiment (Table VI.6) were calculated as follows. Column A shows all 23 carbons of 15, while 139 column B displays their Be chemical shifts . Column C shows the total integration value oftheBe signal for each resonance of15 from this feeding experiment. The integration value ofthe uncoupled Be signal for each resonance of 15 is shown in column D. The percentages ofthe relative enrichments from the intact incorporations appearing in column E were calculated by dividing the total integration value of 15 (column C) by the integration values ofthe uncoupled Be signals (column D), and multiplying by 100. Feeding Experiments. A. Sodium {JJ3C, 1801]Acetate Feeding Experiment. 18 Sodium [1-BC, 0 2]acetate (375 mg) was mixed with unlabelled sodium acetate (622 mg) and administered to 3 x 1-L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of8.4 mg oflabeled 15 was isolated from the crude organic extract. The Be NMR spectral data and % relative enrichment of 15 are summarized in Table VI.2. B. Sodium [2-13C]Acetate Feeding Experiment. Sodium [2-BC]acetate (187 mg) was mixed with unlabelled sodium acetate (311 mg) and fed to 3 x 1-L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of6.1 mg of labeled 15 was isolated from the crude organic extract. The Be NMR spectral data and % relative enrichment of15 are summarized in Table VI.2. 13 C. Sodium [1,2- C1]Acetate Feeding Experiment. Sodium [1,2-BC2]acetate (208 mg) was mixed with unlabelled sodium acetate (415 mg) and provided to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of 10.0 mg oflabeled 15 was isolated from the crude organic extract. The Be NMR spectral data and % relative enrichment of 15 are summarized in Table VI.2. 13 8 Table VI.5. Calculation of[l- C/ 0 2]Acetate Incorporation into Curacin A (15). A • c D E F G H I J K L ... N 0 p Q R s T -l:zocl Normall:td NormaiiJM No.rmai!Hd Normati'ZM lntograthln lm.gratlon tntetratlon lntf-Qratlon lnttgratum f v.~u..r Value Of Valu. of V•lut of Vllut of ...... , Normailza1:lon Normllll:utlon Honnallzatlor """""ll:utl., Enrlcn.d Enrlchtd Enrlchtd Eftfl<:htd .._,., Nltur!Jl Foetor Factor F..tor .. Factor Foetor Tott~llnt•gral curaetn A X lnrlctvntnt curaeln A SErirtmntnt curac:tnA %EnrletttMnt cuncln A ~ Enrichment CuratlnA X Enrlc:hrn.nt .,.,...... , <:ompar.. to --.,.,.... .,.,...... ,. .,.,...... Compar.d to C~redto Comp1rtd to Compredto compartd to Compar.clto Compa~ndto C~to .... <:ompamtol of EnriChed - to _I_ a"c Curtc:tn A 1 C-2 c-3 C-17 '"""::"-7' c-1 C-2 C-2 c-3 C-3 c-11 C-17 OCH3 JL I 40.35 -4.3 1.0 1.0 0.9 1.5 2.9 100.0 2.5 115.0 3.1 94.0 2.5 116.1 z.s 113.1 -1017 .A. 2 74.73 4.2 1.0 1.0 0.9 1.5 29 100.0 3.1 81.8 2.5 101.0 2.6 98.4 -93.& -· 87.0tl ..L 3 131.71 4.6 1.1 1.1 1.0 1.6 106.3 122.2 3.3 100.0 2.7 2.8 120.3 114.$ .. 4 131.25 3.9 0.9 0.9 o.8 1.4 1.21.6~ 90.4 2.3 104.0 2.8 85.0 2.3 2.3 17.4 5 28.54 4.4 1.0 1.0 1.0 1.5 1.4 152.0 2.6 174.7 3.2 142.9 2.6 ?? I tn• 1&3.1 I 33.52 4.6 1.1 1.1 1.0 1.6 1.4 2.8 3.1 92.2 2.7 106.0 3.3 116.7 2.6 $. 2.7 104.3 99.3 7 131.75 3,9 0.9 0.9 0.8 1.3 1.2 3.6 2.6 136.6 2.3 2.8 121.5 2.3 151.5 2.3 163.4 14.0 126.!12 4.0 0.9 0.9 0.9 1.4 1.3 4.4 2.7 1&1.8 2.4 I 186.0 2.9 152.2 2.4 187.9 2.4 183.0 174.2 10• 136.83 1.5 0.3 0.3 0.3 0.5 0.5 0.8 1.0 77.1 0.9 i 86.7 1.1 12.5 0.9 89.6 0.9 87.3 83.0 11 36.18 4.4 1.0 1.4 4.& 3.0 163.2 2.6 176.1 3.2 144.1 2.6 177.9 2.7 173.3 1M.9 12 32.56 4.3 0 0.9 1.4 3.0 2.9 101.0 I 2.5 116.1 3.1 95.0 2.5 117.3 2.6 114.3 1011.7 13 0.8 1.3 1.2 4.2 163.5 I 2.2 187,9~ 153.8 2.2 189.9 2.3 184.9 175.0 14 ~1.0 1.0 0.9 1.5 1.4 3.0 102.7 2.6 118. 1 96.6 2.5 i 119.3 2.6 11&.2 110.fl 15 0.7 0.7 0.6 1.0 0.9 2.9 146.3 1.7 168,1 1 137.6 1.7 i 169.9 1.8 11>5.5 157.6 _n 16 ~117.18 3.9 0.9 0.9 I o.8 1.3 1.2 2.3 ~ 87.7 2.3 100,8 2.8 82.5 2.3 99.2 94.4 17 16.97 2.9 0.7 1.0 0.9 1.7 2.0 86.1 1.7 99,0 2.1 tn.o 1.7 ~ 18 168.8 1.0 0.2 0.2 0.2 0.3 0.3 1.0 0.7 151.7 0.6 174.3 0.7 1-42,6 0.6 171.6 163.3 3.9 0.9 0,9 0.8 1.4 1.2 2.5 2.6 94.7 2.3 108.8 2.8 89.0 2.3 107.1 101.• 4.5 1.1 1.1 1.0 1.6 1.4 3.2 3.1 103.9 2.7 119,5 3.3 97.7 2.7 g 117.6 111.9 4.3 1.0 1.0 0.9 1.5 1.4 4.9 2.9 188.7 2.5 193.9 3,1 158.6 2.5 195.9 2:.& 190.8 181.8 3.3 0.8 0.8 0,7 1.2 1.1 2.0 2.3 90.0 2.0 103.4 2A "84.6 2.0 t04.S 101.8 96.1 - 7 3.1 0.7 0.1 0.7 1.1 1.0 1.9 2.1 88.4 1.ll 101.6 2.3 83.t 1.8 102.7 100.0 96.2 141 Table VI.6. Calculation of[I,2-13C2]Acetate Incorporation into Curacin A (15). A I B c 0 E 1 Curacin A from [ 1.2-13C2]Acetate Feeding Experiment Total Integration Integral of Value for %Relative Enriched Uncouple Enrichment from 2 Carbon# a"c Curacin A Signal Intact lncorporatio; 3 5 28.54 59.5 45.8 129.9 4 6 33.52 60.0 46.3 129.6 5 9 125.92 49.8 39.3 126.7 6 10 136.84 18.9 15.1 125.6 7 11 36.18 58.7 46.3 126.9 8 12 32.56 59.1 45.1 131.0 9 13 80.33 53.0 41.6 127.5 10 14 38.43 59.1 44.6 132.4 11 15 135.73 35.6 28.0 127.1 117.18 54.9 49.7 110.5 168.80 11.4 9.7 117.7 4 20.51 45.7 36.8 124.3 15 20 14.61 47.7 33.6 141.8 16 21 16.37 53.7 42.2 127.2 17 22 12.72 47.7 35.4 134.5 "18 2 13 D. Sodium {1-IJC, 2- H3]Acetate Feeding Experiment. Sodium [1- C, 2 2H3]acetate (187 mg) was mixed with unlabelled sodium acetate (311 mg) and provided to 3 x 1-L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of6.0 mg oflabeled 15 was isolated from the crude organic extract. The 13C spectral data and the measured ~-isotope induced shifts of15 are summarized in Table VI.3. 3 1 13 E. Sodium f2} C, 2- H3]Acetate Feeding Experiment. Sodium [2- C, 2 2H3]acetate (187 mg ) was mixed with unlabelled sodium acetate (311 mg) and supplied to 3 x 1-L ofcultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of3.9 mg oflabeled 15 was isolated from the crude organic extract. The 2H decoupled 13C NMR spectral data and the measured a-isotope induced shifts of15 are summarized in Table VI.3. F. [2}3C]-DL-Mevalonolactone Feeding Experiment. [2-13C]-DL mevalonolactone (100 mg) was provided to 3 x 1-L ofcultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of7.4 mg of15 was isolated from the 142 crude organic extract. The Be NMR spectrum of 15 from this feeding experiment showed no enhancement of any carbon. G. [Jf1C]Glycine Feeding Experiment. [1-13e]glycine (300 mg) was supplied to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of 10.8 mg of labeled 15 was isolated from the crude organic extract. The Be NMR spectrum of 15 from this feeding experiment showed enhancement ofsignal at o131.71 (e3, 132%) (Table Vl.4). H. [2-13C,15N]glycine Feeding Experiment. [2- 13 e, 1 ~glycine 225 mg) was supplied to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of 10.2 mg of labeled 15 was isolated from the crude organic extract. The 13e NMR spectrum of 15 isolated from this feeding experiment showed enhancement of signals at o16.97 (e17, s; 413%), 40.35 (el, d; 1Jcc =29.9 Hz, 253%), 57.70 (OCH3, s; 1 1 1 309%) and 74.73 (e2, d; JCN =2.6 Hz; and dd; Jcc =29.9 Hz and JCN =2.6 Hz; 199%) (Table VI.4). G. [methyl-13C]methionine Feeding Experiment. [methy/-13e]-L-methionine (60 mg) was fed to 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of 11.2 mg oflabeled 15 was isolated from the crude organic extract. The 13e NMR spectrum of15 isolated from this feeding experiment showed enhancement ofsignals at o16.97 (201%) and 57.7 (187%) (Table V1.4). H. L-[JJ3C]Leucine and L-[2-13C]Leucine. L-[1-13C]Leucine (180 mg) andL [2-13e]Ieucine (120 mg), were provided to 3 x 1-L cultures on days 3, 6 and 8, and all three cultures were harvested on day 10. A total of8.2 mg of 15 was isolated from the crude organic extract. The 13e NMR spectrum of15 isolated from this feeding experiment showed no enhancement ofany resonance. L 4(S)-L-[5-13C]Leucine. 4(S)-L-[5)3e]Leucine (180 mg) was added to 3 x 1-L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of 143 12.5 mg of 15 was isolated from the crude organic extract. The Be NMR spectrum of15 from this feeding experiment showed no enhancement at any resonance. J. 4(R)-L-{5-13C]Leucine. 4(R)-L-[5-BC]Leucine (180 mg) was added to 3 x 1 L cultures on days 3, 6, and 8, and all three cultures were harvested on day 10. A total of 15.1 mg of 15 was isolated from the crude organic extract. The Be NMR spectrum of 15 isolated from this feeding experiment showed no enhancement ofany resonance. K. Sodium 3(R)-{4-13C}Methylbutanoate. Sodium 3(R)-[4-Be]methylbutanoate (82 mg) was supplied to the 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of 7.7 mg of 15 was isolated from the crude organic extract. The 13e NMR spectrum of 15 isolated from this feeding experiment showed no enhancement ofany resonance. L. 3(R)-{4-13C]Methylbutanoyl N-Acetylcysteamine (NAC). The NAe thioester of3(R)-[4-13e]methylbutanoate (145 mg) was supplied to the 2 x 1-L cultures on days 3, 6 and 8, and both cultures were harvested on day 10. A total of3.6 mg of15 was isolated from the crude organic extract. The 13e NMR spectrum of 15 isolated from this feeding experiment revealed no enhancement ofany resonance. 144 CHAPTER VII CONCLUSION During the past 50 years, the studies ofmarine natural products have yielded an abundance ofsecondary metabolites from organisms that range from ascidians to primitive prokaryotic organisms. Among the productive sources, marine cyanobacteria, particularly Lyngbya majuscula, have shown to be rich in structurally unique bioactive natural products. Moreover, recent literature has indicated that many compounds that were tso· 1ate d fr om mo11 usks and sponge are a 1so o f cyanob actena . I ongms. . . 37' 39 ' st ' 123-124 This research has focused on the chemical and biochemical studies ofL. majuscula metabolites. The chemical investigations involve the isolation and structure determination ofnovel metabolites using spectroscopic methods. The biochemical studies utilized stable isotope tracer techniques and high field NMR to establish the primary precursors ofnatural products and their biosynthetic sequence. Grenadadiene, debromogrenadadiene, and grenadamide are three new structurally unique cyclopropyl-containing metabolites from a Grenada collection ofL. majuscula. Grenadadiene is an inhibitor of fatty acid amide hydrolase (F AAH), an enzyme that catalyzes the hydrolysis ofbioactive amides, and possesses an interesting profile of cytotoxicity in the NCI 60 cell line assay. Grenadamide displayed a strong brine shrimp toxicity and cannabinoid binding activity. Grenadadiene is biosynthetically related to 2 bromopropenyl2,5-dimethyldodecanoate59 in that it derives from a fatty acid that is esterified with a bromine-containing alkene unit. The chemical examination ofa Fiji extract ofL. majuscula revealed two novel brine shrimp toxic depsipeptides, yanucamides A and B. Both compounds contain the 2,2-dimethyl-3-hydroxy-7-octynoic acid, a (3-hydroxy fatty acid that has only been 145 described in k.ulolide-1 and k.ulokanalide-1, metabolites from the marine mollusk 82 83 Phi/inopsis speciosa. ' Scheuer and co-workers speculated that the kulolides and their related compounds were ofcyanobacterial origins, which were transferred through the food chain via the sea hare Sty/ocheilus longicaudus, an herbivore that consumes on cyanobacteria. Therefore, the isolation ofthe yanucamides from the field-collected cyanobacteria clearly supports this hypothesis. The studies ofa highly brine shrimp toxic extract ofa Fiji collection ofL. 38 89 majuscula led to the isolation ofthe active components, curacins A and 0. • Upon further examinations, the algal extract was also found to contain two new and structurally unrelated depsipeptides, clairamide and carliamide. Clairamide possessed 3-amino-2 methylpentanoic acid, a unit that is unique to cyanobacterial metabolites. Carliamide contains 3-amino-2-methyl-7-octynoic acid, a (3-amino fatty acid that has been found only in onchidin A,91 a metabolite isolated from the marine mollusk Onchidium sp. Hence, the finding ofcarliamide indicated cyanobacteria are the probable producers of the onchidins. The biosynthesis ofbarbamide, a molluscicidal metabolite from L. majuscula, was investigated using various stable isotope precursors. Results from the feeding experiments conducted with the cultured L. majuscu/a have shown that the biosynthesis ofbarbamide involves chlorination ofunactivated pro-S methyl group ofleucine. The incorporation studies also revealed the other primary precursors ofbarbamide including phenylalanine, cysteine and an acetate unit. The carbons ofthe NCH3 and OCH3 were shown to originate from the Ct pool. The biosynthetic studies ofcuracin A, potent antimitotic agent, was also conducted with the same strain ofcultured L. majuscula that produces barbamide. Results from the feeding experiments have established the polyketide origin ofcuracin A. The assembly ofthe polyketide chain appears to be cysteine-initiated. In addition to the 146 lipid chain, labeled acetate units were also incorporated into the five-carbon fragment (C18-C22) ofcuracin A with a labeling pattern that suggests its originating from a triketide-derived precursor or mevalonic acid. The proposed mechanism for the formation ofthe former intermediate is also reported in the biosynthesis of virginiamycin, 114 116 myxovirecin, and oncorhyncolide. - Finally, the methoxy carbon and C 17 on the polyketide chain (C4-C16) were shown to arise from methionine. 147 H 0~ Grenadadiene R =Br Grenadamide Debromogrenadadiene R = H - ( '( ,..LL-Hiv ~0 Y o~··· ~ ~ f f ~ 0 ~"-/"-/X'h I Yanucamide A Yanucamide B Clairamide Carliamide ~~~ OCHs CH3 ~~so 'b/ Curacin A Barbamide 148 BIBLIOGRAPHY 1. Cragg, G. M.; Newman, D. J.; Snader, K. M. J. Nat. Prod 1997, 60, 52-60. 2. Faulkner, D. J. Nat. Prod. Rep. 1986, 3, 1-33. 3. Faulkner, D. J. Nat. Prod Rep. 1987, 4, 539-576. 4. Faulkner, D. J. Nat. Prod Rep. 1988,5,613-663. 5. Faulkner, D. J. Nat. Prod. Rep. 1990, 6, 269-309. 6. Faulkner, D. J. Nat. Prod. Rep. 1991, 8, 97-147. 7. Faulkner, D. J. Nat. Prod. Rep. 1992,9,323-364. 8. Faulkner, D. J. Nat. Prod. Rep. 1993,10,497-539. 9. Faulkner, D. J. Nat. Prod. Rep. 1994, 11, 355-394. 10. Faulkner, D. J. Nat. Prod. Rep. 1995, 12,223-269. 11. Faulkner, D. J. Nat. Prod. Rep. 1996, 13,75-125. 12. Faulkner, D. J. Nat. Prod. Rep. 1997, 14, 259-302. 13. Faulkner, D. J. Nat. Prod Rep. 1998,15, 113-158. 14. Faulkner, D. J. Nat. Prod Rep. 1999,16, 155-198. 15. Faulkner, D. J. Nat. Prod Rep. 2000, 17,7-55. 16. Moore, R. E.; Scheuer, P. J. Science 1971, 172, 494-498. 17. Simidu, U.; Kita-Tsukamoto, K.; Y asumoto, Y.; Yotsu, M. International Journal ofSystematic Bacteriology 1990, 40, 331-336. 18. Kao, C. Y.; Levinson, S. R. Tetrodotoxin, saxitoxin and the molecular biology of the sodium channel.; New York Academy ofScience: New York, 1986. 19. Tachibana, K.; Scheuer, P. J.; Tsukitani, Y.; Kikuchi, D.; Van Engen, D.; Clardy, J.; Gopichand, Y.; Schmitz, F. J. J. Am. Chem. Soc. 1981, 103,2469-2471. 20. Yasumoto, T.; Murata, M.; Oshima, Y.; Sano, M.; Matsumoto, G. K.; Clardy, J. Tetrahedron 1985,41, 1091-1025. 149 21. Conen, P.; Holmes, C.; Tsukitani, Y. Trends in Biomedical Sciences 1990, 15, 98 102. 22. Pettit, G.; Herald, C. J. Am. Chem. Soc. 1982, 104, 6846-6848. 23. Shu, Y.-Z. J. Nat. Prod. 1998, 61, 1053-1071. 24. Moore, R. E. J. Indus!. Microbio/.1996, 16, 134-143. 25. Namikoshi, M.; Rinehart, K. L. J. Indus!. Microbiol. 1996, 17, 373-384. 26. Andersen, R. J.; Burgoyne, D . L.; Williams, D. E.; Kong, F.; talde Silva, E. D.; Miao, S.; Allen, T. M.; Holmes, C. F. B.; Chen, D.; Kent, M. Gazz Chim. Ita/. 1993, 123, 293-300. 27. Rinehart, K. L.; Harada, K.-I.; Namikoshi, M.; Chen, C.; Harvis, C. A.; Munro, M. H. G.; Blunt, J. W.; Mulligan, P. E.; Beasley, V. R.; Dahlem, A. M.; Carmichael, W. W. J. Am. Chem. Soc. 1988, 110, 8557-8558. 28. Dilip de Silva, E.; Williams, D. E.; Anderson, R. J.; Klix, H.; Holmes, C. F. B.; Allen, T. M. Tetrahedron Lett. 1992,33, 1561-1564. 29. Carmichael, W. W. J. Appl. Bact. 1992, 72, 78-86. 30. Carmichael, W. W. Scientific American 1994, 270, 78-86. 31. Sims, J. K.; Zandee Van Rilland, R. D. Hawaii Med. J. 1981,40,243-248. 32. Sakai, S.-1.; Hitotsuyanagi, Y.; Aimi, N.; Fujiki, H.; Suganuma, M.; Sugimura, T.; Endo, Y.; Shudo, K. Tetrahedron Lett. 1986,27, 5219-5220. 33. Moore, R. E.; Blackman, A. J.; Cheuk, C. E.; Mynderse, J. S.; Matsumoto, G. K.; Clardy, J.; Woodward, R. W.; Craig, J. C. J. Org. Chem. 1984, 49, 2484-2489. 34. Carte, B. K. Bioscience 1996, 46,271-286. 35. Smith, C. D.; Zhang, X.; Mooberry, S. L.; Patterson, G. M. L.; Moore, R. E. Cancer Res. 1994, 54, 3779-3784. 36. Panda, D.; DeLuca, K.; Williams, D.; Jordan, M.A.; Wilson, L. Proc. Nat/. Acad. Sci. USA 1998,95,9313-9318. 37. Kobayashi, M.; Kurosu, M.; Ohyabu, N.; Wang, W.; Fujii, S.; Kitagawa, I. Chem. Pharm. Bull. 1994, 42, 2196-2198. 150 38. Gerwick, W. H.; Proteau, P. J.; Nagle, D. G.; Hamel, E.; Blokhin, A.; Slate, D. J. Org. Chem. 1994, 59, 1243-1245. 39. Harrigan, G. G.; Yoshida, W. Y.; Moore, R. E.; Nagle, D. G.; Park, P. U.; Biggs, J.; Paul, V. J.; Mooberry, S. L.; Corbett, T. H.; Veleriote, F. A. J. Nat. Prod 1998, 61, 1221-1225. 40. Pettit, G. R.; Kamano, Y.; Kizu, H.; Dufresne, C.; Herald, C. L.; Bontems, R. J.; Schmidt, J. M.; Boettner, F. E.; Nieman, R. A. Heterocycles 1989,28, 553-558. 41. Ishida, K.; Matsuda, H.; Yamaguchi, K. J. Nat. Prod 1997, 60, 184-187. 42. Pettit, G. R.; Kamano, Y.; Herald, C. L.; Duffresne, C.; Cerny, R. L.; Herald, D. L.; Schmidt, J. M.; Kizu, H. J. Am. Chem. Soc. 1989,111, 5015-5017. 43. Mason, C. P.; Edwards, K. R.; Carlson, R. E.; Pignatello, J.; Gleason, F. K.; Wood, J. M.; Mason, C. P.; Edwards, K. R.; Carlson, R. E.; Pignatello, J.; Gleason, F. K.; Wood, J. M. Science 1982, 215,400-402. 44. Gleason, F. K.; Porwell, J.; Flippen-Anderson, J. L.; George, C. J. Org. Chem. 1986,51, 1615-1616. 45. Moore, B. S. Nat. Prod Rep. 1999, 16, 653-674. 46. Moore, R. E.; Chen, J. L.; Moore, B. S.; Patterson, G. M. L.; Carmichael, W. W. J. Am. Chem. Soc. 1991, 113, 5083-5084. 47. Ditmann, E.; Meibner, K.; Boner, T. Phycologia 1996, 35, 62-67. 48. Shimizu, Y. Chem. Rev. 1993, 93, 1685-1698. 49. Hemscheidt, T.; Puglisi, M.P.; Larsen, L. K.; Patterson, G. M. L.; Moore, R. E.; Rios, J. L.; Clardy, J. J. Org. Chem. 1994, 59, 3467-3471. 50. Floss, H. G.; Beale, J. M. Angew. Chem. Int. Ed. Engl. 1989,28, 146-177. 51. Carmeli, S.; Moore, R. E.; Patterson, G. M. L. J. Nat. Prod 1990, 53, 1533 1542. 52. Carmeli, S.; Moore, R. E.; Patterson, G. M. L.; Yoshida, W. Y. Tetrahedron Lett. 1993,34,5571-5574. 53. Orjala, J.; Gerwick, W. H. J. Nat. Prod 1996, 59, 427-430. 54. Cardellina II, J. H.; Dalietos, D.; Marner, F.-J.; Mynderse, J. S.; Moore, R. E. Phytochemistry 1978, 17,2091-2095. 151 55. Cardellina II, J. H.; Marner, F.-J.; Moore, R. E. J. Am. Chem. Soc. 1979, 101, 240-242. 56. Loui, M. S. M.; Moore, R. E. Phytochemistry 1979, 18, 1733-1734. 57. Cardellina II, J. H.; Moore, R. E. Tetrahedron 1980, 36, 993-996. 58. Entzeroth, M.; Mead, D. J.; Patterson, G. M. L.; Moore, R. E. Phytochemistry 1985,24,2875-2876. 59. Hodder, A. R.; Capon, R. J. J. Nat. Prod. 1991, 54, 1668-1671. 60. Jiao, Y.; Teruhiko, Y.; Ishikuri, S.; Uchino, H.; Ichihara, A. Tetrahedron Lett. 1996, 37, 1039-1042. 61. Enzeroth, M.; Blackman, A. J.; Mynderse, J. S.; Moore, R. E.J. Org. Chem. 1985, 50, 1255-1259. 62. Fenical, W.; Sleeper, S. L.; Paul, V. J.; Stallard, M. 0.; Sun, H. H. Pure Appl. Chem. 1979, 51, 1865-1874. 63. Pennings, S. C.; Paul, V. J. Marine Bioi. 1993, 117, 535-546. 64. Bisogno, T.; Melck, D.; Petrocellis, L.; De Bobrov, M.; Yu Gretskaya, N. M.; Bezuglov, V. V.; Sitachitta, N.; Gerwick, W. H.; DiMarzo, V. Biochem. Biophy. Res. Comm. 1998, 248, 515-522. 65. Wise, M.; Soderstorm, K.; Murray, T. F.; Gerwick, W. H. Experientia 1996,52, 88-92. 66. Meyer, B. N.; Ferrigni, J. L.; Putnam, J. E.; Jacobson, L. B.; Nichols, D. E.; McLaughlin, J. L. Planta Med. 1982, 45, 31-34. 67. Pettit, G. R.; Xu, J.-P.; Williams, M.D.; Hogan, F.; Schmidt, J. M. Bioorg. Med. Chem. Lett. 1997, 7, 827-832. 68. Pettit, G. R.; Kamano, Y.; Brown, P.; Gust, D.; Inoue, M.; Herald, C. L. J. Am. Chem. Soc. 1982, 104,905-907. 69. Pettit, G. R.; Kamano, Y.; Holzapfel, C. W.; van Zy1, W. J.; Tuinman, A. A.; Herald, C. L.; Baczynskyi, L.; Schmidt, J. M. J. Am. Chem. Soc. 1987, 109, 7581-7582. 152 70. Pettit, G. R.; Kamano, Y.; Herald, C. L.; Tuinman, A. A.; Boettner, F. E.; Kizu, H.; Schmidt, J. M.; Baczynskj, L.; Tomer, K. B.; Bontems, R. J. J. Am. Chem. So~ 1987,109,6883-6885. 71. Pettit, G. R.; Sigh, S. B.; Hogan, F.; Lloyd-Williams, P.; Herald, D. L.; Burkett, D. D.; Clewlow, P. J. J. Am. Chem. Soc. 1989, 111, 5463-5465. 72. Matou, T.; Kondo, T.; Ojika, M.; Yamada, K. J. Org. Chem. 1996, 61, 6340 6345. 73. Pettit, G. R.; Kamano, Y.; Herald, C. L.; Dufresno, C.; Bate, R. B.; Schmidt, J. M.; Cerny, R. L.; Kizu, H. J. Org. Chem. 1990,55,2989-2990. 74. Pettit, G. R.; Kamano, Y.; Dufresne, C.; Cerny, R. L.; Herald, C. L.; Schmidt, J. M. J. Org. Chem. 1989, 54, 6005-6006. 75. Sone, H.; Nemoto, T.; Ojika, M.; Yamada, K. Tetrahedron Lett. 1993,34, 8445 8448. 76. Pettit, G. R.; Kamano, Y.; Herald, C. L.; Fujii, Y.; Kizu, H.; Boyd, M. R.; Boettner, F. E.; Doubek, D. L.; Schmidt, J. M.; Chapuis, J.-C. Tetrahedron 1993, 49,9151-9170. 77. Ojika, M.; Nagoya, T.; Yamada, K. Tetrahedron Lett. 1995,36, 5057-5058. 78. Pettit, G. R.; Xu, J.-P.; Hogan, F.; Cerny, R. L. Tetrahedron Lett. 1993,34, 8449 8452. 79. Pettit, G. R.; Xu, J.-P.; Hogan, F.; Cerny, R. LHeterocycles 1998, 47,491-496. 80. Pettit, G. R.; Xu, J.-P.; Hogan, F.; Williams, M.D.; Doubek, D. L.; Schmidt, J .. Cerny, R. L.; Boyd, M. R. J. Nat. Prod 1997, 60, 752-754. 81. Harrigan, G. G.; Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Nagle, D. G.; Paul, V. J.; Mooberry, S. L.; Corbett, T. H.; Valeriote, F. A. J. Nat. Prod 1998, 61, 1075-1077. 82. Reese, M. T.; Gulavita, N. K.; Nakao, Y.; Hamann, M. T.; Yoshida, W. Y.; Coval, S. J.; Scheuer, P. J. J. Am. Chem. Soc. 1996, 118, 11081-11084. 83. Nakao, Y.; Yoshida, W. Y.; Szabo, C. M.; Baker, B. J.; Scheuer, P. J. J. Org. Chem. 1998, 63, 3272-3280. 84. Marfey, P. Carlsberg Res. Commun. 1984,49,591-596. 153 85. Orjala, J.; Nagle, D. G.; Hsu, V. L.; Gerwick, W. H. J. Am. Chem. Soc. 1995, 117, 8281-8282. 86. Orjala, J.; Nagle, D. H.; Gerwick, W. H. J. Nat. Prod. 1995, 58, 764-768. 87. Orjala, J.; Gerwick, W. H. Phytochemistry 1997, 45, 1087-1090. 88. Hooper, G. J.; Orjala, J.; Schatzman, R. C.; Gerwick, W. H. J. Nat. Prod. 1998, 61' 529-533. 89. Marquez, B.; Verdier-Pinard, P.; Hamel, E.; Gerwick, W. H. Phytochemistry 1998,49,2387-2389. 90. Biskupiak, J. E.; Ireland, C. M. J. Org. Chem. 1983, 48, 2302-2304. 91. Rodriguez, J.; Fernandez, R.; Quinnoa, E.; Riguera, R.; Debitus, C.; Bouchet, P. Tetrahedron Lett. 1994, 35, 9239-9242. 92. Fernandez, R.; Rodrigeuz, J.; Quinoa, E.; Riguera, R.; Munoz, L.; Fernandez; Suarez, M.; Debitus, C. J. Am. Chem. Soc. 1996, 118, 11635-11643. 93. Neidleman, S. L.; Geigert, J. Biohalogenation: Principal, Basic Roles and Applications; 1st ed.; Ellis Horwood Limited: West Sussex, 1986 .. 94. Butler, A.; Walker, J. V. Chem. Rev. 1993, 93, 1937-1944. 95. Unson, M. D.; Faulkner, D. J. Experientia 1993, 49, 349-353. 96. Dumdei, E. J.; Simpson, J. S.; Garson, M. J.; Byriel, K. A.; Kennard, C. H. L. Aust. J. Chem. 1997, 50, 139-144. 97. Fu, X.; Zeng, L.-M.; Su, J.-Y.; Pais, M. J. Nat. Prod. 1993, 56, 637-642. 98. Rossi, J. V.; Roberts, M. A.; Yoo, H.-D.; Gerwick, W. H. J. Appl. Phycol. 1997, 9, 195-204. 99. Rossi, J. V. Masters Thesis, Oregon State University, 1997. 100. Kelly, N.; M.; Reid, R. G.; Willis, C. L.; Winton, P. I. Tetrahedron Lett.1995, 36, 8315-8318. 101. Kelly, N. M.l Sutherland, A.; Willis, C. L. Nat. Prod. Rep. 1996, 205-219. 102. Kazlauskas, R.; Lidgard, R. 0.; Wells, R. J. Tetrahedron Lett. 1977, 3183-3186. 103. Kazlauskas, R.; Murphy, P. T.; Wells, R. J. Tetrahedron Lett. 1978,4945-4948. 154 104. Fu, X.; Zeng, L.-M.; Su, J.-Y.; Pais, M. J. Nat. Prod. 1997, 60,695-696. 105. Fu, X.; Ferriera, M. L.; Schmitz, F. J.; Kelly-Borges, M. J. Nat. Prod. 1998, 61, 1226-1231. 106. Gebreyesus, T.; Yosief, T.; Carmely, S.; Kashman, Y. Tetrahedron Lett. 1988, 29, 3863-3864. 107. Hofheinz, W.; Oberhansli, W. E. Helv. Chim. Acta 1977, 60,660-669. 108. Lee, G. M.; Molinski, T. F. Tetrahedron Lett. 1992,33, 7671-7674. 109. Vederas, J. C. Nat. Prod Rep. 1987, 277-337. 110. Seto, H.; Watanabe, H.; Furihata, K. Tetrahedron Lett. 1996, 37, 7979-7982. 111. Martin, G. E.; Crouch, R. C. J. Heterocyclic Chern. 1995, 32, 1665-1669. 112. Williamson, R. T.; Sitachitta, N.; Gerwick, W. H. Tetrahedron Lett.1999, 40, 5175-5178. 113. Nagle, D. G.; Gerald, R. S.; Yoo, H.-D.; Gerwick, W. H.; Kim, T.-S.; Nambu, M. White, J.D. Tetrahedron Lett. 1995, 36, 1189-1192. 114. Kingston, D. G. I.; Kolpak, M. X.; LeFevre, J. W.; Borup-Grochtmann, I. J. Am. Chern. Soc. 1983, 105, 5106-5110. 115. Trowitzsch, W.; Gerth, K.; Wray, V.; Hofle, G. J. Chern. Soc., Chern. Commun. 1983, 1174-1175. 116. Needham, J.; Andersen, R. J.; Kelly, M. T. J. Chern. Soc., Chern. Commun. 1992, 1367-1369. 117. Rohmer, M. Nat. Prod Rep. 1999,16,565-574. 118. Stainer, R. Y.; Cohen-Bazire Annu. Rev. Microbial. 1977, 31, 225-274. 119. Abell, C.; Staunton, J. J. Chern. Soc. Chern. Commun. 1981, 856-858. 120. Leeper, F. J.; Staunton, J. J. Chern. Soc. Perkin Trans. 1984, 2919-2925. 121. Abell, C.; Doddrell, D. M.; Garson, M. J.; Laue, E. D.; Staunton, J. J. Chern. Soc., Chern. Commun. 1983, 694-696. 122. Carmeli, S.; Paik, S.; Moore, R. E.; Patterson, G. M. L. Yoshida, W. Y. Tetrahdron Lett. 1993, 34, 6681-6684. 155 123. Williams, D. E.; Burgoyne, D. L.; Rettig, S. J.; Anderson, R. JJ. Nat. Prod. 1993, 56, 545-551. 124. Harrigan, G. G.; Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Nagle, D. G.; Paul, V. J. J. Nat. Prod. 1999, 62, 655-658. 156 APPENDICES 157 APPENDIX A 158 THE ISOLATION OF MICROCOLINS A AND B FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA Abstract Through a collaborative study between our laboratory and Dow AgroSciences, three previously described metabolites, microcolins A (1), B (2}, and C (3) were isolated from a Cura~ao collection ofLyngbya majuscula. Compounds 1-3 were discovered to be potently active against tobacco bud worm (Heliothis virescens). Comparisons ofthe 1H and 13C NMR data ofthe pure compounds to those ofreported in the literatures . . confinned the structures of 1-3. 159 Introduction Microcolins A (1) and B (2) were first isolated from a Venezuelan collection of Lyngbya majuscula as potent inhibitors ofthe murine mixed lymphocyte response and murine P-388 leukemia in vitro. 1 Subsequently, microcolin C (3) was identified from a 2 Curayao collection ofL. majuscula as a potent brine shrimp toxic metabolite. The 3 4 absolute stereochemistries ofcompounds 1 and 2 were established by total syntheses. ' The microcolins are structurally related to two cytotoxins from L. majuscula, 5 majusculamide D (4) and deoxymajusculamide D (5). In screening ofalgal extracts for insecticidal agents in the Insect Management Program through our collaboration with Dow Agro Science, a lipophilic extract ofL. majuscula collected from Playa Kalki, Curayao, inhibited the hatching oftobacco budworm larvae. Bioassay-guided fractionation ofthis extract by size exclusion chromatography followed by a reversed phase (Cis) VLC provided a bioactive fraction containing a mixture ofthe microcolins. Further purification ofthe microlins mixture by a reversed phase HPLC yielded the insecticidal agents 1-3. 160 Results and Discussion Lyngbya majuscula was collected from Playa Kalki, Curayao and extracted with (2:1) CH2Ch/MeOH to give 9.06 g oflipophilic extract. A portion ofthe lipid extract (1.46 g) was fractionated using size exclusion chromatography (Sephadex LH-20, EtOAc/MeOH (1:1). Fractions that showed similar TLC profiles were combined and submitted for biological evaluation. A bioactive fraction was further fractionated by reversed-phase (C 18) vacuum liquid chromatography using stepwise gradient elution from 60% MeOH in H20 to 100% MeOH to provide an active fraction containing a mixture of the microcolins. This fraction was further purified by reversed-phase HPLC to yield 13.1 mg ofmicrocolin A (1), 9.8 mg ofmicrocolin B (2) and 5.1 mg ofmicrocolin C (3). The structures of1 and 2 were confirmed by comparisons ofthe 1H and 13C NMR spectral data ofthe pure compounds to the previously reported data. I Microcolins A (1), B (2) and C (3) inhibit the hatching oftobacco budworm larvae with LDso values of0.1 ppm. ~~~~~\4-~CHa CHa CHa 0 A CHa 0 0 ·y OR2 0 1 R1 = OH, R2 = Ac 2 R1 =H, R2 =Ac 3 R1 = OH, R2 = H ~ CH3 4 R=OH 5 R=H 161 Experimental GeneraL Nuclear magnetic resonance (NMR) spectra ofwere recorded on Broker AC300 instrument at a proton frequency of300.13 MHz and a carbon frequency of75.46 MHz. Proton spectra were referenced to the residue signal ofCHCb at 8 7 .26. Carbon spectra were referenced to the centerline ofCDCb at 8 77.0. HPLC separations were performed with a Waters M-6000A pump, a Rhoedyne 7010 injector, and a Waters Lambda-Max 480 UV detector. Merck aluminum-backed thin layer chromatography (TLC) sheets (silica gel60 F2s4) were used for TLC. Compounds were detected by UV illumination or by heating plates sprayed with a 50% H2S04 solution. Reversed-phase vacuum liquid chromatography (RP-HPLC) was performed on Baker Bonded Phase octadecyl (Cis). All solvents were distilled from glass prior to use. Collection. Lyngbya majuscula was collected by SCUBA at depth of 7 to 10 m. from Playa Kalki, Curayao. The algal material was immediately preserved in isopropanol for transportation and kept at -20°C until extraction. The voucher specimen is available as NAK-11 Aug 94-02. Isolation and Purification. The defrosted alga (506.8 g, dry wt) was filtered through cheesecloth from the preserving solvent and extracted with CH2Ch/MeOH (2: 1). The CH2Ch/MeOH extract was evaporated in vacuo to give crude extract (9.06 g). A portion ofthe crude extract (1.46 g) was fractionated utilizing size exclusion chromatography [Sephadex LH-20, EtOAc/MeOH (1:1), 2.51.D. x 60 em]. Fractions that exhibited similar TLC profiles were combined and submitted for biological evaluation. A bioactive fraction was further fractionated by reversed-phase (Cis) vacuum liquid chromatography using step-wise gradient elution from 60% MeOH in H20 to 100% MeOH, providing an active fraction containing a mixture ofthe microcolins. The fmal purification ofthe microcolins mixture was accomplished using the ODS(2) HPLC [Phenomenex Spheresorb ODS(2), 10 J.L, 250 x 10 mm, MeOHIH20 ( 4: 1), flow rate 4 162 mVmin, detection at 215 nm, tR =17 min for 1, tR = 21 min for 2, and tR = 26 min for 3] to yield 13.1 mg of1, 9.8 mg of2, and 5.1 mg of3. References 1. Koehn, F. E.; Longley, R. E.; Reed, J. K. J. Nat. Prod. 1992, 55,613-619. 2. Yoo, H.-D. Ph.D. Thesis, Oregon State University, 1997. 3. Decicco, C. P.; Grover, P. J. Org. Chern. 1996, 61, 3534-3541. 4. Andrus, M. B.; Li, W.; Keyes, R. F. J. Org. Chern. 1997, 62,5542-5549. 5. Moore, R. E.; Entzeroth, M. Phytochemistry 1988,27,3101-3103. 163 APPENDIXB 164 THE ISOLATION OF MALYNGAMIDE F ACETATE, MALYNGAMIDES H, I, K AND INDOLE DERIVATIVES FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA Abstract A chemical investigation ofa lipophilic extract ofLyngbya majuscula collected from Barcadera Pariba, Aruba, led to the isolation of five previously described metabolites, malyngamide F acetate (1), malyngamides H (2), I (3), K (4), and N dimethylindole-3-carboxaldehyde (5). The structures ofthe purified metabolites were confirmed by the 1H and 13C NMR comparison with previously reported data. 165 Introduction The malyngamides comprise the largest class oflipopeptides produced by the marine cyanobacterium Lyngbya majuscu/a. Most malyngamides are N-substituted amides of7(S)-methoxytetradec-4(E)-enoic acid that frequently possess a terminal chloromethylene group.l-8 Fractionation ofthe crude lipid extract from an Aruba collection ofL. majuscula yielded four previously reported malyngamides; malyngamide F acetate (1),2 malyngamide H (2), 5 I (3), 7 and K (4).8 In addition to the malyngamides, 9 the extract also provided N-dimethylindole-3-carboxaldehyde (5), the only aldehyde previously reported from L. majuscu/a. 166 Results and discussion A collection the marine cyanobacterium Lyngbya majuscula from Barcadera Pariba, Aruba, was extracted with (2:1) CH2Ch/MeOH to provide 9.63 g oflipophilic extract. Moderately polar fractions from the normal-phase vacuum liquid chromatography (VLC, 8.5 I. D. x 5 em) ofthe crude extract were rich in UV-active and brown-charring spots. These fractions were combined and further purified by silica gel flash column chromatography and NP-HPLC, respectively to yield malyngamide F acetate (1, 12.5 mg, malyngamides H (2, 7.1 mg), I (3, 7.7 mg), K (4, 4.5 mg) and N methylindole-carboxyladehyde (5, 26.8 mg). Structures ofthe purified metabolites were verified by comparing their 1H and 13C NMR data with those ofthe reported in the literatures. Malyngamide F acetate (1) was described to show a slight antimicrobial activity 2 5 7 against Staphylococcus aureus, whereas malyngamides H (2), I (3) and K (4)8 were reported to be mildly ichthyotoxic against goldfish Carassius auratus. The ubiquitous and ichthyotoxic nature ofthe malyngamides suggests that these compounds may play an important role in the natural defense ofL. majuscula from herbivores. 167 1 3 ytHO CHa CHa 5 168 Experimental General. Nuclear magnetic resonance (NMR) spectra were recorded on a Broker AC300 instrument at a proton frequency of300.13 MHz and a carbon frequency of75.46 MHz. Proton spectra were referenced to the residue signal ofCHCh at 8 7.26. Carbon spectra were referenced to the centerline ofthe CDCh signal at 8 77.0. HPLC separations were performed with a Waters M-6000A pump, a Rhoedyne 7010 injector, and a Waters Lambda-Max 480 UV detector. Merck aluminum-backed thin layer chromatography (TLC) sheets (silica gel60 F2s4) were used for TLC. Compounds were detected by UV-illumination or by heating plates sprayed with 50% H2S04 solution. Vacuum liquid chromatography was performed with Merck Silica Gel G for TLC. Flash chromatography utilized Merck Kieselgel 60, 230-400 mesh). All solvents were distilled from glass prior used. Collection. The sample ofLyngbya majuscula was collected from Barcadera Pariba, Aruba (-1 to-2m). The algal material was immediately preserved in isopropanol for transportation and kept at -20°C until extraction. The voucher specimen is available as NBP-10 May 96-01. Isolation and Purification. The defrosted alga (300.5 g, dry wt) was filtered through cheesecloth from the preservation isopropanol and extracted with (2:1) CH2Ch/MeOH. The CH2Ch/MeOH extract was evaporated in vacuo to give 9.63 g of the lipophilic extract. A portion ofthe crude extract (9.43 g) was fractionated using NP VLC by increasing the polarity ofthe eluting solvents (EtOAc/hexanes). Fractions that showed a similar TLC profile were combined. TLC analysis ofthe combined fractions eluting with 80% EtOAc/hexanes contained several UV-active and brown-charring metabolites. Two successive purifications by flash silica gel chromatography using hexanes/EtOAc/MeOH (65:33:2) and a stepped gradient beginning with 100% CH2Chto 100% MeOH, respectively, provided a pure purple-charring compound, N-methylindole 169 carboxyladehyde (5, 26.8 mg) and a fraction containing several UV-active and brown charring compounds. This latter fraction was subjected to NP-HPLC [Alltech Versapack Silica lOJ.t, 300 mm x 4.1 mm, Hex/EtOAc/MeOH (65:33:2)] to yield malyngamide F acetate (1, 12.5 mg), malyngamides H (2, 7.1 mg), I (3, 7.7 mg), and K (4, 4.5 mg). References 1. Cardellina II, J. H.; Marner, F.-J.; Moore, R. E. J. Am. Chern. Soc. 1979, 101, 240-242. 2. Gerwick, W. H.; Reyes, S.; Alvarado, B. Phytochemistry 1987, 26, 1701-1704. 3. Kan, Y.; Fujita, T.; Nagai, H.; Sakamoto, B.; Hokama, Y. J. Nat. Prod 1998, 61, 152-155. 4. Mynderse, J. S.; Moore, R. E. J. Org. Chern. 1978, 43, 4359-4363. 5. Orjala, J.; Nagle, D. H.; Gerwick, W. H. J. Nat. Prod 1995, 58, 764-768. 6. Praud, A.; Valls, R.; Piovetti, L.; Banaigs, B. Tetrahedron Lett. 1993, 34, 5437 5440. 7. Todd, J. S.; Gerwick, W. H. Tetrahedron Lett. 1995,36,7843-7840. 8. Wu, M.; Milligan, K. E.; Gerwick, W. H. Tetrahedron 1997,53, 15983-15990. 9. Todd, J. S.; Gerwick, W. H. J. Nat. Prod. 1995, 58, 586-589. 170 APPENDIXC 171 THE ISOLATION OF KALKITOXIN AND CYMOPOL FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA Abstract Through a collaborative study between our laboratory and Professor Robert Jacobs laboratory at University ofCalifornia, Santa Barbara (UCSB) we searched for new toxic agents from marine algae. This led to the isolation ofa known metabolite, kalk:itoxin (1, IC50 = 10 ng/ml), from a Grenada collection ofLyngbya majuscu/a. In addition to kalkitoxin, the extract also yielded a brine shrimp toxic compound, cymopol (2, LD50 = 1 ppm). Interestingly, cymopol had been previously isolated from a calcified green alga Cymopolia barbata. 172 Introduction A screening program ofour marine algal extracts at the University ofCalifornia Santa Barbara (UCSB) using a sea urchin assay led to the isolation ofa potent toxin, kalkitoxin (1, ICso = 10 ng/ml). Kalkitoxin was first discovered by our laboratory from a collection ofL. majuscula collected from Playa Kalki, CuraQao, as a brine shrimp toxic 1 agent and an ichthyotoxin. In addition to the sea urchin assay, we also utilized the brine shrimp assay to guide the fractionation ofthe same extract to obtain to another bioactive metabolite, cymopol (2, LDso = 1 ppm). Interestingly, cymopol was reported as a green 2 algal (Cymopolia barbata) metabolite. 173 Results and Discussion Lyngbya majuscula was collected from Jendy Point, Grenada in July 1995. The lipid extract (2: 1 CH2Ch/MeOH) exhibited activity in the sea urchin assay and possessed brine shrimp toxicity. Fractionation ofthe extract was achieved by silica gel vacuum liquid chromatography (NP-VLC) using an increasing gradient ofEtOAc in hexanes. The brine shrimp toxic fraction eluting with 30% EtOAclhexanes was further fractionated by reversed-phase (Cts) vacuum liquid chromatography (RP-VLC) to provide 27.7 mg of pure cymopol (2). A more polar VLC fraction (50% EtOAclhexanes) by size exclusion chromatography, RP-VLC and RP-i:IPLC to give 1.5 mg ofpure kalkitoxin (1). Kalkitoxin (1) and cymopol (2) were evaluated in the sea urchin assay at UCSB and brine shrimp toxicity assay at OSU. Compound 1 exhibited a toxic activity (ICso 10 ppb) and brine shrimp toxicity (LDso =30 ppb), whereas 2 showed only brine shrimp toxicity at LD50 = 10 ppm. Compound 2 may not be L. majuscula metabolite since this collection ofL. majuscu/a was found to be contaminated with calcified algae, and 2 was originally reported as metabolite green algal Cymopo/ia barbata. ~ 0 1 2 174 Experimental GeneraL Nuclear Magnetic Resonance (NMR) spectra were recorded on Bruk.er AC300 instrument at a proton frequency of300.13 MHz and a carbon frequency of75.46 MHz. Proton spectra were referenced to the residue signal ofDMSO (a 2.50) or CHCh (a 7.26). Carbon spectra were referenced to the centerline ofsolvent, DMSO-d6 (a 39.51) or CDCh (a 77.0). Vacuum liquid chromatography was performed on Merck Silica Gel for TLC or Baker Bonded Phase-octadecyl (Cts). Merck aluminum-backed thin layer chromatography (TLC) sheets (silica ge160 Fzs4) were used for TLC. HPLC separations were performed with Waters M-6000A pump, a Rhodyne 7010 injector, and a Waters Lambda-Max 480 UV detector. Compounds were detected by UV-illumination or by heating plates sprayed with a 50% HzS04 solution. All solvents were distilled from glass prior to use. Collection. Lyngbya majuscula, which was found growing with calcified algae, was collected from Jendy Point, Grenada, at -0.5 to-1m. The algal material was immediately preserved in isopropanol for transportation and kept at -20°C until - . extraction. The voucher specimen is available as GJP-29 Jul95-0l. Isolation and Purification. The defrosted alga (87 .1 g, dry wt) was filtered through cheesecloth from the preserving solvent and extracted with CHzCh/MeOH (2: 1) and dried in vacuo to give a lipophilic extract (3.61 g). A portion ofthe crude extract (1.36 g) was fractionated by silica vacuum liquid chromatography using an increasing gradient ofEtOAc in hexanes. The fraction eluting with 30% EtOAc/hexanes (brine shrimp toxic LDtoo 1 ppm) was subjected to reversed-phase (Cts) vacuum liquid chromatography using a stepwise gradient elution from 60% MeOHIHzO to 100% MeOH and yield cymopol (2, 27.7 mg). Successive purification ofthe more polar VLC fraction (50% EtOAclhexanes), which showed toxic activity and brine shrimp toxicity (LDso = 0.1 ppm), by size exclusion chromatography (Sephadex LH-20, 1:1 EtOAc/MeOH), 175 reversed~phase (Ct8) vacuum liquid chromatography (stepwise gradient elution from 60% MeOHIH20 to 100% MeOH), and RP~HPLC [Phenomenex Spherisorb ODS (2), 90 % MeOHIH20, detection at 215 nm, flow rate 3 ml/min] to provide kalkitoxin (1, 1.5 mg). Brine Shrimp Toxicity Assay. In a method slightly modified from the original description,3 about 15 newly hatched brine shrimp (Artemia salina) in ca. 0.5 mL artificial seawater were added to each well containing different concentrations ofthe sample in 50 J.tL EtOH and 4.5 mL artificial seawater to make a total volume ofca. 5 mL. Samples and controls were run in duplicate. After 24 h at 28°C, the brine shrimps were observed and counted with a dissecting light microscope. The percentage oflive brine shrimp versus total brine shrimp was used to determine LD5ovalues. References 1. Wu, M. Maters Thesis, Oregon State University, 1996. 2. Hogberg, H.-E.; Thomson, R. H.; King, T. J. J. Chem. Soc. Perkin Trans I 1976, 1696-1701. 3. Meyer, B. N.; Ferrigni, J. L.; Putnam, J. E.; Jacobson, L. B.; Nichols, D. E.; McLaughlin, J. L. Planta Med 1982,45, 31~34.