AN ABSTRACT OF THE THESIS OF
Hye-Dong Yoo
for the degree of
Doctor of Philosophy
in
Pharmacy presented on October 13. 1997
Title: Bioactive Secondary Metabolites from Marine Algae and Study of
Oxidized Anandamide Derivatives
Redacted for Privacy
Abstract approved:
William H. Gerwick
My investigations of the natural products of marine algae have resulted in
the discovery of several new secondary metabolites. Bioassay-guided
fractionation led to the isolation of these new compounds and spectroscopic
analysis was utilized in their structural characterization.
Two new and potent antimitotic metabolites, curacins B and C, were
isolated from a Curacao collection of Lyngbya majuscula. In addition, four
curacin A analogs were prepared by semisynthetic methods. The structures of
the new curacins and the curacin A analogs were determined by spectroscopic
analysis in comparison with curacin A. The biological properties of the new
natural products and synthetic derivatives of curacin A were examined.
Investigations of another Curacao collection of L. majuscula revealed a
new cytotoxic lipopeptide, microcolin C. Microcolin C was found to have an
interesting profile of cytotoxicity to human cancer-derived cell lines.
A new metabolite, vidalenolone, was isolated from an Indonesian red
alga Vidalia sp. The structure of this new cyclopentenolone-containing
compound was determined by a combination of spectroscopic methods.
Filamentous cells isolated from female gametophytes of the brown alga
Laminaria saccharina were cultured in flasks or bioreactors. These cultures
produced a variety of w6-lipoxygenase metabolites: 13-hydroxy-9,11­
octadecadienoic acid (13-HODE), 13-hydroxy-6,9,11,15-octadecatetraenoic
acid (13-HODTA), and 15-hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE).
Five oxidized anandamide derivatives were prepared from anandamide
through autoxidation in an exploration of ligand binding to the cannabinoid
receptor. Their structures were determined by a combination of NMR
spectroscopy and GC-MS. The cannabinoid receptor binding affinity of these
derivatives was evaluated. This study revealed the following trend in activity:
anandamide > 15- > 9- > 8- > 11- > 5-hydroxyanandamide.
Copyright by Hye-Dong Yoo
October 13, 1997
All Rights Reserved
Bioactive Secondary Metabolites from Marine Algae and Study of Oxidized
Anandamide Derivatives
by
Hye-Dong Yoo
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Doctor of Philosophy
Completed October 13, 1997
Commencement June 1998
Doctor of Philosophy thesis of Hye-Dong Yoo presented on
October 13, 1997
APPROVED:
Redacted for Privacy
Major Professor, representing Pharmacy
Redacted for Privacy
Dean of the College of Pharmacy
Redacted for Privacy
Dean of Grad
e School
I understand that my thesis will become part of the permanent collection of
Oregon State University libraries. My signature below authorizes release of my
thesis to any reader upon request.
Redacted for Privacy
Hye-Dong Yoo, Author
ACKNOWLEDGEMENTS
I wish to thank my major advisor, Dr. William H. Gerwick, for his generous
support and guidance throughout my graduate studies. His unlimited patience
and mentorship led me to completion of this project.
I must thank my labmate Mary Ann Roberts for teaching me about algae
and for encouragement during my research.
I would like to thank Dr. George Constantine for his thoughtful advice and
encouragement; Dr. Max Deinzer and Dr. Michael Hyman for accepting to be on
my committee and their time; Dr. Mark Zabriskie and Dr. Philip Proteau for their
insightful suggestions; Dr. Mitchell L. Wise for a long time friendship; Dr. Dale
Nagle for his assistance in my early research years; Dr. Yongfu Li and Dr. Anna
Boulanger for helpful discussions; Brian Arbogast for his mass spectral work;
Roger Kohnert for NMR spectroscopy.
I am also indebted to Dr. Greg Hooper, Mary Roberts, Brian Marquez, Lik
Tan Tong and Namthip Sitachitta for their assistance with this manuscript.
I would like to acknowledge Dr. Pete K. Freeman, Dr. Henry Schaup, and
Dr. Christian Haugen for their unconditional support and friendship.
My past and present labmates are also appreciated for their support and
helpful disscussions.
Finally, I wish to express my gratitude to my family for their patience and
support.
Funding for the work in this thesis was supported by the National Cancer
Institute and Oregon Sea Grant.
CONTRIBUTION OF AUTHORS
Chapters V and VI of my thesis were collaborative efforts performed with
two research groups. Dr. Greg Rorrer research group in the Department of
Chemical Engineering was involved in bioreactor design and start-up and
cultivation of brown alga, Laminaria saccharina female gametophyte cells in
chapter V.
Ken Soderstrom in the laboratory of Dr. Thomas Murray was responsible
for the competitive receptor ligand displacement assay with a variety of
hydroxyanandamides for Chapter VI.
TABLE OF CONTENTS
Page
CHAPTER I. GENERAL INTRODUCTION
1
MARINE MICROORGANISMS
BIOACTIVE METABOLITES FROM CYANOBACTERIA
OXYLIPINS FROM MARINE ALGAE
CHAPTER II. CURACINS, ANTIMITOTIC NATURAL PRODUCTS
FROM THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
23
ABSTRACT
23
INTRODUCTION
24
RESULTS AND DISCUSSION
27
EXPERIMENTAL
50
CHAPTER III. MICROCOLINS, NEW ANTINEOPLASTIC NATURAL
PRODUCTS FROM THE MARINE CYANOBACTERIUM
LYNGBYA MAJUSCULA
59
ABSTRACT
59
INTRODUCTION
60
RESULTS AND DISCUSSION
63
EXPERIMENTAL
76
CHAPTER IV. VIDALENOLONE, A NOVEL SECONDARY
METABOLITE FROM THE INDONESIAN RED MARINE ALGA
VIDALIA SR
79
ABSTRACT
79
INTRODUCTION
80
RESULTS AND DISCUSSION
82
EXPERIMENTAL
89
CHAPTER V. ANALYSIS OF OXYLIPIN PRODUCTS FROM
GAMETOPHYTE CELL CULTURE OF BROWN ALGA
LAMINARIA SACCHARINA
91
ABSTRACT
91
INTRODUCTION
92
RESULTS AND DISCUSSION
97
EXPERIMENTAL
112
CHAPTER VI. PRODUCTION OF OXIDIZED ANANDAMIDE
DERIVATIVES AND EVALUATION OF THEIR BINDING
AFFINITY TO THE CANNABINOID RECEPTOR (CB1)
115
ABSTRACT
115
INTRODUCTION
116
RESULTS AND DISCUSSION
118
EXPERIMENTAL
132
CHAPTER VII. CONCLUSION
135
BIBLIOGRAPHY
138
APPENDIX
146
List of Figures
Figure
Page
1.1
Biosynthesis of the products of arachidonic acid in mammals.
15
11.1
Structures of selected secondary metabolites from
Lyngbya majuscula.
25
Mean graphs of curacins B and C (95:5) mixture obtained
from the National Cancer Institute 60 cell line in vitro screen.
28
11.3
Flow chart of fractionation of Lyngbya majuscula.
29
11.4
Comparison of 1H NMR spectra of curacins A, B, and C.
30
11.5
Geometrical configuration of conjugated diene of curacin B (14)
determined by NOESY in CDCI3.
33
11.6
Thermal isomerization of curacins A, B, and C.
34
11.7
GC-MS traces for thermal isomeration of curacins A, B, and C.
35
11.8
Natural antimitotic agents and synthetic derivatives of
the curacins.
37
11.2
11.9
11-1 NMR spectrum of cyclopropyl ring opened curacin A (18)
in C6D6.
39
1H-1H COSY spectrum of cyclopropyl ring opened
Curacin A (18) in C6D6.
40
11.11
1H NMR spectrum of 19-epi curacin A (19) in C6D6.
41
11.12
1H-1H COSY of 19-epi curacin A (19) in C6D6.
42
11.13
1H NMR spectrum of thiazoline ring opened curacin A
11.10
(20) in C6D6.
43
11.14
HMBC plot of thiazoline ring opened curacin A (20) in C6D6.
11.15
1H NMR spectrum of derivative 21 in C6D6.
45
11.16
'H NMR spectrum of curadisulfide A (22) in C6D6.
46
44
List of Figures (continued)
Figure
Page
Flow chart for the isolation of microcolins.
64
111.2
Structures of microcolins A-D (7-10).
66
111.3
Mass fragmentations for microcolin isomers.
67
111.4
Comparison of 11-1 NMR spectra of microcolins A, B, and
C (7-9) at 300 MHz in CDCI3.
69
111.5
TOCSY contour plot of microcolin C (9) at 600 MHz (CDCI3).
70
111.6
Enzymatic hydrolysis of microcolin B.
72
111.7
1H spectrum of microcolin D (10) at 600 MHz (CDCI3).
74
111.8
1H-1H COSY contour plot of microcolin D (10) at 600 MHz
111.1
(CDCI3).
75
IV.1
1H NMR spectrum of vidalenolone (13).
83
IV.2
Partial structures of vidalenolone (13).
85
IV.3
Structure of vidalenolone (13).
85
IV.4
Proposed biogenetic pathway of vidalenolone (13).
88
V.1
Expected fragment patterns of various oxylipins.
93
V.2
Overview of the identification and analysis of oxylipins
in L. saccharina cell culture.
98
V.3. Total Ion Chromatogram (TIC) of TMS-hydroxy fatty acids
produced from L. saccharina gametophyte cell suspension.
V.4
V.5
99
Mass spectra and structures of TMS ether derivatives
of methyl 13-HODTA (2) and methyl 15-HETE (3).
103
Mass spectra and structures of TMS ether derivatives of
(A) methyl 13 -NODE (1) and (B) of co-eluted methyl ricinelaidic
acid standard (4) and methyl 13 -NODE.
104
List of Figures (continued)
Figure
V.6
Page
Hydroxy fatty acid production from precursors by 15-LOX/
peroxidase metabolism.
106
V.7 Yields of 13-HODTA from cultures added with different
precursors.
108
V.8 Yields of 13-HODE from cultures added with different
precursors.
109
V.9 Yields of 15-HETE from cultures added with different
precursors.
110
V.10 Comparison of the yields of three oxylipins from feeding
experiment.
VI.1
Preparation and purification of various hydroxyanandamides.
VI.2
Structures of various hydroxyanandamides.
VI.3
Mass spectra of (A)15-hydroxyanandamide, (B)
11-hydroxyanandamide, (C) 9-hydroxyanandamide,
(D) 8-hydroxyanandamide.
121
VI.4 Proposed structures of the ions at m/z 315, 327, 355, 407
formed due to rearrangement process corresponding to 8-(A),
9-(B), 11-(C), and 15-hydroxyanandamides (D), respectively.
122
VI.5
11-1 NMR spectrum of 15-hydroxyanandamide (3).
123
VI.6
1H-1H COSY of 15-hydroxyanandamide (3).
124
VI.7 1H NMR data assignments for 15-(3) and 5­
hydroxyanandamide (7).
125
VI.8
1H NMR spectrum of 5-hydroxyanandamide (7).
126
V1.9
1H-1H COSY of 5-hydroxyanandamide (7).
127
VI.10 Inhibition of the specific binding of [3H]CP-55940 to rat
neuronal membrane by anandamide and oxidized anandamide
derivatives.
131
List of Tables
Figure
Page
11.1
1H and 13C NMR data for curacins A (13), B (14), and C (15).
31
11.2
Biological activities of curacins on cell growth, tubulin
polimerization, coichicine binding, and brine shrimp toxicity.
49
'H and 13C NMR data for 9 and 10.
71
Tumor cell growth inhibitory activity and cytotoxicity of microcolin C
in the National Cancer Institute in vitro 60-cell line screen.
73
IV.1
1FI and 13C NMR data for vidalenolone (13).
84
V.1
Identified compounds for the TIC chromatogram shown
in Figure V.3.
101
111.1
111.2
V.2
Effects on the yields of oxylipin from culture fed with
six precursors.
107
List of Abbreviations
AA
Arachidonic Acid
CC
Column Chromatography
CD
Circular Dichroism
CIMS
Chemical Ionization Mass Spectrometry
COSY
1H-1H
DEPT
Distortion less Enhancement by Polarization Transfer
DMSO
Dimethylsulfoxide
EC
Effective Concentration
El
Electron Impact
EtOAc
Ethyl Acetate
FAB
Fast Atom Bombardment
FT
Fourier Transform
FTIR
Fourier Transformed Infrared Spectroscopy
GI
Growth Inhibition
GC
Gas Chromatography
HETCOR
Heteronuclear Correlation Spectroscopy
HEPE
Hydroxyeicosapentaenoic Acid
HETE
Hydroxyeicosatetraenoic Acid
HMBC
Heteronuclear Multiple-Bond Coherence Spectroscopy
HMQC
Heteronuclear Multiple-Quantum Coherence
HPETE
Hydroxyperoxyeicosapentaenoic Acid
HPLC
High Performance Liquid Chromatography
HRCIMS
High Resolution Chemical Ionization Mass Spectrometry
IC
Inhibitory Concentration
Chemical Shift Correlation Spectroscopy
IPA
Isopropyl Alcohol
IR
Infrared Spectroscopy
LD
Lethal Dose
LT
Leukotriene
LX
Lipoxin
LO
Lipoxygenase
MS
Mass Spectrometry
NCI
National Cancer Institute
NMR
Nuclear Magnetic Resonance
NOE
Nuclear Overhauser Effect
NOESY
Nuclear Overhauser Exchange Spectrometry
PG
Prostaglandin
TGI
Total Growth Inhibition
TLC
Thin Layer Chromatography
TMS
Tetramethylsilane
TMSi
Trimethylsilyl
TOCSY
Total Correlation Spectoscopy
UV
Ultraviolet Spectroscopy
XHCORR
Heteronuclear Chemical Shift Correlation Spectoscopy
BIOACTIVE SECONDARY METABOLITES FROM MARINE ALGAE AND
STUDY OF OXYDIZED ANANDAMIDE DERIVATIVES
CHAPTER I : GENERAL INTRODUCTION
Natural products have been a valuable source of medicine for centuries.
Especially, natural products from terrestrial higher plants have been used
extensively for practical pharmaceuticals from the 19th century until the late
1940s. A recent literature review summarized that about 120 drugs have been
discovered from plants.' Remarkable examples include compounds with
anesthetic (cocaine), skeletal muscle relaxant (tubocurarine), antimalarial
(Cinchona alkaloids), antigout (colchicine), cardiac glycoside (Digitalis
glycoside), and anticancer activity (taxol, vincristine, and podophyllotoxin).
Since the collection of terrestrial samples is so heavily carried out, the
study of natural products from this area has been extensively explored. Thus, it
is highly difficult to discover novel compounds from terrestrial sources. On the
other hand, the sea has been relatively unexplored. Since approximately three
quarters of the earth's surface is covered by oceans, containing 1.8 million
different species, marine species therefore exhibit a great diversity of life forms.'
These different species have evolved over millions of years in a stable
environment and natural selection has allowed the development of different
chemicals as unique defense weapons.
2
Since the pioneering study of Bergmann on antibacterial agents from
marine organisms in 1949,3 marine natural products have emerged as a
valuable source of structurally novel compounds which may evolve to be
potential drugs. Marine natural products research in the 1950s focussed on the
study of organisms easily accessible from intertidal and shallow subtidal
environments.'
In the 1960s, the introduction of SCUBA (self-contained underwater
breathing apparatus) allowed researchers to investigate organisms living at a
variety of depths and locations. Today, an even deeper range of collections is
achieved, to nearly 1000 m, by using manned submersibles (such as that
designed by the Harbor Branch Oceanographic Institution (Fort Pierce, FL)).
Clearly, collection of specimens at different depths increases the chance to
discover potential new drugs
.
Prior to 1970, most research on marine natural products was limited
basically to structural studies which included simple compounds such as mono-
and sesquiterpenoids. From the late 1970s, considerable collaborative effort
between marine chemists and pharmacologists has opened up biomedical
applications for marine natural products targeting various neurotoxins,
anticancer agents, antiviral agents, tumor promoters, and antiinflamatory
agents. In the 1990s, due to the major development and swift growth of
molecular biology, biochemical pharmacology, and genomic analysis, the
research has become much more advanced and a new trend in drug discovery
3
programs has introduced the utilization of receptor or mechanism-based assays
that evaluate very specific and selective activity:1'5
Chapter I of this thesis will provide a brief overview of selected fields
including marine microorganisms, and natural products from cyanobacteria,
followed by oxylipin chemistry. Chapters II, Ill, and IV respectively will describe
in detail the discovery of secondary metabolites isolated from the marine
cyanobacteria, Lyngbya majuscula and an Indonesian red alga, Vidalia sp. The
following chapter V will focus on the quantitative analysis of oxylipin products
from gametophyte cell cultures of a marine brown alga Laminaria saccharina.
Chapter VI will concern semisynthesis of oxidized anandamide derivatives, new
ligands for the cannabinoid receptor.
Marine Microorganisms
Microorganisms have had a significant influence on the development of
therapeutic science since the finding that they not only cause infections, but
also biosynthesize metabolites that can heal infections.4 Since the discovery of
penicillin in 1929, microbial drug discovery focused on soil-derived bacteria
and fungi has been the subject of intensive study that has led to the isolation of
approximately 50,000 natural products.6 More than 10,000 of them are shown
to have biological activity and over 100 microbial products are approved for use
today as chemotherapeutic agents in the treatment of human and animal
diseases.
4
Due to recent major advancements in screening methods and molecular
biology, the study of microorganisms is recognized as a notable resource for
drug discovery. Examples are the discovery of the bacterial metabolite FK-506
(1), an important immunosuppressant agent, and the fungal metabolite
lovastatin (2), a compound for the treatment of high serum cholesterol.
Although soil derived microorganisms have provided a fruitful source for many
antibiotics and other bioactive metabolites, investigation of analogous
microorganisms found in the marine environment has not been explored. This
is mainly due to the lack of taxonomic data and difficulty in the isolation and
culture of marine bacteria and fungi. Since marine sediments resemble soil,
there must be huge diversity of bacteria present. In addition, the surfaces and
internal spaces of marine plants and invertebrates provide microorganisms with
unique habitats which are rich in nutrients and organic substrates.6
1
OCH3
2
5
Early studies on marine bacteria indicated the production of antimicrobial
agents.' The first effort to elucidate marine bacterial metabolites revealed a
highly brominated pyrrole antibiotic (3) isolated from the genus Alteromonas by
Burkholder and co-workers. This seawater-derived bacterial compound was
shown to have a high in vitro antibiotic activity against Gram-positive bacteria
with minimum inhibitory concentrations (MICs = 6.3 200 ng/mL). However,
further investigation showed that the compound (3) was not active in in vivo
assays. Other studies revealed that it possessed some antitumor activity.8
Oncorhycolide (4), a structurally unique lactone, was produced by a seawaterderived bacterium obtained from a specimen taken from around a Chinook
salmon (Oncorhyncus tshawytscha) net-pen farm.9
4
A pioneering and comprehensive study of marine bacteria has been
initiated by the Fenical group at the Scripps Institution of Oceanography. His
group has explored the isolation and cultural requirements of marine bacteria.
Intensive efforts to obtain chemically novel bioactive compounds from a deepseawater bacteria have led to the isolation of a novel class of cytotoxic and
antiviral macrolides, the macrolactins A (5) and B (6).b0 This Gram-positive
6
bacterium, which could not be identified by any taxonomic methods, grew
slowly in a salt-based medium. The aglycon, macrolactin A (5), inhibits B16
F10 murine melanoma cancer cells in in vitro assay (IC50 = 3.5 ug/mL) and
shows strong inhibition of mammalian Herpes simplex viruses (types I and II)
(IC50 = 5.0 gg/mL) as well as human immunodeficiency virus (HIV) (IC50 = 10
gg/mL).
CH3
5
CH3
6
Chemical investigation of a cultured marine-derived actinomycete of the
genus Streptomyces, collected from the surface of a Sea of Cortez gorgonian
octocoral Pacifigorgia sp., resulted in the isolation of the novel metabolites
octalactins A (7) and B (8)." Octalactin A showed cytotoxicity against murine
and human cancer cell lines (IC50 = 7.2 ng/mL, 500 ng/mL, respectively).
Octalactin B is devoid of cytotoxic activity, suggesting that the epoxy
functionality is necessary for this activity. In a continuing investigation,
7
0
0
H3C"ss'
0
'
CH3
'''
H3C"'s
'o
QH
CH3
CH3
CH3
8
7
H
`CH3
%,, H 0
(3)\HH3ce
Ws'
N
H
K.,
OH
CH3
CH3
H3C4:-­
H
CH3
10
the Fenical group reported work with an actinomycete of the genus
Streptomyces isolated from the surface of the jellyfish Cassiopeia xamachana.
This surface associated bacterium produces in culture the complex bicyclic
depsipeptides, salinamides A (9) and B (10).12 These compounds possess
antibiotic properties with selective inhibition against Gram-positive bacteria.
8
However, the study of marine bacteria has given rise to another issue.
Microorganisms involved in specific symbiotic relationships with
macroorganisms may be responsible for the origin of some of the bioactive
metabolites that have been isolated from such organisms. Several studies
have demonstrated that some well-known marine toxins are in fact produced by
microorganisms. The potent neurotoxin, tetradotoxin (TTX, 11) , has been
known for many years to cause cases of fatal food poisoning around the world
and was believed to be produced by pufferfish of the family Tetraodonitidae. A
recent study revealed that TTX is truly produced by various unicellular marine
bacteria.'15 Similarly, saxitoxin (12) and its derivatives, the well-known
causative toxin of paralytic shellfish poisoning (PSP), have been known to be
produced by several dinoflagellates.16 Although this fact has been generally
accepted, another study reported that the true producer of 12 is a marine
bacterium.' Since the discovery that 11 and 12 block the nerve membrane
sodium channel, both toxins have become valuable tools for the study of
neurophysiology and neuropharmacology.
+
NH2
HOH
12
9
Bioactive Metabolites from Cyanobacteria
Cyanobacteria serve as an example of an evolutionary stage in the
progress from primitive prokaryotes to eukaryotes. Cyanobacteria are similar to
bacteria in their lack of organized nuclei and in their way of cell division, but are
closer to higher plants in their ability to carry out oxygenic photosynthesis.'
Members of the division Cyanobacteria appear ubiquitously, including
freshwater, ocean, damp soil, pond, polar icecap, and even hot spring habitats.
They are individually microscopic but some species aggregate to form
macroscopic colonies.
It has long been recognized that cyanobacteria produce toxins. Francis
in 1878 reported in Nature the first intellectual delineation of the potential
poisons produced by cyanobacteria, and livestock that drink at the surface of
the water contaminated by cyanobacteria may become ill and even die.'
Freshwater cyanobacteria produce numerous toxic metabolites.
Microcystis aeruginosa, which belongs to the order Chroococcales, is a more
primitive species than filamentous cyanobacteria such as Lyngbya majuscule.
Study of blooms of this alga in eutrophic lakes have led to the characterization
of microcystin (13)20 which was later designated microcystin-LR.21 This
causative agent, a cyclic heptapeptide, is a powerful hepatotoxin which causes
dose dependent chronic liver injury resulting in death in animals,22 and severe
liver damage in humans.'
10
13
HN
NH2
CH3
14
HN
NH2
Extraction of the freshwater cyanobacterium Nodularia spumigena
collected in the Baltic Sea and in New Zealand,' gave nodularin (14), a
microcystin-related toxin which has turned out to be a liver carcinogen.25
Hapalosin (15), a novel cyclic depsipeptide, was isolated from the terrestrial
cyanobacterium Hapalosiphon welwitschii.26 This metabolite was found to be a
lead compound for a new class of potential anti-multidrug resistance (MDR)
agents. Probably one of the most important discoveries of bioactive marine
natural products is cryptophycin A (16).27 Cryptophycins are one of the largest
11
class of cyanobacterial secondary metabolites with 38 members. Cryptophycin
A, the most active compound, was originally reported with gross structure as an
antifungal agent by a group at Merck.
16
However, the Merck group decided not to study this compound further due to its
extreme toxicity, an undesirable property for use as an antifungal agent.
Cryptophycins were later revealed to be powerful antitumor agents exhibiting
exceptional activity against a broad spectrum of solid tumors when inoculated
into mice.28 Recent studies have demonstrated that cryptophycins are
antimitotic agents which depolymerize microtubules and interact with the Vinca
alkaloid binding domain of tubulin.29
12
On the other hand, marine cyanobacteria have also proved to be the
source of many important bioactive secondary metabolites. The highly proinflammatory prenylated cyclic dipeptide, lyngbyatoxin A (17), was isolated from
toxic strains of Lynbya majuscula collected in Hawaii and shown to be the
causative agent of swimmer's itch.'
18
Majusculamide C (18), a novel cyclic nonadepsipeptide, was isolated
from a deep water variety of Lynbya majuscula found in the Marshall Islands.'
This peptide is a potent cytotoxin and showed a broad spectrum of antifungal
activity toward plant pathogens. A Caribbean cyanobacterium, Hormothamnion
enteromorphoides, is a producer of hormothamnin A (19), another cyclic
13
undecapeptide with moderate ichthyotoxic activity (LD50 - 5 j.tg/mL to the
goldfish Carassius carassius).32 Notably, this peptide is closely related to the
laxaphycin A (20) , an antifungal and cytotoxic cyclic peptide, isolated from
Anabaena laxa (Nostocaceae), in both structural and biological features.33
19
20
Oxylipins from Marine Algae
During the past three decades, many classes of marine organisms have
produced eicosanoid-type natural products. The eicosanoids, a term referring
to C-20 polyunsaturated fatty acids, are derived biosynthetically from the three
most generally occurring C-20 polyunsaturated fatty acids found in
invertebrates, vertebrates, and plants: arachidonic (eicosatetraenoic, 20:3 co6),
14
dihomo-gamma-linolenic (eicosatrienoic, 20:4 w6), or eicosapentanoic acid
(20:5 w3).34
The eicosanoids play important roles in fundamental physiological
regulation at the cellular level, and are particularly involved in various
pathophysiological responses in mammals. For example, prostaglandins
influence renal salt and water excretion by alterations in renal blood flow and by
direct effects on renal tubules.35 They also participate in neoplastic diseases,
regulate cellular and humoral immunity, as well as contract or relax many
smooth muscles beside those of the vasculature.36 Leukotrienes cause
hypotention, and contraction of most smooth muscles.' The major fatty acid for
eicosanoid formation in mammalian cells is arachidonic acid which is released
by the action of phospholipase A2.38 Arachidonic acid is metabolized through
one of three notable pathways: the prostaglandin H synthase (cyclooxygenase)
pathway, the lipoxygenase pathway or the cytochrome P450. This leads to
three groups of metabolites, the cyclooxygenase products (prostaglandins and
thromboxanes), the lipoxygenase products (hydroxyperoxy and
hydroxyeicosatetraenoic acid, leukotrienes, and lipoxins), and cytochrome
P450 products (hydroxyeicosanoids). Figure (1.1) presents an overview of the
biosynthesis of the arachidonic acid cascade.
Weinheimer and Spraggins in 1969 reported the pioneering discovery of
the eicosanoid, 15-R-prostaglandin A2 (21), from the gorgonian Plexaura
15
Membrane Phospholipids
Phospholipase A2
N.,
COOH
Arachidonic Acid
PGH Synthase
5-
12-
Lipoxygenase
Lipoxygenase
PGH2
/ \
Prostaglandins
5-HPETE
12-HPETE
HHT's
Leukotrienes
Hepoxilins
P450
Epoxide
15-HPETE
/\
15-HETE
Thromboxane
Figure 1.1
Cytochrome
15-
Lipoxygenase
Leukotrienes
A
HETE's diHETE's
diHETE's
Lipoxins
The Arachidonic Acid Cascade. PG = Prostaglandins;
HHT = Hydroxyheptadecatrienoic acid; HPETE/HETE =
Hydroperoxy/Hydroxyeicosatetranoic acid
diHETE = dihydroxyeicosatetranoic acid;
Biosynthesis of the products of arachidonic acid in mammals.
PG = Prostaglandins; HHT = Hydroxyheptadecatrienoic acid; HPETE/HETE =
Hydroperoxy/Hydroxyeicosatetranoic acid; diHETE = dihydroxyeicosatetranoic
acid
Figure 1.1
homomalla.' Subsequently, more prostaglandins4"1 were isolated from P.
homomalla and from a soft coral, respectively. Since the success in finding
eicosanoids and prostanoids in corals, significant efforts were undertaken to
explore the discovery of this class of chemistry in other marine organisms. As a
result, a large number of oxidized metabolites of arachidonate and eicosanoate,
16
along with fatty acids of various chain lengths (e. g., C18, C20, C22) have been
isolated from marine organisms. In order to cumulatively describe all of the fatty
acid metabolites above, the new term "oxylipin" has been created for including
all chain lengths of oxidized eicosanoid-related compounds.42
Since the first discovery of oxylipins from algae, namely PGE2 (22) and
PGF2c, (23)43, from the red alga Gracilaria lichenoides, the marine algae have
continued to yield an impressive array of oxylipins. Much of the pioneering
work on marine oxylipins has been carried out by Gerwick and co-workers. The
diversity and novelty of this class of chemistry isolated from marine organisms
have been comprehensively reviewed by Gerwick.44-48
COON
21
OH
COON
22
OH
Relatively little has been reported on the production of oxylipins from
cyanobacteria. The first oxylipin to be obtained from a cyanobacterium was the
oxygenated fatty acid metabolite (24) from Lyngbya majuscule (Cardellina et
al.).49 This compound (24) contains the C14 methoxytetradecenoic acid which
17
has been a common lipophilic constituent of many secondary metabolites found
in marine cyanobacteria from the Pacific and Caribbean oceans. Similarly,
malyngic acid (25), 9(S), 12(R), 13(S)-trihydroxyoctadeca-10(E), 15(2)-dienoic
acid, was isolated from L. majuscule.' The structure of malyngic acid (25) was
determined based on chemical and spectral data for the natural product and by
its conversion to trihydroxystearic acid and degradation to deoxy-D-ribitol.
QCH3
COOH
24
OH
25
The red algae (Rhodophyta) appear to be a prolific source of numerous
unusual oxylipins. Over the last decade, the Gerwick group has demonstrated
the ability of these algae to produce novel oxylipins. Two cyclopropyl and
lactone containing oxylipins, constanolactones A (26) and B (27), were isolated
from a red alga Constantinea simplex.' The absolute stereochemistries of
these metabolites were determined by chiral separation of degraded fragments,
CD analysis of the separately formed mono and dibenzoate derivatives and 1H
NMR studies.
18
HO"""..
HO""
27
26
Another three new oxylipins (28-30), structurally similar dihydroxy
oxylipins, were discovered from an Oregon coast red alga Farlowia moffis.'
COON
COOH
OH
28
OH
OH
29
30
Study on a Washington red alga Rhodymenia pertusa has led to the
isolation of two major metabolites, dihydroxy-eicosanoic acid (31) and
dihydroxy-eicosapentaenoic acid (32).46 The absolute stereochemistry of this
diHETE and diHEPE was established by optical comparison to the known four
diastereomers produced by synthesis.53 (-) Jasmonic acid (33) and 7+)­
isojasmonic acid (34) were discovered from a Black Sea red alga
19
COOH
COOH
32
COON
0
34
Gelidium latifolium. In plants, these compounds play an important role as
phytohormones, elicitors, and signal transducers.' They are biosynthesized
through several enzymatic transformations.
The green algae (Chlorophyta) have also produced some oxylipins but
are generally less rich in C20 fatty acids. Two oxylipins, 12R-hydroxy-9(Z),
13(E), 15(Z)-octadecatrienoic acid (35) and 9(S)-hydroxy-10(E), 12(Z), 15(2)­
octadecatrienoic acid (36), were isolated from the freshwater unicellular green
algae Duna liella
COOH
COON
He
35
36
Two C10 metabolites (37-38) and five other C18 oxylipins (39-43) were
obtained from an Oregon marine green algae Acrosiphonia coalita.56 The EZE­
20
conjugated triene (38) was found to rapidly convert to the EEE-conjugated
triene (37) under UV-irradiation. Observing this fact led to the hypothesis that
the EZE-conjugated triene (38) is the true natural product and metabolite (37)
is an artifact produced during isolation.
COOH
COOH
COOH
It was not until 7 years ago that brown algae (Phaeophyta) were found to
be prolific producers of this class of chemistry. Examination of the temperate
brown alga Cymathere triplicate collected from the Pacific Northwest revealed
the C18 metabolite, cymathere ether A (44), and C20 metabolite, cymathere
ether B (45).57 C. triplicate is an intertidal edible kelp in Washington and has an
unique cucumber odor which arises from 2E-nonenal and 2E, 6Z-nonadienal,
21
both produced in cucumbers by the sequential actions of lipoxygenase and
hydroperoxide lyase on fatty acids.' The absolute stereochemistries of these
metabolites were predicted on the basis of biogenetic rational, CD analysis of
w
the p-bromobenzoate derivative and co-isolation of the known oxylipins, 12S­
HETE and 10S -NOTE.
00
HOliii....­
COOH
44
HOmi....­
COON
COON
Extracts of another temperate brown alga, Egregia menziesii, collected at
Seal Rock State Beach, Oregon contained three new chlorinated and
carbocyclic oxylipins, egregiachlorides A-C (46-48).59 A biogenetic pathway of
22
egregiachlorides A (46) was proposed where a 13-lipoxygenase introduces a
hydoperoxide into the C18 precursor followed by the formation of an epoxy
allylic carbocation. This intermediate cyclized to form a new cyclopentyl cation
which was attacked by chloride anion at w3 carbocation to yield
egregiachlorides A (46).
23
CHAPTER II.
CURACINS B AND C, NEW ANTIMITOTIC NATURAL PRODUCTS FROM
THE MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
ABSTRACT
Two new and potent antimitotic metabolites, curacins B and C, were
isolated from a Curacao collection of Lyngbya majuscula. In addition, four
curacin A analogs were prepared by semisynthetic efforts. The structures of the
new curacins, along with the curacin A analogs, were determined by detailed
spectroscopic analysis in comparison with curacin A. The absolute
stereochemical configurations of curacins B and C were deduced to be 2R, 13R,
19R, 21S from interconversion with curacin A via a thermal rearrangement. The
biological and biochemical properties of the new natural products and synthetic
derivatives of curacin A were examined.
24
INTRODUCTION
Different isolates of the marine cyanobacterium Lyngbya majuscula
produce an amazingly wide array of biologically-active secondary metabolites;
in this respect, reminescent of the rich biosynthetic capacity of some species of
Streptomyces. The potent tumor promoters aplysiatoxin (1) and
debromoaplysiatoxin (2),60 the antifungal lipopeptides majusculamides A (3)
and B (4),61 and the structually unusual pyrrolic compounds, pukeleimides A-F
(5-10)62'63 are examples of the range of metabolites isolated from this organism
(Figure 11.1).
Our survey for the discovery of bioactive secondary metabolites from
marine algae has focussed on a Curacao (Caribbean) collection of Lyngbya
majuscula which has resulted in the isolation of several novel compounds.
Antillatoxin (11) is a powerful ichthyotoxic cyclicdepsipeptide from L. majuscula
collected at Curacao (Caribbean).64 The absolute stereochemistry of 11 was
determined on the basis of a combination of analysis by chiral phase TLC,
molecular modeling, NMR and CD spectroscopy. An extract of L. majuscula
from Playa Kalki, Curacao was the source of kalkitoxin (12), a thiazoline
containing lipid.65 This compound (12) shows potent ichthyotoxicity (LC50 =
0.05 lig/mL to Carassius auratus).
During the past five years, a major effort in our laboratory has been the
isolation, structure elucidation, and biological evaluation of curacin A (13).66,67
25
CH3
CH3
o;;;-------o
OH
0
4cH3
T
T
Q
CH3
8H3
0
OH
1. R= Br
OH
2. R=H
H3C
CH3
Figure II. 1 Structures of selected secondary metabolites from Lyngbya majuscula.
26
H3C-N
H
H
i
CH3
CH3
0
CH3
CH3
0
12
In initial screening at Syntex, a crude extract of L. majuscula was highly
cytotoxic against a monkey Vero cell line. Following evaluation of this extract in
the National Cancer Institute (NCI) in vitro 60-cell line screen which showed a
potent antiproliferative and cytotoxic activity (selective for colon, renal and
breast cancer-related cell lines), curacin A was isolated as the major bioactive
species. Further study has revealed curacin A's ability to inhibit tubulin
polymerization, inhibit colchicine binding to tubulin, and inhibit cell
proliferation. 67
In continuing investigations of this organism for minor metabolites of
related structure, we have isolated and characterized two new natural products,
curacins B (14) and C (15).68 Both of these are toxic to brine shrimp,
demonstrate strong cytotoxicity against murine L1210 leukemia and human
CA46 Burkitt lymphoma cell lines, and in the National Cancer Institute in vitro
27
60-cell line assay (Figure 11.2), show a potent antiproliferative activity to many
cancer-derived cell lines in a manner characteristic of antimitotic agents.
RESULTS AND DISCUSSION
Isopropanol preserved Lyngbya majuscula collected from Curacao was
extracted (CHCI3 /Me0H, 2:1) and chromatographed over silica gel to give a
mixture of curacins A-C (13-15) contaminated by fatty acids and phytol (Figure
11.3). These contaminents were removed by methylation (CH2N2) and acetylation
(Ac20/pyridine) of the curacin-containing fraction followed by NP-HPLC to yield
two peaks. The first peak was curacin A (13) (10.3%) while the second was a
mixture of two new curacins by1H NMR spectroscopy and GC-EIMS analysis.
Repeated RP-HPLC (C-8) of this latter fraction eventually gave pure curacin B
(14) and C (15) as 4.7 % and 0.3 % of the extract, respectively.
Compounds 14 and 15, both analyzing for C23H35NOS by HRFABMS
(positive ion, [M+H]+), displayed 1H NMR and 13C NMR bands indicative of four
17
CH3
H2
9
OCH3
S
7
13
OCH3
22
CH3
18
H
20
CH3
OCH3
HC
15
CH3
co
lethality).
50%
for
required
concentration
LC50,
inhibition;
growth
total
for
required
concentration
TGI,
inhibition;
growth
50%
for
required
concentration
(G150,
screen
vitro
in
line
cell
60
Institute
Cancer
National
the
from
obtained
mixture
(95:5)
C
and
B
curacins
of
graphs
Mean
11.2
Figure
1
4
1
1
1
.3
1
I
2 I
.1
41
.2
41
1
.3
I
I
4
la
343
^
La IN
1
1
I
4 1
31
.1
43
J
4 1
4 1
.1
.2
.3
I
Doh
MD
MO
4.23
400 020 7.11
55V
.---­
..
-
--..--­
---------
.......
...-....-.--...­
<
AIDA.
4017
>
a a
-741
.140 440
111.349
44 44
4.0
434
1.470
144411431
on
420
0
400
4
CM
143
4
MCP7MDR.R1111
a
40
.31
.73 100
40/14410231/A1CC
a
410
4.00
0
---.....-.....
Caw
Sam
...........-......-----.....-......-.......-...--­
-...­
........
--
.....
---...-­
----...--
...............
­
.....
.,....-
1
4.11
14CP7
441
4.74
131.1443
o
4.00
ICJ
-------
--
.....
.----­
-...----------­
-....---­
..............
....-­
-.----.-----­
Cao
10501
44D
a
4.0
R30.451
4.40
4.7
.441
420
23
JO
NNE
447
Ma
420
ACM
40
CAK1.1
412 JIM
4.
1
4.33 473
81112C
owEmomor
4.113
4.0 403
0
7.00
144.31 11440
a
403
400
MI
4.41
7364
-
------.­
a-----
-
..
-----
--------
Cam
hal
o
4431
OVCA1741
4.4
AM
4.00
4.00
i
41.03
OVC/414
,
443
........
..--.--------.
.-­
--..---,---­
-
--­
--.-.--­
UACC.41
114144.4
4.30
443
nom
4.03 -7.20
1.1ACC.4.17
4.44
sc-mstAs
4.0)
4.
403 4.0)
s 000
0Moneerml
-.--...--....­
1OROVI
442 400
4.40 4.14
OVCARol
40
a
4.31
OVCAR4
0 0 0
40
Ea
443
4401.3
am
4/73
NI
4.2
4.1.0114
4
4.00
4
RAVI
1.32
420 420 4131
)
4.1 <La
4.11
104.3411L4
41
....-..-.-...
---
......
.4...o..
-
-
.......
-
.
-
-
Mama.
40)
a .0
320
0 0
4.14 4.05
imm.
gm
443
470.11
44
411143
7.51 4.11
1/231
.
402 401
a
40
---
- ..
......
.
-...­
Caw
CRS
­
-.
-­
.....
1014
.741 403
4.03
40.393
s
4
1444
a
410
400
41
443
-7.73
403
<
4.00
.7.13
NCI,
1 III
RCT4
1.311
wrn
.7.34 .7.11
14M12
.4
Mali
4.1
1.,
443
1404901
203
COLO
-...­
-...­
......--..­
....-
Cana
han
AZ
.7
4.00
400
40
NCI4423
4.0
4.15
141441304
4.01 403
ma
14141400
434
4.
<
14141122
110043
110044
443 403
4.10
41
4.00
a
400
...
-
..
-
--
..
.­
war
401
-
­
---..---
.­
--------
-.
­
-
......
--­
.-
-­
.....
­
-
­
440
.
.
420
C
Las
Cen
Madan
L
MOM=
.4.0)
4.13
11124V14
4.00 4.00
..
.711 410
CO 4.44
14041120
4.01
4.00
4.30
4102
111.40070
>
OZR4C114
403 403
4.00 4.40
420 410
441 4.00 403 4.03
MOLT.
440 401
0011414301
La*
LCD
LC-14
Lei.
TOI
1,4.701
OM
OM
Log.
Una
hal/Con
29
Lyngbya majuscula (Curacao collection)
543 mgs of 2.39 gms total
Vacuum Chromatography
5% E/H
5%110%110%115%1201:Y0 30%150%175%175%100°AI
...._.._.,
314 2 mgs
Vacuum Chromatography
1% E/H
1c/011.5cY
1
1.5%1 2%
13% 14% 1
5°%
0110°% 01100cYcl
i
ca. 200 mgs
1
158 mgs
CH2N2
1
HPLC
4% EtOAc /Hexanes
minor
Curacin B & C
(29 mgs)
major
Curacin A
(44.2 mgs)
Ac20/Pyr.
HPLC, C-8 column
64% CH3CN/H20
Curacin B
20 mgs
Curacin C
1.1 mgs
Figure II. 3 Flow chart of fractionation of Lyngbya majuscula
H-8
H-9
11-7
Curacin A
144t,
I
Curacin B
5'5
6.0
5.5
5.0
4.t 5
4.0
3.5
3.0
25
n
Figure 114 Comparison of 1F1 NMR spectra of curacins A, B, and C.
Table 11.1
11-1 and 13C NMR data for curacins A (13), B (14), and C (15)
Curacin A
Curacin B
Curacin C
Position
1H(Jin Hz)a
la
2.79 (dd, 10.7, 10.0)
3.09 (dd, 10.7, 10.0)
5.11 (m)
5.69 (dd, 10.7, 8.7)
5.45 (bdt, 10.7, 7.1)
2.11 (m)
2.11 (m)
5.58 (bdt, 15.0, 7.0)
6.38 (dd, 15.0, 10.8)
6.02 (d, 10.8)
b
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20a
2.23 (m)
1.70 (m)
3.09 (m)
2.26 (m)
5.87 (m)
5.07 (m)
1.70 (s)
13cb
39.96
74.31
131.32
130.84
28.14
33. 12
131.32
127.86
125.50
136.40
35.77
32.15
79.91
38.02
135.32
116.80
16.56
1H(Jin Hz)a
2.79 (dd, 10.7, 10.0)
3.09 (dd, 10.7, 8.4)
5.12 (m)
5.69 (bdd, 10.9, 7.3, 0.9)
5.45 (bdt, 10.9, 7.3, 0.9)
2.11 (m)
2.26 (m)
5.40 (dt, 9.4, 7.4)
6.35 (m)
6.35 (m)
2.23 (m)
1.69 (m)
3.09 (m)
2.26 (m)
5.87 (m)
5.07 (m)
1.69 (s)
168.36
1.65 (m)
0.75 (td, 8.2, 4.2)
1.20 (m)
21
0.98 (m)
22
1.21 (d, 6.3)
2') (OMe) 3.18 (s)
0.75 (td, 8.1, 4.4)
15.98
12.39
56.31
1.21 (m)
0.99 (m)
1.22 (d, 6.3)
3.19 (s)
a300 MHz in C6D6, b75 MHz in C6D6
1H (J in Hz)a
13Cb
39.99
2.80 (dd, 10.5, 10.1)
3.09 (dd, 10.5, 8.6)
5.11 (m)
5.69 (dd, 10.5, 8.9)
5.45 (m)
2.11 (m)
2.11 (m)
5.57 (bdt, 15.0, 6.6)
8.51 (dd, 15.0, 10.8)
5.97 (d, 10.8)
39.94
74.33
131.40
130.91
28.15
27.79
128.83
125.84
120.74
138.73
36.26
32.26
79.95
38.06
135.32
116.84
16.39
1.65 (m)
20.13
14.24
15.99
12.35
56.37
74.31
131.30
1.70 (m)
130.85
28.16
33.12
131.30
127.67
126.42
136.62
28.16
32.16
79.70
37.85
135.23
116.84
23.65
169.34
20.11
0.75 (td, 8.1, 4.1)
14.21
2.38 (m)
1.62 (m)
3.09 (m)
2.24 (m)
5.85 (m)
5.07 (m)
1.74 (s)
168.42
19.87
13.82
b
13cb
1.23 (m)
0.98 (m)
1.22 (d, 6.1)
3.20 (s)
15.98
12.31
56.23
32
bonds and one C=X functionality. Therefore, in consideration of the molecular
formula, both contained two rings in similarity to curacin A (13). Comparison of
the 1H and 13C NMR features of the two new compounds with that of curacin A
revealed a close comparability of all three metabolites (Figure 11.4 and Table II.
1). The only NMR regions of significant difference between them was in that
assigned to the chemical shifts and coupling patterns for the C-8 to C-11 diene
and the associated C-17 methyl group. Hence, it was concluded that curacins B
(14) and C (15) represented geometrical isomers of curacin A in this diene
region of the molecule.
In our prior work on curacin A (13),66 the geometry of the C-8 to C-11
diene was established as 7(E), 9(E) on the basis of a 15 Hz coupling between
H-7 and H-8 (Table II. 1) and a strong nOe interaction between the C-17 methyl
group and H-8. Correspondingly, in curacin B (14), a Z geometry for the C-7
olefin was deduced by observing a 9.4 Hz coupling between H-7 and H-8, an E
geometry for the C-9 olefin by observing the upfield 13C chemical shift for C-17
methyl (16.4 ppm)69 and an nOe between C-17 methyl and H-8 (Figure 11.5). In
curacin C (15), the combination of a 15.0 Hz coupling between H-7 and H-8
with 1H and downfield 13C nmr chemical shifts for C-17 methyl (23.6 ppm)
indicated it was the 7(E), 9(Z) isomer.
The absolute configuration of 14 was defined through thermal
interconversion with 13,7° the absolute configuration of which we recently
defined as 2R, 13R, 19R, 21S (Figure 11.6). Refluxing a 95:5 mixture of 14 and
15 in 2,2,4-trimethylpentane (99° C) for 3 h yielded a 2:1:1 mixture of all three
33
Figure IL5 Geometrical configuration of conjugated diene of curacin B (14)
determined by NOESY in CDCI3.
34
curacins (13-15) by GC-EIMS (Figure 11.7). It is intriguing that all curacins have
unusual peak shapes in GC. Although there is no clear explanation for this, it is
likely that the unstable cis cyclopropyl moiety of the structure might undergo
thermal rearrangment during the GC-MS process. This prediction is based on
the experimental observation where 19-epi-curacin A (19) possessing the trans
cyclopropyl ring showed a single peak in GC. The compound 13 produced in
this experiment was purified by HPLC and had identical spectral properties to
authentic 13 (1H-NMR, GC-EIMS, UV) and a comparable optical rotation ([425
+77°, c = 0.1; lit. [425 +86°, c= 0.64, CHCI3). Therefore, curacin B (14)
possesses the same 2R, 13R, 19R, 21S stereoconfiguration as curacin A (13).
/
OCH3
OCH3
Curacin A (13)
Isooctane, reflux (99 °C)
OCH3
CH3
CH3
Curacin B (14)
\
H3C z
E
Curacin C (15)
Figure II. 6 Thermal isomerization of curacins A, B, and C
Additionally, because HPLC-purified 14 and 15 had comparable optical
rotations (see Experimental), it is likely that 15 also possesses 2R, 13R, 19R,
21 S absolute stereochemistry.
CH3
q. n.ance
TIC:
1.
/0-5-2.0
200000
. mice
TIC: 111-38.0
1000000
800000
150000
600000
1
100000
400000
50000
-
200000
8.00
10.00
12.00
14.00
0
16.00
18.00
ice ->
curacins B and C (95:5), before refluxing
a)
A
......18.
0
Imo ->
TIC: I/-51-1.0
kNiNeese
8.00
10.00
14.00
16.00
18.00
20.00
22.00
refluxing B and C in TMP for 1 h
once
700000
12.00
TIC: III-38-2.0
600000
600000
500000
500000
400000
400000
300000
300000
200000
200000
C­
100000
0
Time ->
8.00
c)
10.00
C­
100000
0
12.00
14.00
16.00
18.00
refluxing B and C for 90 min.
20.00
ice ->
8.00
10.00
12.00
14.00
16.00
18.00
refluxing B and C for 2 h
Figure 11.7 GC-MS traces for thermal isomeration of curacins A, B, and C.
20.00
36
Three experiments were conducted to confirm that 14 and 15 are true
natural products of this cyanobacterium and not extraction/isolation artifacts.
First, a cold extraction of the preserved cyanobacterium was shown by GC­
EIMS to contain the same proportions of 13, 14, and 15 as obtained from the
above warm extraction. Second, direct heating of pure 13 in CH2Cl2 -MeOH
(2:1) at 35° C for 10 h did not produce detectable quantities of 14 or 15. And
third, pure 13 refluxed in aqueous CH2Cl2 -MeOH (2:1) for 90 min. also did not
give 14 or 15. It should be noted that in these latter two experiments, 13 was
heated at a temperature lower than that shown to affect isomerization of the
14/15 mixture into 13.
In the National Cancer Institute (NCI) in vitro 60-cell line screen, a 95:5
mixture of curacins B and C showed potent antiproliferative activity to many
cancer derived cell lines in a manner characteristic of antimitotic agents.68'71
Some structurally diverse compounds, most of which originated from natural
products, have the ability to bind to tubulin or microtubules and inhibit cell
proliferation by acting on spindle microtubules. These microtubule-active
agents can be divided into two classes: those that inhibit the polymerization of
spindle microtubules or depolymerize existing spindle microtubules and cause
the loss of cellular microtubules, and those that stablize microtubule
polymerization and cause the formation of intracellular microtubule bundles.
Example of the former includes colchicine, the vinca alkaloids, rhizoxin,72
dolastatins,73 and curacins. The latter includes taxol, related taxanes and
epothilones.74 Many antimitotic drugs bind to the colchicine (16)-binding
37
0
11
NHCCH3
H
140
N
H
SS
le O
H3COOC
N
16
OCH3
17
CH3
OCOCH3
HO COOCH3
CH3
H2
OCH3
18
SCH3
OCH3
NH
(:)
21
CH3
H2
H3C
CH3
.........0 H2
S -S
OCH3
0
NH
HN
CH3
H3C
0
OCH3
22
Figure II. 8 Natural antimitotic agents and synthetic derivatives of the curacins
38
domain and inhibit the binding of colchicine to tubulin while others bind to the
vinblastin (17)-binding domain and inhibit the binding of vinblastine to tubulin
(Figure 11.8). Most known colchicine-type tubulin inhibitors possess a common
topographical similarity which provide a structure-activity relationship of that
binding site. Curacin A is an example of the former case that inhibits tubulin
assembly by binding to the colchicine site on tubulin.' This is intriguing
because the structure of curacin A has little similarity to other drugs that bind to
the colchicine-site. Consequently, in order to better understand how curacin A
resembles these other drugs and to get further insight into the molecular
mechanism of tubulin-ligand interaction at this site, some curacin A derivatives
(18-22) were prepared.'
Heat treatment of curacin A produced cleavage of the cyclopropyl ring to
yield cyclopropyl ring opened (CPRO) curacin A (18). CPRO-curacin A was
obtained as an optically active pale yellow oil which gave a molecular formula
of C23H36ONS by HRFABMS. The 13C NMR and HMQC identified resonances
for all 23 carbons. The 1H NMR and COSY showed features similar to that of
curacin A except for the absence of cyclopropyl resonances in 18 and extra
olefin signals at 8 4.96 (ddd, J . 10, 1.7, 1.1, Hz, H-22b), 8 5.02 (ddd, J . 17, 1.7,
1.7 Hz, H-22a), and 8 5.76 (m, H-21) (Figure 11. 9 and 10).
Another curacin A derivative, 19-epi-curacin A (19) was produced
accidentally by the treatment of curacin A with acetic anhydride and pyridine. In
fact, the original intention of this reaction condition was to remove the undesired
impurity, phytol, from the mixture of curacins B and C.
j
LI
6.0
5.5
5.0
4.5
4.G
PPM
3.5
3.0
2.5
Figure 11.9 1H NMR spectrum of cyclopropyl ring opened curacin A (18) in C6D6
2.0
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
PPM
Figure 11.10 1H -1H COSY spectrum of cyclopropyl ring opened Curacin A (18) in
C6D6
0
6.5
6.0
5.5
5.0
4.5
4.0
1
3.5
PPM
3.0
2.5
2.0
1.5
Figure 11.11 1F1 NMR spectrum
of 19-epi curacin A (19) in C6D6.
1.0
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
PPM
Figure 11.12 1H-1H COSY of 19-epi curacin A (19) in C6D6
1.0
6.5
6.0
5.5
5.0
4!5
4.0
3.5
PPM
3.0
2.5
2.0
1.5
1.0
Figure 11.13 11-1 NMR spectrum of thiazoline ring opened curacin A (20) in C6D6
10
11
ct)
.5
.04
1I
140
160
180
7.0
6.5
I
6.0
/0.
5.5
5.0
4.5
4.0
PPM
3.5
3.0
2.5
4
2.0
1.5
1.0
1
Figure 11.14 HMBC plot of thiazoline ring opened curacin A (20) in C6D6
PPM
6.5
1
6.0
5.5
1
5.0
4.5
4.1
0
3.I 5
PPM
3.0
2.5
2.0
Figure 11.15 'H NMR spectrum of derivative 21 in C6D6
1.5
1.0
.A.2 Athdukl,
6.5
6.0
1
5.5
,
5.0
1
4.5
4.0
3.5
3.0
2.5
PPM
Figure 11.16 11-I NMR spectrum of curadisulfide A (22) in C6D6.
2.0
1.5
47
Compound 19 was prepared as a pale yellow oil. The molecular ion [M +
was observed at m/z 374, which is consistent with a molecular formula of
C23H35ONS by HRFABMS. Its 11-1 NMR spectrum showed essentially the same
signals as those present in curacin A. The observed differences between the
1H NMR spectrum of 13 and 19 were consistent with 19 being the C-19 epimer
of curacin A, where only the methyl and methine resonances of the cyclopropyl
ring were shifted upfield or downfield (e. g., H-19 at 6 1.65 to 6 1.55; H-20a at 8
0.75 to 8 0.43, H-20b at 6 1.20 to 8 1.34; H-21 at 6 0.98 to 8 1.49; H-22 at 6 1.21
to 8 0.85) (Figure 11.11 and Figure 11.12).
Thiazoline ring opened (TRO) curacin A (20) was obtained from the
treatment of curacin A with acetic anhydride and D20. TRO curacin A (20) was
also obtained a pale yellow oil. This derivative exhibited a molecular formula of
C25H40NO3S based on its HRFABMS data. The IR absorptions at 1678 and
1650 cm-1 were attributed to an amide or thio ester carbonyl group. The
structure of compound 20 was identified by comparing its 1H and 2D NMR
spectra with those of curacin A (Figure 11.13 and Figure 11.14). The main
differences in the 11-1 NMR data of 20 and 13 were the presence in compound
20 of the methyl resonance (8 1.58) and the upfield shift of H19 (8 1.8), and the
downfield shifts of H20a (8 0.65), H20b (8 1.11), and H22 (8 1.12). In addition,
13C NMR spectroscopy revealed the presence of amide carbonyl (8 167.99) and
thio ester carbonyl (8 196.96) resonances in the derivative.
48
The cyclopropyl ring was further cleaved by treating 20 with methyl
lithium in THF, followed by methyl iodide to give the methyl thio ether derivative
of curacin A (21). Compound 21 was isolated as an oil that gave a molecular
ion at m/z 365 in GC-EIMS corresponding to a molecular formula of
C21H3502NS. Analysis of 1H- NMR data for 21 revealed the absence of
cyclopropyl and thiazoline rings, and the presence of amide carbonyl and
methyl thio group. These were further indicated by GC-EIMS by observing
fragments at m/z 318 ([M - SCH3] +), 304 ([M - SC2H5] +), 233 ([M - C5F1,00NS] +).
The 11-1 NMR spectrum showed signals at 8 1.55 (3H, s), 8 1.91 (3H, s), and 8
4.70 (1H, br) for the vinyl methyl, methyl thio, and amide protons, respectively
(Figure 11.15).
Curadisulfide A (22) was produced by treating 20 with sodium hydroxide
in methanol. The FABMS of 22 gave a [M + Hr at m/z 701. A detailed mass
spectrum revealed its dimeric nature by observing a fragment ion at m/z 305.
The 1H NMR spectrum of 22 was very similar to that of 20, except for the
absence of resonances for the methyl cyclopropyl moiety (Figure 11.16).
Using a purified tubulin preparation, these derivatives were examined, as
well as pure curacins A and B, for their effects on the inhibition of tubulin
polymerization, and radiolabeled colchicine binding to tubulin, and cytotoxicity
to a cancer cell line (Table 11.2).71 Curacin B exhibited strong inhibitory activity
on purified tubulin assembly and it was half as inhibitory of colchicine binding to
tubulin as that of curacin A. Curacin B was approximatly 10-fold less active in
49
Compound
Inhibition tubulin
Polymerization
IC50 ± S.D. (pM)
Inhibition colchicine Inhibition of MCF-7 Brine
Binding
Growth
Shrimp
`3/0 Inhibition
IC50 ± S.D. (M)
IC50 (nM)
Curacin A
0.72 ± 0.2
94 ± 2
0.038 ± 0.01
8
Curacin B
0.82 ± 0.2
56 ± 2
0.32 ± 0.06
38
Curacin C
2.3± 1
10 ± 2
3.6 ± 0.8
54
18
0.92 ± 0.2
83 ± 6
0.30 ± 0.2
19
0.77 ± 0.2
88 ± 11
0.09 ± 0.01
20
1.5 ± 0.4
9±8
4.2 ± 2
>1000
21
3.3 ± 0.8
12 ± 15
5.2 ± 3
>1000
22
120 ± 20
8
Colchicine
1.4
15
700
1.7 ± 0.09
Podophyllotoxin 1.6 ± 0.03
83
1.7 ± 0.09
Table 11.2 Biological activities of curacins on cell growth, tubulin polimerization,
colchicine binding, and brine shrimp toxicity.
its inhibitory effects on the growth of MCF-7 breast cancer cells than curacin A.
Curacin C turned out to be remarkably less active than curacin A in the tubulin­
based assay. In the inhibition of tubulin polymerization, it was almost 3-fold less
active and showed only weak inhibitory capability for the binding of
[3H]colchicine to tubulin. On the other hand, it had IC50 values almost 100­
timeshigher than that of curacin A to the growth of MCF-7 cancer cells. Based
on these results, it is apparent that geometrical isomerization of the C-7 and C-9
olefins has a pronounced effect on the ability of these compounds to inhibit
50
tubulin polymerazation, to inhibit colchicine binding to tubulin, and to increase
cytotoxicity. Cyclopropyl ring opened (CPRO) curacin A (18), unlike our
expectation, was slightly less active than curacin A as an inhibitor of both
tubulin polymerization and colchicine binding, but was 10-fold less cytotoxic
than curacin A on MCF-7 cancer cells.
The activities of 19-epi-curacin A (19) were basically identical to those of
curacin A. Thiazoline ring opened (TRO) curacin A (20) still possessed good
inhibitory properties of tubulin assembly and colchicine binding, but was about
9-fold less active than curacin A at inhibiting the growth of cancer cells.'
Surprisingly, the methyl thio ether derivative of curacin A (21) retained
good activity as an inhibitor of tubulin assembly as it was only 4-5 fold less
potent than curacin A.
However, compound 21 was also a very weak inhibitor
of colchicine binding to tubulin and of cancer cell proliferation.
Curadisulfide A (22) was the least active of the curacin A derivatives in
the inhibition of tubulin polymerization. Nevertherless, it retained some activity
as an inhibitor of both MCF-7 cell growth and colchicine binding.
EXPERIMENTAL
General Experimental Procedures. NMR spectra were recorded on
Bruker AM 400 and AC 300 spectrometers in the solvent specified. Chemical
shifts were referenced to solvent (C6D6 signal at 7.2 ppm for 1H NMR and 128
ppm for 130 NMR). Mass spectra were recorded on Kratos MS 50 TC and
Finnigan 4023 mass spectrometers. Gas chromatography/mass spectrometry
51
(GC/MS) was carried out utilizing a Hewlett-Packard 5890 Series II GC
connected to a Hewlett-Packard 5971 mass spectrometer. UV spectra were
obtained on a Hewlett-Packard 8452 A UV-VIS spectrophotometer and IR
spectra on a Nicolet 510 spectrophotometer. All solvents were distilled from
glass prior to use.
Collection, Extraction, and Isolation Lyngbya majuscula was collected
on 13 December 1991 from CARMABI beach, Curacao and preserved in i-PrOH
at low temperature until workup. Isopropanol preserved Lyngbya majuscula
was extracted with CHC13/Me0H (2:1). Gradient vacuum liquid chromatography
of the crude extract (543 mg) with EtOAc /hexanes gave a mixture of curacins
contaminated by fatty acids and phytol. These fatty acids were treated with an
excess of ethereal CH2N2 for 5 min and then concentrated
in vacuo.
Subsequently, the mixture was treated with 0.5 mL of acetic anhydride in
pyridine (0.5 mL). The resulting mixture was then ready for HPLC.
Isolation of Curacins B (1 4) and C (1 5). Repetitive HPLC on a
Phenomenex Si column (10 mm, 500 x 10 mm) of the methylated mixture with
4% (v/v) EtOAc /hexanes resulted in curacin A (44.2 mg) and B (29 mg)
containing C. The latter was subjected to further HPLC on a Phenomenex
phenosphere C8 column (10 mm, 250 x 4.6 mm) with 64% CH3CN/H20 to
afford curacin B (20 mg, 69%) and C (1.1 mg, 4%).
Curacin B showed the following: [a]25D +620 (c = 0.84, CHCI3); UV Amax
(MeOH) 242 nm (E. = 22000); IR vmax (film): 2975, 2930, 2874, 1615, 1441,
52
1379, 1093, 1075, 1056, 998, 967, 914 cm-1; LREIMS m/z (rel. int. °A)) 373 (3),
342 (9), 332 (13), 274 (8), 181 (28), 180 (100), 166 (16), 140 (14) 99 (17), 91
(27), 85 (20), 79 (31), 67 (17), 55 (15); HR FAB MS (positive ion) obs. [M + F1]-+'
at m/z 374.2526 (0.9 mmu error for C23H36ONS);
Curacin C showed the following: [a]25D +62° (c= 0.84, CHCI3); UV ,max
(MeOH) 242 nm (E = 25000); IR 'Umax (film): 2970, 2927, 2885, 1617, 1438,
1095, 1074, 1054, 998, 964, 914 cm-1; LREIMS m/z (rel. int. %) 373 (4), 342 (9),
332 (12), 274 (8), 181 (29), 180 (100), 166 (16), 140 (14) 99 (17), 91 (26), 85
(25), 79 (30), 67 (18), 55 (16); HRFAB MS (positive ion) obs. [M + H]+ at m/z
374.2532 (1.5 mmu error for C23H36ONS);
Thermal lsomerization of Curacin B to A and C. To 6 mg of a 95:5
mixture of curacins B and C was added 2.5 mL of trimethylpentane. The
reaction mixture was refluxed for 3 hr and then concentrated in vacuo
.
The
concentrate was diluted with 4% EtOAc /hexanes, filtered through on sintered
glass, and subjected to HPLC (Phenomenex Maxil 10 µm Si, 500 x 10 mm,
eluted in 4% (v/v) EtOAc /hexanes, flow rate at 9 ml/min, UV detection at 254
nm) yielding curacin A (0.9 mg, 15% yield) and a mixture of curacins B and C
(0.8 mg, 13%)
Confirmation That Curacins B and C are Natural Products. First,
preserved Lyngbya majuscula was cold-extracted by grinding in a mortar with
liquid N2. Solvent (CH2Cl2 -MeOH, 2:1, 30 mL) was added to the algal powder
53
and stirred for 1 h. The algal material was removed by filtration and the solvent
reduced in vacuo, dissolved in minimal 4% EtOAc /hexanes, passed through a
plug of Si gel to remove polar compounds, and analyzed by GC-EIMS. This
gave a similar profile of curacins A-C to that from the hot extracted material. A
second approach involved heating pure curacin A (2.5 mg) at 35° in CH2C12­
Me0H (2:1, 4 mL) for 10 h, followed by GC-EIMS analysis. Finally, curacin A (2
mg) was refluxed in CH2Cl2 -MeOH (2:1, 3 mL) containing 4 drops H2O for 90
min, and then analyzed by GC-EIMS. These latter two experiments showed
only the presence of curacin A.
Preparation of CPRO-curacin A (1 8). To 15 mg of curacin A (40 limo')
was added 2.5 mL of toluene. The reaction mixture was heated at reflux for 4
hr. The resulting solvent was allowed to cool down to RT and removed in
vacuo. The resulting residue was dissolved in 4% EtOAc /hexanes and filtered
through the sintered glass. The filtrate was purified by HPLC (Phenomenex
Maxil 10 Si, 10 mm, 500 x 100 i. d., 4% (v/v) EtOAc /hexanes, flow rate at 9
mL/min, UV detection at 254 nm) to afford CPRO-curacin A (18) (7.3% yield)
which showed the following: [a]2210 +1100 (CHCI3, c = 0.1); UV Xmax (MeOH) 242
nm (e = 25000); IR Dmax: 2977, 2925, 2868, 1621,1437, 1096, 963, 913 cm ' ;IH
NMR (300 MHz, C6D6) 8 6.38 (dd, 1H, J= 15, 10.8 Hz, H-8), 6.05 (d, 1H, J=
10.8 Hz, H-9), 5.87 (m, 1H, H-15), 5.76 (m, 1H, H-21), 5.62 (dd, 1H,J = 10.7, 8.4
Hz, H-3), 5.6 (dt, 1H, J= 15, 7 Hz, H-7), 5.46 (dt, 1H,J =10.7, 7.1 Hz, H-4), 5.14
(m, 1H, H-2), 5.08 (m, 2H, H-16), 5.02 (ddd, 1H,J = 17, 1.7, 1.7 Hz, H-22a), 4.96
54
(ddd, 1H,J = 10, 1.7, 1.1 Hz, H-22b), 3.18 (s, 3H, -OCH3), 3.09 (m, 1H, H-13)
3.09 (dd, 1H,J =10.7, 8.4 Hz, H-lb), 2.77 (dd, 1H, J . 10.7, 8.7 Hz, H-1a), 2.5 (m
2H, H-19), 2.42 (m, 2H, H-20), 2.25 (m, 2H, H-14), 2.25 (m, 2H, H-11), 2.15 (m,
2H, H-6), 2.15 (m, 2H, H-5), 1.70 (s, 3H, H-17), 1.65 (m, 2H, H-12); 13C NMR (75
MHz, Cal)) 8 168.68 (C18), 137.27 (C21), 136.50 (C10), 135.35 (C15), 131.32
(C7), 131.14 (C3), 130.78 (C4), 127.93 (C8), 125.53 (C9), 116.78 (C16), 115.39
(C22), 79.97 (C13), 74.53 (C2), 56.31(-0CH3), 39.83 (C1), 38.06 (C14), 35.80
(C11), 33.98 (C19), 33.13 (C6), 32.19 (C12), 31.62 (C20), 28.19(C5), 16.59
(C17); LREIMS m/z (rel. int. %) 373 (4), 332 (15), 181 (32), 180 (100), 140 (16)
119 (17), 105 (20), 93 (20), 91 (27), 85 (20), 79 (33), 67 (16), 55 (14);
HRFABMS (positive) obs. [M + F1]-1- at m/z 374.2494 (2.4 mmu error for
C23H36ONS)
Preparation of 19-epi-curacin A (1 9). To 20 mg of curacin A was added
1 mL of pyridine and 1 mL of reagent grade acetic anhydride. The reaction
mixture was stirred at rt overnight. The solvents were then removed in vacuo.
The concentrate was diluted with 4% EtOAc /hexanes, and filtered through
sintered glass. HPLC (Phenomenex Maxil 10 Si, 10 mm, 500 x 100 i. d., eluted
in 4% (v/v) EtOAc /Hexanes, flow rate at 9 mi./min, UV detection at 254 nm) gave
19-epi-curacin A (4.2 mg, 21% yield) which showed the following: [a]25D +1220
(CHCI3, c = 0.635); UV X, (MeOH) 242 nm (e = 25000); IR u max: 2928, 2867,
2853, 1614, 1441, 1379, 1096, 1080, 999, 964, 912 cm-1; 1H NMR (300 MHz,
55
C6D6) 6 6.38 (dd, 1H,J= 15, 10.8 Hz, H-8), 6.02 (d, 1H, J. 10.8 Hz, H-9), 5.87
(m, 1H, H-15), 5.61 (dd, 1H,J= 10.5, 9.0 Hz, H-3), 5.58 (dt, 1H,J= 15, 10 Hz, H­
7), 5.45 (dt, 1H,J= 10.5, 7.1 Hz, H-4), 5.12 (ddt, 1H, J = 9.0, 8.7, 1.1 Hz, H-2),
5.06 (m, 2H, H-16), 3.18 (s, 3H, -OCH3), 3.09 (m, 1H, H-13), 3.09 (dd, 1H, J
=10.8, 8.2 Hz, H-lb), 2.75 (dd, 1H,J = 10.7, 8.7 Hz, H-1a), 2.26 (m, 2H, H-14),
2.23 (m, 2H, H-11), 2.11 (m, 2H, H-6), 2.11 (m, 2H, H-5), 1.70 (s, 3H, H-17), 1.65
(m, 2H, H-12), 1.55 (m, 1H, H-19), 1.49 (m, 1H, H-21) 1.34 (dt, 1H,J= 8.6, 4.4
Hz, H-20a), 0.85 (d, 3H,J= 5.7Hz, H-22), 0.48 (m, 1H, H-20b); 13C NMR (75 MHz,
C6D6) 8 170.84 (C18), 136.07 (010), 134.99 (C15), 131.03 (C3), 130.65 (C7),
130.59 (C4), 127.54 (C8), 125, 55.97 (- OCH3), 39.25 (C1), 37.69 (C14), 35.44
(C11), 32.81 (C6), 31.82 (C12), 19 (C9), 116.45 (C16), 79.59 (C13), 73.85 (C2)
27.81 (C5), 23.21 (C19), 17.74 (C22), 17.50 (C21), 17.34 (C20), 16.25 (C17);
LR EIMS m/z (rel. int. %) 373 (5), 342 (13), 332 (13), 274 (13), 181 (27), 180
(100), 166 (30), 98 (20), 91 (24), 85 (16), 79 (28), 67 (14), 55 (12); HRFABMS
(positive) obs. [M + H]+ at m/z 374.2534 (1.7 mmu error for C23H36ONS).
Preparation of TRO-curacin A (2 0). A sample of curacin A (18 mg, 48.3
1.1mol) was dissolved in 1.5 mL of acetic anhydride and 3 drops of D20. The
reaction mixture was stirred and allowed to stand at rt overnight. The solvents
were removed under vacuum. The residue was diluted with 25 % (v/v)
EtOAc /hexanes, and filtered through sintered glass. The filtrate was
chromatographed on a Phenomenex Maxil 10 gm silica column (50 x 10 cm)
56
developed at 9 mL/min with 25 % (v/v) EtOAc /hexanes to obtain TRO-curacin A
(7.6 mg, 17.6 gmol, 42 % yield). IR uma. (film) 3270, 2929, 1678, 1650, 1547,
1536, 1435, 1371, 1096, 963, 913 cm-1; [a]22D +580 (c 0.1, CHCI3); UV Xmax
(MeOH) 240 nm (log E = 4.49); 'H NMR (C6D6, 300 MHz) 8 6.42 (dd, 1H, J =15.0,
10.8, H-8), 6.02 (d, 1H, J=10.8, H-9), 5.87 (m, 1H, H-15), 5.62 (dt, 1H, J = 15.0,
10.0, H-7), 5.48 (dt, 1H, J=10.5, 7.1, H-4), 5.2 (dd, 1H, J=10.5, 9.0, H-3), 5.10
(m, 1H, H-2), 5.07 (m, 2H, H-16), 3.18 (s, 3H, -OMe), 3.09 (m, 1H, H-13), 3.04
(m, 2H, H-1), 2.35 (m, 2H, H-5), 2.26 (m, 2H, H-14), 2.23 (m, 2H, H-11), 2.11 (m,
2H, H-6), 1.8 (m, 1H, H-19), 1.7 (s, 3H, H-17), 1.65 (m, 2H, H-12), 1.58 (s, 3H,
Me-CO), 1.12 (d, J=6.1, 3H, H-22), 1.11 (m, 1H, H-20b), 0.96 (m, 1H, H-21),
0.66 (m, 1H, H-20a); 13C NMR (C6D6, 75 MHz) 8 196.95 (C18), 167.99 (Me-
CONH), 136.19 (C10), 135.37 (C15), 132.89 (C4), 131.58 (C7), 129.21 (C3),
129.21 (C3), 127.0 (C8), 125.64 (C9), 116.75 (C16), 79.95 (C13), 56.29 ( -OMe),
47.40 (C2), 38.05 (C14), 35.77 (C11), 33.59 (C1), 33.09 (C6), 32.14 (C12),
28.16 (C5), 22.97 (C19), 19.39 (C21), 16.60 (C17), 16.15 (C20), 11.95 (C22);
GC EIMS (% rel. int.) obs. [M] + m/z 433 (1), 318 (4), 259 (4), 200 (5), 185 (5),
161 (10), 119 (20), 111 (11), 105 (18), 91 (19), 85 (21), 84 (34), 83 (100), 79
(20), 55 (22); HRFABMS (positive) obs. [M+H]+ at M/Z434.2728 (C25H40NO3S,
deviation of 0.1 mmu).
Preparation of the methyl thio ether derivative of curacin A (2 1). A
solution of the TRO-curacin A (7.0 mg, 19 pmol) in dry THE (2 mL) was added
into MeLi (1 gL, 34.1 gmol) at - 78 °C. In 1.5 h, Mel (1.3 gL, 20.4 limo') was
57
added into the mixture, and the solution was stirred at - 78 °C for 1 h. The
solvents were removed under vacuum, and the residue was diluted with 100 %
EtOAc and filtered through sintered glass. The filtrate was chromatographed on
a Phenomenex Maxil 10 gm silica column (50 x 10 cm) developed at 9 mL/min
with 100 % EtOAc. The first peak was repurified with 30 % EtOAc /hexanes to
obtain the methyl thio ether derivative of curacin A (0.4 mg, 1.1 gmol, 5.7 %
yield). 1H NMR (C6D6, 300 MHz) 8 6.40 (dd, J = 14.9, 10.7, 1H, H-8), 6.02 (d,
1H, J =10.7, H-9), 5.88 (m, 1H, H-15), 5.60 (m, 1H, H-7), 5.52 (m, 1H, H-4), 5.19
(dd, J=10.4, 9.0, 1H, H-3), 5.10 (m, 1H, H-16), 5.10 (m, H, H-2), 4.70 (br, 1H,
NH), 3.18 (s, 3H, -OMe), 3.09 (m, 1H, H-13), 2.46 (d, J =5.8, 2H, H-1), 2.40 (m,
2H, H-5), 2.25 (m, 4H, H-14, 11), 2.18 (m, 2H, H-6), 1.91 (s, 3H, -SMe), 1.69 (s,
3H, H-17), 1.65 (m, 2H, H-12), 1.55 (s, 3H, Me -CO); 13C NMR (DEPT 135°, C6D6,
75 MHz) 8 135.37, 132.88, 131.43, 129.51, 127.67, 125.56, 116.81, 79.96,
56.32, 45.70, 39.89, 38.06, 35.80, 33.12, 32.16, 28.31, 22.94, 16.62, 15.89; GC
EIMS (`)/0 rel. int.) obs. [M] + m/z 365 (1), 333 (12), 318 (2), 304 (1), 233, (8), 234
(9), 227 (14), 213 (25), 207 (18), 185 (38), 171 (22), 161 (73), 159 (43), 145
(40), 133 (46), 119 (100), 117 (48), 111 (55), 105 (92), 94 (32), 85 (89), 77 (42),
69 (66), 55 (47).
Preparation of the curadisulfide A (2 2). The TRO curacin A (20) (14.0
mg, 38 gmol) in Me0H (2 mL) was treated with 0.5 M NaOH. After being stirred
at rt overnight, the reaction mixture was acidified to pH 3 by adding 5N HCI.
The solvents were removed under vacuum, and the residue was solubilized
58
with 100 `)/0 EtOAc and filtered through sintered glass. The filtrate was
chromatographed on a Phenomenex Maxil 10 µm silica column (50 x 10 cm)
operated at 9 mUmin with 90 % EtOAc /hexanes to obtain curadisulfide A (7.6
mg, 10.8 gmol, 54 % yield). 1H NMR (C6D6, 300 MHz) 8 6.42 (dd, 1H, J =15.0,
10.8, H-8), 6.02 (d, 1H, J=10.8, H-9), 5.87 (m, 1H, H-15), 5.62 (dt, 1H, J = 15.0,
10.0, H-7), 5.52 (m, 1HH-4), 5.48 (m, 1H, H-3), 5.25 (m, 1H, H-2), 5.12 (m, 2H,
H-16), 3.21 (m, 1H, H-lb), 3.20 (s, 3H, -OCH3), 3.09 (m, 1H, H-13), 2.75 (dd, J =
13.5, 6.5, 1H, H-1a), 2.35 (m, 2H, H-5), 2.26 (m, 2H, H-14), 2.23 (m, 2H, H-11),
2.11 (m, 2H, H-6), 1.78 (s, 3H, Me-CO), 1.71 (s, 3H, H-17), 1.65 (m, 2H, H-12).
59
CHAPTER III
MICROCOLIN C, A NEW ANTINEOPLASTIC LIPOPEPTIDE FROM THE
MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
ABSTRACT
A new cytotoxic peptide, microcolin C along with three known
compounds, microcolins A, B, and D, were isolated by bioassay-guided
fractionation of the organic extract from Curacao and Granada collections of
Lyngbya majuscula. The structure of microcolin C was assembled on the basis
of extensive 2D-NMR spectroscopy, mass spectrometry, and comparison with
the data sets for microcolins A and B. The absolute stereochemistry of
microcolin C was deduced from spectral data analysis and chemical
interconversion with microcolin B. Microcolin C was found to have an
interesting profile of cytotoxicity to human cancer-derived cell lines.
60
INTRODUCTION
Only few species rival the marine cyanobacterium Lyngbya majuscula in
the structural diversity and biological activity of its secondary metabolites.76 For
example, various collections of L. majuscula have yielded such unique
metabolites as (+)-a- (S)-butyramido-y-butyrolactone (1), a C8 metabolite
containing five-membered lactone,n malyngamide A (2), a chlorine-containing
amide metabolite,18 and ypaoamide (3), a new herbivore antifeedent
compound.79 Interestingly, most of these diverse structures may be considered
as lipopeptides, pointing out a general trend in L. majuscula secondary
metabolism.
Recently, the advent of organ transplantation has resulted in new
medical breakthroughs. This revolutionary event has been particularly due to
the availibility of immunosuppressive agents such as cyclosporin A (CsA) (4),80
FK506 (5),81 and rapamycin (6).82 Although these compounds were isolated in
the early 1970s, it was not until the 1980's that useful immunosuppressive
properties were recognized.
Cyclosporin A (4) is a cyclic undecapeptide which was isolated from the
fungus Tolypocladium inflatum Gams. Cyclosporin A (4) is a lipophilic
compound, and thus, cyclosporine must be solubilized for clinical
administration. FK506 (5) is a 23-membered macrolide antibiotic which was
extracted from a fermentation broth of the soil microorganism Streptomyces
tsukubaensis. Rapamycin (6) is produced by Streptomyces hygroscopicus, and
61
structurally, it resembles that of FK506. The mechanism of action of
cyclosporine (CsA) and FK506 is that both agents inhibit Ca2+ dependent
signaling pathways in many cell types. Alternatively, rapamycin inhibits Ca2+
independent signaling pathways.83 -84
While cyclosporin A (CsA), FK506, and rapamycin are either used
clinically or hold promise for their role in the prevention of organ transplant
rejection and treatment of autoimmune disorders, there is still a critical need for
new agents in this pharmacological class. Our recent interest in this regard has
0
OCH3
0
62
focused on the microcolin metabolites ,85-88 which are of particular note due to
their extremely potent immunosuppressive activity.
0
4
63
RESULTS AND DISCUSSION
In the course of investigating a Curacao collection of Lyngbya majuscula,
the crude organic extract of this species was found by TLC to show several
major metabolites with typical peptide characteristics (e. g., no H2SO4 charring
reaction). This crude extract also was active in the brine shrimp toxicity assay.
Using brine shrimp bioassay-guided fractionation of this extract (Figure 111.1),
we isolated a new metabolite, microcolin C (9) (2.0 mg, 0.05 % of extract) as
well as two known compounds, microcolins A (7) (140 mg, 3.2% of extract), B
(8) (64.4 mg, 1.5% of extract). Also, continuing search for the microcolin type
chemistry into a Granada collection of L. majuscula resulted in the isolation of
microcolin D (10) (2.8 mg, 0.14%) (Figure 111.2).
The existence of a hydroxyl group in 9 was suggested by an IR
absorption at 3200-3700 (br) cm-1. The molecular formula was determined as
C37H6307N5 by HRFABMS. The LRFABMS of microcolin C (9) showed
fragmentation patterns similar to that of microcolin A (7). A careful analysis of
these patterns revealed that the molecular ion as well as several fragment ion
peaks of 9 lacked one (-Ac) group as compared to 7 (Figure 111.3). These data,
together with the 1H NMR spectra of compound 9, suggested that it was a
deacetyl derivative of microcolin B (8).
The 1H NMR spectrum of microcolin C (9) (Table 111.1) was very similar to
those of microcolins A (7)and B (8) (Figure 111.4). For example, the two singlets
at 8 2.93 and 3.06 arising from two N-MeLeu and N-MeVal, the signals
64
Lyngbya (NAC 13 Aug/94-02), WG-EXT- 595,
4.39 g
52.8 mg saved
4.34 g
VLC
I
I
4% E/H 4% E/H
595A
595B
25% E/H
595C
I
50% E/H 50% E/H
595D
595E
85% E/H
595F
(414 mg)
390 mg
B S Toxic
30% E/M 100% M
100%E
595G
595H
5951
VLC
20 mg saved
60% E/H
595F-1
80% E/H
595F-2
(365 mg)
100% E
595F-3
Column Chromatography
B S Toxic
5% M/CHCI3 10% M/CHCI3 50% M/CHCI3
595F-2A 595F-2B 595F-2C 595F-2D
(299 mg)
HPLC (C-18, 80% ACN/H20)
B SIToxic
595F-2B1 595F-2B2 595F-2B3 595F-2B4 595F-2B5 595F-2B6 595F-2B7
(140 mg)
(1.9 mg)
(1.0 mg)
Microcolin A (7) Microcolin C (9)
B S Toxic
(4 ng/ml)
Figure III.
1
B S Toxic
(200 ng/ml)
(64.4 mg)
MicrocolinB (8)
B S Toxic
(400 ng/ml)
Flow chart for the isolation of microcolins (BS = brine shrimp)
65
Lyngbya (GMC-28 Jul/95-01)
1.96 g
VLC [
I
I
I
I
5% E/H 10% E/H 20% E/H 30% E/H 40% E/H 60% E/H 80%E/H 100% E
G
100% M
H
L-y.----./
(103 mg)
B S Toxic
HPLC (C-18, 80% CH3CN/H20)
I
I
(13 mg)
G-1
(2.8 mg)
G-2
1
(29 mg)
(22 mg)
(18 mg)
G-3
G-4
G-5
Microcolin 0 (10) Microcolin A (7) MicrocolinB (8)
Figure 111.1
Continued
66
H3
30
0
17
18
H3
31
15
22
CH3 CH3 CH3 0
44
34
43
6
H3
29
H3
42
H3
CH3 0
H
N
19CH3 0
24
OAc
H3
8
30 CH3
0
H3C'
40
34
18
28
17
H3
29
H3
CH3 0
H
27
15
33 N 26
N' 14
I
CH3 CH3 CH3 0
I
42
41
32
19CH3 0
H3
6
H3
23 OH
24
9
Figure 111.2 Structures of microcolins A-D (7-10)
67
A 425
B 425
C 383
D 383
0
N
N
CH3 CH3 CH3 0...;
A 282
H3
B 282
C 282
D 282
CH3 0
OR1
A 538
B 538
A 650
A 747
B 634
B 731
C 496
C 592
C 689
D 496
D 608
D 705
Microcolin A (7)
R1 = Ac, R2 = OH
Microcolin B (8)
Microcolin C (9)
R1 = Ac, R2 = H2
R1 = H,
R2 = H2
Microcolin D (10) R1 = H, R2 = OH
Figure 111.3. Mass fragmentations for microcolin isomers
resonating at 5 4.5-5.5 attributed to the a protons of the amino acid residues,
and a downfield doublet at 5 6.96 corresponding to an exchangeable NH
proton, all resembled those of 7 and 8. The most significant feature
differentiating 9 from 7 and 8 was the chemical shift (5 4.08) of the H-23
methine proton, and the loss of the acetate methyl group, indicating the loss of
the acetate ester in 9. In addition, the a hydrogen (H-26) at 5 5.21 and 13
hydrogens (H-28) at 5 1.66 of the leu residue in 9 showed triplet and dd splitting
patterns, while the corresponding hydrogens in 7 and 8 were a dd and ddd
68
pattern, respectively. Apparently, the absence of the acetyl moiety in microcolin
C allows for rotational freedom of the isobutyl group of the leu residue.
The analysis of 1H-1H COSY, HMBC, and TOCSY spectra of microcolin C
(9) revealed the presence of the structural units, 4-methylpyrrolidinone, proline,
N-Me-valine, threonine, N-Me-leucine, and dimethyloctanoic acid amide. The
presence of 4-methylpyrrolidinone was determined by 1H-1H COSY correlations
(cross peaks: H-2/H-3, H-2/H-4, H-3/H-4, H-4/H-6) and a comparison of
chemical shifts with the proton and carbon spectra of microcolins A (7) and B
(8). The proline, N-Me-valine, threonine, N-Me-leucine units were defined
based on 1H-1H COSY correlations (proline: H-8/H-9, H-9/H-10, H-10/H-11, N-
Me-valine: H-14/H-16, H-16/H-17, H-16/H-18, threonine: H-21/H-23, H-23/H-24,
and N-Me-leucine: H-26/H-28, H-28/H-29, H-29/H-30, H-29/H-31) and TOCSY
cross peaks (Figure 111.5) (proline: H-8 (8 5.47) - H-9 (8 2.40, 8 1.87) - H-10 (8
2.00)
H-11 (8 3.80, 6 3.70), N-Me-valine: H-19 (8 3.09)
2.29)
H-17 (8 0.81) - H-18 (8 1.01), threonine: H-22 (8 6.96) - H-21 (8 4.75) - H­
23 (8 4.08), N-Me-leucine: H-32 (82.96) - H-26 (8 5.21)
1.43) - H-30 (80.89)
H-14 (8 5.02)
H-16 (8
H-28 (8 1.66)- H-29 (8
H-31 (8 0.94)). Although the cross peak corresponding to
the Thr-Me (8 1.05) in the TOCSY was not observed, the Me resonance at 8
1.05 (H-24) showed strong COSY correlation to the Thr-13 proton resonance (8
4.08) indicating that this Me group belonged to the Thr component. The
presence of the dimethyloctanoic acid amide moiety (C33
C42) was
established by 1H-1H COSY and LRFABMS observing a fragment ion peak at
69
Figure 111.4 Comparison of 1F1 NMR spectra of microcolins A, B, and C (7-9)
at 300 MHz in CDCI3.
7.0
6.5
7.0
6.5
3,5
4.5
6.0
5.5
5.0
4.5
4.0
01714
3.
65
60
5'5
4.5
2.5
1.0
2.5
1.0
1'11
70
No
C.1
I
0
0.
0
C0 OC)
0
0
OI
i
1
.1,
10­
(7 o'
6
o
as
00
@
o
0
00
0
0
0
3
0
0
0
OCO
4
0
Val-a4
o
0
Leu dif.)
-5
0-1-11
P
0
id
-6
4
Thr-a
I
PPm
7
6
5
-7
(1
IT
AI
3
, '"
ppm
I
2
I
1
Figure III. 5 TOCSY contour plot of microcolin C (9) at 600 MHz (CDCI3).
71
Table 111.1. 1H and 13C NMR data for 9 and 10
Microcolin C (9)
Microcolin D (10)
Position
1H (J in Hz)a
1
2
3
4
6
7
8
9 (a)
(b)
10
11 (a)
(b)
13
14
16
17
18
19
6.08 (dd, 6.0, 1.7)
7.23 (dd, 6.0, 1.7)
4.75 (m)
1.47 (d, 6.7)
5.47 (dd, 8.5, 4.8)
1.87 (m)
2.40 (m)
2.00
3.70 (m)
3.80 (m)
5.02 (d,11.0)
2.29 (m)
0.81 (d, 6.7)
1.01 (d, 6.6)
3.09 (s)
20
21
22
23
24
25
26
28
29
30
31
32
33
34
35 (a)
(b)
36
37 (a)
(b)
38
39
40
41
42
4.75 (m)
6.96 (d, 9.3)
4.08 (qd, 6.4, 1.9)
1.05 (d, 6.3)
5.21 (dd, 7.8, 7.8)
1.66 (dd, 7.6, 7.6)
1.43 (m)
0.89 (d, 6.7)
0.94 (d, 6.7)
2.96 (s)
--­
2.80 (m)
1.12 (m)
1.87 (m)
1.33 (m)
1.07 (m)
1.25 (m)
1.24 (m)
1.25 (m)
0.88 (t, 6.8)
1.10 (d, 6.7)
0.82 (d, 6.5)
13Cb
169.0
125.5
153.9
58.1
17.2
171.9
60.1
28.9
24.6
48.0
--­
168.4
59.4
27.6
18.4
19.0
30.8
172.8
52.2
--­
67.3
18.9
171.4
53.9
36.1
24.9
21.8
23.2
30.5
178.2
33.9
41.9
--­
30.8
37.1
--­
29.2
22.9
14.1
18.3
19.6
1H (J in Hz)c
6.04 (dd, 6.0, 1.8)
7.24 (dd, 6.0, 1.8)
4.76 (m)
1.41 (d, 6.7)
5.60 (dd, 8.6, 4.9)
2.00 (m)
2.45 (m)
4.35 (m)
3.92 (m)
3.76 (m)
4.98 (d,11.0)
2.25 (m)
0.77 (d, 6.7)
0.99 (d, 6.7)
3.05 (s)
4.70 (m)
6.99 (d, 9.2)
4.06 (qd, 6.4, 1.9)
0.95 (d, 6.3)
5.16 (dd, 7.8, 7.8)
1.62 (dd, 7.6, 7.6)
1.39 (m)
0.82 (d, 6.7)
0.89 (d, 6.7)
2.94 (s)
--­
2.80 (m)
1.12 (m)
1.83 (m)
1.26 (m)
1.07 (m)
1.24 (m)
1.22 (m)
1.23 (m)
0.83 (t, 6.9)
1.08 (d, 6.8)
0.82 (d, 6.5)
13Cd
169.9
125.3
154.2
58.1
16.9
174.6
58.5
36.5
-
71.9
57.1
-­
169.2
59.3
27.4
18.3
18.8
30.6
172.8
52.1
- -­
67.2
18.9
171.3
53.8
36.0
24.9
21.8
23.1
30.5
178.2
125.5
41.8
30.7
37.0
29.1
22.9
14.1
18.2
19.5
a300 MHz in CDCI3, b75 MHz in CDCI3, C600 MHz in CDCI3, d150 MHz in CDCI3
72
m/z 282 which is consistent with the same linear portion as 7 and 8. Complete
assignments of these partial structures were accomplished by HMBC. The
methine proton resonance at S 4.75 (H-4) showed a HMBC correlation to C-7 (8
171.9) indicating that the pyrrolidinone and proline groups formed an amide
linkage. HMBC cross peaks between a methine signal at 8 5.47 (H-8) and C-13
(8 168.4), between methylene signals at 8 3.80, and 8 3.70 (H-11) and C-13 (8
168.4) established a connection between a proline and valine unit through an
amide bond. Similarly, HMBC correlations between a methine proton at 8 5.02
(H-14) and C-20 (8 172.8), between a methine resonance at 8 4.75 (H-21) and
C-25 (8 171.4), between a methine peak at 8 5.21 (H-26) and C-33 (8 178.2)
determined the linkages of the valine-threonine-leucine-dimethyloctanoic acid
amide residues.
The absolute stereochemistry of 9 was deduced to be 4S, 8S, 14R, 21S,
26R, 34R, and 38R, identical to that of 8 whose absolute configuration was
assigned in a recent study.89 In fact, microcolin C could be synthetically
produced through the enzymatic hydrolysis90-92 of microcolin B (Figure 111.6).
3
H3
CH
0
H3C
0
Esterase
/\/\/\AN
N
E
HZ\ ./\(\)\
3cr,2
Phosphate buffer
CH3 orb cm, ocA cc m3 o
E
E
N
I
CH3 CH3 CH3 0 A CH3 0
0
8
'OH
H30-
9
Figure 111.6 Enzymatic hydrolysis of microcolin B
73
Table 111.2. Tumor cell growth inhibitory activity and cytotoxicity of microcolin C
in the National Cancer Institute in vitro 60-cell line screen.
Panel/Cell line
Log 10GI50a
Logi oTG 1 b
Log10LC50c
Leukemia
CCRF - CEM
MOLT-4
SR
< -8.60
< -8.60
< -8.60
< -8.60
> -4.60
-8.09
< -8.60
> -4.60
Non-Small Cell Lung Cancer
NCI-H322M
< -8.60
< -8.60
-8.51
< -8.60
< -8.60
-8.47
< -8.60
< -8.60
< -8.60
-8.08
-8.08
-7.94
-7.76
-8.18
-5.50
< -8.60
-8.59
-8.05
< -8.60
-7.96
-5.13
< -8.60
-8.59
-8.40
-7.66
-6.13
-5.76
-7.85
0.75
2.75
-6.65
1.95
4.00
-5.55
2.96
Colon Cancer
HCT-15
HT29
KM12
SW-620
Melanoma
LOX IMV1
Prostate Cancer
PC-3
Breast Cancer
MDA-MB-231/ATCC
HS 578T
MG MIDd
Deltae
Range(
-4.98
3.91
aGI50, concentration required for 50% growth inhibition; bTGI, concentration required for
total growth inhibition; cLC50, concentration required for 50% lethality; dMG-MID, the
calculated mean panel GI50 TGI, or LC50; eDelta, the number of log10 units by which the
value of the most sensitive line(s) of the panel differs from the corresponding MG-MID;
(Range, the number of log10 units by which the value of the most sensitive line(s) of the
panel differs from the value of the least sensitive lines.93
i
4
3
2
Figure 111.7 11-1 spectrum of microcolin D (1 0) at 600 MHz (CDC13).
1
TTTTTTT
ppm
6
5i
4
3
Figure III. 8 1H-1H COSY contour plot of microcolin D (1 0) at 600 MHz (CDCI3).
76
Treatment of compound 8 with porcine liver esterase (PLE) in phosphate buffer
gave compound 9 which was confirmed to be identical to the natural product
microcolin C by comparison of optical rotation, mass spectra, and 'H and 13C
NMR spectra.
Desacetylmicrocolin A (10) was previously prepared as a semi-synthetic
derivative by the Harbor Branch Oceanographic Research Group' as part of
their evaluation of the immunosuppressive activity of this series. Comparison of
NMR (Figure 111.7 and 8) and mass spectral data for compound 10 isolated from
L. majuscula and the semi-synthetic desacetyl microcolin A showed them to be
identical.
Microcolin C (9) was investigated in the National Cancer Institute in vitro
60-cell line screen. Several cancer cell panels showed growth inhibition (Table
111.2).
Microcolins A-C (7-9) also exhibited potent toxicity to brine shrimp (LC5O
of 0.004 gg/ml, 0.4 pg /ml, and 0.2 µg /ml, respectively).
EXPERIMENTAL
General Experimental Procedures. NMR spectra were recorded on
Bruker DRX 600, AM 400, and ACP 300 spectrometers in the solvent specified.
Chemical shifts were referenced to solvent CDCI3 signal at 7.27 ppm for 1H and
77.0 ppm for 13C NMR. Mass spectra were recorded on Kratos MS 50 TC and
Finnigan 4023 mass spectrometers. Gas chromatography/mass spectrometry
77
(GC/MS) was carried out utilizing a Hewlett-Packard 5890 Series II GC
connected to a Hewlett-Packard 5971 mass spectrometer. UV spectra were
obtained on a Hewlett-Packard 8452 A UV-VIS spectrophotometer and IR
spectra on a Nicolet 510 spectrophotometer. All solvents were distilled from
glass prior to use.
Biological Material. The cyanobacterium L. majuscula was collected
from shallow water (0.5 - 2.1 m) on August 13, 1994, at CARMABI, Curacao,
Netherlands Antilles, and stored in i-PrOH at low temperature until work-up. A
voucher specimen is available from W.H.G. (collection NAC-13 Aug 94-2).
Extraction and Isolation. A total of 305 gm (dry wt) of Lyngbya majuscula
was extracted with CHCI3/MeOH (2:1 v/v) three times to give 4.39 gm of crude
extract. Gradient vacuum chromatography of the crude extract with
EtOAc /hexanes gave a brine shrimp toxic fraction containing the microcolins
(414 mg). This fraction was subjected to vacuum liquid chromatography on
silica (60%-100% EtOAc /hexanes) to give brine shrimp toxic active fraction (365
mg) which was further chromatographed on silica (5%-50% Me0H/CHC13) to
yield 299 mg. Finally, this brine shrimp toxic fraction (5% Me0H/CHC13) was
fractionated by preparative C18 HPLC (Hibar LiChrosorb RP-18,7 jim 250 x 10
mm) and eluted with 80% CH3CN/H20 using UV detection at 254 nm to yield
microcolins A (140 mg, 3.2% of extract), B (64.4mg, 1.5% of extract), and C (2.0
mg, 0.05% of extract).
78
Microcolin C (9): Microcolin C was isolated as a colorless oil and
showed the following: [a]25D -169° (c 0.5, EtOH); UV km (MeOH) 210 nm (E =
32000); IR um (Film): 3200-3700 (br), 2935, 2335, 1735, 1645, 1250 cm-1; LR
FABMS obs. M+ at m/z (rel. int. %) 689 (1.1), 592 (1.0), 496 (22.7), 478 (15.2),
383 (4.5), 365 (4.5), 282 (67), 100 (100); HR FAB MS (positive ion) obs. [M +
H]+ at m/z 690.4807 (0.2 mmu error for C37H6407N5).
Enzymatic Hydrolysis of Microcolin B (8). Compound 8 (6 mg, 8.2 p.mol)
was dissolved in methanol (0.1 mL) and the resulting solution was added to 0.2
M phosphate buffer (6 mL) containing porcine liver esterase (PLE, 0.5 mg, 100
units Sigma). The solution was stirred for 5.5 hr at room temperature and then
repeatedly extracted with ether (3 x 10 mL). The extracts were dried over
MgSO4, the solvent removed in vacuo, and the crude product subjected to
preparative C18 HPLC (Hibar LiChrosorb RP-18, 7 p.m 250 x 10 mm, 80%
CH3CN/H20, UV detection at 254 nm) to yield microcolin C (4.3 mg, 72% yield);
[a]25D -175° (c 1.1, EtOH); UV
'max
(MeOH) 210 nm (E = 33000).
79
CHAPTER IV
VIDALENOLONE, A NOVEL SECONDARY METABOLITE FROM THE
INDONESIAN MARINE RED ALGA VIDALIA SP.
ABSTRACT
A new secondary metabolite, vidalenolone, was isolated from an
Indonesian red alga Vidalia sp. The structure of this new cyclopentenolone­
containing compound was determined by a combination of spectroscopic
methods.
80
INTRODUCTION
Approximately 5500 species of red algae (Rhodophyta) have been
taxonomically identified.' Many red algae have been recognized as prolific
producers of halogenated monoterpenes and brominated phenols.95 Halomon
(1), a novel antitumor metabolite, and related compounds (2-8),96 isolated from
the red alga Portieria homemannii from different geographical locations, are
good examples of bioactive terpenoids. Although these halogenated
monoterpenes look structually simple, the structure determination of these
compounds is not due to the difficulty in obtaining good mass data and the
difficulties in placing halogen substituents.
CI
CI
2
ci
3
CI
CI
6
CI
7
CI
CI
81
In contrast to other genera of red algae, there has been relatively little
exploration of secondary metabolites from the genus Vidalia. Only a few
compounds have been found in this genus. These include the antiinflammatory vidalols A (9) and B (10),97 a new amino acid (11),98 2-amino-5
trimethylammoniopentanoate, and a halogenated cyclopentenediol (12)."
In our continuing search for bioactive compounds from Indonesian
marine red algae, we have isolated a new metabolite, vidalenolone (13) (2.0
mg), from Vidalia sp. Here, we report the isolation, structure determination and
biogenesis of vidalenolone (13).
OH
HO
HO
HO
OH
OH HO
OH
OH
10
9
NH2
/CH3
+
....,..........,..--......-N-CH3
00C
0H3
11
82
RESULTS AND DISCUSSION
A collection of Vidalia was extracted with CHCI3 and Me0H (2:1 v/v) and
the resulting crude extract was fractionated by Si gel vacuum liquid
chromatography. The fraction eluting with 80% EtOAc /hexanes was further
chromatographed to give fractions which were finally purified by normal phase
HPLC to yield a new cyclopentenolone compound 13
.
The molecular formula
was determined to be C13H1404 by HR CIMS (obs. [M + H]+ at m/z 235.0971, A
0.1 mmu). Hence compound 13 possessed seven degrees of unsaturation,
four of which were due to a phenolic ring, one due to an olefin, and one due to a
ketone functionality. The remaining one degree of unsaturation required 13 to
contain one ring.
The IR spectrum showed the existence of a hydroxyl group (3450 cm-1),
and an a, 13 - unsaturated carbonyl group (1625 cm-1). The presence of a para­
hydroxybenzyl moiety was deduced from the positive CI mass spectrum which
gave a major peak at m/z 107, compatible with C7H70. This was supported by
1H NMR data 8 6.9 (2H, d, J = 8.3 Hz), 6.6 (2H, d, J = 8.3 Hz) (Figure IV.1) and
13C NMR data (8 155, 131 (2C), 125, 114 (2C)). The 1H NMR spectrum in
DMSO-d6 revealed signals for two methylenes (8 2.42, 2.71), one 0-methyl
group (8 3.05), a methine (8 6.3), four protons in the aromatic region, and two
deshielded hydroxyl signals (8 9.2, 9.5) (Table IV.1). Additionally, the 13C NMR
was in good agreement with the presence of a carbonyl carbon at C-1
i
Ppm
u
I
9
I
8
I
7
6
I
1
5
1
4
1
3
2
Figure IV.1 11-I NMR spectrum of vidalenolone (13).
1
84
Table IV.1 11-I and 13C NMR data for vidalenolone (13)
Position
1H (J in Hz)b
13cc
1
202.79
2
153.00
HMBC correlations
3
6.30 (t, 3.1)
129.66
Cl, C4, C5
4
2.40 (dd, 5.7, 3.0)
30.19
Cl, C3, C5
5
80.37
6
2.69 (d, 5.6)
7
40.72
C1, C3, C4, C5, C7, C8
125.63
8
6.9 (d, 8.3)
131.09
C6, C7, C9, C10
9
6.6 (d, 8.3)
114.70
C7, C8, C10
10
155.00
11 ( -OCH3)
3.05 (s)
50.95
C4, C5
(H-O-Ph-)
9.25
C9, C10
(H-0-)
9.55
C1, C2, C3
a Spectra were recorded in DMSO-d6. b 1 1-1 NMR at 300 MHz,
cl3C NMR at 75 MHz
85
OH
0
n
c
b
a
Figure IV.2 Partial structures of vidalenolone (13)
OH
OH
Figure IV.3
Structure of vidalenolone (13)
d
86
(8 202.79), a polarized olefin (8 153.13, 129.66), a methoxy carbon (8 50.95),
and two methylene carbons (8 40.72, 30.19).
Detailed spectral analysis of 13 provided the partial structures, a, b, c
and d (Figure IV.2). The partial structure a was deduced from 1H-1H COSY
correlations between a methine proton at 8 6.9 (H-8) and 8 6.6 (H-9). The
methylene proton resonance at 8 2.69 (H-6) showed HMBC correlations to C-7
and C-8 suggesting its connection to a hydroxylbenzyl moiety. The methylene
proton resonance at 8 2.40 (H-4) in the structural fragment b showed COSY
correlation to a methine resonance 8 6.30 (H-3). The deshielded 13C chemical
shift of the quaternary carbon resonance 8 153.00 (C-2) indicated that it bore
the free hydroxyl group. The location of this free hydroxyl group was further
supported by HMBC correlations to C-1, C-2, and C-3. The partial structure c
was defined based on the methoxy proton resonance 8 3.05 (H-11) showing
HMBC correlation to a deshielded quatenary carbon resonance 8 80.3 (C-5).
The carbon chemical shift at 8 202.79 (C-1) indicated the presence of the
ketone functionality (partial structure d). Complete assignments from these
partial structures were achieved by further HMBC analysis. The methylene
protons at 8 2.69 (H-6) showed HMBC correlations to C-5 connecting partial
fragments a and c. HMBC correlations between methylene protons at 8 2.40
(H-4) and C-5, between a methine proton at 8 6.30 (H-3) and C-5 enabled
87
connection of b and c. Finally, HMBC corelations between methylene protons 8
2.69 (H-6) and C-1, and a hydroxyl proton 8 9.55 and C-1 completed the planar
structure of vidalenolone (13) (Figure IV.3).
The stereochemistry of 13 is yet to be established. Stereochemical study
of 13 was attempted using CD analysis. However, the results were not
conclusive because of an abnormality of observed peaks. The biogenetic origin
of vidalenolone (13) is presented in Figure IV.4. Tyrosine (14) and succinyl
-
CoA (17) could be reasonable major precursors in this biosynthetic pathway.
Similar condensation of succinyl-CoA and hydroxyketone was proposed for the
assembly of a multifunctional mC7N moiety of the munumycin group of
antibiotics.lw In this proposed pathway, tyrosine may be transformed by the
action of transaminase to a-keto acid (15). Subsequently, the carboxylic acid
in 15 could be reduced to the corresponding alcohol (16) which further reacts
with succinyl-CoA (17) to form a new carbon-carbon bond (18).
Five-
membered ring formation can take place by decarboxylation of (18) to give the
a, 13-dihydroxyketone (19). The tent- hydroxyl group of 19 could undergo 0­
methylation to yield 20. Subsequently, the a-hydroxyl of 20 could be oxidized
to give the diketone (21) which enolizes to form the final product, vidalenolone
(13).
88
OH
OH
HO
15
14
SCoA
H3C
HO
v
OH
0
20
HO
HO
19
OH
HO
HO
21
13
Figure IV.4 Proposed biogenetic pathway of vidalenolone (13)
89
EXPERIMENTAL
General Experimental Procedures. NMR spectra were recorded on a
Bruker ACP 300 spectrometer in the solvent specified. Chemical shifts were
referenced to the solvent DMSO signal at 2.50 ppm for 1H and 39.51 ppm for
13C NMR. Mass spectra were recorded on Kratos MS 50 TC and Finnigan 4023
mass spectrometers. UV spectra were obtained on a Hewlett-Packard 8452 A
UV-VIS spectrophotometer and IR spectra on a Nicolet 510 spectrophotometer.
CD measurement was obtained on a Jasco 41A spectropolarimeter. All
solvents were distilled from glass prior to use.
Plant Material. The marine red alga Vidalia sp. (Rhodophyta;
Rhodomelaceae; Ceramales) was collected on November 5, 1994, at Ang
Island, Indonesia. It was found growing subtidally on volcanic substrate. The
alga was preserved in i-PrOH at low temperature until extraction. A voucher
specimen (collection IAI-5-Nov-94-1) is stored in the natural products laboratory
of the College of Pharmacy, Oregon State University.
Extraction and Isolation. The i-PrOH preserved alga Vidalia was
extracted with CHCI3 and Me0H (2:1 v/v) three times at room temperature. The
solvent was evaporated to dryness under reduced pressure at 30 °C to provide
a crude extract (6.02 g) which was fractionationed by Si gel vacuum liquid
column chromatography with a gradient solvent system (hexanes- EtOAc). The
fraction eluting with 80% EtOAc /hexanes was further chromatographed to give
90
eight fractions. Fraction 5 was subjected to repetitive HPLC on a Al !tech
Versapack Si column (10 gm, 2 x 300 x 4.1 mm) with 25% hexanes/ EtOAc to
afford pure vidalenolone (13).
Vidalenolone (13): [a]25D -950 (c 0.31, MeOH); UV Xmax (MeOH) 209 nm
(E = 32000), 224 (E = 35000), 266 (E = 12000); IR uma, = (film) 3450 (br), 2935,
2335, 1700, 1625 cm-1, LR CIMS (rel. int. %) obs. [M + H]+ at miz 235 (18), 234
(13), 219 (2), 217 (2), 205 (18), 204 (21), 203 (100), 157 (6), 135 (6), 129 (45),
128 (90), 107 (78); HR CI MS (positive ion) obs. [M + H]+ at nilz 235.0971 (0.1
mmu error for C13F11404); CD (MeOH): AE = -3.3, -1.3, -1.8 (Am.), = 220, 250, 270
nm).
91
CHAPTER V.
Analysis of Oxylipin Products from Gametophyte Cell Cultures of the Brown
Alga Laminaria saccharina
ABSTRACT
Filamentous cells isolated from female gametophytes of the brown alga
Laminaria saccharina were cultured in flask or bioreactor culture. These
cultures produced a variety of 15-lipoxygenase metabolites. Analysis by GC­
MS of TMS ethers of the methylated hydroxy acids identified the presence of
three oxylipins: 13-hydroxy-9,11-octadecadienoic acid (13-HODE), 13-hydroxy­
6,9,11,15-octadecatetraenoic acid (13-HODTA), and 15-hydroxy-5,8,11,13­
eicosatetraenoic acid (15-HETE). Feeding experiments were carried out to
attempt the enhancement of these 15-lipoxygenase metabolites using linoleic
acid, a-linolenic acid, y-linolenic acid, arachidonic acid, jasmonic acid and
calcium ionophore antibiotic. The presence of linoleic acid, y-linolenic acid and
jasmonic acid increased the production of the three oxylipins 2 to 4 times over
the controls. On the other hand, addition of antibiotic did not increase the yield,
while a-linolenic acid suppressed production of these metabolites. Arachidonic
acid addition gave rise to a variety of HETE products via autoxidation.
92
INTRODUCTION
Macrophytic marine algae or seaweeds have proved to be a rich source
of pharmacologically active compounds.' They are able to biosynthesize
eicosanoids and related oxylipins through different pathways of the arachidonic
acid (AA) cascade (see Figure 1.1 in Chapter 1).'
In mammalian systems, precursors such as arachidonic acid lead to a
number of pharmacologically important metabolites known cumulatively as
eicosanoids which include the HETEs, prostaglandins, thromboxanes,
hepoxilins, and leukotrienes.1'
Many of the secondary metabolites isolated from marine macroalgae
show potential as biomedicines. Of particular interest are marine oxylipins.
Many of the bioactive oxylipins isolated from different classes of marine
macroalgae are not found in mammals. This is especially true of some genera
of red and brown algae.
Marine oxylipins appear to be the product of the lipoxygenase pathway.
For example, the red algae tend to metabolize C20 acids via the 12­
lipoxygenase pathway, green algae utilize C18 acids at C-9 and C-13, and
brown algae metabolize both C18 and C20 acids as substrates. Figure V.1
represents a variety of oxylipins produced by various lipoxygenase and
significant fragment ions from their derivatives as observed by GC-EIMS.
A major problem in drug discovery is that the yields of secondary
metabolites from natural sources are usually in such low amounts that an
adequate supply can be serious a problem, not only for preliminary in vivo drug
93
02CH3
5-HETE [M]+ 406
8-HETE
1.- 225 11-HETE
9-HETE
CO2CH3
CO2CH3
..:
,
OTMS i
295
--ff'
12-HETE
14-HETE
15-HETE
Figure V.1 Expected fragment patterns of various oxylipins
cr
,ce
\+
eb
01/
og-
r';')4)
'',
/
S
1/
95
GO 2.0-,3
GO P-4,3
0020-A3
"0110:,
3eA.----
1 3-IA001N \M1+ g78
311-*-­
A3-1-3tODE. ot NOTE klK 382 ov 380
CO2.04
96
evaluations, but also for the commercial market. There are several approaches
which enable an adequate supply of bioactive compounds for commercial
products. 102,103 The traditional approach is to recollect larger quantities of the
organism. However, this approach has been frequently criticized by ecologists
and environmentalists because ocean resources are often over utilized in the
name of science.' For example, over 1 ton of the sea hare Dolabella
auricularia was collected to obtain 29 mg of dolastatin 10, a potent anticancer
compound.' Subsequently, to provide material for clinical trials, about 700
tons of the sea hare would be required. The anti-ovarian-cancer drug taxol
offers another good example of the supply issue. The amount of tree bark from
the Pacific yew tree needed to obtain 1 kg of pure taxol is about 30,000 pounds,
while the annual medicinal need for treating patients in the U.S. is over 25 kg. 102
Another approach is to try the total synthesis of the desired target
compound. One drawback of this approach is that it may not be suitable for
some compounds due to the difficulty of controlling stereochemistry and the
poor yield of the desired compounds. For example, a group at OSU reported
the total synthesis of curacin A, however, with only a 0.6 % overall yield."
A third approach is to culture and harvest the organism. This approach
has been successfully carried out with culturable marine microorganisms in the
laboratory. 107,108 But, the scale-up culture of a marine microorganism for clinical
trials or pharmaceutical demands has not been reported."
A relatively new approach is to utilize biotechnology such as bioreactors
to cultivate marine organisms. Undifferentiated cells are placed in a liquid
97
suspension culture, followed by stimulation to yield the target secondary
metabolites in the culture. 110,111 Advantages of this approach over traditional
culture include: fast and controlled cultivation of cells in a bioreactor and the
opportunity to increase the production of the desired compound through a fine
level of control in changing the culture conditions."' However, there are few
reports regarding cell culture studies using bioreactors. In this work, the brown
alga L. saccharina was cultivated in a bioreactor and isolation, identification,
and quantitative analysis of oxylipins was described. This work was a
collaborative effort with the laboratory of Dr. Rorrer in the Department of
Chemical Engineering at O.S.U. Dr. Rorrer's laboratory was responsible for the
bioreactor design, and start-up and cultivation of L. saccharina female
gametophyte cells."'
RESULTS AND DISCUSSION
This work describes the capability of the brown alga L. saccharina to
produce bioactive marine oxylipins utilizing gametophyte cell suspension
cultures. The procedures for the isolation and analysis of oxylipins produced
from the L. saccharina female gametophyte cell biomass are shown in Figure
V.2 and described in the experimental section.
Prior to the quantitative analysis of the oxilipins produced in these
cultures, the qualitative analysis was carried out using GC-MS. Fresh wet
tissue was homogenized in 20 mL of CH2C12/Me0H (2:1 v/v) (Figure V.2).
98
Fresh wet tissue
homogenize
add internal standard
pool extracts
1
Methylation of extract (CH2N2)
Partial puification of oxylipins by prep-TLC
scrape the spot containing desired compounds
and internal standard
1
Trimethylsilyl (TMS) derivatization
Hexamethyldisilazane (HMDS)
Chlorotrimethylsilane (TMS-CI)
Pyridine
1
vacuum
1
centrifugation
GC-Mass spectrometry
Figure V.2 Overview of the identification and analysis of oxylipins
in L. saccharina cell culture
99
Ricinelaidic acid was added to the extract as an internal standard for
quantitative analysis. The extracts were pooled, filtered, and resulting solvent
was dried using a Buchi rotary evaporator. A portion was then methylated with
CH2N2 in Me0H. This methylated extract was subjected to Prep-TLC using
silica gel 60
F254
in 25% EtOAc /hexanes. The spot containing the desired
methylated hydroxy acids was broadly scraped off the TLC plate. Several
methylated oxylipins were used for references to locate the methylated hydroxy
acid fraction of the extract. The resulting powdered gel was extracted with Et20
and evaporated by rotavaporation. Tissue extraction was derivatized with
appropriate reagents to be TMS ethers of the methyl esters of the hydroxy fatty
acids prior to the injection into GC-MS. Then, derivatized materials were
analyzed by GC-MS (Figure V.3).
TIC: 3- 107 -4.D
3
2000000­
4
5
1500000­
12
2
1000000
6
20
500000­
11
13 16
mor10
ime-->7.00
8.00
9.00
10.00
1-;
11.00
12.00
13 00
14.00
Figure V.3. Total Ion Chromatogram (TIC) of TMS-hydroxy fatty acids produced
from L. saccharina gametophyte cell suspension.
100
The significant peaks were identified by matching the averaged mass spectra
for the eluted peak to mass spectra of standards using the library program of the
HP Data station (Table V.1).
Detailed analysis of GC peaks, especially, in the range of 12.6 - 15.7 min
(Figure V.3), resulted in the identification of three oxylipins derived from action
of a 15-lipoxygenase: 13-hydroxy-9,11-octadecadienoic acid (13-HODE) (1),
13-hydroxy-6,9,11,15-octadecatetraenoic acid (13-HODTA) (2), and 15­
hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE) (3).
The identities of these TMS ether derivatives of the methyl esters of 13­
HODE (1), 13-HODTA (2), 15-HETE (3) were indicated by their distinct
fragmentation patterns at m/z 311, 309, and 335 (Figure V.4 and 5),
respectively, as well as matching their GC retention times to that of the
corresponding authentic standard compounds.
The yield of oxylipins in the sample was calculated by employing the
internal standard method using the total peak areas of the corresponding
oxylipin, the peak area of the standard (ricinelaidic acid), the concentration of
the internal standard added to the analyzing sample, and the detector response
factor.
During the course of quantitative analysis of these oxylipins, two factors
made this study challenging. First, the derivatized oxylipins rearranged during
GC and showed separate peaks. This phenomenon of rearrangement has
been commonly observed when these kinds of oxylipins are injected into a
Gc.12 Second, the internal standard, ricinelaidic acid, co-eluted with 13­
101
# Peak
Quality (%)3
1
78
Palmitic acid Me ester
2
99
Palmitoleic acid Me ester
3
95
Palmitic acid Me ester
4
81
6,9,12-Octadecatrienoic acid, methyl ester
5
99
Oleic acid Me ester
6
99
Oleic acid Me ester
7
91
Si lane, [(3,7,11,15-tetramethy1-2-hexadecenyl)oxy]trimet
8
99
Docosane
9
93
Arachidonic acid Me ester
10
95
Methyl 15-HODTA, TMS ether
11
12
Si lane, dimethy1-2-propenyl(tetradecyloxy)­
12
90
Ricinelaidic acid, TMS ether (Internal standard) + Me 13-HOD
13
53
Methyl 14-HEPE, TMS ether
14
91
Methyl 15-HODTA, TMS ether (rearranged)
15
73
Methyl 13-HODE, TMS ether
16
50
Hexanedioic acid, mono (2-ethylhexyl) ester
17
32
Pregn-17(20)-en-16-one, (5 alpha., 17Z)­
18
95
Methyl 15-HETE, TMS
19
25
1,3,5-Triazine-2,4-diamine, 6-(3-methylpheny1)­
20
64
Bis (2-ethylhexyl) phthalate
21
80
Methyl 15-HETE, TMS ether (rearranged)
ID
Table V.1. Identified compounds for the TIC chromatogram shown in Figure V.3
a Quality ( %) means purity of all m/z fragments of being analyzed
compound relative to all m/z fragments of standard compound.
102
HODE (1) (Figure V.5). Thus, the detector response factors used to quantify 13­
HODE production were based on peak areas calculated from mass-selective or
mass extract ions at m/z 311 for 13-HODE and at m/z 187 for the ricinelaidic
acid. Detector response factors used to quantify 13-HODTA (2) and 15-HETE
(3) were determined based on peak areas obtained from TIC data.
In an attempt to enhance the yield of oxylipins derived from the 15­
lipoxigenase pathway, six precursor or elicitor feeding experiments were
carried out. These included linoleic acid (6,9-octadecadienoic acid), a-linolenic
acid (9,12,15-octadecatrienoic acid), y-linolenic acid (6,9,12-octadecatrienoic
acid), jasmonic acid, antibiotic (Ca-ionophore), and arachidonic acid. The
relationship of these precursors as well as several other and their
corresponding 15-lipoxigenase products is depicted in Figure V.6.
At the 20th day of cultivation, 2 mL of a 10 g/L stock solution containing a
given precursor or elicitor was added to the liquid suspension culture. Then,
the culture was incubated for an additional 10 days to give a total cultivation
time of 30 days.
The toxic effect of exogenously added precursors to cell suspension
cultures was also examined. It was found that both the structure of the fatty acid
and its concentration in the culture medium are critical to its toxicity. The
general trend of toxicity for these fatty acids is a-linolenic acid > linoleic acid > y­
linolenic acid. The maximum tolerable precursor concentration for the culture
was found to be 200 mg/L and concentrations over 1000 mg/L were toxic to
103
bundance
Average of 12.747 to 12.769 min.: 3-105-7.D (+,
73
)
309
CO2CH3
90001
8000
\OTIA4
309,
7000
13-HODTA [Mr 378
6000
5000
4000­
3000
91
1291-45
2000­
187
59
1000­
49
0
/Z
243
219
,
->
50
100
150
Abundance
200
-AL
1
280
,
250
't
ti
34'7363
I
"'
,
,
300
350
,
400
#40: Methyl 15 -HETE, TMS (*)
7 3
90001
CO2CH3
80001
7000:
4bTMi'f
335
225
6000
15-HETE
5000­
4000:
3000=
91
129
2000
155173
1000, 55
199
251
0
M/Z
->
50
100
150
200
250
335
316
291
349
300
350
391406
400
Figure V.4 Mass spectra and structures of TMS ether derivatives of the 13­
HODTA and 15-HETE
104
(A) Abundance
#13: Methyl 13 -RODE, TMS Ether (*)
7 3
90001
CO2CH3
80001
70001
311.
130
6000:
tTMiA
13 -NODE R,A1+382
50001
4000:
311
30001
225
93
20001
55
382
155171
10001
199
M/Z
->
111TH
50
100
(B) bundance
150
269 292
200
250
335
300
367
350
Average of 12.940 to 12.952 min.: 3-104-7.D
187
(+,
418
.,1.,
400
)
9000 1
CO2CH3
8000 1
OTMS
7000 1
187
6000
Ricineladcadd
5000 1
40001
73
3000
20001
103
270
311
1000:
55
129
0
4
50
1551711
44140'rmli-14"4
100
150
200
299
225239
.
,-"
.
250
300
337
3682
r,
350
Figure V.5 Mass spectra and structures of TMS ether derivatives of (A) methyl
13-HODE (1) and (B) of co-eluted methyl ricinelaidic acid standard (4) and
methyl 13-HODE.
105
cultures in the 10 day incubation. The production of 13-HODTA, 13 -NODE, and
15-HETE following addition of the six precursors or elicitors is shown in Table
V.2 and Figure V.7 through Figure V.10.
Based on absolute yields, all culture experiments produced 13 -NODE in
the greatest quantity, followed in order by 13-HODTA and 15-HETE. Figure V.7
shows that the yield of 13-HODTA was significantly enhanced by the addition of
y-linolenic acid, resulting in a yield over three times higher than that of controls.
The yield of 13 -NODE was increased in three precursor feeding experiments
(linoleic acid, a-linolenic acid, and y-linolenic acid). But, the enhancements
were not highly significant as less than two times the value of the controls were
observed (Figure V.8).
The yield of 15-HETE was significantly increased by several of the
feeding precursor experiments (Figure V.9). Cultures provided with arachidonic
acid led to a significant increase in 15-HETE production. However, careful
analysis utilizing extract ion chromatography in GC-MS revealed that the
exogenous addition of arachidonic acid resulted in the production of various
HETE's due to non-enzymatic autoxidation. This fact was established by
observing significant ion fragments at m/z 295 (12-HETE), 265 (8-HETE), 255
(9-HETE)225 (11-HETE), and 335 (15-HETE).
It is notable that the feeding experiments with a- linolenic acid gave yields
of oxylipins (15-HODTA (2) and 13-HODE (1)) lower than control experiments.
One possible explanation for this observation is that, from the toxic screening
106
Fatty Acid Precursor
Product
COOH
COOH
15-LOX
peroxidase
''OH
linoleic acid (LA)
COOH
13 -NODE (1)
COON
15-LOX
peroxidase
OH
H
a-linolenic acid (a-LA)
13-HOTE
COOH 15-LOX
COOH
peroxidase
OH
rlinolenic acid ( y-LA)
13-HOTE ( y)
COON 15-LOX
COOH
peroxidase
OH
stearidonic acid
13-HODTA (2)
OOH
15-LOX
peroxidase
arachidonic acid
SOH
15-HETE (3)
Figure V.6 Hydroxy fatty acid production from precursors by 15-LOX/peroxidase
metabolism
107
13-HODTA
13-HODE
15-HETE
#1. L s control
123.2 ± 39.4
43.2 ± 10.5
13.2 ± 8.3
#2. Ds control
120.2 ± 36.8
49.9 ± 26.8
18.1 ± 16.3
#3. linoleic acid
#4. a- linolenic acid
#5. y-linolenic acid
219.3 ± 93.2
97.6 ± 37.5
37.4 ± 28.5
78.9 ± 20.2
31.5 ± 4.0
36.7a
234.4 ± 115.6
144.9 ± 108.3
77.2 ± 55.4
246.7a
80.0'
62.8a
105.8a
42.8a
20.3a
#6. jasmonic acid
#7. antibiotic
#8. arachidonic acid
199'
336.6"
All numbers are ppm unit, a single run, others average of two or three runs,
b
sum of various HETE's including 5-,8-,11-,12-,15-HETE
Table V.2 Effects on the yields of oxylipin from culture fed with six precursors
experiment, a-linolenic acid might decrease production through sample toxicity
to the algal cells. Jasmonic acid increased the production of three oxylipins,
while addition of Ca-ionophore had no effect on the yield of oxylipins.
The standard deviations in the oxylipin production observed in these
studies could be due to the following reasons: firstly, biological experiments
have unpredictable results because of the complexities inherent in living
systems: secondly, experimental variability contributes to this deviation,
including production of mass of cell density, partial oxylipin purification, oxylipin
derivatization, and GC-MS analysis.
#1. Ls control
#2. Ds control
#3. linoleic acid added
#4. alpha-linolenic acid added
#5. gamma-linolenic acid added
#6. jasmonic acid added
#7. antibiotic added
i
2
3
4
5
6
7
Experiment
Figure V.7 Yield of 13-HODTA from cultures added with different precursors.
400
#1. Ls control
#2. Ds control
300
#3. linoleic acid added
#4. alpha-linolenic acid added
200
#5. gamma-linolenic acid added
#6. jasmonic acid added
100
#7. antibiotic added
#8. arachidonic acid added
0
1
2
3
4
5
6
7
8
Experiment
Figure V.8 Yield of 13 -NODE from cultures added with different precursors.
400
sum of various HETE's
(not truely quantifiable)
#1. Ls control
#2. Ds control
300
#3. linoleic acid added
#4. alpha-linolenic acid added
200
#5. gamma-linolenic acid added
#6. jasmonic acid added
100
#7. antibiotic added
#8. arachidonic acid added
0
2
3
4
5
6
7
8
Experiment
Figure V.9 Yield of 15-HETE from culture added with different precursors.
0
#1. Ls control
400
#2. Ds control
#3. linoleic acid added
300
E
#4. alpha-linolenic acid added
ca
a_
#5. gamma-linolenic acid added
200
#6. jasmonic
#7. antibiotic added
100
#8. arachidonic acid added
13-HODTA
ra 13 -NODE
0
1
2
3
4
5
6
7
8
15-HETE
Experiment
Figure V.10 Comparison of the yields of three oxylipins from feeding experiment
112
EXPERIMENTAL
Partial Purification of Oxylipins. Fresh wet tissue was homogenized by a
Potter-Elvehjem tissue homogenizer in 20 mL of CH2C12/Me0H (2:1 v/v) (three
times). Ricinelaidic acid was added to the extract as an internal standard for
quantitative analysis. The extracts were pooled, filtered, and resulting solvent
was dried using a Buchi rotary evaporator. A portion was then methylated with
CH2N2 in Me0H at room temperature. This methylated extract was subjected to
Prep-TLC using silica gel 60 F254 in 25% EtOAc /hexanes. The spot containing
the desired methylated hydroxy acids was broadly scraped off the TLC plate.
Several methylated oxylipins were used for references to locate the methylated
hydroxy acid fraction of the extract. The resulting powdered gel was extracted
with Et20, filtered, and evaporated by rotavaporation.
Gas Chromatography-Mass Spectrometry (GC-MS). A portion of each
oxylipin sample was derivatized to trimethylsilyl (TMS) ethers with 2 drops of
hexamethyldisilazane (HMDS) and chlorotrimethylsilane (TMS-CI) in pyridine
for 2 h. The reaction mixture was concentrated in vacuo and the residue was
resuspended in hexanes. After centrifugation, the supernatant was injected into
the GC-MS. The TMS ethers of the methyl esters of the hydroxy fatty acids were
chromatographed by GC-MS using a Hewlett-Packard HP 5890 Series II gas
chromatograph and HP 5971 mass spectrometer. For the injection, 14 of the
sample volume was injected into a Hewlett-Packard HP-1 crosslinked methyl
113
silicone capillary column (12.5 m x 2 mm x 0.33 µm) operating with helium as
the carrier gas at a temperature program of 100 °C initial, 10 °C/ min ramp, 240
°C final, 250 °C injector, and 280 °C detector.
Calibration Curves and Detector Response Factors. Different stock
solutions of ricinelaidic acid internal standard and each oxylipin were prepared
from stock solutions in hexanes at concentrations of 1-100 g,g/j1L by serial
dilution. For the injection, 14 of each was injected into the GC-MS in triplicate.
Total ion peak areas of 13-HODTA and 15-HETE were measured and plotted
for each concentration using Cricket graph while extract ion peak areas (m/z =
311 for 13-HODE, m/z = 187 for ricinelaidic acid) were used to resolve the
problem. The detector response factors (Drf) for each oxylipin to internal
standard was determined from the following relationship (1):
Dry=
Mis x Ao
Mo x Ais
(1)
Muo -
Mis x Auo
Drf x Ais
(2)
where Drf is the detector response factor, Mis is the concentration of the internal
standard, ricinelaidic acid, Ao is the total or extract ion peak area of oxylipin
while Mo is the concentration of the oxylipin, and Ais is the the total or extract
ion peak area of standard compound.
Once the detector response factor of each oxylipin was obtained, the
quantity of oxylipin, expressed as ppm, was calculated from the modified
114
formula (2); where, Muo is the concentration of unknown oxylipin in the sample,
and Auo is the total or extract ion peak area of unknown oxylipin.
115
CHAPTER VI.
PRODUCTION OF OXIDIZED ANANDAMIDE DERIVATIVES AND EVALUATION
OF THEIR BINDING AFFINITY TO THE CANNABINOID RECEPTOR (CB1)
ABSTRACT
Five oxidized anandamide derivatives were prepared from
arachidonylethanolamide (anandamide) through autooxidation, followed by
reduction using triphenyiphosphine. Their structures were determined by a
combination of NMR spectroscopy and GC-MS. The cannabinoid receptor
binding affinity of these derivatives was evaluated. This study revealed the
following trend in activity: anandamide > 15-hydroxyanandamide > 9­
hydroxyanandamide > 8-hydroxyanandamide > 11-hydroxyanandamide > 5­
hydroxyanandamide.
116
INTRODUCTION
Anandamide (arachidonylethanolamide) (1) is a mammalian
endogenous ligand and hydrophobic brain constituent that binds with high
affinity to brain cannabinoid receptors (CB1).113 This lipid is produced and
released from brain neurons' and is decomposed by amidohydrolase activity
to give arachidonate and ethanolamine.115 The first discovery of this naturally
occuring brain component was described by Devane et al. in 1992.11'
Until the late 1980's, the misleading hypothesis was dominated that the
endogenous substrate for this receptor existed as a water soluble component.'
In 1990, the finding of cannabinoid receptors"' expressed at significant levels
in distinct regions of the brain suggested the existence of an endogeneous
ligand for this receptor. Then, in the early 1990's, alternate ideas that an
endogenous cannabimimetic compound might be a lipophilic constituent like
THC (A9-tetrahydrocannabinol) (2), the psychoactive principle in cannabis,
resulted in the successful isolation and characterization of an
endocannabinoid.111
During the past five years, researchers have revealed that anandamide
shares pharmacological properties with THC and other cannabinoids in both
the central nervous systems (CNS) and peripheral systems.
A few structure-activity relationship (SAR) studies on anandamide have
been conducted to gain insight into how anandamide and THC bind to
cannabinoid receptors. These studies which would allow the development of
117
better ligands for these receptors possibly useful in therapeutical applications.
Recent studies have shown that both chain length and degree of unsaturation of
the fatty acid chain are critical in determining the affinity of the ligands for
binding to receptors."' Hydroxylation of anandamide by lipoxygenase at
different olefinic positions also resulted in changes in receptor affinity. 119,120 In
this chapter, I will describe the preparation and structure elucidation of several
hydroxy anandamides produced in a non-specific manner (autoxidation),
followed by a discussion of the binding affinities of these derivatives to the CBI
cannabinoid receptor. Ken Soderstrom in the laboratory of Dr. Thomas Murray
was responsible for the competitive receptor ligand displacement assay with a
variety of hydroxyanandamides.
0
H
N
H
CH3
1
CH3
118
RESULTS AND DISCUSSION
A small quantity of anandamide was oxidized (see experimental). Half of
this aliquot was reduced to gain some insight into the potential activity of these
derivatives on the CB1 receptor prior to large scale synthesis. Both portions of
the hydroperoxy and hydroxy derivatives exibited activity in the receptor binding
assay. Consequently, a large scale oxidation of anandamide was performed,
followed by reduction in the presence of triphenylphosphine. Separation of
these products was achieved using a Sep-Pak column containing 500 mg of
silica gel, followed by normal phase HPLC. Figure VI.1 presents a general
scheme for the production and the elution profile of oxidized anandamides in
normal phase HPLC.
Identification of each peak was determined using a combination of NMR
spectroscopy and GC-EIMS spectroscopy. The peaks 2, 3, 4, 6, and 7 were
identified as unreacted starting material, anandamide (1), 15­
hydroxyanandamide (3), 11-hydroxyanandamide (4), mixture of 9- and 8­
hydroxyanandamides (5) (6), and 5-hydroxyanandamide (7), respectively
(Figure VI.2).
GC-EIMS analysis of the methyl ester-TMS ether derivative of peak 3
identified its structure to be 15-hydroxyanandamide. Its mass spectra exhibited
distinct ions at m/z 436, 335, and 173 (Figure VI. 3). The fragment ion at m/z
407 was produced from a rearrangement of the trimethylsilazane moiety and
subsequent cleavege of the C14-C15 bond. This is consistent with a previous
119
0
Anandamide
1. thin film
2. 02, 5 min, rt
Mixture of ROOH's
3. reduction with triphenyl phosphine
4. partial separation by Sep-Pak
Mixture of hydroxy Anandamides
5. normal phase HPLC
Figure VI.1 Preparation and purification of various hydroxyanandamides
120
0
1'
.........,..............F.://...OH
15W)
N
H
11
OH
15-hydroxyanandamide (3)
O
NOH
H
HO
11-hydroxyanandamide (4)
0
9-hydroxyanandamide (5)
8-hydroxyanandamide (6)
5-hydroxyanandamide (7)
Figure VI.2 Structures of various hydroxyanandamides
121
(A)
Abundance
12000
Average of 18.890 to 19.086 min.: 1V90-3.D
713
173
10000
8000­
173
6000
15-hydroxyanandamide (3)
4000
407
2000­
260
113
50
p/z-->
(B)
100
Abundance
250001
150
200
250
492
507
367
314
300
436
350
400
450
500
Average of 18.211 to 18.408 min.: 1V90-4.D
215
arms
20000
15000
11-hydroxyanandarnide (4)
100001
384
5000,
355
17116
155
204
260
29412
417
446
0
/2-->
50
100
Abundance
(C)
150
200
250
300
350
400
492
507
450
500
Average of 18.231 to 18.259 min.: IV91-6B.D ( +, *)
25000
20000
15000
10000
5000
0
8/2
(D)
100
50
Abundance
150
200
250
300
350
400
450
Average of 18.853 to 18.884 min.: 1V91-6A.D (+,*)
205.­
8000
.
7000
344
6000
8-Indronenanclanwle(8)
5000
4000
254
3000
211
2000
103
75
->
:46
281 315
I.
0
4/2
164
6
1000
SO
100
150
200
250
300
370
350
492
417
515
400
450
500
Figure V1.3 Mass spectra of (A)15-hydroxyanandamide, (B) 11­
hydroxyanandamide, (C) 9-hydroxyanandamide, (D) 8-hydroxyanandamide.
122
0
OTMS
TMS
A
m/z = 407
OTMS
m/z = 355
TMS
B
m/z = 355
0
NOTMS
TMS
C
m/z = 327
0
NOTMS
TMS
D
m/z = 315
Figure VI.4 Proposed structures of the ions at m/z 407, 355, 327, 315
formed due to rearrangement process corresponding to 15- (A), 11- (B), 9­
(C), and 8-hydroxyanandamides (D), respectively.
H-5
H-6
H-8
H-9
H-11
H-17
H-18
H-19
H-2
H-20
H-1'
H-2'
H-4
H-7
H -3
H-12
H-10
H-14
H-13
H-15
H-16
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
PPM
Figure VI.5 1H NMR spectrum of 15-hydroxyanandamide (3).
2.0
1.5
1.0
H -16
H-20
H-17
H 18
H-19
4.5
4.0
PPM
3.5
3.0
2.5
2.0
i.5
Figure VI.6 1H-1H COSY of 15-hydroxyanandamide (3).
4=,
125
0
5.40
5.40
2.98a
5.40
5.40
6.56
3.42
1.73
2.12
2.21
1.55
1.31
3.70
CH3
0.89
4.18
5.40
5.98
1.31
1.31
5.72
OH
OH
15-hydroxyanandamide (3)
Coupling constant for 3 (400 MHz, CDCI3): J2_3 = 7.4, J3-4 = 7.4, J11-12 = 11.0
= 4.7
J13-14 = 15.2, J14-15 = 6.3, J19-20 = 7.0,
0
OH
5.40
5.69
5.98
6.5
2.82a
4.20
2.98a
5.40
5.40
5.40
3.43
1.75
5.40
1.51
2.28
2.05
1.31
OH
3.72
CH3
0.89
1.31
1.31
5-hydroxyanandamide (7)
Coupling constant for 7 (400 MHz, CDCI3): J2_3 = 7.4, J3.4 = 7.4, J8_8= 6.4
J6-7 = 15.2, J8.9 = 11.0, J19-20 = 7.0, J1,_2. = 4.7
Figure VI.7. 1H NMR data assignments for 15- (3) and 5-hydroxyanandamides (7)
alnterchangeable.
H-20
H-17
H-18
H-19
H-9
H-11
H-2
H-2'
H-12
H-14
H-15
H-4
H-1'
H-13
H-16
H-10
H-3
H-8
H-7
H-5
H-6
It\
7.0
6.5
6.0
5.5
5.0
4.5
4,0
PPM
3.5
3.0
2.5
2.0
Figure VI.8 1H NMR spectrum of 5-hydroxyanandamide (7).
15
1.0
H-2
H-3 H-4
H-5
cmo
O
03
a
O
7.0
6.5
6.0
5.5
5.0
4.5
4.0
PPM
3.5
3.0
2.5
2.0
Figure VI.9 11-1-1I-1 COSY of 5-hydroxyanandamide (7).
1.5
1.0
128
report (Figure VI.4). 119,120 Additional analysis of the one- and two-dimensional
NMR spectra of this compound confirmed it as 15-hydroxyanandamide (3)
(Figure VI.5). The 1H NMR spectrum showed the presence of one 0-bearing
methine resonance at 5 4.18 and conjugated olefinic proton resonances at S
6.56, 5 5.98, 5 5.72, and 5 5.40. The presence of two methylene proton
resonances at 5 2.82 and at 5 2.98 in the 1H NMR spectrum of this product
suggested that it was possibly either 5- or 15-hydroxyanandamide. Other
possible derivatives, including 8-, 9-, 11-, and 12-hydroxyanandamides, could
be eliminated because these derivatives do not have two methylene proton
resonances but rather contain a single proton resonance around 5 2.82 to S
2.98. Further analysis of the 1H-1H COSY spectrum indicated the identity of
peak 3 as 15-hydroxyanandamide (3) (Figure VI.6). The methine proton
resonance at S 4.18 bearing the hydroxyl group showed COSY correlation to
the methylene proton resonance at 8 1.55 (H-16) which was correlated to three
methylene proton resonances at 8 1.31 (H-17, H-18, H-19). Furthermore, the 1H
NMR spectrum of this compound was compared to that of 15­
hydroxyanandamide previously reported and it proved to be identical.119 The 1H
NMR data assignments for 15- (3) and 5-hydroxyanandamides (7) are shown in
Figure VI.7.
Peak 4 was identified as 11-hydroxyanandamide (4) on the basis of GC­
EIMS, observing significant fragment patterns at m/z 384 and 225. A fragment
129
ion peak at m/z 355 was also produced from a rearrangement of 11­
hydroxyanandamide (4).
Peak 6 was found to be mixture of the 9-and 8-hydroxyanandamides as
determined by 11-I NMR spectroscopy. Repeated normal phase HPLC of this
fraction eventually yielded separation of the two derivatives. The first portion
was identified as 9-hydroxyanandamide (5). The identity of this product was
supported by GC-EIMS, exhibiting ions at m/z 356 together with 327 as a
fragment ion from a rearrangement process. The second portion was verified
as 8-hydroxyanandamide (6) which was characterized by ions at m/z 344 and
265. The peak observed at m/z 315 was due to a rearrangement of 6.
Peak 7 was identified as 5-hydroxyanandamide (7) based on detailed
analysis by 1H NMR spectroscopy (Figure VI.8). Analysis of the 1H NMR
suggested either 5- or 15-hydroxyanandamide based on the presence of the
two methylene proton resonances at 6 2.97 and 6 2.85. The 1H NMR spectrum
of this peak showed the presence of a methine (8 4.2, H-5) bearing a secondary
hydroxyl group whose location, 5-hydroxy, was shown by spin couplings in 1H­
1H COSY spectrum (Figure VI.9). In the 1H-1H COSY, the methine proton
resonance at 8 4.2 bearing the hydroxyl group showed a correlation to the
methylene proton resonance at 8 1.51 (H-4) which was correlated to another
methylene proton resonance at 6 1.75 (H-3). This methylene at 8 1.75 (H-3)
was further correlated to the methylene protons at 6 2.28 (H-2) proximate to the
amide ketone group.
130
The Yamamoto group in 1995 reported inhibitory effects on electricallyevoked contraction of mouse vas deferens assay with 12S- and 15S­
hydroxyanandamides prepared from the incubation of [14C]anandamide with
several purified lipoxygenases.119 The 15-hydroxyanandamide was active with
an IC50 of 0.63 !AM as well as anandamide (0.17 gM), while the 12­
hydroxyanandamide was inactive ( > 10 liM). However, the cannabinoid
binding assay was not performed with these derivatives.
At the same time, the Bornheim group reported a CB1-binding assay with
11-, 12-, and 15-hydroxyanandamides.12° The 12-hydroxyanandamide was the
most active with Ki (affinity of ligand for receptor) of 0.031 p,M. Anandamide was
the next most active with Ki of 0.071 liM, followed by the 15­
hydroxyanandamide (0.418 p,M), and 11-hydroxyanandamide (1.1 p,M).
In this study, we explored the effect of hydroxylation of anandamide on
the activity in the CB1 receptor binding assay. Anandamide was the most active
with Ki of 0.51 µM in this assay (Figure VI.9). The fact that 15­
hydroxyanandamide (3) was active (0.74 p,M) while 11-hydroxyanandamide (4)
was inactive (3.3 p,M) is in good agreement with the results described above.
The significance of this study is the discovery that the 9- (5) and 8­
hydroxyanandamide (6) are active at 2.6 pM and 2.1 gM, respectively. The 5­
hydroxyanandamide (7), a new compound, was found to be relatively inactive
(8.24 p.M). All of these hydroxy derivatives are racemic. Further study is
100
90
80
70
60
50
40
30
20
Anandamide
Ki = 0.51 RM
0 15-Hydroxyanandamide
KI = 0.74 piM
Mixture of 9- and 8-Hydroxyanandamide
Ki = 1.83 WI
11-Hydroxyanandamide
Ki = 3.30IAM
A 5-Hydroxyanandamide
Ki = 8.24 p.M
10
0
-8
1
-7
-6
1
-5
Iog(Competitor)
I
-4
,
-3
Figure VI.10 Inhibition of the specific binding of [3H]CP -55940 to rat neuronal
membrane by anandamide and oxidized anandamide derivatives.
Specific Binding
132
ongoing to attempt chiral separation of these mixtures, and to investigate which
form of the isomers is active to CB1 receptor binding.
EXPERIMENTAL
General Experimental Procedures. NMR spectra were recorded on a
Bruker AM 400 spectrometer in the solvent specified. Chemical shifts were
referenced to solvent CDCI3 signal at 7.27 ppm for 1H NMR. Gas
chromatography/mass spectrometry (GC/MS) was carried out utilizing a
Hewlett-Packard 5890 Series II GC connected to a Hewlett-Packard 5971 mass
spectrometer.
Production and Isolation of Hydroxyanandamides. Anandamide (8 mg)
purchased from Cayman Chemicals (Ann Arbor, MI) was added into a round
bottom flask (250 mL). Pure molecular oxygen was charged for five minutes
into the flask which was slowly rotated on a rotary evaporater to make a thin film
on the surface of the flask. The reaction was terminated after 4 hr. The resulting
hydroperoxide derivatives were reduced with triphenylphosphine (5.0 mg) in
2.0 mL Et20, and the reaction mixture was stirred at rt for 2 hr. The solvent was
reduced by a rotary evaporater. The concentrate was diluted with 50%
EtOAc /hexanes, and passed through a Si Sep-Pak column containing 500 mg
of silica gel to remove triphenylphosphine. The desired products were further
eluted by 10% Me0H/CHCI3 on Sep-Pak. Repetitive HPLC on a Phenomenex
133
Si column (10 mm, 500 x 10 mm) of the oxidized products at 9 mL/min with
65:32:3 (v/v) EtOAc /hexanes /MeOH resulted in the elution of 15­
hydroxyanandamide (0.5 mg), 11-hydroxyanandamide (0.2 mg), 9­
hydroxyanandamide (0.2 mg), 8-hydroxyanandamide (0.2 mg), and 5­
hydroxyanandamide (0.4 mg), respectively.
Gas Chromatography-Mass Spectrometry (GC-MS). A portion of
oxidized anandamide derivatives was derivatized to trimethylsilyl (TMS) ethers
with 2 drops of hexamethyldisilazane (HMDS), 2 drops chlorotrimethylsilane
(TMS-CI), and 2 drops pyridine for 2 h. The reaction mixture was concentrated
in vacuo and the residue was resuspended in hexanes. After centrifugation, the
supernatant was injected into the GC-MS. The TMS ethers of the
hydroxyanandamides were chromatographed by GC-MS using a HewlettPackard HP 5890 Series II gas chromatograph and HP 5971 mass
spectrometer. For the injection, li_tl_ of the sample volume was injected into a
Hewlett-Packard HP-1 crosslinked methyl silicone capillary column (12.5 m x 2
mm x 0.33 p.m) operating with helium as the carrier gas at a temperature
program of 100 °C initial, 10 °C/ min ramp, 240 °C final, 250 °C injector, and 280
°C detector.
Evaluation of hydroxyanandamides as ligands to the cannabinoid
receptor Ken Soderman in the laboratory of Dr. Thomas Murray was
134
responsible for the competitive receptor ligand displacement assay with a
variety of hydroxyanandamides.
135
CHAPTER VII
CONCLUSION
Until recently marine cyanobacteria have not been explored extensively
for their potential uses as pharmaceuticals. However, they are emerging as a
new source of novel biologically active metabolites. Particulary, various
collections of L. majuscula have provided diverse bioactive secondary
metabolites. In this regard, my interest has focused on the investigation of L.
majuscula.
As a part of our research project aimed at the discovery of new bioactive
compounds from marine algae, we isolated two new and potent antimitotic
metabolites, curacins B and C, from a Curacao collection of Lyngbya majuscula.
In addition, four curacin A analogs were prepared by semisynthetic efforts. The
biological and biochemical properties of the new natural products and synthetic
derivatives of curacin A were examined.
The isolation and identification of curacins from L. majuscula prompted
us to investigate another collection of this species. Consequently, a new
cytotoxic lipopeptide, microcolin C along with three known compounds,
microcolins A, B, and D, were isolated by bioassay-guided fractionation of the
organic extract from Curacao and Granada collections of Lyngbya majuscula.
Microcolin C was found to have an interesting profile of cytotoxicity to human
cancer-derived cell lines. Microcolins A-C also exhibited potent toxicity to brine
shrimp (LC50 of 0.004 µg /ml, 0.4 µg /ml, and 0.2 µg /ml, respectively).
136
Investigation of an Indonesian red alga Vidalia sp. yielded a new
secondary metabolite, vidalenolone. The structure of this new
cyclopentenolone-containing compound was determined by a combination of
spectroscopic methods.
Filamentous cells isolated from female gametophytes of the brown alga
Laminaria saccharina were cultured in a flask or bioreactor culture. These
cultures produced a variety of 15-lipoxygenase metabolites. Feeding
experiments were carried out to attempt the enhancement of these 15­
lipoxygenase metabolites. The presence of linoleic acid, y-linolenic acid and
jasmonic acid increased the production of the three oxylipins, 13-hydroxy-9,11­
octadecadienoic acid (13-HODE), 13-hydroxy-6,9,11,15-octadecatetraenoic
acid (13-HODTA), and 15-hydroxy-5,8,11,13-eicosatetraenoic acid (15-HETE),
2 to 4 times over the controls. On the other hand, addition of antibiotic did not
increase the yield, while a-linolenic acid suppressed production of these
metabolites. Arachidonic acid addition gave rise to a variety of HETE products
via autooxidation.
I described the preparation and structure elucidation of several hydroxy
anandamides including 5-, 8-, 9-, 11-, and 15-hydroxyanandamide. The aim of
this work was to develop better ligands for these receptors which may be of
some use in therapeutical applications. This study on the cannabinoid receptor
binding affinity revealed the following trend in activity: anandamide > 15­
hydroxyanandamide > 9-hydroxyanandamide > 8-hydroxyanandamide > 11­
hydroxyanandamide > 5-hydroxyanandamide.
137
In conclusion, the overall hypothesis of this work is based on the
published literature that marine organisms, especially marine algae, have been
an important source of structurally novel classes of bioactive compounds.
Hence, continued investigations of these life forms for this purpose seems
reasonable and warrented. My investigations reported in this thesis of various
marine algae have yielded several new and bioactive metabolites, such as
curacins B and C, microcolin C, and vidalenolone. Thus, these findings indicate
that marine algae are a reliable source for bioactive metabolites and there is
more work to be investigated with these marine organisms as a source of
potential drugs.
138
BIBLIOGRAPHY
1.
Farnsworth, N. R. Screening plants for new medicines. In: Biodiversity
(Eds. Wilson, E. 0. and Peters, F. M.), Academic Press, New York, 1988,
61-73.
2.
Barth, R. H.; Broshears, R. E. The invertebrate world, CBS College
Publishing, New York, New York. 1982, 646.
3.
Bergmann, W. J. Mar. Res. 1949, 8, 137-176.
4.
Carte, B. K. Bio Science 1996, 46, 271-286.
5.
Clark, A. M. Pharmaceutical Research, 1996, 13, 1133-1141.
6.
Fenical, W. Chem. Rev. 1993, 93, 1673-1683.
7.
8.
Burkholder, P. R.; Pfister, R. M.; Leitz, F. P. Appl. Microbiol. 1966, 14,
649.
Laatsch, H.; Pudleiner, H. Liebigs. Ann. Chem. 1989, 9, 863.
9.
Needham, J.; Andersen, R.; Kelly, M. T. Tetrahedron Lett. 1991, 32, 315.
10.
Gustafson, K.; Roman, M.; Fenical, W. J. Am. Chem. Sco. 1989, 111,
7519-7524.
11.
Tapiolas, D. M.; Roman, M.; Fenical, W. J. Am. Chem. Soc. 1991, 113,
4682-4683.
12.
Trischman, J. A.; Tapiolas, D. M.; Jensen, P. R.; Dwight, R.; Fenical, F. J.
Am. Chem. Soc. 1994, / /6, 757-758.
13.
Kobayashi, J.; Ishibashi, M. The Alkaloids , Brossi, A.; Cordell, G. A.
Academic Press: San Diego, 1992, 41, 41-124.
14.
Yasumoto, T; Yasumura, D.; Yotsu, M.; Endo, A.; Kotaki, Y. Agric. Bio.
Chem. 1986, 50, 793-795.
15.
Do, H. K.; Kogure, K.; Simidu, U. Applied and Environmental
Microbiology, 1990, 56, 1162-1163.
16.
Ikawa, M.; Wegner, K.; Faxall, T. L.; Sasner, J. J., Jr. Toxicon 1982, 20,
747.
17.
Kodama, M.; Ogata, T.; Sato, S. Agric. Biol. Chem. 1988, 52, 1075.
139
18.
Moore, R. E. Constituents of blue-green algae in Marine natural products:
Chemical and biological perspectives, Vol. IV. P. J. Scheuer, ed.
Academic Press, New York. 1981, 1-52.
19.
Francis, G. Nature 1878, 18, 11-12.
20.
Bishop, C. T.; Anet, E.; Gorham, P. R. Can. J. Biochem. Physiol. 1959,
37, 453-471.
21.
Carmichael, W. W.; Beasley, V.; Bunner, D. L.; Eloff, J. N.; Falconer,
(.;Gorham, P.; Harada, K.; Krishnamurthy, T.; Min-Juan, Y.; Moore, R. E.;
Rinehart, K.; Runnegar, M.; Skulberg, 0. M.; Watanabe, M. Toxicon
1988, 26, 236-241.
22.
Carmichae, W. W. J. Appl. Bacteriol. 1992, 72, 445-449.
23.
Falconer, I. R.; Beresford, A. M.; Runnegar, M. T. C. Med. J. Chem. 1983,
1, 511-514.
24.
Rinehart, K.; Harada, K.; Namikoshi, M.; Chen, C.; Harvis, C. A. J. Am.
Chem. Sco. 1988, 110, 8557-8558.
25.
Ohta, T.; Sueoka, E.; lida, N.; Komori, A.; Suganuma, M.; Nishiwaki, R.;
Tatematsu, M.; Kim, S-J.; Carmichael, W. W.; Fujiki, H. Cancer Res.
1994, 54, 6402-6406.
26.
Stratmann, K.; Burgoyne, L.; Moore, R. E.; Patterson, G. M. J. Org. chem.
1994, 59, 7219-7226.
27.
Trimurtulu, G.; Ohtani, I.; Patterson, G. M.; Moore, R. E.; Corbett, T. H.;
Valeriote, F. A.; Demchik, L. J. Am. Chem. Soc. 1994, 116, 4729-4737.
28.
Trimurtulu, G.; Ogino, J.; Heltzel, C. E.; Husebo, T. L.; Jensen, C. M.;
Larsen, L. K.; Patterson, G. M.; Moore, R. E.; Mooberry, S. L.; Corbett, T.
H.; Valeriote, F. A.J. Am. Chem. Soc. 1995, 117, 12030-12049.
29.
Smith, C. D.; Zhang, X. J. Biol. Chem. 1996, 271, 6192-6198.
30.
Cardellina, J. H.; Marner, F-J.; Moore, R. E. Science, 1979, 204, 193­
195.
31.
Carter, D. C.; Moore, R. E.; Mynderse, J. S.; Niemczura, W. P.; Todd, J. S.
J. Org. Chem. 1984, 49, 236-241.
140
32.
Gerwick, W. H.; Jiang, J. D.; Agarwal, S. K.; Farmer, B. Tetrahedron,
1992, 48, 2313-2324.
33.
Frankmolle, W. P.; Knubel. G.; Moore, R. E.; Patterson, G. M. L. J.
Antibiotics, 1992, 45, 1458-1466.
34.
Mayer, A. M. S.; Jacobs, R. Manoalide: An antiinflammatory and
analgesic marine natural product In Biomedical importance of marine
organisms, (Fautin, D.) California Academy of Sciences, San francisco,
1988, 133-142.
35.
Campbell, W. B.; Halushka, P. V. Lipid-Derived Autocoids: Eicosanoids
and Platelet-Activating Factor In The Pharmacological Basis of
Therapeutics, 9th ed.; Hardman, J. G.; limbird, L. E.; Molinoff, P. B.;
Ruddon, R. W.; Gilman, A. G., Ed; Mcgraw-Hill: New York, 1996, 601­
616.
36.
Cohen, M. M. Biological protection with prostaglandins, Vol. I. CRC
Press, Inc., Boca Raton, FL. 1985.
37.
Feuerstein, G.; Hallenbeck, J. M. FASEB, 1987, 1, 186-192.
38.
Piomelli, D. Arachidonic Acid in Cell Signalling, Champman & Hall,
1996, 15-53.
39.
Weinheimer, A. J.; Sparggins, R. L. Tetrahedron Lett. 1969, 59, 5185­
5188.
40.
Coll, J. C. Chem. Rev. 1992, 92, 613-631.
41.
Carmely, S.; Kashman, Y.; Loya, Y.; Benayahu, Y. Tetrahedron Lett.
1980, 21, 875-878.
42.
Gerwick, W. H.; Moghaddam, M.; Hamberg, M. Archives of Biochemistry
and Biophysics, 1991, 290, 436-444.
43.
Gregson, R. P.; Marwood, J. F.; Quinn, R. J. Tetrahedron Lett. 1979, 46,
4505-4506.
44.
Gerwick, W. H.; Nagle, D. G.; Proteau, P. J. Topics in Current Chemistry,
Marine Natural Products-Diversaty and Biosynthesis, (Scheuer, P. J. Ed),
Springer-Verlag Berlin Heidelberg, 1993, Vol. 167, 117-180.
45.
Gerwick, W. H. Chem. Rev. 1993, 93, 1807-1823.
141
46.
Gerwick. W. H.; Bernart, M. W. in Pharmaceutical & Bioactive Natural
Products. Marine Biotechnology, Vol 1 (Attaway, D. W.; Zaborsky, 0. R.
Eds), Plenum Press, New York,1993, 101-152.
47.
Gerwick, W. H. Biochimica et Biophysica Acta, 1994, 1211, 243-255.
48.
Gerwick, W. H. Lipid, 1996, 31, 1215-1230.
49.
Cardellina Ii, J. H.; Dalietos, D.; Marner, F-J.; Mynderse, J. S.; Moore, R.
E. Phytochemistry, 1978, 17, 2091-2095.
50.
Cardellina II, J. H.; Moore, R. E. Tetrahedron, 1980, 36, 993-996.
51.
Nagle, D. G.; Gerwick. W. H. Tetrahedron Lett. 1990, 31, 2995-2998.
52.
Solem, M. L.; Jiang, Z. D.; Gerwick, W. H. Lipids, 1989, 24, 256-260.
53.
Kugel, C.; Lellouche, J-P.; Beaucourt, J.-P. Tetrahedron Lett. 1989, 30,
4947-4959.
54.
Gundlach, H.; Muller, M.; Kutchan, T. M.; Zenk, M. H. Proc. Natl, Acad.
Sci. USA, 1992, 89, 2389-2393.
55.
Pollio, A.; Greca, M. D.; Monaco, P.; Pinto, G.; Previtera, L. Biochim.
Biophys. Acta 1987, 963, 53-60.
56.
Bernart, M. W.; Gerwick. W. H. J. Nat. Prod. 1992, 56, 245-259.
57.
Proteau, P. J.; Gerwick, W. H. Tetrahedron Lett. 1992, 33, 4393-4396.
58.
Gardner, H. W. Biochim. Biophys. Acta 1991, 1084, 221-239.
59.
Todd, J. S.; Proteau, P. J.; Gerwick, W. H. Tetrahedron Lett. 1993, 34,
7689-7692.
60.
Kato, Y; Scheuer, P. J. Pure Appl. Chem. 1975, 41, 1.
61.
Cardellina, J. H.; Marner, F.-J.; Moore, R. E. Science, 1979, 204, 193.
62.
Cardellina, J. H.; Moore, R. E. Tetrahedron Lett., 1979, 22, 2007-2010.
63.
Simmons, C. J.; Marner, F.-J.; Cardellina, J. H.; Moore, R. E.; Seff, K.
Tetrahedron Lett., 1979, 22, 2003-2006.
64.
Orjala, J.; Nagle, D. G.; Hsu, V. L.; Gerwick, W. H. J. Am. Chem. Soc.
1995, 117, 8281-8282.
142
65.
Wu, M., Ph.D. Thesis, Oregon State University, Corvallis, 1996, 18-47.
66.
Gerwick, W.H.; Proteau, P.J. ; Nagle, D.G.; Hamel, E. ; Blokhin, A.; Slate,
D.L. J. Org. Chem. , 1994, 59, 1243-1245.
67.
Hamel, E.; Blokhin, a. V.; Nagle, D. G.; Yoo, H. D.; Gerwick, W. H. Drug
Dev. Res. 1995, 34, 110-120.
68.
Yoo, H. D.; Gerwick, W. H. J. Nat. Prod. 1995, 58, 1961-1965.
69.
Couperus, P. A.; Clague, A. D. H.; Dongen, J. P. C. M. Org. Magn. Reson.
1976, 8, 426-431.
70.
Nagle, D. G.; Gerald, G. S.; Yoo, H. D.; Gerwick, W. H.; Kim,T. S.; Nambu,
M.; White, J. D.Tetrahedron Lett., 1995, 36, 1189-1192.
71.
Blokhin, A. V.; Yoo, H. D.; Geralds, R. S.;Nagle, D. G.; Hamel, E.; Gerwick,
W. H. Molecular pharmacology 1995, 48, 523-531.
72.
Takahashi, M.; Iwasaki, S.; Kobayashi, H.; Okuda, S.; Murai, T.; Sato, Y.
Biochim. Biophys. Acta 1987, 929, 215-223.
73.
Pettit, G. R.; Kamano, Y.; Herald, C. L.; Tuinman, A. A.; Boettner, F. E.;
Kizu, H.; Schmidt, J. M.; Baczynskyj, L.; Tomer, K. B.; Bontems, R. J. J.
Am. Chem. Soc. 1987, 109, 6883-6885.
74.
Bollag, D. M.; McQueney, P. A.; Zhu, J.; Hensens, 0; Koupal, L.; Liesch,
J.; Goetz, M.; Lazarides, E.; Woods, C. M. Cancer Res., 1995, 55, 2325­
2333.
75.
Pinard, P. V.; Lai, J. Y.; Yoo, H. D.; Nagle, D. G.; Marquez, B.; Falck, J. R.;
Gerwick, W. H.; Hamel, E.; Nambu, M; White, J. D. Molecular
Pharmacology in press 1997.
76.
Moore, R.E.; Patterson, G.ML.; Carmichael, W.W. Memoirs Calif. Acad.
Sci, No. 13, San Francisco, 1988, pp 143-150; Carmichael, W.W., J.
Appl. Bact., 1992,72, 445-459.
77.
Marner, F.-J.; Moore, R. E. Phytochemistry, 1978, 17, 553-554
78.
Cardellina, J. H.; Marner, F.-J.; Moore, R. E. J. Am. Chem. Sco. 1979,
101, 240.
79.
Nagle, D. G.; Paul, V. J.; Robert, M. A. Tetrahedron Lett. 1996, 37, 6263­
6266.
143
80.
Petcher, T. Y.; Weber, H. P.; Ruegger, A. He Iv. Chim. Acta. 1976, 59,
1480- 1489.
81.
Tanaka, H.; Kuroda, A.; Marusawa, H.; Hatanaka, H.; Kino, T.; Goto, T.;
Hashimoto, M. J. Am. Chem. Sco. 1987, 109, 5031-5033.
82.
Vezina, C.; Kudelski, A.; Sehgal, S. N. J. Antibiotics, 1975, 28, 721-726.
83.
Kay, J. E.; Benzie, C. R.; Goodier, M. R.; Wick, C. J.; Doe, S. E. A.
Immunology 1989, 67, 473-477.
84.
Kay, J. E.; Moore, A. L.; Doe, S. E. A.; Benzie, C. R.; Schonbrunner, R.;
Schmid, F. X.; Halestrap, A. P. Transplantation Proc. 1990, 22, 96-99.
85.
Koehn, F. E.; Long ley, R. E.; Reed, J. K. J. Nat. Prod., 1992, 55, 613-619.
86.
Koehn, F. E.; 0. J. McConnell, 0. J.; Long ley, R. E.; Sennett, S. H.; Reed,
J. K. J. Med. Chem., 1994, 37, 3181-3186.
87.
Mattern, R. -H.; Gunasekera, S.; McConnell, 0. Tetrahedron, 1996,52,
425-434.
88.
Decicco, C. P.; Grover, P. J. Org. Chem., 1996,61, 3534-3541.
89.
Andrus, M. B., Li, W.; Keyes, R. F. J. Org. Chem., 1997, 62, 5542-5549.
90.
Laumen, K; Schneider, M. Tetrahedron Lett., 1985, 26, 2073-2076.
91.
Deardorff, D.; Mathews, A. J.; McMekin, D. S.; Craney, C. L. Tetrahedron
Lett.,
1986, 27, 1255-1256.
92.
Herzig, J. J. Org. Chem., 1986, 51, 727-730.
93.
Boyd, M. R.; Paull, K. D. Drug Development research, 1995, 34, 91-109.
94.
Hoek, C. V. D.; Mann, D. G.; Jahns, H. M. "Algae, an introduction to
phycology" Cambridge University Press, 1995, 48-95
95.
Stallard, M. 0.; Faulkner, D. J. Chemical Constituents of the Digestive
Glandin of the Sea Hare Aplysia californica-I. Importance of Diet.Comp.
Biochem. Physiol. B, 1974, 49, 25-35.
96.
Fuller, R. W.; Cardellina, J. H.; Jurek, J.; Scheuer, P.; Lindner, B. A.;
McGuire, M.; Gary, G. N.; Steiner, J. R.; Clardy, J.; Menez, E.; Shoemaker,
R. H.; Newman, D. J.; Snader, K. M.; Boyd, M. R. J. Med. Chem. 1994,
37, 4407-4411.
144
97.
Wiemer, D. F.; Idler, D. D.; Fenical, W. Experientia, 1991, 47, 851-853.
98.
Patti, A.; Morrone, R. J. Nat. Prod. 1992, 1, 53-57.
99.
Kazlauskas, R.; Murphy, P. T.; Wells, R. J. Aust. J. Chem. 1982, 35, 219­
220.
100. Thiericke, R.; Zeeck, A.; Nakagawa, A.; Omura, S.; Herrold, R. E.; Wu, S.
T. S.; Beale, J. M.; Floss, H. J. Am. Chem. Soc. 1990, 112, 3979-3987.
101. Clissold, D; Thickitt, C. Natural Product Report, 1994, 611-637.
102. Cragg, G. M.; Schepartz, S. A.; Suffness, M.; Greyer, M. R. J. Nat. Prod.
1993, 56, 1657-1668.
103. Rouhi, A. M. C&EN 1995, 42-44.
104. Anderson, I New Scientist 1995, 2005, 5.
105. Pettit, G. R.; Kamano, Y.; Herald, C. L.; Tuinman, A. A.; Boettner, F. E.;
Kizu, H.; Tomer, K. B.; Bonterms, R. J. J. Am. Chem. Soc. 1987, 109,
6883-6885.
106. White, J. D.; Kim, T. S.; Nambu, M. J. Am. Chem. Soc. 1995, 117, 5612­
5613.
107.
Rossi, J. V.; Roberts, M. A.; Yoo, H.-D.; Gerwick. W. H. J. Appl. Phycol.
1997, 0, 1-10.
108. Hirosawa, T.; Wolk, P. J. Gen. Microbiol. 1979, 114, 433-441.
109. Armstrong, J. F.; Janda, K. E.; Alvarado, B.; Wright, A. E. J. Appl. Phycol.
1991,3, 227-282.
110. Panda, A. K.; Mishra, S.; Bisaria, V. S.; Bhojwani, S. S. Enzyme and
Microbial technology, 1989, 11, 386-397.
111. Rorrer, G. L.; Modrell, J.; Zhi, C.; Yoo, H.-D.; Nagle, D. G.; Gerwick, W. H.
J. Appl. Phycol. 1995, 7, 187-198.
112. Personal communication with Dr. W. H. Gerwick.
113.
Devane, D. A.; Hanus, L.; Breuer, A.; Pertwee, R. G.; Stevenson, L. A.;
Griffin, G.; Gibson, D.; Mandelbaum, A.; Etinger, A.; mechoulam, R.
Science, 1992, 258, 1946-1949.
114.
Marzo, V. D.; Fontana, A.; cadas, H.; Schinelli, S.; Cimino, G.; Schwartz,
J. C.; Piomelli, D. Nature, 1994, 372, 686-691.
145
115.
Desarnaud, F.; Cadas, H.; Piomelli, D. J. Biol. Chem., 1995, 270, 6030­
6035.
116. Evans, D. M.; Lake, J. T.; Johnson, M. R.; Howlett, A. C. J. Pharmacol
Exper. Ther. 1994, 268, 1271-1277.
117. Matsuda, L. A.; Lolait, S. J.; Brownstein, M. J.; Young, A. C.; Bonner, T. I.
Nature, 1990, 346, 561-564
118. Felder, C. C.; Briley, E. M.; Axelrod, J.; Simpson, J. T.; Mackie, K.;
Devane, W. A. Proc. Natl. Acad. Sc!. USA 1993, 90, 7656-7660.
119. Ueda, N.; Yamamoto, K.; Yamamoto, S.; Tokunaga, T.; Shirakawa, E.;
Shinkai, H.; Ofawa, M.; Sato, T.; Kudo, I.; Inoue, K,; Takizawa, H.;
Nagano, T.; Hirobe, M.; Matsuki, N.; Saito, H. Biochimica et Biophysica
Acta 1995, 1254, 127-134.
120. Hampson, A.; Hill, W. A. G.; Zan-Phillips, M.; Makriyannis, A.; Leung, E.;
Eglen, R. M.; Bornheim, L. M. Biochimica et Biophysica Acta 1995, 1259,
173-179.
146
APPENDIX
Patellamide C
I
.1
.
8.0
7.5
7.0
6.5
i
6.0
5.5
5.0
4.5
4.0
PPM
3.5
3.0
Figure A.1 11-1 NMR spectrum of patellamide C in CDCI3.
180
170
160
150
140
130
120
110
go
9Q
PPM
80
70
60
50
40
Figure A.2 13C NMR spectrum of patellamide C in CDCI3.
30
20
10
0