AN ABSTRACT OF THE THESIS OF
Bingnan Han for the degree of Doctor of Philosophy in Pharmacy presented on
June 3. 2005.
Title: Natural Products Studies of the Marine Cyanobacteria
Lynqbya semiplena and Lyngbya majuscula
Abstract
Redacted for Privacy
This thesis details my investigations of marine cyanobacterial natural
products that resulted in the discovery of eighteen new secondary metabolites.
Isolation and characterization of these unique molecules were carried out using
different chromatographic techniques and by careful analyses of ID and 2D NMR
data, respectively.
Preliminary bioassays of the crude organic extract of the marine
cyanobacterium Lyngbya semiplena collected from Papua New Guinea in 1999,
showed good activity in the brine shrimp toxicity model at 10 ppm. Guided by this
assay, seven new anandamide-like fatty acid amides, semiplenamides A to G,
together with four cyclic depsipeptides, wewakpeptins A to 0, were identified. Due
to the structural resemblance of the novel ethanolamide derivatives
(semiplenamide A-G) with anandamide, (N-arachidonoyl-ethanolamine), an
endogenous agonist of cannabinoid receptors compounds, semiplenamide A-G
were tested on the well characterized proteins of the endocannabinoid system: 1)
the "central" cannabinoid GB1 receptors; 2) the anandamide membrane transporter
(AMT), which is responsible for anandamide cellular uptake; and 3) the fatty acid
amide hydrolase (FAAH), which catalyses anandamide hydrolysis. Three showed
modest potency in displacing radiolabeled anandamide from the cannabinoid
receptor (GB 1), and one was a modest inhibitor of the anandamide membrane
transporter (AMT). The wewakpeptins were tested for cytotoxicity to NCI-H460
human lung tumor and neuro-2a mouse neuroblastoma cells. Intriguingly,
wewakpeptin A and B were approximately 10-fold more toxic than C and 0 to
these cell lines.
Lyngbya majuscula has been recognized as a chemically and biologically
rich strain. Five new lyngbyabellin analogs, lyngbyabellins E-1, along with the
known compound dolabellin, originally isolated from sea hare Dolabella auricularia,
were identified from a 2002 Papua New Guinea collection of the marine
cyanobacterium Lyngbya majuscula. The lyngbyabellins were tested for
cytotoxicity to NCI-H460 human lung tumor and neuro-2a mouse neuroblastoma
cells and had
LC50
values between 0.2 and 4.8 pM. Intriguingly, lyngbyabellin E
and H appeared to be more active against the H460 cell line with
0.4 pM and 0.2 pM, respectively, compared to
LC50
LC50
values of
values of 1.2 and 1.4 pM in the
neuro-2a cell line. Lynbyabellin I was the most toxic to neuro-2a cells (LG50 0.7
pM), whereas lyngbyabellin G, was the least cytotoxic of all compounds to either
cell line. On the basis of this limited screening, it appears that lung tumor cell
toxicity is enhanced in the cyclic representatives, and overall potency is increased
in those containing an elaborated side chain.
Additionally, two new cytotoxins, aurilides B and C, which are closely
related to aurilide, originally isolated from the sea hare Do/abe/Ia auricularia, have
been identified from the same extract where the lyngbyabellins E-1 were isolated.
Aurilides B and C were tested for cytotoxicity to NCI-H460 human lung tumor and
neuro-2a mouse neuroblastoma cells. Interestingly, aurilide B was approximately
4-fold more toxic than C to these cell lines. The
LC50
for Aurilide B was 0.01 pM
and 0.04 pM for neuro-2a and H460 cells, respectively, and 0.05 pM and 0.13 pM,
respectively, for aurilide C. Aurilide B was evaluated in the NCI 60 cell line panel
and found to exhibit a high level of growth inhibition in leukemia, renal, and
prostate cancer cell lines with a Gl
less than 10 nM. Aurilide B showed net tumor
cell killing activity in the NCI's hollow fiber assay, an in vivo model for assessing a
chemical's anticancer activity.
©Copyright by Bingnan Han
June 3, 2005
All Rights Reserved
NATURAL PRODUCTS STUDIES OF THE MARINE CYANOBACTERIA
LYNGBYA SEMIPLENA AND LYNGBYA MAJUSCULA
by
Bingnan Han
A THESIS
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Doctor of Philosophy
Presented June 3, 2005
Commencement June, 2006
Doctor of Philosophy thesis of Bingnan Han presented on June 3, 2005.
APPROVED:
Redacted for Privacy
Major Professor, representing Pharmacy
Redacted for Privacy
Dean of the CoIIge of Pharmacy
Redacted for Privacy
Dean of the' ràdi1ate 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
Bingnan Han, Author
ACKNOWLEDGMENTS
I would like to express my gratitude to my major advisor, Dr. William
Gerwick, and my committee members, Dr. Philip Proteau, Dr. George Constantine,
Dr. Michael Schimerlik, and Dr. Claudia Maier for their support and academic
guidance throughout my graduate career at Oregon State University. I would also
like to acknowledge Brian Arbogast and Jeff Morre for their extensive mass
spectral work, and Roger Kohnert for his valuable NMR assistance. My
appreciation also goes to Dr. Vincenzo Di Marzo (Endocannabinoid Reserch
Group, Istituto di Chimica Biomolecolare, Italy), and Dr. Susan Mooberry
(Southwest Foundation for Biomedical Research, San Antonio, Texas).
I am very grateful to my current and former group colleagues for their
mentorship, insight, and most importantly, friendship. I would especially like to
acknowledge Dr. Kenneth Milligan, Dr. Patricia Flatt, Dr. Kerry McPhail, Dr.
Dougals Goeger, and Dr. Harald Gross for their continual inspiration and
motivation, Dr. Omar Mohamed Sabry, and Eric Andrianasolo for their friendship
and encouragement throughout the years. Completion of this degree would have
been impossible without them all.
Finally, I would like to express my gratitude to my wonderful family,
especially My Mother, Ruoli Wang, for their support and encouragement. I am
forever indebted to them for their patience and love through these challenging
years.
CONTRIBUTION OF AUTHORS
Anandamide cellular uptake assay, CB1 receptor binding assay, and fatty
acid amide hydrolase assay were carried out by Dr. Vincenzo Di Marzo
(Endocannabinoid Reserch Group, Istituto di Chimica Biomolecolare, Italy).
Cytotoxicity bioassays against human lung cancer and mouse neuro-2a cell lines
were carried out by Dr. Douglas Geoger in our laboratory. Actin-disrupting assay
was performed by Dr. Susan Mooberry (Southwest Foundation for Biomedical
Research, San Antonio, Texas). Collision induced ESI-MS/MS experiments in
chapter three of my thesis were performed by Dr. Claudia Maier at the Department
of Chemistry, OSU. Dr. Kerry McPhail and Dr. Harald Gross provided various
assistance in chapter four and five of my thesis.
TABLE OF CONTENTS
Page
CHAPTER ONE: GENERAL INTRODUCTION
HIGHLIGHTS OF MARINE NATURAL PRODUCTS .................................. I
BIOACTIVE SECONDARY METABOLITES FROM
MARINE CYANOBACTERIA............................................................... 14
THE SYMBIOSIS OF MARINE INVERTEBRATES AND MARINE
CYANOBACTERIA ........................................................................... 22
GENERAL THESIS CONTENTS ......................................................... 26
REFERENCES ................................................................................ 30
CHAPTER TWO: SEMIPLENAMIDES A G, FATTY ACID AMIDES FROM A
PAPUA NEW GUINEA COLLECTION OF THE MARINE CYANOBACTERIUM
LYNGBYA SEMIPLENA
ABSTRACT ...................................................................................... 38
INTRODUCTION .............................................................................. 38
RESULTS AND DISCUSSION ........................................................................ 40
EXPERIMENTAL ............................................................................................. 53
REFERENCES ................................................................................ 57
CHAPTER THREE: THE WEWAKPEPTINS, CYCLIC DEPSIPEPTIDES FROM A
PAPUA NEW GUINEA COLLECTION OF THE MARINE CYANOBACTERIUM
LYNGBYA SEMIPLENA
ABSTRACT .................................................................................... 58
INTRODUCTION .................................................................................. 59
RESULTS AND DISCUSSION ............................................................ 60
EXPERIMENTAL ............................................................................. 76
REFERENCES ................................................................................ 80
CHAPTER FOUR: ISOLATION AND STRUCTURE OF LYNGBYABELLIN
DERIVATIVES FROM A PAPUA NEW GUINEA COLLECTION OF THE MARINE
CYANOBACTERIUM LYNGBYA MAJUSCULA
ABSTRACT.................................................................................... 82
TABLE OF CONTENTS (Continued)
Page
INTRODUCTION .............................................................................. 83
RESULTS AND DISCUSSION ............................................................. 85
EXPERIMENTAL .............................................................................. 102
REFERENCES ................................................................................. 106
CHAPTER FIVE: ISOLATION, STRUCTURE, AND BIOLOGICAL ACTIVITY OF
AURILIDE B AND C FROM A PAPUA NEW GUINEA COLLECTION OF THE
MARINE CYANOBACTERIUM LYNGBYA MAJUSCULA
ABSTRACT .................................................................................... 108
INTRODUCTION .............................................................................. 109
RESULTS AND DISCUSSION ............................................................ 110
EXPERIMENTAL .............................................................................. 120
REFERENCES ................................................................................. 123
CHAPTER SIX: CONCLUSION ................................................................... 125
BIBLIOGRAPHY ........................................................................................ 129
APPENDICES ........................................................................................... 136
APPENDIXA ................................................................................... 137
APPENDIXB ................................................................................... 148
LIST OF FIGURES
Page
1.1
Marine toxins
1.2
Largest non-proteinaceous marine toxin .................................................. 4
1.3
Marine toxins ...................................................................................... 5
1.4
Anticancer marine natural products ........................................................ 7
1.5
Anticancer marine natural products ....................................................... 10
1.6
Antiinflammatory marine natural products ............................................... 11
1.7
Anti-Alzheimer marine natural products .................................................. 12
1.8
Antimalarial marine natural products...................................................... 13
1.9
Anti-HIV marine natural products .......................................................... 14
1.10
Chemical structures of natural products 30-35 ......................................... 20
1.1 1
Chemical structures of natural products 36-46 ......................................... 21
1.12
Chemical structures of natural products 47-54 ......................................... 26
1.13
Chemical structures of natural products 55-63 ......................................... 27
11.1
1H and 13C NMR spectra of semiplenamide A (1) ...................................... 42
11.2
COSY spectrum of semiplenamide A (1) ................................................. 43
11.3
HSQC spectrum of semiplenamide A (1) ................................................. 44
11.4
HMBC spectrum of semiplenamide A (1) ................................................. 45
11.5
Determination of the geometry at C6-C7 by a 1H-1H
.3
decoupling experiment ........................................................................ 46
11.6
Determination of the geometry at C2-C3 by a 1 D NOE experiment............... 46
111.1
1H NMR spectrum of wewakpeptin A (1) ................................................. 61
111.2
13C NMR spectrum of wewakpeptin A (1)................................................ 61
111.3
HSQC spectrum of wewakpeptin A (1) .................................................... 62
111.4
HMBC spectrum of wewakpeptin A (1) ................................................... 63
111.5
COSY spectrum of wewakpeptin A (1) ................................................... 64
111.6
HSQC spectrum of wewakpeptin A (1) ................................................... 65
111.7
Key fragments from collisionally induced ESI-MS/MS experiments
with wewakpeptin A (1).and wewakpeptin C (3) ......................................... 68
111.8
1H NMR spectrum of wewakpeptin B (2) .................................................. 70
111.9
13C NMR spectrum of wewakpeptin B (2) ................................................. 70
111.10
1H NMR spectrum of wewakpeptin C (3) ................................................. 71
LIST OF FIGURES
(Continued)
Page
111.11
13C NMR spectrum of wewakpeptin C (3) ................................................ 72
111.12
1H NMR spectrum of wewakpeptin D (4) ..................................................72
111.13
13C NMR spectrum of wewakpeptin D (4) .................................................... 72
111.14
Cytotoxicity of wewakpeptins A-D (1-4) ..................................................... 76
IV.1
H and 13C spectra of Iyngbyabellin E (1) .................................................... 87
IV.2
HSQC spectrum of
IV.3
HMBC spectrum of lyngbyabellin E (1) .................................................... 90
lV.4
HSQC-TOCSY spectrum of lyngbyabellin E (1) ......................................... 91
IV.5
COSY spectrum of
IV.6
Diagram of all possible rotamers for the two possible epimers
lyngbyabellin
lyngbyabellin
E (1) ..................................................... 89
E (1) ....................................................
92
26S,27S (A series) and 26R,27S (B series) of lyngbyabellin E (1) ................. 94
IV.7
1H and 13C spectra of lyngbyabellin F (2) ................................................. 96
lV.8
1H and 13C spectra of
lyngbyabellin
G (3) ................................................. 96
IV.9
1H and 13C spectra of
Iyngbyabellin
H (4) .................................................. 99
IV.10
1H and 13C of spectra of
IV. 11
Effect of Iyngbyabellin E (1) on the actin
V.1
1H and 13C spectra of aurilide B (1) ........................................................ 111
V.2
COSY spectrum of aurilide B (1) ........................................................... 113
V.3
HSQC spectrum of aurilide B (1) ........................................................... 114
V.4
HMBC spectrum of aurilide B (1) ........................................................... 115
V.5
HSQC-TOCSY spectrum of aurilide B (1) ................................................ 116
V.6
Key COSY, HMBC, and NOE correlations for 1
V.7
'5
V.8
1H and 13C spectra of aurilide C (2) ......................................................... 119
(s-R)
lyngbyabellin
1(5) ................................................. 99
cytoskeleton
of A-I 0 cells .............. 102
........................................ 117
values (ppm) of MTPA esters of aurilide B (1) obtained in CD3CN ....... 118
LIST OF
TABLES
Page
1.1
Selected marine natural products currently in clinical trials ......................... 7
11.1
1H and 13C NMR data for semiplenamides A (I) and B (2) in CDCI3 ............ 47
11.2
1H and 13C NMR data for semiplenamide C (3) in CDCI3 ........................... 49
11.3
1H and 13C NMR data for semiplenamides D (4) and E (5) in CDCI3 ............ 50
11.4
1H and 13C NMR data for semiplenamides F (6) and G (7) in CDCI3 ............ 51
111.1
NMR data for wewakpeptins A (1) and B (2) in CDCI3 ..............................66
111.2
NMR data for wewakpeptins C (3) and D (4) in CDCI3 .............................. 73
IV.1
NMR data for lyngbyabellin E (1) in CDCI3 ............................................. 88
IV.2
NMR data for lyngbyabellins F (2) and G (3) in Cod3 .............................. 97
IV.3
NMR data for Iyngbyabellins H (4) and 1(5) in CDCI3 .............................. 100
lV.4
Cytotoxicity of compounds 1-5 ........................................................... 101
V.1
NMR data for aurilides B (1) and C (2) in C6D6 ...................................... 112
LIST OF ABBREVIATIONS
AEA
Arachidonoylethanolamide (Anandamide)
Ala
Alanine
AMT
Anandamide membrane transporter
Ara-A
Adenine arabinoside
Ara-C
Cytosine arabinoside
br
Broad
CB1
Central cannabinoid receptors
Cl
Chemical Ionization
CNS
Central Nervous System
COSY
1H-1H Chemical Shift Correlation Spectroscopy
d
Doublet
OCAMO
7 ,7-dichloro-3-acyloxy-2-methyloctanoate
Dhiv
a, 13-d ihydroxyisovaleric acid
Dhoaa
2,2-Dimethyl-3-hydroxy-octanoic acid
Dhoya
2,2-Dimethyl-3-hydroxy-7-octynoic acid
Dmhha
2,2-Dimethyl-3-hydroxy-hexanoic acid
OMSO
Dimethylsulfoxide
ElMS
Electron-Impact Mass Spectrometry
EtOAc
Ethyl Acetate
FAAH
Fatty acid amide hydrolase
FAB
Fast Atom Bombardment
FDAA
1 -fluro-2,4-dinitrophenyl-5-L-alaninamide
GC
Gas Chromatography
Gly
Glycine
HETLOC
Hetero half-filtered TOCSY
Hila
2-hydroxyisoleucic acid
Hiva
2-Hydroxy-isovaleric acid
HIV
Human Immunodeficiency Virus
H MBC
Heteronuclear Multiple-Bond Correlation Spectroscopy
Hmha
5-Hydroxy-3-methyl-heptanoic acid
HPLC
High Performance Liquid Chromatography
HSQC
Heteronuclear Single Quantum Coherence
LIST OF ABBREVIATIONS (Continued)
IC
Inhibitory Concentration
lie
Isoleucine
lR
Infrared Spectroscopy
Lac
Lactic acid
LD
Lethal Dose
m
Multiplet
Maba
2-Methyi-3-amino-butanoic acid
Mapa
2-Methyl-3-amino-pentanoic acid
Me
Methyl
MeOH
Methanol
MS
Mass Spectroscopy
MIT
3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide
NMR
Nuclear Magnetic Resonance
NOESY
Nuclear Overhauser Exchange Spectroscopy
NRPS
Non-Ribosomal Peptide Synthetase
Phe
Phenylalanine
PKS
Polyketide Synthase
PMSF
Phenylmethylsuiphonylfluoride
Pro
Proline
q
Quartet
ROESY
Rotating Overhauser Exchange Spectroscopy
RI
Room Temperature
s
Singlet
SCUBA
Self-Contained Underwater Breathing Apparatus
sPLA2
secreted Phospholipase A2
SRi 4171 6A
I H-Pyrazole-3-carboxamide, 5-(4-chlorophenyl)- I -(2,4-dichlorophenyl)-4-
methyl-N-i -piperidinyl-, monohydrochloride
t
Triplet
TEA
Trifluoroacetic acid
TOCSY
Total Correlation Spectroscopy
tR
Retention Time
Tyr
Tyrosine
UV
Ultraviolet Spectroscopy
Val
Valine
VLC
Vacuum liquid chromatography
NATURAL PRODUCTS STUDIES OF THE MARINE CYANOBACTERIA
LYNGBYA SEMIPLENA AND LYNGBYA MAJUSCULA
CHAPTER ONE
GENERAL INTRODUCTION
Nature has continuously provided mankind with broad and structurally diverse
natural products which have a variety of purposes including use as food, fragrances,
pigments, insecticides, and medicines. The use of natural products for the treatment
of human ailments has its beginnings dating back to ancient times in both Eastern
and Western cultures. Traditionally, higher plants and, since the discovery of the
penicillins, terrestrial microorganisms have proven to be the richest sources of natural
drugs that are indispensable, especially for the treatment of fatal diseases such as
cancer and microbial infections. Currently, there is a pressing need for new
therapeutic agents and this is due mainly to the resurgence of pathogenic
microorganisms and parasites that have developed resistance to traditional
chemotherapies. However, due to the increasing rates of deforestation and
rediscovery of known compounds, scientists are beginning to focus their attention on
the marine realm as an alternative source of novel and useful natural products.
Oceans comprise over 70% of our world's surface and harbor a tremendous
variety of flora and fauna. Although still quite young by many standards, the field of
marine natural products has already proven to be a productive source for biologically
active natural products. Coupled with the development of SCUBA in the 1960s, and
more recently submersible vehicles, these have allowed relatively easy accessibility
of both shallow and deep-water marine organisms for studies by natural products
chemists.
In contrast to their terrestrial counterparts, many marine derived secondary
metabolites are structurally complex with unique functionalities. This is due in part to
the differences in the physicochemical nature of the two environments where higher
pressures, lower temperatures, lack of light as well as high ionic concentrations
present in sea water may account for the biosynthesis of highly functionalized and
unusual molecules in these organisms. It is because of these structural novelties that
natural products chemists hope to uncover new classes of therapeutic agents with
unique pharmacophores which may lead to defining novel sites of action in proteins
and enzymes that are involved in human illnesses.
HIGHLIGHTS OF MARINE NATURAL PRODUCTS
Historically, the first recorded research on marine natural products began in
the early 1950s with the pioneering work of Bergman, which led to the discoveries of
the nucleosides spongouridine (1) and spongothymidine (2) from marine sponges
collected off the coast of
Florida.1
This initial serendipitous discovery led to the
development of Ara-A (3) and Ara-C (4), which are based on these sponge
nucleosides and are clinically available today. Ara-C (4) is an anticancer drug used
for the treatment of acute myelocytic leukemia and non-Hodgkin's lymphoma while
the antiviral drug, Ara-A (3), is used for the treatment of herpes
HO
HO
OH
o
HO
HO
OH
Spongouridine (1)
Spongothymidine (2)
j
HO
OH
Ara-A (3)
infections.2'3
HO
oJ
HO
OH
Ara-C (4)
Since the early 1950s, more then 10,000 secondary metabolites have been
defined from marine
organisms.4
At the early stage, pioneering research was mainly
concerned with marine toxins, in part because of the numerous poisonings of
American soldiers in the Pacific during World War II.
Many of the marine toxins
appeared to be macromolecules, belonging mainly to the polyether structural class.
Some of the well known marine polyether toxins are: brevetoxin B (5), isolated from
the dinofiagellate Gymnodinium breve;6 ciguatoxin (6), initially reported from moray
eels and a toxic constituent implicated in ciguateric seafood poisoning;7 maitotoxin
(7), isolated from Gambierdiscus toxicus;8'9 and okadaic acid (8), a potent tumor
promoter and major causative agent of diarrhetic shellfish poisoning, found from two
3
species of dinoflageflates, Prorocentrum lima and
Dinophysis species.10'11
Of these
macromolecules, maitotoxin is the largest (MW 3422), and most likely, the most lethal
non-proteinaceous toxin (LD
of Ca. 50 ng/kg in mice) characterized from a marine
microorganism to date. Palytoxin (9), a Complex polyol, is another well known toxic
macromolecule isolated from the zoanthid, Palythoa
toxicus.12
Many of these toxic
macromolecules have also proven to be invaluable as molecular tools for dissecting
cellular processes by molecular
pharmacologists.13'14
Brevetoxin B (5)
Figure I. 1. Marine toxins.
Meanwhile, the increasing need for drugs able to control new illnesses or
resistant strains of microorganisms stimulated the search for new biomedicals from
marine organisms. Thus, a main objective of most marine natural products programs
is to discover secondary metabolites with a broad spectrum of therapeutic utilities,
such as: antibiotic, antifungal, cytotoxic, neurotoxic, antimitotic, antiviral, and
antineoplastic activities. In more recent years, new targets have been added to the
general screening, for example: AIDS, immunosuppression, anti-inflammation,
Alzheimer's disease, aging processes and some tropical
diseases.15
This is
especially so in the area of cancer research where numerous marine natural products
are being screened. Although none of the discoveries since the 1970's has led to a
clinical product, there are a number of potential candidates currently under intense
investigation.
Maitotoxin (8)
Figure I. 2. Largest non-proteinaceous marine toxin with a
MW of 3422.
5
Among the more promising anticancer natural products are ecteinascidin 743
(10) and bryostatin 1 (11) (Tablel. I). The marine alkaloid ecteinascidin 743 (ET-743)
was isolated from the Caribbean ascidian, Ectinascidia turbinata,16'
17
and is by far the
most advanced compound in the late stages of phase II clinical trials as an anti-
cancer drug. Ecteinascidin's structure is consistent with a natural microbial origin
(e.g. the saframycins). Indeed, there are two patents for bacterial symbionts of the
tunicate E. turbinata. The first focuses on the isolation of the ET-743 producing
microbe,18
while the second uses 16S rDNA sequences to identify the endosymbiont
as Endoecteinascidia frumentensis, the apparent producer of the ecteinascidins.19
Okadaic acid (8)
I-lu
),,
HO
LJOH
'OH
OH
HO
HO
OH
0H
Palytoxin (9)
Figure I. 3. Marine toxins.
More recently, ET-743 has been reported to bind in the minor groove of DNA to
induce an unprecedented bend in the DNA helix towards the major groove.20 The
multi-faceted mechanism of action of ET-743 includes interference with the cellular
transcription-coupled nucleotide excision repair to induce cell death, and cytotoxicity
which is independent of p53 status yet occurs with MDR
elicitation.21'22
Overall,
advanced ovarian, breast and mesenchymal tumors which had been heavily
pretreated with platinum/taxanes showed greatest response to ET-743 in phase I
trials.23'24
In phase II trials, ET-743 was most effective in patients with refractory soft
tissue sarcoma (STS), ovarian and breast cancer. However, difficulties in
establishing the drug's efficacy in STS prevented its approval in 2003. Meanwhile,
EU's Committee for Proprietary Medicinal Products has granted ET-743 orphan drug
status for the treatment of refractory ovarian cancer. It is still in Phase Ill trials in the
US and the EU, and it will probably be the first marine-derived anti-tumor agent to be
commercialized.
Bryostatin 1 (11) is a polyketide first isolated in minute quantity from the
bryozoan,
Bugula neritina
collected from the Gulf of California in the 1970's.25 It is
currently in phase Il clinical trials for the treatment of various leukemias, lymphomas,
melanoma, and solid tumors. The biological profiles of bryostatin 1 (11) are
interesting in that not only does it inhibit tumor growth but it stimulates bone marrow
growth, increasing red blood cell
production.26
Recent research has found that this
cytotoxic macrolide shows potential in the treatment of ovarian and breast cancer and
it also enhances lymphocyte survival in patients undergoing radiation
treatment.27
To
date, eighteen bryostatins have been defined and some members of this family (e.g.
bryostatin 5 and 8) exhibit biological activities equivalent to or better than bryostatin 1
(11).28
Bryostatin 1 (11)
Ecteinascidin 743 (10)
OHO
0
NH
N
0
NH
0
0
OCH3
Dehydrodidemnin B (12) (= Aplidine)
Figure I. 4. Anticancer marine natural products.
One major obstacle that may prevent the development of bryostatin 1 (11) to a
clinically useful anticancer agent is the supply issue. The yield of bryostatin 1 (11)
from
B. neritina is
organism.29
low, in the range of about 1.4 pg per gram of wet weight of the
Another complication is the unpredictable production of the molecule by
the bryozoan. One of the ways to meet the demand issue is through aquaculture
which was undertaken by CalBioMarine Technologies in California, and has since
proven to be a commercially viable source of the molecule. However, a recent report
presented evidence that suggests a symbiotic bacterial origin for bryostatin 1 (11)
production.30
This finding may open the way for genetic manipulation of the
biosynthetic genes and fermentation for increased production of this potent
compound.
Table I. 1. Selected marine natural products currently in clinical trials21'31'32
Compound (Figure number)
Source Organism
Ecteinascidin 743
(Yondelis®)
Dolastatin 10
Ecteinascidia turbinate (tunicate) (possible
bacterial source)
Dolabella auricularia / Symploca sp.
(mollusc/cyanobacterium)
Bugula neritina (bryozoan)
Bryostatin 1
Synthadotin (1LX651,
Dolastatin 15 derivative)
Kahalalide F
Squalamine
Dehydrodidemnin B
(Aplidine®)
Cemadotin (LU 103793,
Dolastatin 15 derivative)
Soblidotin (TZT-1027,
Dolastatin 10 derivative)
E7389 (Halichondrin B
derivative)
NVP-LAQ824 (Psammaplin
derivative)
Discodermolide
HTI-286 (Hemiasterlin
derivative)
LAF-389 (Bengamide B
derivative)
KRN-7000 (Agelasphin
derivative)
Dolabella auricularia/ Symploca sp
(synthetic analog)
Elysia rufescens I Bryopsis sp.
(mollusc/green alga)
Squalus acanthias (shark)
Current
Status
Phase Il/Ill
Phase II
Phase II
Phase II
Phase II
Phase II
Trididemnum solidum (tunicate, synthetic)
(possible bacterial/cyanobacterial source)
Dolabella auricularial Symploca sp
(synthetic analog)
Dolabella auricularia/ Symploca sp
(synthetic analog)
Halichondria okadal (sponge, synthetic)
Phase I
Psammaplysilla sp. (sponge, synthetic)
Phase I
Discodermia dissolute (sponge)
Phase I
Cymbastella sp. (synthetic analog of
sponge metabolite)
Jaspis digonoxea (sponge, synthetic)
Phase I
Agelas mauritianus (sponge, synthetic)
Phase I
Phase II
Phase II
Phase II
Phase I
The dolastatins are a series of cytotoxic peptides that were originally isolated
in very low yield by Pettit's group as part of their work on marine invertebrates from
the Indian Ocean mollusk, Do/abe/Ia auricuIaria.2lM.
Due to the potency and
mechanism of action of dolastatin 10 (13), a linear depsipeptide which was shown to
be a tubulin interactive agent, the drug entered Phase I clinical trials in the 1990s
under the auspices of the NCI.'37 Dolastatin 10 progressed through to Phase II trials
as a single agent, but although tolerated at the doses used, which were high enough
to give the expected concentrations in vivo to inhibit cell growth, it did not
demonstrate significant antitumor activity in a Phase II trial against prostate cancer in
men.
As a result, many derivatives of the dolastatins have been synthesized with
TZT-1027 (Auristatin PE or Soblidotin) (14) now in Phase II clinical trials in Europe,
Japan and the USA. Two derivatives of dolastatin 15 (15), a peptide related to
dolastatin 10 in structure and origin, are known as Cemadotin or LU-I 03793 (16) and
Synthadotin or ILX65I (17). These were entered into Phase I clinical trials by Abbott
GmBH under the Knoll division for treatment of breast cancer.21 By using tritiumlabeled dolastatin 15 as a bioprobe, Hamel's group at NCI
recently reported that
the ymca domain in tubulin may well be composed of a series of overlapping domains
rather than being a single entity.
Another important marine-derived compound, aplidine (= dehydrodidemnin B)
(12), is a synthetic analog of a cyclic antiproliferative depsipeptide, didemnin B.
Didemnin B was isolated from the Caribbean tunicate Trididemnum solidum,4° and
was the first marine natural product to enter clinical trial as an antitumor agent.41
However, it was dropped due to a disappointing performance in phase II testing. The
dehydro-derivative, dehydrodidemnin B (12) is ten times more effective than didemnin
B, and is less toxic. Apladin®, a registered trademark of aplidine (12), was found to
be more selective towards leukemia and lymphoma cells than towards normal cells.
In addition, the activity of Apladin® was found to be independent of other anti-cancer
drugs commonly used in leukemia and lymphoma, suggesting that Apladin® may be
effective in cases that have proved unresponsive to other agents. The success of
aplidine in phase I trials has led to its current evaluation in Phase II trials against solid
tumors.
10
NNNN
I
H
N
I
OCH
N' S
\=1
DolastaUn 10 (13)
I
OHO
0CH30
TZT-1027 (14)
(Auristatin PE or Soblidotin)
Nj
OCH3
0
Dolastatin 15 (15)
LU 103793 (16)
(Cemadotin)
1LX651 (17)
(Synthadotin)
Figure I. 5. Anticancer marine natural products: the dolastatins.
In addition to the discovery of marine natural products with anticancer
properties, marine organisms have also yielded a number of important antiinflammatory agents. Many of these compounds act as phospholipase A2 inhibitors,
preventing release of arachidonic acid from the lipids of cell membranes.42
11
Arachidonic acid is involved in the biosyntheses of many proinflammatory compounds
via enzymes such as cyclooxygenase, lipoxygenase, or monooxygenase. The
quintessential compound from marine sources exhibiting activity as an anti-
inflammatory agent was the sponge metabolite manoalide (18). This compound was
originally reported by Scheuer's group as an antibiotic isolated from the marine
sponge Luffariella
variabilis.43
The groups of Jacobs and Dennis
independently
established that this compound was a potent inhibitor of the enzyme phospholipase
A2. The original compound was licensed to Allergan and placed into clinical trials as
a topical antipsoriatic with a company code name of AGN-1 90093. It advanced to
Phase II, but work on the natural product stopped as sufficient quantities of the
compound would not pass through the skin using the formulation developed for the
trials.21
In addition, pseudopterosins A (19) and E (20), from the Caribbean
gorgonian, Pseudopterogorgia elisabethae, are two diterpene glycosides showing
promising in vitro and in vivo activities as anti-inflammatory agents.46'47
LL1o\
0
io
Manoalide (18)
OH
Pseudopterosin A (19)
Pseudopterosin E (20)
Figure I. 6. Antiinflammatory marine natural products.
In 1971, Kern et al. reported the isolation of hoplonemertine toxin, a
compound that subsequently became known as anabaseine (21 )48 A variety of
synthetic analogues of the basic structure were made by Kem's group, and one of
12
which, GTS-21 (22), has been shown to have cytoprotective and memory enhancing
effects, perhaps due to the ability to displace the binding of nicotinic ligands and to
affect the function of the a413 and a7 subtypes of this
pharmacokinetic
trials,50
receptor.49
Following
GTS-21 was licensed to the Japanese company, Taiho, by
the University of Florida for clinical trials as a potential anti-Alzheimer's agent and is
currently in Phase I trials in both Europe and USA under the auspices of Taiho.
(LN
K)
Anabaseine (21)
GTS-21 (22)
Figure I. 7. Anti-Alzheimer marine natural products.
One of the emerging trends in many infectious organisms is their resistance to
the standard drugs. This is especially so in malarial infections by the protozoan
parasite, Plasmodium species. Many resistant strains of this parasite are being
uncovered and treatments of such infections by traditional drugs are proving
ineffective. However, recent research has suggested that the next generation of new
antimalarial chemotherapeutic agents might come from marine organisms. In an
initial screening of extracts from marine sponges for antimalarial properties,
Acanthella klethra
yielded an active component, axisonitrile 3 (23), which is a bicyclic
spiro-sesquiterpene with an isonitrile
group.51
This compound possesses potent
in
vitro antimalarial activity with no detectable cytotoxicity. It was further revealed that
marine natural products containing NC, NCS, and NCO functionalities exhibited
antimalarial
properties.51
A new class of 13-carboline alkaloids from marine sponges,
known as the manzamines, is also gaining attention as potent antimalarial agents.
Four members of this series, manzamine A (24), its enanatiomer (26), 8hydroxymanzamine A (25), and a manzamine dimer, neo-kauluamine (27) gave
13
encouraging results in preliminary in vivo testing. These preliminary data also
indicated that the manzamines are more active and less toxic than the current
available drugs artemisinin and chloroquine.52'53
Axisonitnle 3 (23)
Manzamine A (24) R = H
8-Hydroxymanzamine A (25) R = OH
ent-8-Hydroxymanzamine A (26)
neo-Kauluamine (27)
Figure I. 8. Antimalarial marine natural products.
In the area of antiviral research, especially anti-HIV (human immunodeficiency
virus), two main leads have been found from marine sources. The marine natural
compound ilimaquinone (28) is currently being investigated as a chemotherapeutic in
HIV infection.M.55 This compound acts by targeting at the RNase function of the
reverse transcriptase enzyme encoded by HIV. The first natural product that inhibits
HIV integrase enzyme was reported from a species of marine sponge belonging to
the genus
Lame!!aria.
The metabolite, lamellarin alpha 20-sulfate (29) was found to
inhibit early steps of HIV
replication.56
14
OCH3
NaO3SO
0
0
H3C0N\
H3CO
Ilimaquinone (28)
Figure I.
H3COT'OCH3
Lamellarin a 20-sulfate (29)
9. Anti-H IV marine natural products.
BIOACTIVE SECONDARY METABOLITES FROM MARINE CYANOBACTERIA
The predominant marine organisms studied by natural products chemists are
sponges, soft corals, marine algae, sea squirts, and bryozoans. Marine algae are
rich in secondary metabolites proven to serve multiple roles through field and
laboratory bioassays. They are loosely divided into two groups, macroalgae, which
include red (Rhodophyta), brown (Phaeophyceae), and green algae (Chlorophyta)
and microalgae, including cyanobacteria (blue-green algae) as well as dinoflagellates
(Dinoflagellatea, Protozoa). However, cyanobacteria, the most ancient (ca. 2 x
1O9
years) of the microalgae, have been the richest producers of novel and bioactive
secondary metabolites.
Cyanobacteria, also known as blue-green algae, are ancient photosynthetic
prokaryotes which inhabit a wide diversity of habitats including open oceans, tropical
reefs, shallow water environments, terrestrial substrates, aerial environments such as
in trees and rock faces, and fresh water ponds, streams and
puddles.57
The rich
elaboration of biologically-active natural products has assisted some of these
organisms to survive in predator-rich tropical reef ecosystems. As a result, tropical
marine cyanobacteria, particularly the filamentous forms such as Lyngbya sp. or
Symploca sp., have been exciting sources of novel natural products with therapeutic
and biotechnological potential.
a. Tubulin-binding compounds
Symplostatin 1(30)
15
In 1998, the first natural dolastatin 10 analog, symplostatin 1 (30) was
discovered by direct isolation from the cyanobacterium Symploca
hydnoides.
Symplostatin I differs from dolastatin 10 by the addition of a single CH2 in the
dolavaline moiety resulting in a terminal N,N-dimethylisoleucine versus the terminal
N,N-dimethylvaline.58
This compound exhibited potent cytotoxicity (lC= 0.3 ng/mL)
against a human nasopharyngeal carcinoma cell line (KB), compared to < 0.1 ng/mL
for dolastatin 10. Symplostatin 1 was also shown to induce microtubule loss by 80%
at I ng/mL when tested in A-I 0 cells. It has been concluded that the mechanism of
action of symplostatin 1 must be similar if not identical to that determined for
dolastatin 1 0.
In further analysis, symplostatin I displayed efficacy against a
number of cancer cell lines and caused the formation of abnormal mitotic spindles
and accumulation of cells in metaphase. These effects were observed at
concentrations that elicited only minor loss of interphase microtubules. Cell cycle
analysis indicated that symplostatin 1 caused G2/M arrest, a finding consistent with its
effects on mitotic spindles. Evaluation of this aspect revealed potent inhibition of both
endothelial cell proliferation and invasion, with in vivo efficacy against murine colon
38 and murine mammary 16/C cell
lines.59
In addition to the exciting biological
activity, this discovery offered further evidence that these potent peptides and
depsipeptides are of cyanobacterial origin.
Curacin A (31)
Not all antimitotic marine natural products have an overt peptidic nature and
the highly lipophilic natural product curacin A (31) is a good example. Isolated from a
Caribbean collection of the cyanobacterium L. majuscula, the structure of curacin A
consists of an interesting thiazoline ring containing a 14-carbon alkyl chain with
conjugated diene and terminal olefin, and a methyl substituted cyclopropyl moiety.6°
Curacin A was originally shown to be active against a Vero cell line, and subsequent
assessment in the NCI 60 cell line assay revealed potent antiproliferative and
cytotoxicity with some selectivity for colon, renal, and breast cancer cell lines. It was
later shown that treatment of curacin A resulted in the depolymerization of purified
tubulin induced by either glutamate or microtubule-associated proteins with an
IC=
4.0 pM and lC=6.0 pM. Further testing revealed that curacin A (31) binds tightly at
the colchicine site of tubulin. Curacin A stimulates the uncoupled GTPase reaction
16
typical of colchicine site agents, and indirect observations were consistent with
curacin A binding rapidly and dissociating slowly from tubulin.61 Additionally, curacin
A was shown to inhibit formation of the Cys239-Cys354 cross-link in f3-tubulin.62
Furthermore, under conditions ideal for tubulin polymerization, high concentrations of
curacin A induced formation of aberrant tubulin polymers appearing similar to a
twisted cable of fine spiral filaments.63 Based on these results a series of semi-
synthetic analogs were produced to explore the structure activity relations in this new
drug class. The structural modifications included reduction and E-to-Z transitions of
the otefinic bonds in the alkyl side chain; disruption of and configurational changes in
the cyclopropyl ring, thiazoline moiety, and substituent modifications at the Cl 0
methyl- and Cl 3 methoxy groups. This work revealed that the most crucial portions
of curacin A for tubulin interaction seems to be the thiazoline ring and the side chain
at least through C4, and the portions of the side chain including the C9-Cl 0 olefinic
bond, and the ClO methyl group.M Recently, some new directions have emerged
with construction of synthetic combinatorial libraries using solution phase and
fluorous scavenging techniques, resulting in new analogs with improved
bioavailability and
efficacy.65
b. Actin-binding compounds
Hectochlorin (32)
Hectochlorin (32), was isolated from the marine cyanobacterium L. majuscula
collected from Hector Bay, Jamaica, and Boca del Drago Beach, Bocas del Toro,
Panama.66 The planar structure was deduced by one- and two-dimensional NMR
spectroscopy, and X-ray crystallography was used to determine the absolute
stereochemistry. Hectochlorin was initially shown to be a potent antifungal agent in
preliminary antimicrobial bioassays. Further testing showed that Ptk2 cells (derived
from the rat kangaroo, Potorous tridactylus) treated with hectochiorin showed an
increase in the number of binucleated cells as result of arrest of cytokinesis. This
result is consistent with an interference with the actin cytoskeleton. Hectochlorin is
similar in its activity to jasplakinolide (or jaspamide) (33) in its ability to promote
hyperpotymerization of actin.67'68 The main biochemical difference between
jasplakinolide and hectochioriri is that while the former can displace fluorescently
labeled phalloidin from actin polymers, the latter is unable to do so, suggesting that
17
the two have distinct interactions with actin. Hectochlorin was also evaluated against
the in vitro panel of 60 different cancer cell lines (National Cancer Institute) and
showed strong potency towards cell lines in the colon, melanoma, ovarian and renal
sub-panels. It had a flat dose-response curve against most cell-lines which is
characteristic of compounds that are antiproliferative but not directly cytotoxic.66
Lyngbyabellins A (34) and B (35)
The lyngbyabellins (34,35) were isolated from collections of
L.
majuscula from
the South Pacific and Caribbean, and bear structural resemblance to hectochlorin
and dolabellin (36).6971 The structure of lyngbyabellin A was determined using 20
NMR techniques and its absolute stereochemistry was determined by chiral HPLC
analysis. Lyngbyabellin A (34) exhibited IC values of 0.03 pg/mL and 0.50 pg/mL
against KB cells (human nasopharyngeal carcinoma cell line) and LoVo cells (human
colon adenocarcinoma cell line), respectively. It was also shown to disrupt the
microfilament network in fibroblastic A-10 cells at concentrations between 0.01-5.0
pg/mL. However, when treated with a higher concentration of lyngbyabellin A many
cells became binucleate, an observation which is consistent for compounds inhibiting
cytokinesis. The structure of Iyngbyabellin B (35) was determined using a
combination of ID and 20 NMR spectroscopy and its stereochemistry was proposed
using a combination of NMR and chiral GC/MS analysis. Lyngyabeflin B was found to
be less toxic than Iyngbyabellin A with IC values of 0.10 pglmL and 0.83 pg/mL
against KB and LoVo cell lines, respectively. Lyngbyabellin B had the same effect as
hectochlorin on PtK2 cells, in that an increased number of binucleate cells were
observed when the cells were treated with 1 0 pM of the agent. These findings
suggest that actin is the likely cellular target of the lyngbyabellins.
c. Neurotoxic compounds
Antillatoxin (37)
The crude extract of a Curaçao collection of L. majuscula was found to be
highly ichthyotoxic and molluscicidal. Fractionation and subsequent purification led to
the discovery of the potent lipopeptide ichthyotoxin, antillatoxin.72 The structure of
antillatoxin was elucidated using several spectroscopic methods including HR
FABMS, IR and NMR. Antillatoxin (37) is one of the most ichthyotoxic metabolites
18
isolated to date from a marine cyanobacterium (LD
in potency only by the
brevetoxins.73
= 0.05 pglmL), and is exceeded
Initial pharmacological studies showed that
antillatoxin was acutely neurotoxic and rapidly induced morphological changes in rat
cerebellar granule neurons (CGC's), including blebbing of neurite membranes.
However, the toxicity was remarkably reduced when cells were treated with NMDA
receptor antagonists like dextrorphan and MK-801, indicating that the toxicity of
antillatoxin was mediated through an NMDA receptor- dependent mechanism.74
Antillatoxin was later shown to be a powerful activator of voltage-gated sodium
channels and resembled brevetoxin in this respect. But unlike brevetoxin which is
known to bind site 5 on the a- subunit of voltage-gated sodium channels, the binding
site of antillatoxin is unknown and remains to be
identified.75
The unique biological
activity of antillatoxin combined with its unusual structure spurred several synthetic
efforts, and the total synthesis of antillatoxin was reported by two groups.778
However, spectroscopic analyses of the synthetic versions showed that they were
different from "natural antillatoxin". Eventually, the synthetic 4R,5R stereoisomer,
which was identical to natural antillatoxin was reported by the Shioiri group.78 In
addition to resolving some initial stereochemical issues, the syntheses of several
stereoisomers of antillatoxin facilitated SAR studies and a recent report showed that
naturally occurring antillatoxin was about 25-fold more potent than any of its C-4 or C5 stereoisomers.79
Kalkitoxin (38)
Kalkitoxin (38) was first isolated from a Caribbean collection of L. majuscula
and exhibited potent brine shrimp and fish toxicity. Subsequently, the compound was
re-isolated in very small quantities from various Caribbean collections of L.
majuscula.6°
Structural elucidation was accomplished by various 2D NMR methods
and stereochemistry was resolved using Marfey's analysis as well as comparison with
synthetic standards. Kalkitoxin was toxic to rat CGC's and had an LC =3.86 nM and
this effect could be inhibited by the addition of NMDA receptor antagonists.74 There is
also evidence indicating that kalkitoxin is a blocker of voltage-sensitive Na channel in
mouse neuro-2a cells (EC = 1 nM). In an inflammatory disease model, kalkitoxin
was found to inhibit the release of 1L43-induced sPLA2 (secreted phospholipase A2), a
key enzyme in the inflammatory cascade (IC = 27 nM).
19
Jamaicamides A-C (39-41)
A strain of L. majuscula collected from Hector's Bay, Jamaica yielded a series
of novel and functionalized lipopeptides, jamaicamides A-C (39-41 ).81 The structure
of the jamaicamides was deduced using a combination of NMR spectroscopy
including an ACCORD-1,1 -ADEQUATE experiment. Jamaicamaide A was
particularly unique in its presence of an acetylenic bromide functionality. The
jamaicamides exhibited cytotoxicity against H-460 human lung cell line and mouse
neuro-2A neuroblastoma cell lines with an LC50
15 pM for both cell lines and
showed sodium channel blocking activity at a concentration of 5 pM. From a
combined precursor feeding study and a molecular genetics approach a fascinating
mixed NRPS and PKS pathway was deduced.
d. Protein kinase C activators
Lyngbyatoxins A-C (42-44)
Lyngbyatoxin A (42) was isolated from a Hawaiian shallow-water variety of L.
majuscula and is believed to be responsible for a condition known as 'swimmer's itch'.
The structure of lyngbyatoxin A is closely related to the teleocidins, metabolites
produced by several Streptomyces sp.82 The structure of lyngbyatoxin A was
elucidated using high resolution mass spectrometry and NMR analyses and is the first
indole alkaloid to be reported from a marine cyanobacterium. This highly
inflammatory compound had an
LD100
=0.3 mg/kg in mice which was comparable to
the toxicity of the teleocidins. It was also found to be ichthyotoxic and exposure to
0.15 pg/mL resulted in death of fish in 30 minutes. Ensuing pharmacological studies
showed that lyngbyatoxin A (42) and teleocidin were potent tumor promoters in a
manner similar to that of the phorbol esters when tested in mice.83 To confirm
whether lyngbyatoxin was produced by the cyanobacterium or merely associated with
it, collections of L. majuscula from Kahala Beach, Hawaii were re-examined and this
led to the isolation of two related metabolites, lyngbyatoxin B (43) and lyngbyatoxin C
(44)84
In an effort to understand the interaction of lyngbyatoxin A with PKC, several
analogs were synthesized and it was found that the lactam ring is essential for PKC
activation and that the hydrophobic group attached to C-7 actually downregulates
PKC production.85'86 Very recently, the Gerwick group has reported the molecular
20
cloning of the lyngbyatoxin biosynthetic gene cluster from a Kahala Beach, Oahu,
collection of L. majuscula, as well as the biochemical characterization of a novel
aromatic prenyltransferase that transfers a geranyl group as the final step in the
biosynthesis of lyngbyatoxin A (42).87
e. Other bioactive cyanobacterial metabolites
Apratoxin A (45)
bouillonhi isolated from Apra Bay, Guam initially mis-identified as L.
majuscula was found to produce a very potent cytotoxin, apratoxin A (45).88 The
L.
structure of this mixed polyketide-peptide was deduced using a variety of 2D-NMR
techniques, and the amino acid configurations were determined using chiral HPLC
analysis. Apratoxin exhibits an
IC50
= 0.52 nM and 0.36 nM against KB and LoVo
cells, respectively but its exact mode of action at present is uncharacterized.
0
o__-_...
0
H N:*,,
OCH3
i-f' V 'Curacin A (31)
Symplostatin 1 (30)
OH
Ly
il__i
HN
0
"S
BrNZ)%
H
Hectochlorin (32)
CICI
Jasplakinolide (33)
HN
HN
HC70
Lyngbyabeflin A (34)
0
CT8
0
HN
N HN
H(LQ
Lyngbyabellin B (35)
Figure I. 10. Chemical structures of natural products and derivatives 30-35 discussed in text.
21
cI
CI
0
o
OH
0
OCH3
Dolabelljn (36)
N'f'
AntiUatoxjn (37)
Kalkitoxjn (38)
OCH3
Jamaicamjde A (39) R=Br
Jamaicamide B (40) R=H
'NO
cO
OCH3
NO
(J 0
Jamaicamide C (41)
NNOO
HO"'
0
0
NJ.... N-2°
c)
Apratoxin A (45)
I
H
Lyngbyatoxin A (42)
Lyngbyatoxin B (43)
Lyngbyatoxin C (44)
OCH3:
I
N
rrt
N'S
'I
Barbamide (46)
Figure I. 11. Chemical structures of natural products and derivatives 36-46
discussed in text.
The overall unusual structure of apratoxin A and its potent activity, prompted several
synthetic efforts and the total synthesis of apratoxin was reported by Chen et al.89
22
Further pharmacological studies along with the synthesis of several structural analogs
may enable the determination of the biological target of this potent molecule.
Barbamide (46)
Barbamide (46) was obtained from the lipid extract of a Curaçao collection of
L. majuscula and possesses several unique structural elements including a
trichioromethyl
group.9°
Standard spectroscopic techniques were used for the
structural elucidation of barbamide and stereochemistry was deduced by a
combination of degradation and biosynthetic methods. The presence of several
structurally intriguing elements in barbamide led to a detailed examination of its
biosynthesis including the cloning and characterization of the barbamide biosynthetic
gene
cluster.91'92
Preliminary bioassays indicated that barbamide possesses anti-
molluscicidal activity (LC = 10.0 pg/mL). However, the compound was found to be
inactive in other assays and the full extent of its biological properties remains
unknown.
THE SYMBIOSIS OF MARINE INVERTEBRATES AND MARINE
CYANOBACTERIA
Marine invertebrates are largely sessile, filter-feeding organisms that contain
complex assemblages of symbiotic microorganisms. Thus, when biologically active
compounds are isolated from marine invertebrate sources, it is always uncertain if the
invertebrate host or an associated microorganism is the actual producer. Observation
that some metabolites occur in unrelated genera of marine invertebrates, as well as
the isolation of related compounds from microbial sources, provides evidence that
some natural products are actually produced by symbionts rather than their
invertebrate
hosts.30'93
For example, the sea hare, D. auricularia has yielded several
cytotoxic agents that share similarity with cyanobacterial counterparts, including the
anticancer agent dolastatin 10 (13) and the cytotoxic agent, dolabellin (36). The
isolation of a dolastatin-lO and related compounds, including symplostatin 1 (30),
from the marine cyanobacteria Symploca spp., and the dolabellin-like compounds,
23
Iyngbyabellin A and B from L. majuscula, suggests that sea hares obtain and
accumulate bioactive compounds as a result of their cyanobacterial diet.
Similarities between structures of cyanobacterial origin have also been found
in filter-feeding invertebrates as well. For example, the antifungal compound,
majusculamide C (47) was originally isolated from a deep water collection of L.
majuscula and subsequently re-isolated from the sponge, Ptilocau/is trachys.94'95 In
parallel, metabolites closely related to majusculamide C were also isolated from the
sea hare Do/abel/a auricularia, including dolastatin 11(48) and dolastatin 12
(49)97
Furthermore, a study of several Guam collections of L. majuscu/a and L.
majuscula/Sty/ocheilus ca/cico/a yielded a re-isolation of dolastatin 12, its
c-i 5
epimer (50), and two closely related compounds, lyngbyastatin 1 (51) and its C-i5
epimer (52) providing more direct evidence of their cyanobacterial origin.
Many didemnid family ascidians have also been shown to produce biologically
active cyclic peptides, including the patellamides (e.g. patellamide A (53))and
trunkamide (54)9899 Because many of these cyclic peptides resemble cyanobacterial
metabolites and it is known that ascidians house cyanobacterial symbionts of the
genus
Proch!oron,
there is a high likelihood that Prochloron is the producer of these
cyclic peptides. However, chemical localization studies have demonstrated that the
patellamides accumulate in the tunic tissue of the ascidian and are not associated
with the cyanobacterial
symbiont.10°
In light of these contradictory results, it cannot
be ruled out that the compounds are in fact produced and secreted by the
cyanobacterial symbiont where they would then accumulate in the ascidian tissues.
Advancements in our understandings of the biosynthetic pathways required to
produce cyclic pepetides and other secondary metabolites will undoubtedly help to
resolve these conflicting issues. Experiments are quickly being developed to assess
the biosynthetic capacities housed within hosts and their symbiotic partners. In the
case of Lissoclinum pate//a, Schmidt et al, recently identified an NRPS gene
sequence from a mixture of the ascidian and its symbiotic Prochioron spp.
Based
on BLAST sequence homology and overall G/C base content, the NRPS domain
appears to be of cyanobacterial origin, providing genetic evidence that the Proch/oron
symbiont possesses the biosynthetic capacity to make complex metabolites.
Evidence for NRPS-Iike genes in the tunicates has yet to be found.
24
Perhaps the best studied sponge-symbiont relationship involves the sponge of
the genus Dysidea. Dysidea sp. have an abundance of symbiotic cyanobacteria,
predominantly Oscillatoria
spongelliae.102
In fact, the resulting cyanobacterial biomass
can account for up to 40% of the total sponge biomass! Furthermore, within the genus
Dysidea, it appears that different species have an obligate requirement for specific
populations of microbial symbionts. Thacker and Starnes (2003) found that three
species of Dysidea from Guam (D. herbacea, D. chiorea, and D. granulosa) each
contain a genetically distinct dade of 0. spongeiiae
symbionts.102
Concurrently,
biochemical analysis has shown that Dysidea spp. are renowned for their production of
bioactive metabolites that share high structural similarity with cyanobacterial natural
products. The Okinawan sponge D. arenaria yielded the cytotoxic compound
arenastatin A (55) that shows high structural similarity with cryptophycin A (56), a cyclic
peptide isolated from the terrestrial cyanobacterium Nostoc sp.10105 A related sponge,
Dysidea (Lameiodysidea) herbacea and its 0. spongellieae symbiont produces several
chlorinated metabolites, including dysidenin
(57)1o61o8
The trichloromethyl group of
barbamide and the related cyanobacterial metabolites, pseudodysidenin
(58),
dysidenamide (59), nordysidenin (60), and herbamide B (61), strongly resembles that
found in these various Dysidea-derived
metabolites.90'109
Cell separation studies
using flow cytometry analysis has shown that the chlorinated metabolites are
associated with the cyanobacterial symbiont. Most recently, molecular analysis of the
sponge-symbiont biomass have resulted in the isolation of biosynthetic genes
containing high similarity with the barbamide biosynthetic gene cluster (bar).
Fluorescence in situ hybridization (FISH) techniques were used to localize the bar-like
halogenase biosynthetic genes to the 0. sponge!!ieae symboint, demonstrating that
the biosynthetic potential for compound production is housed within the
cyanobacterial symbiont (P.M.Flatt et al, manuscript in press).
Recognition that microbial symbionts are often important contributors to the
biosynthesis of biologically active compounds opens up additional opportunities to
develop sustainable methods for compound production. One could envision the
isolation and culturing of the associated microorganism as a viable option. However,
isolating and culturing a producing microorganism from an invertebrate host has
proven to be a difficult challenge. The microbial communities residing within marine
25
invertebrates are often diverse and many appear to have an obligate requirement for
the host. For example, one report characterized a total of 228 different bacterial
species, 25 fungi, 3 actinomycetes, and 9 cyanobacterial strains from the sponge
Candidaspongia
f/abe/late'1°
In addition, a limited screening discovered 64 bacterial
symbionts from the sponge Aplysina aerophoba and 51 from Theonella swinhoe
representing seven different bacterial
divisions.111
In some cases, bacteria account
for up to 40% of the total sponge biomass.2 Since the early 1990's, several
research groups have explored the stable production of bioactive compounds through
the isolation and culturing of marine invertebrate-derived symbiotic microorganisms,
including unicellular bacteria, filamentous bacteria, and fungi. However, there are
only a few limited examples where the same natural product substances have been
isolated from both an invertebrate macroorganism and their associated prokaryotic
microorganisms, and include makaluvamine A (62) from both Zyzzya sponges
113
and Didymium bahiense myxomycetes,'14 manzamine A (24) from several sponges
and a sponge-derived actinomycete,5 the polychlorinated metabolites, neodysidenin
(63)108 from D. herbacea and its epimer, pseudodysidenin109 from L. majuscula. In
addition, the trichlorinated compound, herbamide B has also been isolated from both
the sponge D. herbacea
116
and the cyanobacterium L. majuscula (unpublished,
McPhail).
Alternatively, the isolation and characterization of biosynthetic genes from
host-symbiont populations is a promising new approach to address the compound
supply problem. The development of genetic techniques to screen sponge-symbiont
metagenomes for genes related to secondary metabolite biosynthesis and then the
use of these for in-situ hybridization experiments is a critical first step in determining
the identity of the producing organism.
However, the abundance of different microbial symbionts within an
invertebrate host makes the isolation and characterization of individual biosynthetic
gene clusters extremely difficult. From a conservative estimate of only 50 symbiotic
species per marine sponge, an estimated 125,000 cosmid-containing colonies
(estimated insert size -40kb) would need to be screened from a total genomic library
to fully cover the genomic sequences of each organism. However, through the
careful selection of metabolites that are likely of microbial origin and a sponge-
26
symbiont system that is limited to a few abundant symbiotic associates, as in the case
of the D. herbaceae - 0. sponge!Iiae system, the screening and isolation of
biosynthetic gene clusters can be feasible.
R1
N
HN1
%%-
Majusculamide C (47)
Dolatstatin 11(48)
Dolastatin 12(49)
Epidolastatin 12(50)
Lyngbyastatin 1(51)
Epilyngbyastatm 1(52)
R2
R3
R4
R5
CH3 H OCH3 CH3
H
CH3 H OCH3 H
CH3 CH3 H
H
H
CH3 H
Cl-I3 H
H
CH3 CH3 H OCH3 H
CH3 H CH3 OCH3 H
H
-
R
R7
H
CH3
CH3
CH3
CH3
CF!3
CF!3
CH3
CH3
CH3
CH3
Cl-I3
oo
(NS
NHN1
HN
NI-I
0
Patellamide A (53)
Trunkamide A (54)
Figure I. 12. Chemical structures of natural products and derivatives 47-54
discussed in text.
27
O>1OHN
OHN
0
J)
0
OCH3
Arenastatin A (55)
I-1N
0
Cryptophycin A (56)
CI3C,-. N-CCI3
H
CI3C-IL(_(CCI3
OCH3
H
CI3C-. N-.CCI3
0
o
O/L
N=(
Pseudodysidenin (58) R = CH3
Nordysidenin (60) R = H
Dysidenin (57)
0
GI3C1NL
N
NR
H2N
Dysidenamide (59)
H
0N
S
'-I
Herbamide B (61)
Makaluvamine A (62)
Neodysidenin (63)
Figure I. 13. Chemical structures of natural products and derivatives 55-63
discussed in text.
GENERAL THESIS CONTENTS
Investigation of marine cyanobacteria for their production of novel and
bioactive chemistry is the focus of our laboratory and of the research presented within
this Ph.D. thesis. The chemistry of marine microalgae, particularly that of the
cyanobacterium Lyngbya sp., is extremely rich and diverse. Our extensive collections
of marine algae are the results of continual expeditions to different parts of the world,
including the Caribbean, Indonesia, Japan, Papua New Guinea, Fiji, South Africa,
Panama, and Madagascar. As such, the various chapters outlined in this thesis have
unifying themes, which entail isolation, structure elucidation (both planar and
stereochemical determinations), and biological characterization of novel secondary
metabolites from marine cyanobacteria of the genus Lyngbya collected from Papua
New Guinea.
This thesis begins with an investigation of semiplenamides A to G, a series of
new anandamide-like fatty acid amides, which were isolated from a 1997 Papua New
Guinea collection of the marine cyanobacterium Lyngbya semiplena. The planar
structures of these lipids were determined using standard 1 D and 2D NMR methods.
The relative stereochemistry of the aliphatic portion of the new metabolites was
deduced from ID NOE data and 1H-decoupling experiments, while the absolute
stereochemistry of the amino alcohol moieties was assigned by chemical
derivatization and chiral GC-MS methods. All of these new metabolites displayed
toxicity in the brine shrimp model system, three showed modest potency in displacing
radiolabeled anandamide from the cannabinoid receptor (CBI), and one was a
modest inhibitor of the anandamide membrane transporter (AMT).
Chapter three presents four new depsipeptides also isolated from the marine
cyanobacterium Lyngbya semiplena collected from Papua New Guinea. The amino
and hydroxy acid partial structures of wewakpeptin A-D were elucidated through
extensive spectroscopic techniques, including HR-FABMS, 1 D 1H and 13C NMR, as
well as 2D COSY, HSQC, HSQC-TOCSY, and HMBC spectra. The sequence of the
residues was determined through a combination of multifaceted approaches including
ESI-MS/MS, HMBC, ROESY, and a modified ID HMBC experiment. The absolute
stereochemistry of each residue was determined by chiral HPLC and chiral GC-MS
methods. The wewakpeptins represent an unusual arrangement of amino- and
hydroxy-acid subunits relative to known cyanobacterial peptides, and possess a bisester, a 2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya) or 2,2-dimethyl-3hydroxyoctanoic acid (Dhoaa) residue, and a diprolyl group reminiscent of dolastatin
15 (15). Wewakpeptin A and B were the most cytotoxic among these four
29
depsipeptides with an LCof approximately 0.4 pM to both the NCI-H460 human
lung tumor and the neuro-2a mouse neuroblastoma cell lines.
A total of five new lyngbyabellin analogues isolated from the marine
cyanobacterium Lyngbya majuscula collected from Papua New Guinea, are
presented in chapter four. The structures of lyngbyabellins E-1 were elucidated
through extensive spectroscopic techniques, including HR-FABMS, ID 1H and 13C
NMR, as well as 2D COSY, HSQC, and HMBC spectra. The absolute configuration
of lyngbyabellin E has been ascertained by chiral HPLC and GC/MS analysis of
degradation products, in combination with NMR experiments. The lyngbyabellins
were tested for cytotoxicity to NCI-H460 human lung tumor and neuro-2a mouse
neuroblastoma cells and had LC values between 0.2 and 4.8 pM. Intriguingly,
lyngbyabellin E and H appeared to be more active against the H460 cell line with LC
values of 0.4 pM and 0.2 pM, respectively, compared to LC values of 1.2 and 1.4
pM in the neuro-2a cell line. Lynbyabellin I was the most toxic to neuro-2a cells (LC
0.7 pM), whereas Iyngbyabellin G, was the least cytotoxic of all compounds to either
cell line. On the basis of this limited screening, it appears that lung tumor cell toxicity
is enhanced in the cyclic representatives, and overa potency is increased in those
containing an elaborated side chain.
In the fifth chapter of this thesis, the structural and stereochemical analyses of
two new aurilides (B and C) isolated from the marine cyanobacterium Lyngbya
majuscula collected from Papua New Guinea, along with their biological activities, will
be discussed.
The thesis ends with a concluding chapter, summarizing the results presented
in the preceding chapters, as well as comments on structural trends that are
emerging in cyanobacterial secondary metabolites. General thoughts on the future
direction of natural products and the pursuit of the marine environment as a viable
source for novel chemistry of medicinal value are also considered.
30
REFERENCE
(1)
Bergman, W.; Feeney, R. J. J. Org. Chem. 1951,
(2)
Bodey, G. P.; Freirich, E. J.; Monto, R. W.; Hewlett, J. S. Cancer
16, 981-987.
Chemother. 1969, 53, 59-66.
(3)
Buchanan, R. A.; Hess, F. Science 1980, 127,
(4)
Jaspars, M. Haivey, A.
(5)
Halstead, B. W. US Government Printing Office, Washington DC.
1965, I,
(6)
L., Ed.;
143-171.
Wiley: New York, 1998,
1-155.
Lin, Y.-Y.; Risk, M.; Ray, S. M.; Van Engen, D.; Clardy, J.; Golik, J.;
James, J. C.; Nakanishi, K. J. Am. Chem. Soc. 1981,
(7)
65-84.
103, 6773-6775.
Murata, M.; Legrand, A.-M.; Ishibashi, Y.; Yasumoto, T. J. Am. Chem.
Soc. 1989, 111, 8927-8931.
(8)
Murata, M.; Nakoi, H.; Iwashita, T.; Matsunaga, S.; Sasaki, M.;
Yokoyama, A.; Yasumoto, T. J. Am. Chem. Soc. 1993,
115, 2060-
2062.
(9)
Norimura, T.; Sasaki, M.; Matsumori, N.; Miata, M.; Tachibana, K.;
Yasumoto, T. Angew. Chem. Int. Ed.
(10)
1996, 35, 1675-1678.
Yasumoto, T.; Oshima, Y.; Sugawara, W.; Fukuyo, Y.; Oguri, H.;
Igarashi, T.; Fujita, N. Nip.
(11)
EngI.
Sui. Gak.
1980, 46, 1405-1411.
Murakami, Y.; Oshima, Y.; Yasumoto, T.
Nip. Sui. Gak.
1982, 48, 69-
72.
(12)
Moore, R. E.; Scheuer, P. J. Science 1971, 172, 495-498.
(13)
Mayer, A. M. S.; Hamann, M. T. Marine Biotechnology 2004,
(14)
Poli,
(15)
Kelecom, A. An Acad Bras Cienc 1999, 71, 249-263.
(16)
Wright, A. E.; Forleo, D. A.; Gunawardana, G. P.; Gunasekera, S. P.;
M. A. Re. Adv.
Mar. Blot.
2002, 7,
6, 37-52.
1-30.
Koehn, F. E.; McConnell, 0. J. J. Org. Chem. 1990, 55,4508-4511.
(17)
Rinehart, K. L.; Holt, T. G.; Fregeau, N. L.; Stroh, J. G.; Kiefer, P. A.;
Sun, F.; Li, L. H.; Martin, D. G. J. Org. Chem. 1990,
(18)
Haygood, M. G.; Salomon, C. E.; Faulkner, D. J.
EP1360337 2003.
55, 4512-4515.
International
Patent
31
(19)
Perez, E.; Beatriz, A.; Perez, T.; Velasco, I.; Henriquez, P.; Munoz, M.;
Moss, C.; Mckenzie, D. International Patent W02004015143 2004.
(20)
Zewail-Foote, M.; Hurley, L. J Med Chem 1999, 42, 2493-2497.
(21)
Newman, D. J.; Cragg, G.M. JNatProd2004, 67, 1216-1238.
(22)
van Kesteren, C.; de Vooght, M.; Mathot, R.; Schellens, J.; Jimeno, J.
M.; Beijnen, J. H. Anti-Cancer Drugs 2003, 14, 487-502.
(23)
Amador, M. L.; Jimeno, J.; Paz-Ares, L.; Cortes-Funes, H.; Hidago, M.
Ann Oncol 2003, 14, 1607-1615.
(24)
Dileo P, C. P., Bacci G, et al. Proceedings of the 2002 ASCO Annual
Meeting. Abstr #1628. 2002.
(25)
Pettit, G. R.; Herald, C. L.; Doubek, D. L.; Herald, D. L.; Arnold, E.;
Clardy, J. J. Am. Chem. Soc. 1982, 104, 6846-6848.
(26)
May, W. S.; Sharkis, S. J.; Esa, A. H.; Gebbia, V.; Kraft, A. S.; Petht,
G. R.; Sensenbrenner, L. L. Proc. Nat!. Acad. Sc!. USA 1987, 84,
8483-8487.
(27)
Kraft, A. S. J. Nat!. Cancer !nst. 1993, 85, 1790-1792.
(28)
Kraft, A. S.; Woodley, S.; Pettit, G. R.; Gao, F.; Coil, J. C.; Wagner, F.
Cancer Chemother. Pharmaco!. 1996, 37, 271-278.
(29)
Schaufelberger, D. E.; Pettit, G. R. J. Nat. Prod. 1991, 54, 1265-1270.
(30)
Haygood, M. G.; Schmidt, E. W.; Davidson, S. K.; Faulkner, D. J. J.
Mo!. Microbiol. Biotechno!. 1999, 1, 33-43.
(31)
Salomon, C. E.; Magarvey, N. A.; Sherman, D. H. Nat Prod Rep 2004,
21, 105-121.
(32)
Simmons, T. L.; Andrianasolo, E.; McPhail, K.; Flatt, P.; Gerwick, W.
H. Mo!. Cancer Ther. 2005, 4, 333-342.
(33)
Pettit, G. R.; kamano, Y.; Herald, C. L.; Tuinman, A. A.; Boettner, F.
E.; Kizu, H.; Schmidt, J. M.; Baczynkyji, L.; Tomer, K. B.; Bontems, R.
J. J. Am. Chem. Soc. 1987, 109, 7581-7582.
(34)
Bal, R.; Pettit, G. R.; Hamel, E. Biochem. Pharmacol. 1990, 39, 19419.
(35)
Bai, R.; Friedman, S. J.; Pettit, G. R.; Hamel, E. Biochem. Pharmacol.
1992, 43, 2637-45.
32
(36)
Bai, R.; Pettit, G. R.; Hamel, E. J. Biol. Chem. 1990, 265, 17141-9.
(37)
Hamel, E. Biopolymers 2002, 66, 142-160.
(38)
Vaishampayan, U.; Glode, M.; Du, W.; Kraft, A.; Hudes, G.; Wright, J.;
Hussain, M. C/in. Canc. Res. 2000, 6, 4205-8.
(39)
Cruz-Monserrate, Z.; Mullaney, J. T.; Harran, P. G.; Pettit, G. R.;
Mullaney, J. T.; Harran, P. G.; Pettit, G. R. Eur. J. Biochem. 2003,
270, 3822-3828.
(40)
Rinehart, K.; Gloer, J.; Cook, J.; Carter, J.; Mizsak, S.; Scahill, T.
JAm Chem Soc 1987, 103, 1857-9.
(41)
Nuijen, B.; Bouma, M.; Manada, C. Anti-Cancer Drugs 2000, 11, 793811.
(42)
Wylie, B. L.; Ernst, N. B.; Grace, K. J. S.; Jacobs, R. S. Prog. Surg.
1997, 24, 146-152.
(43)
de Silva, E. D.; Scheuer, P. Tet. Left. 1980, 21, 1611-14.
(44)
Jacobs, R. S.; Culver, P.; Langdon, R.; O'Brien, T.; White, S.
Tetrahedron 1985, 41, 981-4.
(45)
Deems, R. A.; Lombardo, D.; Morgan, 0. P.; Mihelich, E. D.; Dennis,
E. A. Biochim. Biophys. Acta. 1987, 917, 258-68.
(46)
Look, S. A.; Fenical, W.; Matsumoto, G. K.; Clardy, J. J. Org. Chem.
1986, 51, 5140-5145.
(47)
Roussis, V.; Wu, Z.; Fenical, W.; Strobel, S. A.; Van Duyne, G. D.;
Clardy, J. J. Org. Chem. 1990, 55, 4916-4922.
(48)
Kern, W. R.; Abbott, B. C.; Coates, R. M. Toxicon 1971, 9, 15-22.
(49)
de Fiebre, C. M.; Meyer, E. M.; Henry, J. C.; Muraskin, S. I.; Kern, W.
R.; Papke, R. L. Mo!. Pharmacol. 1995, 47, 164-171.
(50)
Mahnir, V.; Lin, B.; Prokai-Tatrai, K.; Kern, W. R. Biopharm. Drug
Disposit. 1998, 19, 147-151.
(51)
Wright, A. 0.; Konig, G. M. J. Nat. Prod. 1996, 59, 71 0-716.
(52)
Ang, K. K. H.; Holmes, M. J.; Higa, 1.; Hamann, M. T.; Kara, U. A. K.
Antimicrob. Agents Ch. 2000, 44, 1645-1649.
33
(53)
El Sayed, K. A.; Kelly, M.; Kara, U. A. K.; Ang, K. K. H.; Katsuyama, I.;
Dunbar, D. C.; Khan, A. A.; Hamann, M. T. J. Am. Chem. Soc. 2001,
123, 1804-1808.
(54)
Luibrand, R. T.; Erdman, T. R.; Vollmer, J. J.; Scheuer, P. J.; Finer, J.;
Clardy, J.
Tetrahedron 1979,
35, 609-12.
(55)
De Cercq, E. Res. Rev. 2000, 20, 323-349.
(56)
Reddy, M. V.; Rao, M. R.; Rhodes, D.; Hansen, M. S.; Rubins, K.;
Bushman, F. D.; Venkateswarlu, Y.; Faulkner, D. J. J. Med. Chem.
1999, 42, 1901-1 907.
(57)
Whitton, B. A.; Potts, M. In The ecology of cyanobacteria. Their
diversity in time and space., B.A. Whitton
and M.Potts,
eds. (Dordrecht:
Kluwer Academic Publisher), 2000, 1-11.
(58)
Harrigan, G. G.; Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Nagle, D.
G.; Paul, V. J.; Mooberry, S. L.; Corbett, T. H.; Valeriote, F. A. J.
Prod. 1998, 61,
(59)
1075-1077.
Mooberry, S. L.; Leal, R. M.; Tinley, 1. L.; Luesch, H.; Moore, R. E.;
Corbett, T. H. Inter. J. Cancer
(60)
2003, 104,
512-521.
Gerwick, W. H.; Proteau, P. J.; Nagle, D. G.; Hamel, E.; Blokhin, A. V.;
Slate, D. L. J. Org. Chem.
(61)
Nat.
1994,
59, 1243-1245.
Blokhin, A. V.; Yoo, H. D.; Geralds, R. S.; Nagle, D. G.; Gerwick, W.
H.; Hamel, E. Mol Pharmacol 1995, 48, 523-31.
(62)
Luduena, R. F.; Prasad, V.; Roach, M. C.; Banerjee, M.; Yoo, H. D.;
Gerwick, W. H. Drug Deliv. Res.
(63)
40, 223-229.
Hamel, E.; Blokhin, A. V.; Nagle, D. G.; Yoo, H. 0.; Gerwick, W. H.
Drug Deliv. Res.
(64)
1997,
1995,
34, 110-120.
Verdier-Pinard, P.; Lai, J. Y.; Yoo, H. D.; Yu, J.; Marquez, B.; Nagle, D.
G.; Nambu, M.; White, J. D.; Falck, J. R.; Gerwick, W. H.; Day, B. W.;
Hamel, E. Mo!. Pharmacol.
(65)
1998,
53, 62-76.
Wipf, P.; Reeves, J. T.; Day, B. W. Curr. Pharm. Des.
1437.
2004, 10,
1417-
34
(66)
Marquez, B. L.; Watts, K. S.; Yokochi, A.; Roberts, M. A.; Verdier-
Pinard, P.; Jimenez, J. I.; Hamel, E.; Scheuer, P. J.; Gerwick, W. H. J
Nat Prod. 2002, 65, 866-71.
(67)
Zabriskie, 1. M.; Kiocke, J. A.; Ireland, C. M.; Marcus, A. H.; Molinski,
T. F.; Faulkner, 0. J.; Xu, C. F.; Clardy, J. J. Am. Chem. Soc. 1986,
108, 3123-4.
(68)
Crews, P.; Manes, L. V.; Boehier, M. Tetrahedron Lett. 1986, 27,
2797-2800.
(69)
Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J.; Mooberry, S. L.
J. Nat. Prod. 2000, 63, 611-5.
(70)
Sone, H.; Kondo, T.; Kiryu, M.; Ishiwata, H.; Ojika, M.; Yamada, K. J.
Org. Chem. 1995, 60, 4774-81.
(71)
Milligan, K. E.; Marquez, B. L.; Williamson, R. T.; Gerwick, W. H. J.
Nat. Prod. 2000, 63, 1440-3.
(72)
Orjala, J.; Nagle, 0. G.; Hsu, V.; Gerwick, W. H. J. Am. Chem. Soc.
1995, 117, 8281-2.
(73)
Baden, D. G.; Bikhazi, G.; Decker, S. J.; Foldes, F. F.; Leung, I.
Toxicon 1984, 22, 75-84.
(74)
Berman, F. W.; Gerwick, W. H.; Murray, T. F. Toxicon 1999, 37, 16458.
(75)
Li, W. I.; Berman, F. W.; Okino, T.; Yokokawa, F.; Shioiri, T.; Gerwick,
W. H.; Murray, T. F. Proc. Nat!. Acad. Sd. USA 2001, 98, 7599-7604.
(76)
Yokokawa, F.; Shioiri, T. J. Org. Chem. 1998, 63, 8638-8639.
(77)
White, J. D.; Hanselmann, R.; Wardrop, D. J. J. Am. Chem. Soc.
1999, 121, 1106-1107.
(78)
Yokokawa, F.; Fujiwara, H.; Shioiri, T. Tetrahedron 2000, 56, 17591775.
(79)
Li, W. I.; Marquez, B. L.; Okino, T.; Yokokawa, F.; Shioiri, T.; Gerwick,
W. H.; Murray, T. F. J. Nat. Prod. 2004, 67, 559-68.
(80)
Wu, M.; Okino, T.; Nogle, L. M.; Marquez, B. L.; Williamson, R. T.;
Sitachitta, N.; Berman, F. W.; Murray, T. F.; McGough, K.; Jocobs, R.;
35
Colsen, K.; Asano, T.; Yokokawa, F.; Shioiri, T.; Gerwick, W. H. J. Am.
Chem. Soc. 2000, 122, 12041-1 2042.
(81)
Edwards, D. J.; Marquez, B. L.; Nogle, L. M.; McPhail, K.; Goeger, D.
E.; Roberts, M. A.; Gerwick, W. H. Chem. Biol. 2004, 11, 817-33.
(82)
Cardellina, J. H., 2nd; Marner, F. J.; Moore, R. E. Science 1979, 204,
193-5.
(83)
Fujiki, H.; Mori, M.; Nakayasu, M.; Terada, M.; Sugimura, T.; Moore, R.
E. Proc. Nat!. Acad. Sd. USA 1981, 78, 3872-6.
(84)
Aimi, N.; Odaka, H.; Sakai, S.; Fujiki, H.; Suganuma, M.; Moore, R. E.;
Patterson, C. M. J. Nat. Prod. 1990, 53, 1593-6.
(85)
Basu, A.; Kozikowski, A. P.; Lazo, J. S. Biochemistry 1992, 31, 382430.
(86)
Kozikowski, A. P.; Shum, P. W.; Basu, A.; Lazo, J. S. J. Med. Chem.
1991, 34, 2420-30.
(87)
Edwards, D. J.; Gerwick, W. H. J. Am. Chem. Soc. 2004, 126, 1143211433.
(88)
Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J.; Corbett, T. H. J.
Am. Chem. Soc. 2001, 123, 5418-23.
(89)
Chen, J.; Forsyth, C. J. Proc. Nat!. Acad. Sc!. USA 2004, 101, 1206712072.
(90)
Orjala, J.; Gerwick, W. H. J. Nat. Prod. 1996, 59, 427-30.
(91)
Sitachitta, N.; Rossi, J.; Roberts, M.; Gerwick, W. H.; Fletcher, M. D.;
Willis, C. L. J. Am. Chem. Soc. 1998, 120, 71 31-7132.
(92)
Chang, Z.; Flatt, P.; Gerwick, W. H.; Nguyen, V. A.; Willis, C. L.;
Sherman, 0. H. Gene 2002, 296, 235-47.
(93)
Osinga, R.; Armstrong, E.; Burgess, J. G.; Hoffmann, F.; Reitner, J.;
Schumann-Kindel, G. Hydrobio!ogia 2001, 461, 55-62.
(94)
Carter, D. C.; Moore, R. E.; Mynderse, J. S.; Niemczura, W. P.; Todd,
J. S. J. Org. Chem. 1984, 49, 236-241.
(95)
Williams, 0. E.; Burgoyne, 0. L.; Rettig, S. J.; Andersen, R. J.; FathiAfshar, Z. R.; Allen, 1. M. J. Nat. Prod. 1993, 56, 545-551.
36
(96)
In, Y.; Doi, M.; Inoue, M.; Ishida, T.; Hamada, V.; Shioiri, T. Acta
Ctystallogr. C. 1994, 50, 432-434.
(97)
McDonald, L. A.; Ireland, C. M. J. Nat. Prod. 1992, 55, 376-379.
(98)
Degnan, B. M.; Hawkins, C. J.; Lavin, M. F.; McCaffrey, E. J.; Parry, D.
L.; van den Brenk, A. L.; Watters, D. J. J. Med. Chem. 1989, 32, 13491354.
(99)
Fu, X.; Do, T.; Schmitz, F. J.; Andrusevich, V.; Engel, M. H. J. Nat.
Prod. 1998, 61, 1547-1 551.
(100)
Salomon, C. E.; Faulkner, 0. J. J. Nat. Prod. 2002, 65, 689-692.
(101) Schmidt, E. W.; Sudek, S.; Haygood, M. G. J. Nat. Prod. 2004, 67,
1341-1345.
(102)
Thacker, R. W.; Starnes, S. Marine Biology 2003, 142, 643-648.
(103) Trimurtulu, G.; Ohtani, I.; Patterson, G. M.; Moore, R. E.; Corbett, T.
H.; Valeriote, F. A.; Demchik, L. J. Am. Chem. Soc. 1994, 116, 47294737.
(104) Smith, C. D.; Zhang, X.; Mooberry, S. L.; Patterson, G. M.; Moore, R.
E. Cancer Res. 1994, 54, 3779-3784.
(105)
Kobayashi, M.; Kitagawa, K. J. Nat. Toxins 1999, 8, 249-258.
(106)
Dumdei, E. J.; Simpson, J. S.; Garson, M. J.; Bryriel, K. A.; Kennard,
C. H. L. Aus. J. Chem. 1997, 50, 139-144.
(107)
Harrigan, G. G.; Goetz, G. H.; Luesch, H.; Yang, S.; Likos, J. J. Nat.
Prod. 2001, 64, 1133-1138.
(108)
MacMillan, J. B.; Trousdale, E. K.; Molinski, T. F. Org. Lett. 2000, 2,
2721-2723.
(109) Jiménez, J. I.; Scheuer, P. J. J. Nat. Prod. 2001, 64, 200-203.
(110)
Burja, A. M.; Hill, R. 1. Hydrobiologica 2001, 461,41-47.
(111)
Hentschel, U.; Hopke, J.; Horn, M.; Friedrich, A. B.; Wagner, M.;
Hacker, J.; Moore, B. S. App!. Enviro. Micro. 2002, 68, 4431-4440.
(112)
Vacelet, J.; Donadey, C. J. Exp. Mar. Ecol. 1977, 30, 301-314.
(113)
Barrows, L. R.; Radisky, D. C.; Copp, B. R.; Swaffar, D. S.; Kramer, R.
A.; Warters, R. L.; Ireland, C. M. Anti-Cancer Drug Design 1993, 8,
333-347.
37
(114)
Ishibashi, M.; Iwasaki, T.; lmai, S.; Sakamoto, S.; Yamaguchi, K.; Ito,
A.; Iwasaki, T.; lmai, S.; Sakamoto, S.; Yamaguchi, K.; Ito, A.
J. Nat.
Prod. 2001, 64, 108-110.
(115)
Hill, R. T.; Hamann, M. T.; Peraud, 0.; Kasanah, N.
Ref Type: Patent
2003, 4115-1 80 PCT.
(116)
Clark, W. D.; Crews, P.
Tetrahedron Left. 1995,
36, 1185-8.
CHAPTER TWO
SEMIPLENAMIDES A - G, FATTY ACID AMIDES FROM A PAPUA NEW GUINEA
COLLECTION OF THE MARINE CYANOBACTERIUM LYNGBYA SEMIPLENA
ABSTRACT
Semiplenamides A (1) to G (7), a series of new anandamide-like fatty acid
amides, were isolated from a 1999 Papua New Guinea collection of the marine
cyanobacterium Lyngbya semiplena. The planar structures of these lipids were
determined using standard ID and 2D NMR methods. The relative stereochemistry of
the aliphatic portion of the new metabolites was deduced from ID NOE data and 1Hdecoupling experiments, while the absolute stereochemistry of the amino alcohol
moieties was assigned by chemical derivatization and chiral GC-MS methods. All of
these new metabolites displayed toxicity in the brine shrimp model system, three
showed modest potency in displacing radiolabeled anandamide from the cannabinoid
receptor (CBI), and one was a modest inhibitor of the anandamide membrane
transporter (AMT).
INTRODUCTION
Marine cyanobacteria have emerged over the past few years as one of the
richest groups of marine organisms for their bioactive and structurally complex natural
products. Particularly prevalent among these is a rich elaboration upon a cyclic
peptide template using a diversity of standard as well as modified amino
acids.1
However, a growing trend from our investigations of these life forms is the production
of fatty acid amides that combine unusual fatty acids with a variety of "biogenic
amines". Examples include the hermitamides A (8) and B
(9)2
and grenadamide
39
(1O), two types of toxic malyngamide-type natural products from the marine
cyanobacterium Lyngbya majuscula. In this regard, they structurally resemble
anandamide (11) and other endocannabinoids (12, 13) of importance to mammalian
physiology,3
a finding that has stimulated our investigation of their pharmacological
properties in several relevant bioassays.
OCH3
0
H
Hermitamide A (8)
Hermitamide B (9)
Grenadamide (10)
Anandamide (11)
OH
N
H
LN0H
Palmitamide (12)
NH2
Oleamide (13)
40
RESULTS AND DISCUSSION
Collections of a shallow water (1-3 m) strain of L.
semiplena
were made in
Wewak Bay, Papua New Guinea. Preliminary bioassay of the crude organic extract
showed good activity in the brine shrimp toxicity model at 10 ppm.4 Guided by this
assay, the natural products semiplenamide A (1) to G (7) were isolated and purified
by sequential vacuum liquid chromatography (VLC) and HPLC in 0.1-1.5 mg/g yield.
)OR
semiplenamide
A,B
I
R=H
2
R=Ac
J.OH
1'
16
2'
semiplenamide
o
20
4
semiplenamide
o
18
semiplenamide
5
E
OR
18
semiplenamide
F,G
19
6
R=H
7 R=Ac
Semiplenamide A (1) showed an [M+H] peak at m/z 366.3372 for a molecular
formula of C23HNO2 by HR FABMS (3 degrees of unsaturation). The structure of I
41
was established mainly from analysis of ID (1H NMR, 13C NMR, and NOE) and 2D
NMR (COSY, HSQC, and HMBC) spectra (Figure II. 1-4). The 1H NMR spectrum of I
was indicative of a fatty acid amide-type metabolite, with the presence of proton
signals for a long aliphatic chain in the 6 1.25-1.35 envelope (1 8H), and a poorly
defined terminal CH3 triplet (6 0.88), as well as a broad amide singlet (6 6.25)
(Tablell. 1). The three degrees of unsaturation were accounted for by an amide
carbonyl carbon resonance (6 171.5; IR absorption at 1659 cm1) and two olefinic
bonds (6 130.7, 137.4, 129.2, 132.3). In the 1H NMR spectrum of I, a deshielded
methyl singlet at 6 1.89 (C-21) was assigned to a vinyl methyl moiety because it
showed two and three bond HMBC correlations to two olefinic carbons (6 130.7, C-2
and 137.4, C-3) as well as to the amide carbonyl carbon (6 171.5, C-I). An
ethanolamine moiety could be deduced from two mutually coupled (COSY) midfield
proton signals at 6 3.54 (dt, 6.0, 4.0 Hz) and 6 3.82 (t, 5.0 Hz), each of which
integrated to 2 protons, and which were correlated by HSQC with two midfield carbon
resonances at 6 43.6 and 63.8, respectively. The former signal was also coupled to a
broadened amide NH proton (6 6.25). The relative position of the two olefinic bonds
was determined by HMBC (Table I. 1) which showed 2- and 3- bond coupling from
two mutually coupled methylene resonances (6 2.23, H2-4; 2.14, H2-5) to all four
olefinic carbons. Additionally, the olefinic proton resonance at 6 6.45 (H-3) showed a
three bond correlation to the amide carbonyl carbon resonance (6 171.5, C-i), thus
positioning this partial structure relative to the amide motif. The linking of the
ethanolamine moiety to this amide was shown by an HMBC connectivity between the
midfield methylene protons (6 3.54, H2-i') and the C-i carbonyl. Combining these
partial structures accounted for all atoms, and thus completed the planar structure of
semiplenamide A (I).
The geometry of the C-6, 7 disubstituted olefin was determined by a 1H NMR
decoupling experiment (Figure Il. 5), in which the signals for the associated protons
were simplified by decoupling H2-5 and H2-8 (6 2.14, 1.98). In this latter spectrum it
was possible to measure JH-61H.7as 15.0 Hz, thus assigning the C-6 olefin as trans.
The geometry of the C-2, 3 olefin in I was assigned using I D NOE experiments.
Reciprocal NOE enhancements were seen from the vinyl methyl protons (6 1.88, H3-
42
21) to the methylene protons (6 2.23, H2-4) as well as to the amide NH proton, thus
defining the geometry of the C-2 olefin as E (Figure II. 6).
The 1H NMR data for semiplenamide B (2) were similar to those obtained for
1, except for an additional singlet methyl resonance at 6 2.08. HR FABMS data for 2
showed an [M + H] ion at m/z 408.3473 consistent with a molecular formula of
C25H46NO3,
indicating a structure with four degrees of unsaturation. In concurrence
with the 1H NMR data (Table II. 1), the additional degree of unsaturation was
attributed to a terminal acetate motif linked to the ethanolamine; this was confirmed
by 1H, 13C and 2D NMR data (Table II. 1).
7.5
170
7.0
160
Figure II. 1.
150
6.5
140
6.0
130
............
5.5
120
5.0
110
4.5
4.0
100
90
3.5
3.0
80
70
2.5
60
2.0
50
1H and 13C NMR spectra of semiplenamide A (1).
1.5
40
30
I
0.5 ppm
1.0
20
ppm
43
ppm
8.0-
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
Figure II. 2. COSY spectrum of semiplenamide A (1).
2.0
1.5
1.0
0.5
ppm
44
ppm
0
10
20
30
"4
40
50
60
4
70
80
90
100
110
120
130
140
...........................................
9
8
7
6
5
4
3
Figure II. 3. HSQC spectrum of semiplenamide A (1).
2
1
0
ppm
45
ppm
0
20
40
60
0
80
100
120
01
140
160
I
180
200
220
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
Figure II. 4. HMBC spectrum of semiplenamide A (1).
2.0
1.5
1.0
0.5
ppm
46
I
Iis.I
J'"l
-
.JJI
5.5
I.
I
U4
..:.
,.. -
5.4
Figure Il. 5. Determination of the geometry at C6-C7 in semiplenamide A (1) by a 1H
NMR decoupling experiment.
2.23ppm
H 1rH
/
2Oppm
OH
1,5
1.5
E.D
5.5
5.
4.5
4.0
3.5
3.0
2.0
1.5
1,0
0.5 p,:li
Figure II. 6. Determination of the geometry at C2-C3 in semiplenamide A (1) by a 1D
NOE NMR experimen
47
Table II. 1. 1H and 13C NMR Assignments for Semiplenamides A (1) and B (2).
nr
Semiplenamide A (1)
HMBC
1
2
6.45 (t, 7.0)
4
2.23 (dm, 7.0)
5
2.14 (dm, 7.0)
6
5.42 (m)
7
5.46 (m)
8
1.98 (dm, 7.5)
9
1.33 (m)
10-17 1.27(m)
18
1.27 (m)
19
1.30 (m)
20
0.88 (t, 6.5)
21
1.89 (s)
3
1'
2'
(dd,
3.82 (t, 5.0)
6.0,
171.5
130.7
137.4 C-21
29.0 C-3, C-5, C-6
32.3
129.2 C-8
132.3 C-5
33.4
30.6
30.3
32.8
23.8
14.9
12.5
HMBC
6.36 (t, 7.0)
2.23 (dm, 7.0)
2.14 (dm, 7.0)
5.42 (m)
5.46 (m)
1.98 (dm, 7.5)
1.33 (m)
1.27(m)
1.27 (m)
1.30 (m)
C-18, 19
0.88 (t, 6.5)
C-I, C-2, C-3 1.84 (s)
(dd,
43.6 C-i, C-2'
63.8
OAc
N-H
Semiplenamide B (2)
6.25 (br s)
4.21 (t, 5 .0)
2.08 (s)
169.8
131.2
136.4
29.0
32.3
129.2
132.3
33.4
30.6
30.3
32.8
c-i
C-3, c-5
C-3, C-4, C-7
C-8
C-6, C-8
C-6, C-7, c-9
C-8, C-b
C-9
23.8
14.9
6.0,
13.1
C-I, C-2, C-3
39.6
C-2', OAc
63.8
21.30
171.5
c-i', c-i
6.03 (br s)
HR FABMS of semiplenamide C (3) gave an [M + H] peak at 326.3053,
yielding a molecular formula of C20H40NO2with two degrees of unsaturation. As for 1,
the planar chemical structure of 3 was elucidated using an assemblage of 1 D and 2D
NMR (COSY, HSQC, and HMBC) experiments. The 1H NMR spectrum of 3 had
several features strikingly similar to that of I (Table II. 2), including a deshielded
olefinic 1H signal ( 6.38), a broad amide NH resonance ( 5.81) and a deshielded
methyl singlet ( 1.84), as well as proton signals for a long aliphatic chain in the S
1.25-1.35 envelope and a terminal methyl triplet (5 0.88). However, semiplenamide C
(3) lacked resonances for the
double bond found in I and 2, and was therefore a
more saturated derivative. An additional difference in the 1H NMR spectra for I and 3
48
was the presence of a CH resonance at 6 4.13 (H-i '), and CH2 resonances at 6 3.72
and 3.58 (H2-2') in 3 instead of the two CH2 resonances at 6 3.54 (H2-1') and 3.82
(H2-2') in the spectrum of 1. This methine resonance (84.13 m) was coupled to two
mutually coupled methylene resonances (6 3.72 dd, 3.58 dd), a broad amide NH
resonance (6 5.81) and a methyl resonance (6 1.22 d) in the COSY spectrum of 3.
Two- and three- bond HMBC correlations (Table II. 2) from H3-3' to C-i' and C-2'
confirmed the connectivity deduced from COSY, and thus established the structure of
an alaninol moiety in 3 versus the ethanolamine moiety in 1. An L configuration for
this group was determined by chiral GC-MS comparison of the corresponding PFPAAc ester derivative with similarly prepared alaninol standards.5
HR FABMS data for semiplenamide D (4) showed an EM + H] ion at m/z
424.3802 consistent with a molecular formula of C26HNO3, indicating a structure
with three degrees of unsaturation. The 1H and 13C NMR spectra for 4 displayed
signals very similar to those of 3 (Table II. 3), except for an additional singlet methyl
resonance (6H 2.08, oc 21.2). Detailed analysis of the 13C and 2D NMR data assigned
this additional singlet methyl signal to a terminal acetate motif linked to the
ethanolamine (as in 2), which also accounted for the additional unsaturatiori degree.
Consistent with HRFABMS data, compounds 3 and 4 also differed in the lengths of
their 'fatty acid' chains (16 and 20 carbons, respectively).
While the NMR data for semiplenamide E (5, Table Il. 3) were essentially
identical to those obtained for 4, integration of the aliphatic chain proton envelope
was slightly less for 5 than for 4 (24H vs 28H), suggesting that 5 had a shorter
aliphatic chain. HR FABMS data for 5 showed an EM + H] ion at m/z 396.3584
consistent with a molecular formula of C24H46NO3 and an 18-carbon chain, in contrast
to the 20-carbon chain of 4.
HR FABMS of semiplenamide F (6) gave an EM + H] peak at 370.3803,
yielding a molecular formula of C22HNO3 with two degrees of unsaturation. In the
1H NMR spectrum of 6 (Table II. 4), chemical shifts of the alaninol residue protons
were apparent as well as a broad amide NH resonance (6 6.49). Notably, the
deshielded olefinic proton resonance and vinyl methyl singlet in the spectra of 1-5
were replaced by an upfield methine triplet (6 2.89, J = 6.2 Hz) and methyl singlet (6
1.55). The appearance of two midfield signals (6 60.3, 64.4) in the 13C NMR
49
spectrum, taken together with the HMBC data for 6 (Table II. 4), led to the assignment
of an epoxide ring at C-2 ( 60.3) and C-3 (ö 64.4).
Table II. 2. 1H and 13C NMR Assignments for Semiplenamide C (3).
Semiplenamide C (3)
HMBC
number
1
170.9
2
131.1
3
4
5
6-13
6.38 (t, 7.2)
2.13 (dm, 7.2)
1.41 (m)
1.25(m)
14
1.25(m)
15
1.31 (m)
16
17
0.88 (t, 7.0)
1.84 (s)
1'
4.13(m)
2a'
2b'
3'
N-H
137.7 C-i, C-i7, C-4
29.1
C-3, C-5, C-2
29.4 C-4
29.7
32.0
22.8
14.8
12.8
48.7
3.58 (dd, 5.5, 3.5) 68.3
3.72 (dd, 5.5, 6.0) 68.3
1.22 (d, 7.0)
17.3
5.81 (brs)
C-i4, C-15
C-I, C-2, C-3
C-i', C-3'
C-i', C-3'
C-i',C-2'
50
Table II. 3. 1H and 13C NMR Assignments for Semiplenamides D (4) and E (5).
Semiplenamide 0 (4)
nr
HMBC
2
4
5
6-15
HMBC
169.3
1
3
Semiplenamide E (5)
169.3
130.9
130.9
6.33 (t, 7.0)
2.13 (dm, 7.0)
1.41 (m)
1.25 (m)
30.1
6.33 (t, 7.0)
2.11 (dm, 7.0)
1.41 (dm, 7.0)
1.25 (m)
137.1
28.8
29.2
137.1 C-I, C-2, c-ic
28.8
29.2
C-3, C-5
C-4
30.1
16
1.25(m)
32.3
1.25(m)
32.3
17
1.29 (m)
23.1
23.1
18
23.1
14.5
C-17, C-16
19
1.29 (m)
1.29 (m)
1.29 (m)
0.88 (t, 6.0)
23.1
1.82 (s)
13.0
C-i, C-3
20
0.88 (t, 6.0)
14.5
21
1.82(s)
4.33 (m)
13.0 C-i,C-2,C-3
4.33 (m)
4.20 (dd,
5.8)
4.05 (dd,
4.0)
1.20 (d, 7.0)
2.08 (s)
45.1
C-2', C-3'
1'
(dd,
2a'
4.05 (dd,
4.0)
3'
1.20 (d, 7.0)
QAc 2.08(s)
2b'
C-19, C-18
45.1
67.5
67.5
17.8 C-i', C-2'
21.2
171.8
N-H
5.82 (br s)
67.5 C-i', C-3'
67.5
C'-1, C-3'
17.8 C-i', C-2'
21.2
171.8
5.82 (br s)
The 1H NMR data for semiplenamide G (7) were quite similar to those
obtained for 6 (Table II. 4), with the exception again of an additional singlet methyl
resonance at
2.08. HR FABMS data for 7 showed an [M + HJ ion at m/z4i2.3427
consistent with a molecular formula of C24H46N04, indicating a structure with three
degrees of unsaturation. From examination of the HMBC data as well as the 13C NMR
spectrum (Table 4), the additional degree of unsaturation in 7 was attributed to a
terminal acetate group linked to the alaninol motif. While in CDCI3 the H3-19 and H2-4
signals were overlapped, in MeOH-d4 they were resolved and a I D NOE irradiation of
H3-i 9 produced a significant enhancement in the H2-4 resonance (ö 1.58). Therefore,
a
51
Table II. 4. 1H and 13C NMR Assignments for Semiplenamides F (6) and G (7).
Semiplenamide F (6)
HMBC
6c
number
Semiplenamide G (7)
HMBC
6c
2.89 (t, 6.2)
1.58 (m)
1.26 (m)
1.27 (m)
173.2
60.3
64.4 C-4
28.4 C-3
32.0
30.2
2.84 (t, 6.2)
1.58 (m)
1.26 (m)
1.27 (m)
60.4
64.4 C-i, C-2, C-1
28.4 C-3, C-5
32.0 C-4
30.2
16
1.30(m)
23.1
1.30(m)
23.1
17
1.41 (m)
0.88 (t, 6.0)
1.55 (s)
4.01 (m)
26.6
1.41 (m)
14.5 C-17
0.88 (t, 6.0)
13.4 C-I, C-2, C-3 1.51 (s)
48.1 C-I, C-2', C-34.20 (m)
1
2
3
4
5
6-15
18
19
1'
2a'
2b'
3'
(dd,
6.0,
(dd,
4.0,
1.16(d,7.2)
67.5 C-i', C-3'
67.5
C-i', C-3'
17.8
C-i',C-2'
OAc
N-H
6.49 (br s)
173.2
(&1,
(dd,
26.6
14.5
C-17, C-16
13.4 C-i, C-3
45.1 C-2', C-3'
2.0,
2.0,
67.5 C-I', C-3'
67.5 C'-i, C-3'
1.13 (d, 7.2)
17.8
2.08(s)
21.2
171.8
C-1',C-2'
6.40 (br s)
relative stereochemistry of 2S*, 3R* was assigned to the epoxide in 7. Because the
optical rotation values for semiplenamides F and G (6, 7) were very similar (-5.0° and
-3.0°, respectively), we conclude that they likely possess the same relative and
absolute stereochemistry. However, while it is likely that the alaninol moiety of 6 and
7 are of the same L-stereochemistry as in semiplenamides C-E (3-5), the small
amount isolated of these compounds precluded this assignment.
The semiplenamides were evaluated for their biological activity in several
systems. In the brine shrimp
(Artemia sauna)
toxicity assay,4 semiplenamides A (1)
toG (7) showed LDvalues of 1.4, 2.5, 1.5, 18, 19, 1.4, and 2.4 pM respectively.
Semiplenamides E (5) and G (7) are structurally very close except for the
replacement of the olefin moiety in 5 by an epoxide ring in 7; however, the LD value
52
of 5 is eight fold higher than that of 7. Apparently, the epoxide ring produces an
enhanced toxicity in this series of metabolites.
Due to the structural resemblance of the novel ethanolamide derivatives with
anandamide, (N-arachidonoyl-ethanolamine), an endogenous agonist of cannabinoid
receptors compounds,6 we tested them on the best characterized proteins of the
endocannabinoid system: 1) the "central" cannabinoid CB1 receptors; 2) the
anandamide membrane transporter (AMT), which is responsible for anandamide
cellular uptake; and 3) the fatty acid amide hydrolase (FAAH), which catalyses
anandamide hydrolysis! When the semiplenamides were tested for their capability to
displace the high affinity CB1 ligand [3H]SR141716A from cannabinoid CB1 receptors
in rat brain membranes, three semiplenamides (A (1), B (2) and G (7)) exhibited some
affinity. The K values were 19.5±7.8, 18.7±4.6, and 17.9±5.2 for compounds A, B and
G, respectively (means±SD, n=3). Under the same conditions, anandamide exhibited
a K =0.4 pM, and therefore, these three Lyngbya metabolties should be considered
as weak CB1 agonists. This was not surprising because substrate recognition
requires the presence of at least one cis double bond situated at the middle of the
fatty acid carbon chain, indicating a preference for ligands whose hydrophobic tail can
adopt a bent U-shaped
conformation.8
Moreover, fatty acid ethanolamides need to
contain at least three homoconjugated double bonds in order to interact in an optimal
way with GB1 receptors.6 However, it was interesting to note that: 1) semiplenamide
B (2) was as active as its non-acetylated analogue, semiplenamide A (1),
in
agreement with the previous finding that the 2'-hydroxy-group in acylethanolamides is
not necessary for the interaction with
CB1 receptors;6
2) semiplenamide G (7) was
more potent than its non-epoxide analogue, compound E (5); this is the first time that
the affinity for GB1 receptors of a 2,3-epoxide-acyl-ethanolamide derivative was
investigated.
Next, we tested the effect of the novel compounds on [14C]anandamide
hydrolysis, by using membranes from mouse NI8TG2 neuroblastoma cells, which
express high levels of FAAH
(Km
for anandamidel5
j.tM).9
No appreciable inhibitory
effect was found for any of the tested compounds. This was not surprising since
previous experiments had shown that both 2-methyl and 1'-methyl groups on
53
anandamide confer a certain refractoriness to become a substrate for or to
competitively inhibit FAAH.6
Finally, we tested the effect of the semiplenamides on the uptake of
[14C]anandamide by intact RBL-2H3 cells, a cell type in which an anandamide
membrane transporter (AMT) has been partially characterized
ranges from 9.3 to 33
pM).7'1°
(Km
for anandamide
Only semiplenamide A inhibited the AMT
(lC=18.1±3.2 tM). Although more potent AMT inhibitors have been described in the
literature (with lC
values ranging between 0.8 and 10 pM), compound A (1) is the
first of such class of inhibitors to contain at the same time 6(E) and 2-methyl-2(Z)double bond. As these two features should confer increased metabolic stability, the
of semiplenamide A widens
finding
the
possible
chemical
features
that
acylethanolamide derivatives might possess in o rder to inhibit anandamide cellular
uptake.
In conclusion, these data support the importance of natural products as a
possible reservoir for new molecules capable of interacting with proteins of the
endocannabinoid system.
EXPERIMENTAL
General Experimental Procedures.
Optical rotations were measured on a
Perkin-Elmer 141 polarimeter. IR and UV spectra were recorded on Nicolet 510 and
Beckman DU64OB spectrophotometers, respectively. NMR spectra were recorded on
Bruker Avance DPX 400 MHz and Bruker Avance 300 MHz spectrometers with the
solvent CDCI3 used as an internal standard
(oH
at 7.26, Oc at 77.2). Mass spectra
were recorded on a Kratos MS5OTC mass spectrometer. Chiral GCMS analysis was
accomplished on a Hewlett-Packard gas chromatograph 5890 Series II with a
Hewlett-Packard 5971 mass selective detector using an Alltech capillary column
(CHIRASIL-VAL phase 25 m x 0.25 mm). HPLC isolations were performed using a
Waters 515 HPLC pump and Waters 996 photodiode array detector.
Collection. The marine cyanobacterium Lyngbya semiplena (voucher
specimen available from WHG as collection number PNGE12-7Dec99-3) was
collected from shallow waters (1-3 m) in Wewak Bay, Papua New Guinea, on
December 7, 1999. The material was stored in 2-propanol at -20°C until extraction.
54
Extraction and Isolation. Approximately 138 g (dry wt) of the alga were
extracted repeatedly with CH2Cl2/MeOH (2:1) to produce 3.05 g of crude organic
extract. The extract (3.0 g) was fractionated by silica gel vacuum liquid
chromatography using a stepwise gradient solvent system of increasing polarity
starting from 10% EtOAc in hexanes to 100% MeOH. Fractions eluting with 50%,
65%, 75%, and 100% EtoAc in hexanes were found to be the most active at 10 ppm
in the brine shrimp toxicity assay. These fractions were further chromatographed on
Mega Bond RP18 solid-phase extraction (SPE) cartridges using a stepwise gradient
solvent system of decreasing polarity starting from 80% MeOH in H20 to 100%
MeOH. The most active fractions after SPE (85% toxicity at 1 ppm to brine shrimp)
were then purified by HPLC [Phenomenex Sphereclone 5 j ODS (250 x 10.00 mm),
9:1 MeOH/H20, detection at 211 nm] giving compounds 1 (1.3 mg), 2 (0.5 mg), 3 (4.5
mg), 4 (2.5 mg), 5 (0.3 mg), 6 (0.3 mg), and 7 (2.0 mg).
Semiplenamide A (1): white amorphous solid; UV (MeOH)
Amax
206 nm
(c 5100); IR (neat) 3330, 2918, 2850, 1659, 1616, 965 cm1; 1H and 13C NMR data,
see Table 1; HR FABMS m/z [M + H] 366.3372 (calcd for C23HNO2, 366.3375).
Semiplenamide B (2): colorless oil; UV (MeOH)
Amax
203 nm (c 5900); IR
(neat) 3389, 2920, 2851, 1741, 1660, 966 cm1; 1H and 13C NMR data, see Table 1;
HR FABMS m/z [M + H] 408.3473 (calcd for C25H46NO3, 408.3469).
Semiplenamide C (3): white amorphous solid; [aJ26D -5.0° (c 0.3, CHCI3); UV
(MeOH)
Amax
213 nm (c 6500); IR (neat) 3282, 2915, 2847, 1658, 1622, 966 cm1; 1H
and 13C NMR data, see Table 2; HR FABMS m/z [M + H] 326.3053 (calcd for
C20HNO2, 326.3047).
Semiplenamide D (4): white amorphous solid; [a]26D -10.6° (c 0.15, CHCI3);
UV (MeOH) Amax 207 nm (s 5200); IR (neat) 3279, 2916, 2849, 1721, 1622, 955 cm1;
1H and 13C NMR data, see Table 3; HR FABMS m/z EM + H] 424.3802 (calcd for
C26HNO3, 424.3814).
Semiplenamide E (5): white amorphous solid; [a126D -7.1 ° (c 0.28, CHCI3);
UV (MeOH) A
207 nm (c 4300); IR (neat) 3280, 2916, 2849, 1719, 1623, 966 cm1;
1H and 13C NMR data, see Table 3; HR FABMS m/z [M
C24H46NO3, 396.3578).
HJ 396.3584 (calcd for
55
Semiplenamide F (6): white amorphous solid; EU1D -5.0° (c 0.3, CHCI3); UV
(MeOH) A,
205 nm (c 5500); IR (neat) 3295, 2916, 2849, 1646, 1535, 1080 cm1;
1H
and 13C NMR data, see Table 4; HR FABMS m/z [M + H1 370.3803 (calcd
C22HNO3,370.3808).
Semiplenamide G (7): white amorphous solid;
(MeOH)
Amax
[a]26D
3.00 (c 0.6, CHCI3); UV
207 nm (c 6100); IR (neat) 3283, 2917, 2849, 1723, 1650, 1539, 1273
cm1; 1H and 13C NMR data, see Table 4; HR FABMS m/z [M + H] 412.3427 (calcd
for C24H46N04, 412.3428).
Chiral GC-MS Analyses of the Alaninol Moiety in 3. Approximately 0.2 mg
of compound 3 was hydrolyzed with 6 N HCI at 110°C (20 h). The hydrolysate was
evaporated to dryness and resuspended in HCI (0.2N, 100 pL) and heated (110 °C, 5
mm). The residue was reduced to dryness using a stream of N2, and the reaction
mixture was esterified with acetyl chloride (10 pL) at 100 °C (lh) and reduced to
dryness. To the ice cooled esterified mixture, pyridine (0.5 pL), CH2Cl2 (10 pL), and
pentafluoropropionic anhydride (1.5 pL) were added successively. Finally, the
mixture was heated in a sealed vial (2 h, 100 °C), evaporated to dryness and
resuspended in hexane. Racemic mixtures as well as optically pure L- or 0- alaninol
standards were derivatized in similar fashion. Capillary GC-MS analyses were
conducted using a Chirasil-Val column (Alltech, 25 m x 0.25 mm) with the following
conditions: initial oven temperature of 50 °C (4 mm), a 5 °C 1mm ramp from 50 °C to
150 °C, and concluding with a 20 °Clmin ramp from 150 °C to 180 °C. The fragment
derivatized from compound 3 and the derivatized L-alaninol standard both eluted at
13.02 mm. The derivatized D-alaninol standard eluted at 13.58 mm.
Brine Shrimp Toxicity Bioassay. Brine shrimp Artemia sauna toxicity was
measured as previously described.4 After a 24 h hatching period, aliquots of a 10
mglmL stock solution of compounds A-G were added to test wells containing 5 mL of
artificial seawater and brine shrimp to achieve a range of final concentrations from 0.1
to 100 ppm. After 24 h the live and dead shrimp were tallied.
Anandamide cellular uptake assay. The effect of compounds on the uptake
of [14CJAEA by intact rat basophilic leukemia (RBL-2H3) cells was studied by using
5.0 pM (20,000 cpm) of [14C]AEA as described
[14C]AEA for 5 mm at
previously.7
Cells were incubated with
37 °C, in the presence or absence of varying concentrations of the inhibitors.
Residual [14C]AEA in the incubation media after extraction with CHCI3/CH3OH 2:1 (by
vol.), determined by scintillation counting of the lyophilized organic phase, was used
as a measure of the AEA that was taken up by cells. Previous
studies7
had shown
that, after a 5 mm incubation, the amount of [14C]AEA disappeared from the medium
of RBL-2H3 cells is found mostly (>90%) as unmetabolized [14C]AEA in the cell
extract. Non-specific binding of [14CJAEA to cells and plastic dishes was determined in
the presence of 100 pM AEA and was never higher than 30%. Data are expressed as
the concentration exerting 50% inhibition of AEA uptake (lC).
CB, Receptor Binding Assay. Displacement assays for CB1 receptors were
carried out by using [3H]SR141716A (0.4nM, 55Ci/mmol, Amersham) as the highaffinity ligand, and the filtration technique described
previously,11
on membrane
preparations (0.4 mg/tube) from frozen male CD rat brains (Charles River,
Wilmington, MA, U.S.A.), and in the presence of 100 pM PMSF. Specific binding was
calculated with 1 pM SR141716A and was 84.0%. Data are expressed as the
K1,
calculated using the Cheng-Prusoff equation from the concentration exerting 50%
inhibition of [3H]SR141716A specific binding (lC).
Fatty acid amide hydrolase assay.
The effect of compounds on the
enzymatic hydrolysis of AEA was studied as described
previously,9
using membranes
prepared from mouse neuroblastoma NI8TG2 cells, incubated with the test
compounds and [14C]AEA (10 pM, 40,000 cpm) in 50 mM Tris-HCI, pH 9, for 30 mm at
37 °C. [14C]Ethanolamine produced from [14C]AEA hydrolysis was measured by
scintillation counting of the aqueous phase after extraction of the incubation mixture
with 2 volumes of CHCI3/CH3OH 2:1 (by vol.). Data are expressed as the
concentration exerting 50% inhibition of [14CJAEA hydrolysis (lC).
57
REFERENCES
(1)
Gerwick, W. H.; Tan, L. T.; Sitachitta, N. In Alkaloids: Chemistry and
Biology, Cordell, G. A., Ed.; Academic Press: New York, 2001; Vol. 57,
pp 75-184.
(2)
Tan, L. T.; Okino, T.; Gerwick, W. H. J. Nat. Prod. 2000, 63, 952-955.
(3)
Sitachitta, N.; Gerwick, W. H. J. Nat. Prod. 1998, 61, 681-684.
(4)
Meyer, B. N.; Ferrigni, N. R.; Putnam, L. B.; Jacobsen, L. B.; Nichols, D.
E.; McLaughlin, J. L. Planta Med. 1982, 45, 31-3.
(5)
Gerard, J. M.; Haden, P.; Kelly, M. T.; Andersen, R. J. J. Nat. Prod.
1999, 62, 80-85.
(6)
Melck, 0.; Bisogno, T.; De Petrocellis, L.; Chang, H.; Julius, 0.; Bifulco,
M.; Di Marzo, V. B iochem. Biophys. Res. Commun. 1999, 262, 27584.
(7)
Di Marzo, V.; Bisogno, T.; De Petrocellis, L.; Melck, D.; Martin, BR.
Curr. Med. Chem. 1999, 6, 721-744
(8)
Sheskin, T.; Hanus, L.; Stager, J.; Vogel, Z.; Mechoulam, R. J. Med.
Chem. 1997, 40, 659-667.
(9)
Maurelli, S.; Bisogno, 1.; De Petrocellis, L.; Di Marzo, V. FEBS Left.
1995, 377, 82-86.
(10)
Fowler, C.J.; Jacobsson, SO. Prostaglandins Leukot. Essent. Fatty
Acids 2002, 66, 193-200.
(11)
Bisogno, T. ; Maurelli, S. ; Melck, D. ; De Petrocellis, L. ; Di Marzo, V. J.
Biol. Chem. 1997, 272, 3315-23.
CHAPTER THREE
THE WEWAKPEPTINS, CYCLIC DEPSIPEPTIDES FROM A PAPUA NEW GUINEA
COLLECTION OF THE MARINE CYANOBACTERIUM LYNGBYA SEMIPLENA
ABSTRACT
Four new depsipeptides have been isolated from the marine cyanobacterium
Lyngbya semiplena collected from Papua New Guinea. The amino and hydroxy acid
partial structures of wewakpeptin A-D (1-4) were elucidated through extensive
spectroscopic techniques, including HR-FABMS, ID 1H and 13C NMR, as well as 2D
COSY, HSQC, HSQC-TOCSY, and HMBC spectra. The sequence of the residues
was determined through a combination of multifaceted approaches including ESI-
MS/MS, HMBC, ROESY, and a modified I D HMBC experiment. The absolute
stereochemistry of each residue was determined by chiral HPLC and chiral GC-MS
methods. The wewakpeptins represent an unusual arrangement of amino- and
hydroxy-acid subunits relative to known cyanobacterial peptides, and possess a bisester, a 2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya) or 2,2-dimethyl-3hydroxyoctanoic acid (Dhoaa) residue, and a diprolyl group reminiscent of dolastatin
15. Wewakpeptin A and B were the most cytotoxic among these four depsipeptides
with an LC50 of approximately 0.4 pM to both the NCI-H460 human lung tumor and
the neuro-2a mouse neuroblastoma cell lines.
59
INTRODUCTION
Cyanobacteria are an ancient and diverse group of microorganisms which
occupy a broad range of habitats from marine to terrestrial. Marine representatives,
especially those belonging to the genus Lyngbya, are a prolific source of secondary
metabolites with pharmaceutical potential. A particularly prevalent structural theme
among these is a rich elaboration of cyclic peptides using a diversity of standard as
well as modified amino
acids.1
An additional and emerging theme is the production of
acyl amides, as revealed by our recent isolation of a series of new anandamide-like
fatty acid amides, semiplenamides A to G, from the marine cyanobacterium Lyngbya
semiplena Gomont, collected in Papua New Guinea.2 These latter metabolites
displayed modest potency in displacing radiolabeled anandamide from the
cannabinoid receptor (CB1). In the course of this latter effort, we found the most
polar fraction from this organic extract contained a series of four new cyclic
depsipeptides, wewakpeptin A-D (1-4). Herein, we report the isolation and structure
elucidation of this new series of L. semiplena peptides, as well as their cytotoxicity to
tumor cells. Wewakpeptin A and B were the most potent among these four
depsipeptides, with an LC of approximately 0.4 pM to both the NCI-H460 human
lung tumor and the neuro-2a mouse neuroblastoma cell lines.
30
2
4
19
Ii
wewakpeptin A (1)
R=
wewakpeptin C (3)
R
wewakpeptin B (2)
R=
wewakpeptin 0 (4)
R=
RESULTS AND DISCUSSION
Collections of a shallow water (1-3 m) strain of L. semiplena were made in
Wewak Bay, Papua New Guinea. The algae were extracted with CH2Cl2/MeOH (2:1)
and fractioned by silica gel vacuum liquid chromatography. Preliminary bioassay of
the MeOH-eluted fraction showed toxicity in the brine shrimp model (LD
- I ppm).
Guided by this assay, this fraction was further chromatographed over a Mega Bond
RP18 solid-phase extraction (SPE) cartridge and then via reversed-phase HPLC to
afford four new depsipeptides, wewakpeptin A-D (1-4).
The molecular formula of wewakpeptin A (1) was determined as C52H65N7011
on the basis of HR-FABMS and NMR spectral analysis (Table III. 1). From the proton
and carbon NMR data (Figure III. I and 2), nine carbonyls accounted for nine of the
fourteen degrees of unsaturation implied by the molecular formula. The IR spectrum
revealed that I contained both amide (1644 cm1) and ester bonds (1739 cm1),
indicating that it was in fact a peptolide. Its peptidic nature was further supported by
the presence of two amide NH signals (6 8.11, 6.91) and three N-methylamide signals
61
(6 3.02, 2.98, and 2.75) in the 1H NMR spectrum. Additionally, the 13C NMR spectrum
showed two distinctive carbon signals at 6 83.6 and 6 69.3, consistent with a terminal
acetylenic functionality. As previously observed, the carbon at 6 69.3 exhibited weak
HSQC correlations to a methine proton at 6 1.93, but this proton exhibited a
2JcH
HMBC correlation to the quaternary carbon at 6 83.6, confirming the presence of an
acetylene (Figure III. 3 and 4)3
The above functionalities accounted for 11 degrees
of unsaturation, indicating that wewakpeptin A (1) had three rings.
6.5
,:
Figure Ill. 1.
170
160
Figure III. 2.
6.0
5.5
5.0
a.,
4.0
3.5
3.0
2.5
60
50
2.0
1.5
1.0
0.Sppm
1H NMR spectrum of wewakpeptin A (1).
150
140
130
120
110
100
90
80
70
13C NMR spectrum of wewakpeptin A (1).
40
30
20
10 ppm
62
pp
10
S
15
20
25
,
30
gc?
35
40
0
I
45
50 -
0
55
60
65
N
70
75
80
85
B.5
.0
7.5
7.0
6.5
6.0
5.5
5.0 4.5
4.0
i.5
i.0
4.5
Figure III. 3. HSQC spectrum of wewakpeptin A (1).
4.0
1.5 1.0
ppm
63
ppm
0
I
S
UI
2o
1,4
,
I
3o
1
4
I
5:
P
.
II
Ih
6 o
'l.
$b
0
p
80
U.
'
C
9
10 0
11
12
13
14 o
I
15 0
16
17 :
.0
/.
1.0
b.
b.0
.0
.3
.IJ
.j.z
i.0
I.
Figure III. 4. HMBC spectrum of wewakpeptin A (1).
.O
1.
1.1.1
u.
ppm
64
ppm
0.5
1.0
1.5
2.0
ti'1
.
f,.1
2.5
3.0
3.5
4.0
4.5
5.0
5.5.
'P
'.4
6.0
6.5
I
7.0
+
7.5.
8.0-
q
*
I
8.5/
6
5
4
3
2
Figure III. 5. COSY spectrum of wewakeptin A (1).
1
ppm
ppm
.0-
i5..
0
4
0
e
4
00-
0
m
0
::
4,
?%
'
40
I
I
'0
3
.
.0
b.
.IJ
I.
I.0
b.
.0
U
q.
S.
i.0
2.1)
Figure III. 6. HSQC-TOCSY spectrum of wewakpeptin A (1).
i.
J-.0
u.
ppm
Table III. 1. NMR Data for Wewakpeptin A (1) and B (2) in CDCI3.
Wewakpeptin A (1)
Unit
Dhoya
Position
1
2
3
4
5
6
7
8
lie
9
10
NH
11
N-MeAla
12
13
14
15
16
17
174.2
46.4
79.4
29.8
24.7
18.2
83.6
69.3
26.2
23.1
54.9
39.5
17.6
22.9
12.0
172.3
30.6
4.94, dd (10.0, 2.0)
1.90, 1.38
1.40, 1.36
2, 4, 9, 10, 52
3, 5
4, 6
2.16,m
5,7,8
1.93, brt(2.6)
7
1.28, s
1.26, s
6.91, d (8.2)
5.01, dd (8.2, 2.0)
1.71, m
1.04
1.01, 1.36
0.88, t (7.0)
1,2, 10
1, 2,9
1, ii
12, 13, 16
11
11, 12, 14
15
3.02, s
16, 18
5.86, m
1.17, d (6.6)
16, 17, 19, 20, 21
21
49.2
14.9
170.6
30.3
2.98,
20, 22
22
23
24
25
26
27
59.4
27.6
18.4
20.6
170.4
30.2
5.12, d (10.6)
2.46, m
0.81, d (6.8)
0.95, d (6.8)
20, 21, 23, 25, 26
24, 25
2.75, s
26, 28, 32
28
29
30
65.2
28.0
19.5
20.6
169.7
77.4
29.5
4.40, d (10.6)
2.36, m
0.98, d (7.1)
1.24, d (7.1)
26, 27, 29, 32
32, 34, 35, 36, 37
20.1
4.96, d (5.0)
2.26, m
1.07, d (7.0)
17.6
1.04, d (7.0)
3.72, 3.45, q (8.4)
2.26, 1.92
2.10,2.01
4.64, dd (8.0, 3.7)
39, 41
3.82, 3.60, q (8.2)
2.23, 1.94
1.89, 1.37
4.54, dd (7.2, 3.6)
44, 45, 46
43, 45, 46, 47
31
32
Pro-i
33
34
35
36
37
38
39
40
41
Pro-2
54.8
39.6
17.6
22.9
12.1
S
42
43
44
45
46
47
166.8
47.3
25.6
28.9
58.4
169.9
47.4
25.0
29.4
59.7
172.7
4.92
1.72, 1.28
1.29, 1.15
1.26
1.28
0.85c
1.28
1.26
6.89
5.02
1.71
1.04
1.36, 1.01
0.86
3.02
49.3
14.9
170.6
30.3
2.97
59.5
27.5
5.12
2.46
22, 23,25,
18.4
0.81
22, 23, 24
20.6
170.4
30.2
0.94
65.2
28.0
19.5
20.6
169.7
77.6
29.5
4.39
2.36
0.98
33, 34,
20.1
33, 34,37
17.6
1.07
1.04
18,20
Vat-I
Hiv
174.4
46.4
80.2
31.0
26.0
31.7
22.9
14.4
26.2
23.2
172.3
30.6
20
N-MeVai-2
WewakinB (2)
12, 14
18
19
N-Me-
HMBCa
8, (J in I-li)
30, 31
28, 29, 31
28, 29, 30
33
38, 40,41
38,39,41,42
38, 39, 40, 42, 43
43, 44,46, 47
44, 45, 47
166.9
47.3
25.5
28.8
58.4
169.9
47.4
24.9
29.4
59.7
172.7
5.83
1.19
2.75
1.25
4.95
2.26
3.77, 3.45
2.26, 1.92
2.11,2.04
4.65
3.80, 3.62
2.25, 1.94
1.85, 1.34
4.57
67
NH
Val
8.11,d(9.7)
46,47,48
59.0
4.59, dd (9.7,4.6)
47, 49, 50, 51,52
31.5
19.8
2.27, m
48, 51, 52
50
1.07, d (6.8)
48, 49, 51
51
19.1
1.09, d (6.8)
48, 49, 50
52
171.0
48
49
58.7
31.8
19.8
19.0
171.1
8.08
4.64
2.26
1.08
1.09
a
Proton showing HMBC correlation to indicated carbon. b Coupling constants for
wewakpeptin B (2) are identical to those given for A (1) except where noted. C
Resonance H3-8 was a triplet with 3JHCCH = 7.2 Hz.
Two oxygenated
the
13C
sp3 carbons
were indicated by signals at 6 79.4 and 77.4 in
NMR spectrum, suggesting the presence of two hydroxy acids in addition to
several amino acids. Further NMR analysis revealed a 2-hydroxyisovaleric acid
(Hiva) and 7 a-amino acids, including two N-methylvalines (MeVal), two prolines
(Pro), one isoleucine (ILe), one valine (Val), and one N-methylalanine (MeAla)
(Tablelll. 1). Additionally, a 2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya) was
evident based on COSY (Figure III. 5) and HMBC correlations (Figure III. 4). Geminal
dimethyl protons (H3-9 and H3-IO) showed HMBC correlations to a carbonyl carbon at
174.2 ppm (C-I), a quaternary carbon at 46.4 ppm (C-2), and an oxymethine carbon
at 79.4 ppm (C-3). Further, COSY and HSQC-TOCSY (Figure III. 6) allowed
connection of protons H-3 to H-6, the later of which showed long range coupling to H-
8 (J = 2.6 Hz). In the HMBC spectrum, the quaternary carbon of the terminal
acetylene at 6 83.6 ppm (C-7) showed a two-bond correlations with H-8 (6 1.93) and
H-6 (6 2.16, Table Ill. I).
Determination of the sequence and connection of amino acid residues and
other units (Hiva, Dhoya) in I was achieved primarily by long range 13C-1H correlation
experiments (HMBC) with different mixing times and a ROESY experiment. Despite
several proton and carbon resonances in I having close or overlapping chemical
shifts, two fragments accounting for all seven amino acid and two hydroxy acid units
could be constructed from these data (MeVal-MeVal-MeAla-lle-Dhoya; Hiva-Pro-Pro-
Val). However, neither the Dhoya nor the Hiva moiety showed the connectivity
necessary to complete the sequence. Fortunately, a modified decoupled HMBC
pulse sequence (ID HMBC) was effectively employed and showed the necessary
correlations
6,7
By irradiation of C-52 (6 171.0), selective couplings were seen to
protons at 6 8.11(NH), 4.59 (H-48), and 4.94 (H-3). Similarly couplings to protons at 6
4.40 (H-28) and 4.96 (H-33) were observed when C-32 (ö 169.75) was irradiated,
thus completing the sequence of residues in wewakpeptin A (1).
Further evidence supporting this sequence was developed from (MS)
experiments.8'9
Collisionally induced ES I-MS/MS of the mlz 985
[C52H86N7011
+ H] gave ions at m/z
899 (lIe-Dhoya-Val-Pro-Pro-Hiva-MeVaI-MeVal +H), mlz 787 (Ue-Dhoya-Val-Pro-
Pro-Hiva-MeVal +H), mlz 673 (lle-Dhoya-Val-Pro-Pro-Hiva +H), m/z 573 (lIeDhoya-Val-Pro-Pro +H), m/z 460 (Dhoya-Val-Pro-Pro +H), and m/z 294 (Val-ProPro +H) (Figure III. 7). These assignments corroborate the NMR results and
illustrate the power of (MS) experiments in assigning cyclic peptide structures.
m/z 899 (+H)
m/z 871 (+H) a
m/z 787 (+H)
m/z 673 (+1-1)
m/z 573 (+H)
m/z 460 (+H)
m/z 294 (+H)
H3
MeVaI MeVaIHivaPro ProVat DhoyaIIe
Q
wewakpeptin A (1)
m/z 920 (+11)
mlz 891 (+H)'
m/z 835(+H)
m/z 721 (+1-1)
m/z 573 (+H)
m/z 476 (+H)
m/z 343 (+H)
MeAla MeAlaMeVaIPIa---ProProVaI DhoyaIle
wewakpeptin C (3)
Figure Ill. 7. Key fragments from collisionally induced ESI-MS/MS experiments with
a) wewakpeptin A (1) and b) wewakpeptin C (3). Dhoya = 2,2-dimethyl-3-hydroxy-7octynoic acid; Hiva = 2-hydroxyisovaleric acid; Pla = 3-phenyllactic acid; a,b Fragment
ions (= a-type fragment), formed by loss of CO from the respective acylium ions, mlz
899 and m/z 920 (= b-type fragment).
The absolute configuration of I was established by analysis of degradation
products. A small sample was hydrolyzed with 6N HCI to its constituent amino and
rJ
hydroxy acid units. These were analyzed by chiral HPLC as well as chiral GC-MS
and compared with the retention times of authentic standards. All of the amino acids
as well as the Hiva unit were shown to possess L-configuration. While the Dhoya unit
was not stable to acid
hydrolysis,1°
its stereochemistry was determined by comparing
the retention time of authentic 2,2-dimethyl-3-hydroxyoctanoic acid (Dhoaa) with that
obtained through hydrogenation of I followed by acid hydrolysis and chiral GC-MS.
The hydrogenated Dhoya product from I possessed the same retention time as
synthetic R-Dhoaa. Proline amide bonds are known to have cis/trans geometry,
which correlates with the difference of the proline 13 and y 13C NMR values (.6
compound I, the small values of A6
)5
In
(3.3 and 4.5 ppm) seen for Pro-i and Pro-2,
respectively, is indicative of a trans geometry for both proline amide bonds in 1, a
finding that was supported by ROESY interactions observed between H2-43/H-41,
and H2-38/H-33.
Wewakpeptin B (2) was isolated by RP-HPLC from the crude fraction
containing wewakpeptin A (1). Its high structural homology to I was evident by
nearly identical 1H (Figure III. 8) and 13C NMR (Figure III. 9) chemical shifts (Table Ill.
1). However, it displayed an obvious difference in the Dhoya unit. The acetylenic
carbons were absent from the 13C NMR spectrum of 2 and were replaced by two
additional high-field carbons at 6 22.9 and 14.4, suggesting that 2 was likely the
tetrahydro equivalent of I. Indeed, comparison of 1H NMR spectra and FABMS after
reducing the triple bond in I over Pd-C confirmed these assignments (Table III. 1).
Hydrolysis and stereoanalysis of 2 were not undertaken due to its limited quantity;
however, because of the comparable spectroscopic properties of I and 2, we
propose they are of the same enantiomeric series.
70
8.5
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
1.0
ppm
60
50
40
30
20
10 pptn
Figure III. 8. 1H NMR spectrum of wewakpeptin B (2).
170
160
150
140
130
120
110
100
90
80
70
Figure III. 9. 13C NMR spectrum of wewakpeptin B (2).
High-resolution FABMS analysis of wewakpeptin C (3) revealed an [M+H] ion
(m/z 1004.6053) consistent with a molecular formula of CMH82N7O11,thus requiring
18 degrees of unsaturation. The IR spectrum of 3 gave characteristic absorption
bands for esters and amides
at 1741 and 1650 cm1, and the peptidic nature of 3 was again indicated by two
exchangeable NH proton resonances at
8.12 and 7.00 and three distinct NCH3
proton singlets at ö 3.00, 2.95 and 2.66. Of the 54 carbon resonances in its 13C NMR
spectrum, nine amide/ester carbonyls in the ö 165-1 80 range as well as six
71
characteristic low-field aromatic carbon resonances were observed. Two sp carbons
resonating at 6 83.6 and 69.6 suggested a terminal acetylenic functionality in 3, as in
wewakpeptin A (1). Oxygenated sp3 carbons were suspected because of signals at 6
79.3 and 74.0 in the 13C NMR spectrum, suggesting the presence of two hydroxy
acids. Detailed analysis of 1H (Figure Ill. 10) and 13C NMR (Figure III. II), COSY,
HSQC, HSQC-TOCSY, and HMBC NMR experiments (CDCI3) showed that the
structure of 3 was closely related to 1, with seven a-amino acid residues including two
methylalanines (MeAla), two prolines (Pro), one isoleucine (ILe), one valine (Val), and
one methylvaline (MeVal) (Table III. 2). By comparative NMR analysis, one of the
non-amino acid moieties in 3 was the hydroxy acid Dhoya. The other hydroxy acid,
with proton resonances at 6 7.26-7.33 (H-34
H-38), 6 3.21 and 2.99 (H2-32), and 6
5.37 (H-31), was similar to those of phenylalanine; however, the chemical shift of the
a carbon in this residue (C-31, 74.0 ppm) was typical for an oxymethine, thus
indicating that this residue was 3-phenyllactic acid (Pla). Pla is a substructure that
has been observed in several other marine metabolites, including the molluscan
metabolite kulolide-1
(5)11
and the cyanobacterial metabolite symplostatin 3
(6).12
(Table III. 2).
8.5
8.0
I
7.5
1.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
Figure UI. 10. 1H NMR spectrum of wewakpeptin C (3).
2.5
2.0
1.5
1.0
ppm
72
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10 ppm
Figure III. 11. 13C NMR spectrum of wewakpeptin C (3).
8.5
8.0
7.0
1.5
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
50
40
1.5
1.0
ppm
Figure III. 12. 1H NMR spectrum of wewakpeptin D (4).
170
160
Figure III. 13.
150
130
140
130
120
110
100
90
80
70
60
NMR spectrum of wewakpeptin D (4).
30
20
10 ppm
73
TablellI. 2. NMR Spectral Data for Wewakpeptin C (3jand 0 (4) in
Unit
Dhoya
Position
1
2
3
4
lie
5
6
7
8
9
10
NH
11
N-Me-
12
13
14
15
16
17
3H
174.4
46.4
79.3
29.9
HMBC
(.J in Hz)
N-MeVal
2,4,9, 10, 54
18.3
1.38, 1.33
2.12, m
4, 6
5, 7, 8
83.6
69.6
25.9
31.8
22.9
1.89, brt(2.5)
7
14.5
1.24, S
1,2, 10
1.22,s
1,2,9
26.2
23.2
7.00, d (8.2)
4.98, dd (8.0, 2.2)
1.66, m
0.99
10
1.30, 1.01
0.84
15
24.7
26.11
23.11
54.7
39.4
17.3
23.2
12.3
172.1
30.5
3, 5
12, 13, 16
11, 12, 14
12, 14
Pia
3.00,
49.0
14.5
5.82
1.30
2.95, S
20, 22
5.55, m
1.18, d (6.7)
20, 21, 23, 24
2.66, s
24, 26
26
27
28
29
64.9
28.7
4.19, d (10.0)
2.23, m
0.83
24, 25, 27, 30
28, 29
26, 27, 29
1.01
26,27,28
5.37, t (7.0)
3.21, 2.99, dd (14.4,
3.7)
30, 32, 33, 34,39
31, 33, 34, 39
7.26
7.33
7.28
7.33
7.26
33, 35, 36,
32, 34, 36, 37, 38
35, 37, 38
34, 36,
36, 37
40
41
Pro-2
Vai
42
43
44
45
46
47
48
49
NH
50
51
52
53
54
170.1
22,24
136.1
25.1
29.6
59.8
172.6
59.2
31.7
19.1
19.1
170.6
0.86
16, 17, 19, 20, 21
18, 20
49.7
15.5
171.3
30.6
129.9
129.0
127.6
129.0
129.9
166.6
47.2
25.7
28.9
58.4
170.0
47.6
1.68
1.00
1.30, 0.99
5.85, m
1.25
22
23
24
25
33
34
35
36
37
38
39
4.98
3.00
30.2
31
1.30
0.85c
1.27
1.24
6.95
30.5
21
19.8
20.3
169.7
74.0
37.8
1.27, 1.14
1.26
16, 18
S
14.6
170.1
32
Pro-i
49.0
54.7
39.4
17.3
23.1
12.3
4.88
1.73, 1.29
172.1
18
19
20
30
174.5
46.3
80.2
31.0
4.91, dd (7.0, 2.0)
1.37, 1.87
Ala-i
N-MeAla-2
CDCI3.
30.3
2.94
49.8
15.5
171.2
30.7
5.54
64.9
28.8
19.8
20.4
169.7
74.0
37.8
1.21
2.65
4.17
2.22
0.85
1.03
5.34
3.21, 2.98
136.1
1.96,2.24
40,42,43
40,43,44
4.48
40,41,42, 44, 45
3.85, 3.59, q (8.0)
1.71, 2.09
1.89, 1.37
4.54
46, 47
129.9
129.0
127.6
129.0
129.9
166.6
47.3
25.6
28.9
58.4
170.0
47.6
45,47
25.1
8.12, d (9.8)
4.51, dd (9.8, 4.5)
2.20, m
1.02, d (6.8)
1.04, d (6.8)
49
49, 51, 52, 53, 54
50, 52, 53
3.58, 3.12, q (7.7)
1.95, 2.27
41,
45, 46, 48
44, 45, 46, 47, 49
50, 51,53
50, 51, 52
29.5
59.8
172.5
59.0
31.7
19.2
19.0
170.9
7.26
7.33
7.28
7.33
7.26
3.62, 3.10
1.96, 2.29
1.96, 2.24
4.49
3.63, 3.83
1.66, 1.96
1.87, 1.35
4.57
8.10
4.56
2.22
1.04
1.06
74
a
Proton showing HMBC correlation to indicated carbon. bcoupling constants for
wewakpeptin 0 (4) are identical to those given for C (3) except where noted. C
Resonance H3-8 was a triplet with 3JHCCH = 7.2 Hz.
I
OCH
OCH3rj
O"OH
II
0
Symplostatin 3 (6)
Kulolide-1 (5)
Nj
0
0
1OCH3
I
0
O,CH3,O
0
0
0
LU 103793(8)
Dolastatin 15(7)
The sequencing of residues in 3 was achieved primarily by long range 13C-1H
correlation experiments (HMBC) with different mixing times as well as a ROESY
experiment. The modified decoupled HMBC pulse sequence (1D HMBC) was also
employed as in the structure elucidation of 1, and indicated two fragments (MeValMeAla-MeAla-lle-Dhoya; Pla-Pro-Pro-Val) which accounted for all seven amino acid
and the two hydroxy acid units. Collisionally induced ESI-MS/MS of the m/z 1005
[CMHB2N7O11 +
gave ions at m/z
920 (lle-Dhoya-Val-Pro-Pro-Pla-MeVal-MeAla
+H)4, m/z 835 (lle-Dhoya-Val-Pro-Pro-Pla-MeVal +H), mlz 721 (Dhoya-Val-Pro-Pro-
Pla-MeVal +H), mlz 573 (lle-Dhoya-Val-Pro-Pro +H), m/z 476 (lle-Dhoya-Val-Pro-
+H), and m/z 343 (Pro-Pro-Pla
+H)4
(Figure III.
7)
whose assignment corroborated
the NMR results and completed the linkages between the structural units. Chiral
HPLC and GC-MS analysis of the acid hydrolysate showed the presence of L-MeAla,
L-MeVal, L-Pro, L-Val, L-lle, and D-Pla. The configuration of Dhoya was determined
75
as R using the method as described above for 1. Curiously, the Pla residue in 3 has
the opposite stereochemistry compared with the Hiva residue of 1.
HR-FABMS of wewakpeptin D (4) indicated that the molecular weight was four
mass units higher than that of wewakpeptin C (3), and NMR data (Figure III. 12-13)
indicated that it was closely related to 3. However, it also displayed obvious
differences in the Dhoya unit in that it lacked the triplet at 6 1.89 for H-6 and the
acetylenic carbons at 6 83.6 and 69.3. Chemical reduction (Pd-C, H2) of the triple
bond in 3 produced 4, confirming these assignments (Table Ill. 2). Stereoanalysis of
4 was not undertaken due to its limited quantity; however, due to the highly
comparable spectroscopic properties of 4 and 3, we again propose that they also are
of the same enantiomeric series.
The wewakpeptins were tested for cytotoxicity to NCI-H460 human lung tumor
and neuro-2a mouse neuroblastoma cells (Figure Ill. 14). Intriguingly, wewakpeptin A
and B were approximately 10-fold more toxic than C and 0 to these cell lines. The
LC
for wewakpeptin A was 0.49 pM and 0.65 pM for neuro-2a and H460 cells,
respectively, and 0.20 pM and 0.43 pM, respectively, for wewakpeptin B. The LC50 for
wewakpeptin C was 10.7 pM and 5.9 pM for neuro-2a and H460 cells, respectively,
and 1.9 pM and 3.5 pM, respectively, for wewakpeptin D. These cyclic peptides
most likely derive from a nonribosomal polypeptide synthetase (NRPS)
pathway,1
and
thus, the structural variation of the wewakpeptins is intriguing and might suggest that
adenylation domains with relaxed substrate specificity are involved in their
biosynthesis. Hydrolysis of the two ester bonds in the wewakpeptins yields two linear
peptides in each case, and these are structurally related to the dolastatins (e.g.
dolastatin 15 (7), and LU 103793 (8)), and thus, it is conceivable that the
wewakpeptins undergo these reactions in nature to release more potent fragments.
76
I
0
U
0
0
20
C.)
C.)
-8
Dose (log M)
-7
-6
-5
Dose (log M)
I
1'
U
U
.2
.2
C)
C.)
0
-7
-6
-5
.4
-7
Dose (log M)
-6
-5
-4
Dose (log M)
Figure III. 14. Cytotoxicity of wewakpeptins A-D, panels A-D respectively, to H460
cells, solid lines and neuro 2-a cells, broken lines. Cells were exposed to the
indicated concentration of chemical for 48 hr before assessment of toxicity by MU
reduction. Where error bars are not present, they are occluded by the data point.
EXPERIMENTAL
General Experimental Procedures.
Optical rotations were measured on a
Perkin-Elmer 141 polarimeter. IR and UV spectra were recorded on Nicolet 510 and
Beckman DU64OB spectrophotometers, respectively. NMR spectra were recorded on
Bruker Avance DPX 400 MHz and Bruker Avance 300 MHz spectrometers with the
solvent CDCI3 used as an internal standard
(6H
at 7.26, & at 77.4). High resolution
mass spectra were recorded on a Kratos MS-50 TC mass spectrometer. Tandem
mass spectrometric data were obtained on an electrospray ionization (ESI)
quadrupole ion trap mass spectrometer (Finnigan LCQ, San Jose, CA). For ESIMS/MS analysis, samples were injected onto a Cl 8 trap column for desalting and
77
introduced into the mass spectrometer by isocratic elution using 50% acetonitrile
containing 0.1 % formic acid. For MS/MS investigations, the protonated molecular ion
clusters were isolated in the ion trap and collisionally activated with different collision
energies to find optimal fragmentation conditions. For the most intense fragment ions
MS3 and MS4 experiments were performed. Chiral GC-MS analysis was
accomplished on a Hewlett-Packard gas chromatograph 5890 Series II with a
Hewlett-Packard 5971 mass selective detector using an Alltech capillary column
(CHIRASIL-VAL phase 25 mx 0.25 mm). HPLC was performed using Waters 515
HPLC pumps and a Waters 996 photodiode array detector.
Collection. The marine cyanobacterium Lyngbya semiplena (voucher
specimen available from WHG as collection number PNG12-7Dec99-3) was collected
from shallow waters (1-3 m) in Wewak Bay, Papua New Guinea, on December 7,
1999. Taxonomy was assigned by microscopic comparison with the description given
byDesikachary.13
The material was stored in 2-propanol at -20°C until extraction.
Extraction and Isolation. Approximately 138 g (dry wt) of the alga were
extracted repeatedly with CH2Cl2/MeOH (2:1)to produce 3.05 g of crude organic
extract. The extract (3.0 g) was fractionated by silica gel vacuum liquid
chromatography using a stepwise gradient solvent system of increasing polarity
starting from 10% EtOAc in hexanes to 100% MeOH. The fraction eluting with 100%
MeOH was found to be active at I ppm in the brine shrimp toxicity assay. This
fraction was further chromatographed on Mega Bond RP18 solid-phase extraction
(SPE) cartridges using a stepwise gradient solvent system of decreasing polarity
starting from 80% MeOH in H20 to 100% MeOH. The most active fractions after SPE
(85% toxicity at I ppm to brine shrimp) were then purified by HPLC [Phenomenex
Sphereclone 5 p ODS (250 x 10 mm), 9:1 MeOH/H20, detection at 211 nm] giving
compounds 1 (5.0 mg), 2 (0.7 mg), 3 (2.0 mg), and 4 (0.7 mg).
Wewakpeptin A (1): colorless amorphous solid;
UV (MeOH) A
[aJo
-45° (c 0.40, CHCI3);
216 nm (log c 4.6); IR (neat) 3327, 2926, 2875, 1739, 1644, 1454,
1243, 979 cm; 1H and 13C NMR data, see Table 1; HR FABMS m/z[M + H]
984.6262 (calcd for C52H86N7011, 984.6302).
Wewakpeptin B (2): colorless amorphous solid;
UV (MeOH)
Amax
215 nm (log
[a]26D
-53° (c 0.47, CHCI3);
4.6); IR (neat) 3316, 2928, 2873, 1739, 1650, 1461,
1242, 976 cm; 1H and 13C NMR data, see Table 1; HR FABMS m/z[M
HJ
988.6843 (calcd for C52HN7O11, 988.6701).
Wewakpeptin C (3): colorless amorphous solid; [a]260 -56° (c 0.27, CHCI3);
UV (MeOH) Amax 220 nm (log E 4.5); IR (neat) 3340, 2932, 2875, 1741, 1650, 1446,
1242, 980 cm'; 1H and 13C NMR data, see Table 2; HR FABMS m/z EM + H]
1004.6053 (calcd for CH82N7O11, 1004.6072).
Wewakpeptin D (4): colorless amorphous solid; [a]26D -65° (c 0.60, CHCI3);
UV(MeOH)Amax 221 nm(log E4.5); IR (neat) 3343, 2924, 2851, 1741, 1654, 1457,
1244, 989 cm1; 1H and 13C NMR data, see Table 2; HR FABMS m/z [M + H]
1008.6276 (calcd for CH86N7O11, 1008.6380).
Absolute stereochemistry of 1. Wewakpeptin A (1, 500 pg) was
hydrolyzed in 6 N HCI at 105 °C for 16 h, then dried under a stream of N2 and further
dried under vacuum. The residue was reconstituted with 300 pL of H20 prior to chiral
HPLC analysis [Phenomenex Chirex 3126 (D), 4.6 x 250 mm; UV 254 nm detector],
mobile phase I: 100% 2 mM CuSO4 in H20, flow rate 0.7 mL/min; [column,
Phenomenex chirex 3126 (0), 4.6 x 50 mm; UV 254 nm detector], mobile phase II: 2
mM CuSO4 in MeCN/H20 (15:85), flow rate 0.8 mL/min. Mobile phase I elution times
(tR, mm) of authentic standards: L-MeAla (16.7), D-MeAla (17.2), L-MeVal (24.0), D-
MeVal (39.0), L-Pro (28.4), D-Pro (63.4), L-Val (38.5), D-Val (68.9). Mobile phase II
elution times (tR, mm) of authentic standards: L-Hiv (9.2), D-Hiv (14.5). The
hydrolysate was chromatographed alone and co-injected with standards to confirm
assignments of L-MeAla, L-MeVal (2 eq), L-Pro (2 eq), L-Val, and L-Hiv. The
presence of L-lle was confirmed by chiral GC-MS using established methods.14
Determination of the configuration of the Dhoya unit was established through
hydrogenation (10% Pd-C, H2), acid hydrolysis (6 N HCI, 105 °C for 14 h), and chiral
GC-MS analysis of the 2,2-dimethyl-3-hydroxyoctanoic acid (Dhoaa) residue as direct
hydrolysis of wewakpeptin A did not yield a sufficient amount of 2,2-dimethyl-3hydroxy-7-octynoic acid (Dhoya) for analysis.
For chiral GC-MS analysis of methyl-Dhoaa, portions of each standard R- and
S-Dhoaa (obtained as the gift from professor T. Ye at Hong Kong Polytechnic
University) were separately diluted in 50 pL of MeOH and treated with diazomethane
for 10 mm. Excess diazomethane and solvent were removed with a stream of N2,
79
and the residues were resuspended in CH2Cl2. Capillary GC-MS analysis was
conducted using a Chirasil-Val column (Altech, 25 m x 0.25 mm) using the following
conditions: column temperature held at 40 °C to 100 °C at a rate of 3 °Clmin, then
from 100 °C to 150 °C at a rate of 15 °Clmin. The retention time of the methylated
Dhoaa residue derived from hydrogenation and hydrolysis of I and 3, followed by
methylation (CH2N2) matched that of the methylated R-Dhoaa standard (34.6 mm) but
not the methylated S-Dhoaa standard (35.2 mm).
Absolute stereochemistry of 3. The wewakpeptin C (3, 500 pg) hydrosylate
(6 N HCI, 105 °C, 16 h) was worked up and analyzed as described above for the
wewakpeptin A hydrolysate. The following residues co-eluted with wewakpeptin C
hydrosylate peaks; mobile phase I: L-MeAla (16.3 mi, L-MeVal (25.9 mm), 2
equivalent L-Pro (28.3 mm), L-Val (38.2 mm); Mobile phase II: D-Pla (66.0 mm). The
presence of L-lle and R- Dhoya were confirmed by chiral GC-MS as described above
for the wewakpeptin A.
Biological Activity. Brine shrimp
(Artemia sauna)
toxicity was measured as
previously described.15 After a 24 h hatching period, aliquots of a 10 mg/mL stock
solution of compounds A-D were added to test wells containing 5 mL of artificial
seawater and brine shrimp to achieve a range of final concentrations from 0.1 to 100
ppm. After 24 h the live and dead shrimp were tallied.
Cytotoxicity was measured in NCI-H460 human lung tumor cells and neuro-2a
mouse neuroblastoma cells using the method of Alley et.
determined by MU
reduction.17
al'6
with cell viability being
Cells were seeded in 96-well plates at 6000 cells/well
in 180 p1 medium. Twenty-four hours later, the test chemical was dissolved in DMSO
and diluted into medium without fetal bovine serum and then added at 20 pg/well
DMSO was less than 0.5% of the final concentration. After 48 hr, the medium was
removed and cell viability determined.
REFERENCES
(1)
Gerwick, W. H.; Tan, L. T.; Sitachitta, N. In The Alkaloids; Cordell, G.
A., Ed.; Academic Press: San Diego, 2001; Vol. 57, pp 75-184.
(2)
Han, B.; McPhail, K. L.; Ligresti, A.; Di Marzo, V.; Gerwick, W. H. J.
Nat. Prod. 2003, 66, 1364-1 368.
(3)
Sitachitta, N.; Williamson, R. T.; Gerwick, W. H. J. Nat. Prod. 2000,
63, 197-200.
(4)
Wan, F.; Erickson, K. L. J. Nat. Prod. 2001, 64, 143-146.
(5)
Siemion, I. Z.; Wieland, T.; Pook, K. Angew. Chem., Int. Ed. Engi.
1975, 14, 702.
(6)
Meissner, A.; Sorensen, 0. W. Magn. Reson. Chem. 2001, 39, 49-52.
(7)
Nogle, L. M.; Marquez, B. L.; Gerwick, W. H. Org. Left. 2003, 5, 3-6.
(8)
Kuroda, J.; Fukal, T.; Nomura, 1, 1. J. Mass. Spectrom. 2001, 36, 3037.
(9)
Ngoka, L. C. M.; Gross, M. L. J. Am. Soc. Mass Spectrometry 1999,
10, 732-746.
(10)
(a) Nakao, Y. ; Yoshida, W. Y. ; Szabo, C. M.; Baker, B. J.; Scheuer,
P. J. J. Org. Chem. 1998, 63, 3272-3280. (b) Luesch, H. ;Pangilinan,
R. ; Yoshida, W. Y.; Moore. R. E.; Paul, V. J. J. Nat. Prod. 2001, 64,
304-307.
(11)
Reese, M. T.; Gulavita, N. K.; Nakao, Y.; Hamann, M. T.; Yoshida,
W. V.; Coval, S. J..; Scheuer, P. J. J. Am. Chem. Soc. 1996, 118,
11081-11084.
(12)
Luesch, H. ; Yoshida, W. V.; Moore, R. E.; Paul, V. J.; Mooberry, S.
L. ; Corbett, T. H. J. Nat. Prod. 2002, 65, 12-20.
(13)
Desikachary,
T.V.
Cyanophyta;
Indian
Council
of
Agricultural
Research: New Delhi, India, 1959; p686.
(14)
Trimurtulu, G.; Ohtani, I.; Patterson, G. M. L.; Moore, R. E.; Corbett, T.
H.; Valeriote, F. A.; Demchik, L. J. Am. Chem. Soc. 1994, 116, 472937.
81
(15)
Meyer, B. N.; Ferrigni, N. R.; Putnam, L. B.; Jacobsen, L. B.; Nichols,
0. E.; McLaughlin, J. L.
Planfa Med. 1982, 45, 31-3.
(16)
Alley, M. C., Scudiero, D. A. CancerRes. 1988, 48, 589-601.
(17)
Manger, R. L., Leja, L. S. J. AOAC.
mt. 1995, 78,
521-527.
CHAPTER FOUR
ISOLATION AND STRUCTURE OF LYNGBYABELLIN DERIVATIVES FROM A
PAPUA NEW GUINEA COLLECTION OF THE MARINE CYANOBACTERIUM
LYNGBYA MAJUSCULA
ABSTRACT
Five new lyngbyabellin analogs along with the known compound, dolabellin,
have been isolated from the marine cyanobacterium Lyngbya
majuscula
collected
from Papua New Guinea. The structures of lyngbyabellins E-I (1-5) were elucidated
through extensive spectroscopic analysis, including HR-FABMS, 1D and 2D NMR
experiments. The absolute configurations of 1 and 4 were ascertained by chiral HPLC
and GC/MS analysis of degradation products, in combination with NMR experiments.
All five lyngbyabellins showed cytotoxicity to NCI-H460 human lung tumor and
neuro-2a mouse neuroblastoma cell lines with LC50 values between 0.2 and 4.8 j.tM.
INTRODUCTION
Cyanobacteria are phenomenal producers of structurally intriguing and
biologically active secondary
metabolites.1'2
The pantropical marine cyanobacterium
Lyngbya majuscula Gomont (Oscillatoriaceae) is one of the more prolific producers of
interesting secondary metabolites, yielding no fewer than 150 reported compounds.1
As part of an effort to discover new and biologically active natural products, we now
report the isolation and structure elucidation of a series of new lyngbyabellin analogs,
lyngbyabellins E-1 (1-5), and the known compound dolabellin (6), originally isolated
from the sea hare Do/abel/a
auricularia.3
The sea hare, D. auricularia has yielded
several cytotoxic agents structurally similar to cyanobacterial counterparts, including
the anticancer agent dolastatin-1 0 and the cytotoxic agent,
dolabellin.3'4
The
isolation of dolastatin-lO (7) and related compounds, including symplostatin-1 (8),
from the marine cyanobacteria Symploca spp.,5 and the dolabellin-like compounds,
lyngbyabellins A-D (9-12) and hectochlorin(13) 6° from L. majuscula, suggests that
these sea hares sequester bioactive metabolites from their cyanobacterial diet.
Lyngbyabellin G (3) R = OH
Lyngbyabeilin E (1) R = OH
Lyngbyabellin H (4) R = H
Lyngbyabellin F (2) R = OH
Lyngbyabellin I (5) R = H
Dolabellin (6)
N
,,...
I
OCH3
o
OCH3
o
\=J
Dolastatin 10 (7)
XAN1AS -'
N
I
0
OCH3O
OCH3O
Symplostatin 1 (8)
Lyngbyabellin A (9)
Lyngbyabellin B (10)
Lyngbyabellln U (12)
Lyngabyabellin C (11)
Hectochlorin (13)
Results and Discussion
A shallow water (1-3 m) strain of L. majuscula was collected by hand from
Alotau Bay, Papua New Guinea, in 2002. The alga was extracted with CH2Cl2/MeOH
(2:1) and fractionated by silica gel vacuum liquid chromatography. Preliminary
bioassay of the EtOAc/hexanes (9:1) eluted fraction showed toxicity in the brine
shrimp model (LD
- I ppm). Guided by this assay, this fraction was further
chromatographed over a Mega Bond RP18 solid-phase extraction (SPE) cartridge and
then via reversed-phase HPLC to afford five new lyngbyabellin analogs E-1 (1-5), plus
dolabeHin (6).
The molecular formula of Iyngbyabellin E (1) was determined as
C37H51C12N3012S2
on the basis of HR-FABMS and NMR spectral data (Table 1). The
ratio of [M+H1 isotope peaks, 5:4:1, at m/z 862/864/866, clearly indicated the
presence of two chlorine atoms. From the 1H and 13C NMR data (Figure IV. 1), six
carbonyls, two carbon-carbon double bonds, and two additional carbon-heteroatom
double bonds accounted for ten of the thirteen degrees of unsaturation implied by the
molecular formula. Two downfield 1H singlets at 6 8.12 (H-12) and 8.27 (H-18),
HSQC-correlated to carbons at 6 128.7 and 130.1 (Figure IV. 2), were consistent
with the presence of two 2, 4-disubstituted thiazole rings, which assigned two of the
remaining three degrees of unsaturation. The final degree of unsaturation could be
accounted for by an additional ring within the structure of 1.
The two thiazole ring structures were confirmed by HMBC correlations from H-
12 to C-Il
(8c
146.4) and C-13 (6c 165.9), and H-18 to C-17 (3c 145.9) and C-19 (6c
165.0), respectively (Table IV. 1, Figure IV. 3). Furthermore, three-bond correlations
from H-12 and H-18 to conjugated carbonyl carbons at 6c 161.2 (C-b), and 161.9 (C16), respectively, were indicative of carboxylic acid derived functionalities attached to
these heterocycles. HMBC correlations from a 1H methine triplet at 6 6.45 (H-14) to
the thiazole C-I 3 (öc 165.9) and an oxymethylene carbon at 6 64.4 (C-I 5)
established a 2-( I ,2-dihydroxyethyl)thiazole-4-carboxylate unit. The second thiazole
ring formed a 2-(1 ,2-dihydroxy-2-methylpropyl)thiazole-4-carboxylate unit as well
based on HMBC correlations from an oxymethine singlet at 6 5.65 (H-20) to thiazole
C-19 (6c 165.0) and an oxygenated quaternary carbon at 6 72.1 (C-21), which was in
turn showed 3-bond correlations to two methyl singlets (6 1.18 and 1.34, H3-22 and
H3-23, respectively).
Inspection of the 1H NMR spectrum of I revealed a series of upfield and highly
coupled resonances indicative of an aliphatic chain. A downfield methyl singlet at 6H
2.09 (H3-8) showed HMBC correlations to a quaternary carbon at 6c 90.4 (C-7) and a
methylene carbon at 6c 49.3 (C-6). The chemical shift of C-7 was indicative of a
gem-dichioro subtituent as observed in dolabellin (6), hectochlorin, and lyngbyabellins
A-D. HSQC-TOCSY was used to extend this moiety to include an additional six
carbons (C-2 - C-6, and C-8), identifying this unit as 7,7-dichloro-3-acyloxy-2methyloctanoate (DCAMO) (Figure IV. 4). Further analysis of 2D NMR data identified
acetate and butyric acid moieties as well. The remaining unassigned C8H14NO2 was
clearly attached to C-14 via an ester linkage, based on a 3JCH correlation from H-14 to
C-24. A 1JCH from H-26 (6 5.17) to a downfield carbon at 72.6 ppm indicated that this
carbon was oxygenated, while COSY correlation to the methylene protons (H2-25)
and two HMBC correlations from H-26 to carbonyls at 169.2 and 170.6 (C-24, C-36)
established the acetylated structure depicted. An exchangeable amide proton signal
at 6 5.60 (27-NH) showed a strong COSY correlation to H-27 (Figure IV. 5), which
was expanded into a modified leucine unit (C-24 to C-31) based on the extensive 2D
NMR correlations. The HMBC data summarized in Table 1 allowed us to connect the
partial structures and functional groups described above. Although the ester linkage
between C-3 and C-b was not evidenced by the HMBC experiments, the 1H and
13C
data were strongly supportive of this remaining linkage, and were needed to account
for the final degree of unsaturation calculated from the molecular formula. The planar
structure of lyngbyabellin E is thus represented by I, closely related to the recently
reported structure of lyngbyabellin D.9
'I
8.5
8.0
7.5
7.0
6.5
6.0
170
160
150
140
130
120
5.5
110
5.0
100
4.5
90
4.0
80
3.5
70
Figure IV. 1. 1H and 13C spectra of Iyngbyabellin E (1).
3.0
60
2.5
50
2.0
40
1.5
3i
1.0
20
ppm
ppm
Table IV. 1. NMR Data for Lyngbyabellin E (1) in CDCI3.
Lyngbyabeuin E (1)
H (J in Hz)
Position
1
3.00, dq (9.5, 7.5)
5.26, m
1, 3, 9
2, 4
1.80,1.75,m
3,5
1.75, 1.73, m
2.23, 2.15, m
6
5, 7
2.09, s
6, 7
1.25,d(7.5)
1,2,3
8.12,s
10,11,13
6.45, dd (5.4, 7.3)
4.99, dd (11.4, 5.4)
4.50, dd (11.4, 7.3)
13, 15, 24
13, 14, 16
130.1
8.27, s
16, 17, 19
20
165.0
76.0
5.65, s
1, 19,21
21
72.1
3.63, brs
1.18,s
1.34,s
20, 21,22
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
21-OH
22
23
24
25
26
27
27-NH
28
29
30
31
32
33
34
35
36
37
a
174.4
43.3
74.6
31.4
20.9
49.3
90.4
37.6
14.4
159.6
146.4
128.7
165.9
70.6
64.4
HMBCa
160.5
145.9
27.4
26.1
169.1
36.5
72.6
49.2
40.0
25.0
21.6
23.8
2.81, dd (16.0, 4.9)
2.74, dd (16.1, 7.7)
5.17, ddd (4.9, 5.5, 7.7)
4.32, m
5.60, d (9.4)
1.28, m
1.65, m
0.90, d (6.5)
0.93, d (6.3)
28, 30, 31
28, 29, 31
28, 29, 30
2.12, t (7.2)
1.62, m
0.89, t (7.2)
32, 34, 35
32, 33, 35
33, 34,
2.05,s
36
24, 26,27
27, 36
26, 28
27, 29,30
173.1
38.9
19.4
13.9
170.6
21.2
Proton showing HMBC correlation to indicated carbon.
ppm
'C
2C
I
30
'I
40
50
60
70
4
'
80
90
100
110
120
44
130
......................................
9
8
7
6
5
4
3
Figure IV. 2. HSQC spectrum of Iyngbyabellin E (1).
2
1
ppm
ppm
D
..
I
*
10-
I
.
I
4*
1
0.
10-
0-
I,
I
1
C
p
I.
9.0
.5
8.0
1.5
7.0
b.5
b.0 5.5
5.0 4.5 4.0 3.5 3.0 2.5 2.0
Figure IV. 3. HMBC spectrum of Iyngbyabellin E (1).
1.5
1.0
ppm
91
pp."
1.
1,
1.
.5
B.0
7.
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
1.5
Figure IV. 4. HSQC-TOCSY spectrum of Iyngbyabellin E (1).
1.0
0.5
ppm
ppm
1
4
-
2
-
3
4
.
-
S
V
5
I.
6
7
+
8
9
8
9
7
6
5
4
3
2
1
ppm
Figure IV. 5. COSY spectrum of lyngbyabellin E (1).
Compound I posed a number of interesting problems in assignment of
stereochemistry. (S)- and/or (R)- a,13-dihydroxyisovaleric acid (Dhiv) could not be
detected in the acid hydrolyzate by comparison with the retention times of synthetic
standards.11
It appeared that the a,3-dihydroxy acid was unstable under the
conditions of acid hydrolysis. However, ozonolysis and subsequent base hydrolysis
at 90 °C afforded (S)- aj3-dihydroxyisovaleric acid, without any detectable
racemization, as demonstrated by chiral GC-MS analysis of both synthetic (S)- and
93
(R)- a,13-dihydroxyisovaleric methyl esters and the hydrolyzed natural product. Thus,
C-20 in compound I possessed an S configuration, which is identical to that found in
the aj3-dihydroxyisovaleric acid of lyngbyabellin A6 and
hectochlorin.1°
Chiral HPLC
firmly established the stereochemistry of the glyceric acid moiety, and thus
lyngbyabellin E possessed an R configuration at C-14.
There are few reports of and no general method for determining the
configuration of y-amino--hydroxy acids. However, reports in the literature have
indicated that 4-amino-3-hydroxy-5-methylhexanoic acid, such as found in the
lyngbyabellin D, undergoes an epimerization at C-3 via an acid-catalyzed
dehydration/hydration sequence, and that significant quantities of the intermediate
aj3-unsaturated acid is produced upon prolonged
hydrolysis.9'12
This suggested the
absolute configuration of C-27 could be determined by acid hydrolysis, subsequent
ozonolysis, and oxidative workup to D- or L- leucine. Indeed, this reaction sequence
gave leucine, and chiral HPLC showed unambiguously that C-27 was derived from L-
leucine. Knowing the absolute configuration at C-27, the stereochemistry at C-26
was defined using homonuclear coupling constant information and selective I D NOE
experiments in combination with a systematic analysis of all appropriate energetically
favored (staggered) rotamers (Figure IV. 6). Enhancement of the resonances for H26 and H2-25 was observed following low power irradiation at ö 4.32 (H-27), ruling out
model A2, A3, B2 and B3 (Figure IV. 6). The 'H-1H coupling constant of 5.5 Hz
between H-26 and H-27 suggested a gauche relationship between the two protons,
consequently supported ruling out again model A3 and B2. Irradiation at 5.60 (NH27) led to enhancement of the signals for H-26 (ruling out model Al and again B3), H27, H2-28, H-29 and H2-33. The remaining model BI fulfilled all criteria, suggesting a
26R, 27S configuration for the y-amino-3-hydroxy acid residue of lyngbyabellin E (I).
94
35
3LOAc
A
C-28
Ac0,-5H
HN
H C-25
>
C-28
C-25-0Ac
C32-N
HH
Al
H
C-28
H3C-25
C-32-.N
H
I
H OAc
A2
A3
35
3Q
B
HN
OAc
ZJ26
125
C28
Ac0..A..C-25
C32NAJH
HH
BI
C28
H5ç0AC
C-32-...N
i
H
C2I
C 25
C-32NH
H C-25
H OAc
B2
B3
Figure IV. 6. Diagram of all possible rotamers for the two possible epimers 26S,27S
(A series) and 26R,27S (B series) of Lyngbyabellin E (1).
The relative stereochemistry of the polyketide-derived 3-hydroxy acid (DCAO)
was assigned from NOE and hetero half-filtered TOCSY (HETLOC) experiments.
NOE correlations were observed between H-3 and CH3-9, and between CH3-9 and
H2-4. In combination with a large 3JH coupling of 9 Hz between H-2 and H-3, these
data defined an anti relationship between H-2 and H-3, in an S,S or R,R
configuration. In the HETLOC spectrum, a small heteronuclear coupling of 3JH3-C9
=
3.0 Hz was consistent with a gauche arrangement of H-3 and CH3-9, while a larger
2JH2-C3 (5.2
Hz) also implied a gauche relationship between H-2 and the oxygen of the
ester linked (dihydroxyethyl)thiazole carboxylate moiety. Other JCH values for I were
0.3 Hz (H-3/C-2), -3.0 Hz (H-3/C-9), 2.6 Hz (H-9/C-2) and -5.3 Hz (H-9/C-3). All of
these data were comparable to those for hectochlorin (13), the 2S, 3S configuration
of which was determined by X-ray.13 Conformational analysis of both I and 7 by
standard calculation methods (MM2*) based on the crystal structure of 7 resulted in
good structural overlay of this energetically favored 2S, 3S configuration. Therefore,
given also that all other members of this family possess a 3S configuration,3' 610
these data support the stereochemical assignment of lyngbyabellin E (I) as 2S, 3S,
14R, 20S, 26R, and 27S.
95
Lyngbyabellin F (2) was isolated by RP-HPLC from a slightly more polar
fraction than that containing Lyngbyabellin E (1). Its high structural homology to I
was evident from nearly identical 1H and 13C NMR chemical shifts (TablelV. 2, Figure
IV. 7). However, the molecular formula of 2 was established as CH55Cl2N3O13S2 by
HR-FABMS, indicating twelve degrees of unsaturation. These could be ascribed to six
carbonyls, two carbon-carbon double bonds, two additional carbon-heteroatom
double bonds, and two rings. The presence of a methoxy singlet at 395 (H-38) in
conjunction with a free hydroxyl group at C-14 (Table IV. 2), suggested the linear
nature of 2. Moreover, C-24 (169.4 ppm) was clearly attached to the 3-hydroxyl
group of the glyceric acid-derived moiety via an ester linkage, rather than to the a-
hydroxyl group in the structure of I, based on an HMBC correlation from H-15 to C24. Therefore 2 was a linear lipopeptide, structurally similar to the previously isolated
lyngbyabellin D.
High-resolution FABMS analysis of lyngbyabellin G (3) revealed a [M + H1 at
595.0733, consistent with a molecular formula of C23H29C12N2O8S2, thus requiring 10
degrees of unsaturation. Careful examination of ID and 2D NMR data for 3 indicated
that 3 possessed the same structure as I from C-I through C-23, but lacked the
acyclic moiety from C-24 to C-37 (Figure IV. 8). In this respect, the planar structure of
3 was almost identical to previously isolated lynbyabellin C, which has an aj3dihydroxy43-methylpentanoic acid unit instead of the aj3-dihydroxyisovaleric acid
present in 3. Intriguingly, one ester linkage of lyngbyabellin E (1) appeared to be
particularly prone to methanolysis and hydrolysis, as treatment of I with MeOH and
H20 caused it to convert slowly to 3 by regioselective ester cleavage at C-14. Hence,
lyngbyabellin G may be an artifact of I from storage of the collection in ethanol and
seawater. Hydrolysis and stereoanalysis of 2 and 3 were not undertaken due to their
limited quantity; however, because of the comparable spectroscopic properties of I,
2, and 3, we propose that they are of the same enantiomeric series.
.........................................................
6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0
-.-I-.
8.5
8.0
7.5
7.0
140
140
140
4o
90
0
70
60
50
40
30
I
ppm
20
ppa
Figure IV. 7. 1H and 13C of spectra of Iyngbyabellin F (2).
170
160
150
140
130
120
110
100
90
80
70
60
Figure IV. 8. 1H and 13C of spectra of Iyngbyabellin G (3).
50
40
30
20
ppm
Table IV. 2. NMR Data for Lyngbyabellin F (2) and G (3) in
Lyngbyabellin F (2)
Position
1
2
3
4
5
6
7
8
9
10
11
12
13
14
14-OH
15
16
17
18
19
20
21
22
23
24
25
26
27
27-NH
28
29
30
31
32
33
34
35
36
37
38
CDCI3.
Lyngbyabellin G (3)
(J in Hz)
171.3
43.9
75.1
31.5
21.8
49.3
90.4
37.6
13.7
161.2
146.7
129.3
173.4
70.0
68.7
161.9
146.4
128.8
167.7
78.7
72.2
29.9
27.0
169.4
36.9
174.1
3.02, dq (4.8, 7.0)
5.44, m
1.80, m
1.78, m
2.25, 2.22, m
2.12, s
1.29,d(7.0)
8.21, s
5.32, m
5.82, d (5.4)
4.60, dd (11.3, 3.2)
4.46, dd (11.3, 8.7)
69.2
3.16, dq (9.5, 7.3)
5.35, m
1.95, 1.77, m
1.78, m
2.29, 2.17, m
2.12, s
1.30,d(7.3)
8.17, s
5.43, m
6.06, d (10.0)
4.81, dd (11.6,2.3)
4.56, dd (11.6, 2.9)
161.1
8.17, s
6.14, s
1.26,s
1.38, s
2.76, dd (15.5, 6.1)
2.62, dd (15.7, 8.0)
5.20, ddd (3.9, 6.2, 7.9)
73.2
49.1
4.30,m
38.1
5.52, d (9.3)
1.40, 1.35, m
24.9
21.5
23.9
173.6
39.0
19.4
13.9
170.9
21.2
52.7
43.3
75.2
30.9
20.7
49.7
90.5
37.8
15.2
160.3
146.0
128.5
170.5
70.2
1.61,m
0.90, d (6.5)
0.95, d (6.5)
2.17, t (7.2)
1.65,m
0.94, t (7.2)
2.09, s
3.95, s
146.2
130.0
166.2
78.3
72.2
27.2
25.9
8.22, s
5.59, s
1.32, s
1.36, s
HR-FABMS of lyngbyabellin H (4) indicated that its molecular formula of
C37H51C12N3011S2, which is one less oxygen than in lyngbyabellin E (1). 1D NMR
data indicated that lyngbyabellin H (4) was also closely related to I (Figure IV. 9,
Table IV. 3). However, it displayed obvious differences in the a,f3-dihydroxyisovaleric
acid (Dhiv) unit in that the singlet at 6 5.65 for H-20 in I was replaced by a doublet at
6 5.52 in the spectrum for 4. Moreover, the oxygenated quaternary carbon C-21 (6
72.1) in I was replaced by an upfield methine carbon (6 32.5) in 4, thus indicating that
this residue was 2-hydroxyisovaleric acid (Hiva). From extensive 2D NMR analysis,
the planar structure of lyngbyabellin H (4) was otherwise the same as lyngbyabellin E
(I), and conceptually, also a precursor of dolabellin (6), the original compound in the
series discovered from the sea hare Do/abe/Ia auricularia. Ester bond cleavage at C24 in 4 may occurr in the acidic digestive glands of the sea hare to first yield the
cyclized form of dolabellin, and subsquential methanolysis at any stage of the storage
or isolation procedure would yield dolabellin (6). Chemical conversion of diet-derived
metabolites in the digestive glands of sea hares has been observed previously (e.g.
laurinterol into aplysin in Aplysia
californica).12
In partial support of this proposal,
lyngbyabellin I (5), the methyl ester derivative of lyngbyabellin H (4), was also isolated
from this L. ma] uscula crude extract. From NMR data in conjunction with HR FABMS
m/z [M + H] 880.2639 (calcd for C38H56C12N3O12S2, 880.2682), the planar structure of
5 appeared to be nearly identical to lyngbyabellin F (2) (Figure IV. 10), except for a
residue replacement of Dhiv by Hiva as was the case for the structure of 4. The
absolute configurations of 4 at C-14, 20, 26 and 27 were determined as described
above for Iyngbyabellin E (I) by chiral HPLC and GC/MS analysis of degradation
products, in combination with NMR analysis, which assigned the stereochemistry as
14R, 20S, 26R, and 27S. The absolute configurations at C-2 and C-3 of 4 were
determined as 2S, 3S by comparison of coupling constants and chemical shifts with
that of compound I. Due to the limited amount of compound 5, the absolute
stereochemistry was not established by chemical methods; however, we propose
that it belongs to the same enantiomeric series as these other co-occurring
metabolites.
8.5 8.0
170
160
7.5
7.0
150
140
6.5 6.0
130
120
5.2 5.0
110
4.5
100
4.0
90
3.5
60
2.5 2.0
3.0
70
60
50
40
ppm
1.0
1.5
30
20
p
Figure IV. 9. 1H and 13C of spectra of Iyngbyabellin H (4).
8.5
8.0
7.5
10
10
7.0
6.5
....
330
6.0
10
5.5
ill
5.0
100
4.5
90
4.0
3.5
. 0
3.0
2.5
60
Figure IV. 10. 1H and 13C of spectra of Iyngbyabellin I (5).
2.0
1.5
1.0
......
ppm
100
Table IV. 3. NMR Data for Lyngbyabellin H (4) and I (5) in
Lyngbyabellin H (4)
Position
1
2
3
Lyngbyabellin 1(5)
8H
174.2
43.5
74.9
4
5
6
7
8
31.1
9
10
15.1
20.8
49.3
90.4
37.6
(J in Hz)
3.02, dq (9.5, 7.4)
5.35, m
1.79,1.75, m
1.75, 1.73, m
2.26, 2.15, m
2.10, s
1.27,d(7.4)
172.3
43.3
75.0
30.4
21.6
49.5
90.5
37.5
13.1
159.8
146.6
128.6
165.5
70.4
8.09, s
6.39, dd (5.4, 7.3)
70.1
15
64.2
4.77, m
68.8
16
17
18
19
160.7
146.3
129.2
11
12
13
14
160.8
147.1
128.6
172.9
14-OH
20
22
23
24
25
26
72.6
27
27-NH
28
29
30
49.3
31
32
33
34
35
36
37
38
8.20, s
168.1
77.3
32.5
19.2
18.5
169.2
36.5
21
40.0
25.0
21.8
23.7
5.52, d (8.1)
2.47, m
0.92, d (6.5)
1.06,d(6.5)
161.9
146.9
127.7
170.3
77.9
33.6
18.9
17.4
3.12, dq (6.0, 7.0)
5.45, m
1.81, m
1.78, m
2.24, m
2.12, s
1.31,d(7.1)
8.06, s
5.29, m
5.82, d (5.3)
4.58, dd (11.4, 2.7)
4.50, dd (11.4, 8.7)
8.12, s
6.02, d (5.7)
2.44, m
0.95, d (6.5)
0.98,d(6.5)
169.1
2.82, dd (16.0, 4.8)
2.74, dd (16.1, 7.8)
5.18, ddd (4.9, 5.9,
7.8)
4.32, m
5.86, d (9.4)
1.30, m
1.62, m
0.88, d (6.4)
0.94, d (6.4)
173.1
38.9
19.4
13.9
170.6
21.2
CDCI3.
2.14, t (7.2)
1.62, m
0.91, t (7.2)
2.05,s
36.9
73.2
49.1
37.9
24.9
21.5
23.9
173.7
39.0
19.4
13.9
170.9
21.3
52.7
2.80, dd (15.2, 5.7)
2.61, dd (15.4, 8.3)
5.21, ddd (3.9, 5.7, 8.3)
4.32, m
5.56, d (9.4)
1.37, m
1.62, m
0.91, d (6.7)
0.96, d (6.7)
2.19, t (7.3)
1.66, m
0.96, t (7.2)
2.11,s
3.95, s
101
The lyngbyabellins were tested for cytotoxicity to NCI-H460 human lung tumor
and neuro-2a mouse neuroblastoma cells and had LC values between 0.2 and 4.8
pM (Table 4). Intriguingly, lyngbyabellin E (1) and H (4) appeared to be more active
against the H460 cell line with
LC50
values of 0.4 pM and 0.2 pM, respectively,
compared to LC values of 1.2 and 1.4 pM in the neuro-2a cell line. Lynbyabellin I (5)
was the most toxic to neuro-2a cells
(LC50
0.7 pM), whereas lyngbyabellin G (3), was
the least cytotoxic of all compounds to either cell line. On the basis of this limited
screening, it appears that lung tumor cell toxicity is enhanced in the cyclic
representatives, and overall potency is increased in those containing an elaborated
side chain.
Because lygbyabellin A and hectochiorin appeared to be strong actindisrupting agents as previously reported,6'
10
the cytoskeletal-disrupting effects of I
were also tested in this work. Lyngbyabellin E (I), at concentrations of 0.01-6.0 pM,
disrupted the cellular microfilament network in A-b
cells (Figure IV. 11). Additionally,
at the higher concentrations tested, many cells contained two nuclei, consistent with
the inhibition of cytokinesis that often occurs after disruption of the microfilament
network. The effects were specific for microfilaments, as there was no evidence of
microtubule loss at these concentrations.
Table IV. 4. Cytotoxicity of Compounds 1-5
Compound
(pM)
1
2
3
4
5
H460 LC (pM)
0.4
1.0
2.2
0.2
1.0
Neuro-2a
LC50
1.2
1.8
4.8
1.4
0.7
Along with lyngbyabellins E (1) 1(5), we also isolated the known compound
dolabellin (6), which was originally obtained from sea hare Do/abel/a auricularia. Due
to the very minute amount (0.08 mg) of 6 isolated, the identity of dolabellin (6) was
established by direct comparison of 1H NMR spectrum and HR-TOFMS with literature
data.3
102
Biosynthetically, these lyngbyabellins may derive from an assembly by
nonribosomal polypeptide synthetases (NRPS) and polyketide synthases
(PKS),1
and
thus, the structural variation of the lyngbyabellins is intriguing and might suggest that
adenylation
biosynthesis.
domains with relaxed substrate specificity are involved
in
their
Recently, the gene cluster involved in the biosynthesis of a related
compound hectochlorin has been characterized, in which the adenylation domain
involved in incorporation of a,3-dihydroxyisovaleric acid has exhibited relaxed
substrate
specificity.13
The isolation of dolabellin (6) and the related compounds
lyngbyabellins A-I (1-5) from the marine cyanobacterium Lyngbya majuscula suggests
that sea hares are obtaining and accumulating dolabellin and perhaps some other
compounds (eg, dolastatin 10) as a result of their cyanobacterial diet.
Figure IV. 11. Effect of lyngbyabellin E (1) on the actin cytoskeleton of A10 cells. After 24 h cells were processed and exposed to the microfilament
staining reagent TRITC-phalloidin (visualized as red in the figure) and to
the DNA-reactive compound DAPI (visualized as blue). (A) Control cells.
(B) Treatment of the cells with lyngbyabellin F (1) at 60 nM, which caused
complete loss of the cellular microfilament network and generated
binucleated cells.
Experimental
General Experimental Procedures.
Optical rotations were measured on a
Perkin-Elmer 141 polarimeter. IR and UV spectra were recorded on Nicolet 510 and
Beckman DU64OB spectrophotometers, respectively. NMR spectra were recorded on
Bruker Avance DPX 400 MHz and Bruker Avance DPX 300 MHz spectrometers with
the solvent CDCI3 used as an internal standard
(6H
at 7.27,
at 77.2). High
103
resolution mass spectra were recorded on a Kratos MS-50 TO mass spectrometer.
For chiral GC-MS, analysis was accomplished on a Hewlett-Packard gas
chromatograph 5890 Series II with a Hewlett-Packard 5971 mass selective detector
using an Alltech capillary column (CHIRASIL-VAL phase 25 m x 0.25 mm). HPLC
was performed using Waters 515 HPLC pumps and a Waters 996 photodiode array
detector. Asymmetric dihydroxylation reagents, AD-mix-a and AD-mix-13, were
purchased from Aldrich.
Collection. The marine cyanobacterium Lyngbya majuscula (voucher
specimen available from WHG as collection number PNG5-27-02-1) was collected by
hand from shallow waters (1-3 m) in Alotau Bay, Papua New Guinea, on May 7,
2002. The material was stored in 2-propanol at -20°C until extraction.
Extraction and Isolation. Approximately 150 g (dry wt) of the alga were
extracted repeatedly with CH2Cl2/MeOH (2:1)to produce 3.30g of crude organic
extract. The extract (3.2 g) was fractionated by silica gel vacuum liquid
chromatography using a stepwise gradient solvent system of increasing polarity
starting from 10% EtOAc in hexanes to 100% MeOH. The fraction eluting with 100%
MeOH was found to be active at I ppm in the brine shrimp toxicity assay. This
fraction was further chromatographed on Mega Bond RP18 solid-phase extraction
(SPE) cartridges using a stepwise gradient solvent system of decreasing polarity
starting from 80% aqueous MeOH to 100% MeOH. The most active fractions after
SPE (85% toxicity at I ppm to brine shrimp) were then purified by HPLC
[Phenomenex Sphereclone 5 u ODS (250 x 10 mm), 9:1 MeOH/H20, detection at 211
nm] giving compounds 1 (7.0 mg), 2 (0.7 mg), 3 (0.7 mg), 4 (0.7 mg), 5 (0.5 mg), and
6 (0.08 mg).
Lyngbyabellin E (1): colorless amorphous solid; [ctJ26D -31 ° (c 0.70, MeOH);
UV (MeOH) A
240 nm (log c 4.18); IR (neat) 3365, 2926, 2875, 1739, 1644, 1454,
1243, 1030
cm1; 1H and 13C NMR data, see Table 1; HR FABMS mlz EM + H] 864.2373 (calcd
for C37H52C12N3012S2, 864.2370).
Lyngbyabellin F (2): colorless amorphous solid; [a]260 -6.5° (c 0.20, MeOH);
UV (MeOH)
1232, 1098
Amax
238 nm (log
4.0); IR (neat) 3365, 2928, 2873, 1731, 1650, 1455,
104
cm1; 1H and 13C NMR data, see Table 2; HR FABMS mlz [M +
H]4
896.2633 (calcd
for C38H56C12N3O13S2, 896.2631).
Lyngbyabellin 0 (3): colorless amorphous solid;
UV (MeOH)
Amax
[a]26D
-26° (c 0.20, MeOH);
238 nm (log g 3.75); IR (neat) 3398, 2925, 2853, 1732, 1484, 1233,
1097 cm1; 1H and 13C NMR data, see Table 2; HR FABMS m/z [M +
H]4
595.0733
(calcd for C23H29C12N2O8S2, 595.0742).
Lyngbyabellin H (4): colorless amorphous solid;
UV (MeOH)
Amax
[a]260
53° (c 0.08, MeOH);
242 nm (log c4.15); IR (neat) 3305, 2959, 2852, 1735, 1649, 1470,
1238, 1073 cm1; 1H and 13C NMR data, see Table 3; HR FABMS m/z [M + H]
848.2458 (calcd for C37H52C12N3O11S2, 848.2420).
Lyngbyabellin 1(5): colorless amorphous solid;
UV (MeOH)
Amax
[a]26D
-25° (C 0.04, MeOH);
238 nm (log 4.28); IR (neat) 3309, 2960, 2852, 1740, 1650, 1461,
1233, 1096 cm1; 1H and 13C NMR data, see Table 3; HR FABMS m/z [M + H]
880.2639 (calcd for C36HCl2N3O12S2, 880.2682).
Absolute stereochemistry of Lyngbyabellin E (1). A stream of ozone was
bubbled through 0.8 mg of I dissolved in 2 mL of CH2Cl2 at -78 °C until the solution
turned pale blue (Ca. I mm). The solvent was removed under a stream of N2, and 0.3
mg of the sample was hydrolyzed with 2N NaOH at 90 °C for 7 h. This was analyzed
by chiral HPLC [column Phenomenex Chirex 3126 (D), 4.6 x 50 mm; 0.8 mI/mm;
detection at 254 nm]. The retention times (mm) of the commercially available Dglyceric and L-glyceric acid were 8.2 and 6.2 mm, respectively, with a solvent system
of 0.5 mM CuSO4 in MeCN/H20 (5:95). The base hydrolyzate was found to contain D-
glyceric acid (8.2 mm), which was confirmed by coinjection of the appropriate
standard. The presence of S-a,3-dihydroxyisovaIeric acid in I was determined by
chiral GC-MS analysis. Standard R- and S-Dhiv methyl esters were synthesized by
asymmetric dihydroxylation of methyl 3,3-dimethylacrylate using AD-mix-a and ADmix-13, respectively.9'11
Capillary GC-MS analysis was conducted using a Chirasil-Val
column (Altech, 25 m x 0.25 mm) using the following conditions: column temperature
was set at 70 °C for 3 mm, and was increased from 70 °C to 100 °C (3 °C/min), then
100 °C to 200 °C (15 °C/min). The retention time of the methylated Dhiv residue
derived from ozonolysis and saponification of 1, followed by methylation (CH2N2)
105
matched that of the methylated S-Dhiv standard (25.0 mm) but not the methylated RDhiv standard (24.7 mm).
0.5 mg of ozonized sample of I was hydrolyzed at 118 °C for 24 h in 6N HCI.
The acid was removed under a stream of N2 and the residue ozonized for 30 mm in 2
mL of methanol at -78°C. After removal of the solvent, the residue was dissolved in
2:1 98% formic acid and 30% H202, and left to stir overnight before refluxing for I h at
100 °C. The solvent was removed and the sample was analyzed by chiral HPLC
[column Phenomenex Chirex 3126 (D), 4.6 x 250 mm; 0.8 mI/mm; detection at 254
nm]. The retention time (mm, % CH3CN:2 mM CuSO4) of the standards, were L-Leu
(15.2, 15%), D-Leu (16.5, 15%), while the hydrolysate showed a peak for the former
at 15.2 mm.
Absolute stereochemistry of Lyngbyabellin H (4). 0.4 mg of 4 was
ozonized as described above for the lyngbyabellin E (1). Portion of the sample was
hydrolyzed with 2N NaOH, and followed by chiral HPLC analysis to confirm
assignments of D-glyceric acid (8.2 mi, L-Hiva (9.2 mm). The rest of sample was
determined by acid hydrolysis, subsequent ozonolysis, and oxidative workup to give
L- leucine (15.2 mm) from chiral HPLC analysis.
Biological Activity. Brine shrimp (Artemia sauna) toxicity was measured as
previously
described.14
After a 24 h hatching period, aliquots of a 10 mg/mL stock
solution of compounds 1-5 were added to test wells containing 5 mL of artificial
seawater and brine shrimp to achieve a range of final concentrations from 0.1 to 100
ppm. After 24 h the live and dead shrimp were tallied.
Cytotoxicity was measured in NCI-H460 human lung tumor cells and neuro-2a
mouse neurablastoma cells using the method of Alley et.
a115
with cell viability being
determined by MIT reduction.16 Cells were seeded in 96-well plates at 6000 cells/well
in 180 pL medium. Twenty-four hours later, the test chemicals were dissolved in
DMSO and diluted into medium without fetal bovine serum and then added at 20
p1./well DMSO was less than 0.5% of the final concentration. After 48 hr, the
medium was removed and cell viability determined.
Microfilament Disrupting Assay. Lyngbyabellmn E (I) was tested for
microfilament-disrupting activity using rhodamine-phalloidin. A-b
cells (rat aortic
smooth muscle cell line) were grown on glass coverslips in Basal Medium Eagle
106
(BME) containing 10% fetal calf serum. The cells were incubated with the test
compound for 24 h and then fixed with 3% paraformaldehyde for 20 mm,
permeabilized with 0.2% Triton X-100 for 2 mm, and chemically reduced with sodium
borohydride (1 mg/mL in PBS) three times for 5 mm each. Following a 45 mm
incubation with 100 nM TRITC-phalloidin in phosphate buffered saline (to visualize
the actin cytoskeleton), the coverslips were washed, stained with 4,6-diamidino-2phenylindole (DAPI) to visualize DNA, mounted on microscope slides, and examined
and photographed using a Nikon E800 Eclipse fluorescence microscope with a
Photometrics Cool Snap FX3 camera. The images were colorized and overlayed
using Metamorph® software.
REFERENCES
(1)
Gerwick, W. H.; Tan, L. T.; Sitachitta, N. In The Alkaloids; Cordell, G.
A., Ed.; Academic Press: San Diego, 2001; Vol. 57, pp 75-1 84.
(2)
Burja, A. M.; Hill, R. T. Hydrobiologia 2001, 461, 41-47.
(3)
Sone, H.; Kondo, T.; Kiryu, M.; Ishiwata, H.; Ojika, M.; Yamada, K. J.
Org. Chem. 1995, 60,4774-4781.
(4)
Pettit, G. R.; Kamano, Y.; Herald, C. L.; Tuinman, A. A.; Boettner, F.
E.; Kizu, H.; Schmidt, J. M.; Baczynkyji, L.; Tomer, K. B.; Bontems, R.
J. J. Am. Chem. Soc. 1987, 109, 7581-2.
(5)
Harrigan, G. G.; Luesch, H.; Yoshida, W.Y.; Moore, R. E.; Nagle, D.
G.; Paul, V. J.; Mooberry, S. L.; Corbett, T. H.; Valeriote, F. A. J. Nat.
Prod. 1998, 61, 1075-1077.
(6)
Luesch, Hendrik; Yoshida, W. V.; Moore, R. E.; Paul, V. J.; Mooberry,
S. L. J. Nat. Prod. 2000, 63, 611-615.
(7)
a. Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J. J. Nat. Prod.
2000, 63, 1437-1439; b. Milligan, K. E.; Marquez, B. L.; Williamson, R.
1.; Gerwick, W. H. J. Nat. Prod. 2000, 63, 1440-1443.
(8)
Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J. Tetrahedron,
2002, 58, 7959-7966.
107
(9)
William, P. G.; Luesch, H.; Yoshida, W. Y.; Moore, R. E.; Paul, V. J. J.
Nat. Prod. 2003, 66, 595-598.
(10)
Marquez, B. L.; Watts, K. S.; Yokochi, A.; Roberts, M. A.; VerdierPinard, P.; Jimenez, J. I.; Hamel, E.; Scheuer, P. J.; Gerwick, W. H. J.
Nat. Prod. 2002, 65, 866-871.
(11)
The standards (R)-(-)-a,13-dihydroxyisovaleric acid and (S)-(+)-aj3-
dihydroxyisovaleric acid were synthesized by asymmetric
dihydroxylation of methyl 3,3-dimethylacrylate (analogous to:
Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G. A.; Hartung,
J.; Jeong, K. S.; Kwong, H. L.; Morikawa, K.; Wang, Z. M.; Xu, D.;
Zhang, X. L. J. Org. Chem. 1992, 57, 2768- 2771) using AD-mix-a and
AD-mix-f3, respectively.
(12)
Stallard, M. 0.; Faulkner, D. J. Comp. Biochem. Physiol, Part B:
Biochemistry & Molecular Biology 1974, 49, 37-41.
(13)
Ramaswamy, A. V.; Gerwick, W. H. (manuscript submitted)
(14)
Meyer, B. N.; Ferrigni, N. R.; Putnam, L. B.; Jacobsen, L. B.; Nichols,
D. E.; McLaughlin, J. L. Planta Med. 1982, 45, 31-33.
(15)
Alley, M. C., Scudiero, D. A. CancerRes. 1988, 48, 589-601.
(16)
Manger, R. L.; Leja, L. S. J. AOAC. mt. 1995, 78, 521-527.
CHAPTER FIVE
ISOLATION, STRUCTURE, AND BIOLOGICAL ACTIVITY OF AURILIDES B AND
C FROM A PAPUA NEW GUINEA COLLECTION OF THE MARINE
CYANOBACTERIUM LYNGBYA MAJUSCULA
ABSTRACT
The bioassay-guided fractionation of the cytotoxic constituents of the marine
cyanobacterium Lyngbya majuscula collected from Papua New Guinea led to the
isolation of aurilides B (1) and C (2). The planar structures of I and 2 were
established by spectroscopic analysis, including HR-FABMS, ID 1H and 13C NMR, as
well as 2D COSY, HSQC, HSQC-TOCSY, and HMBC spectra. The absolute
stereochemistry was determined by spectroscopic analysis and chiral HPLC analysis
of acid hydrolysates of I and 2. Both aurilide B and C showed in vitro cytotoxicity
toward NCI-H460 human lung tumor and the neuro-2a mouse neuroblastoma cell
lines, with LCvalues between 0.01 and 0.13 pM. Aurilide B (I) was evaluated in the
NCI 60 cell line panel and found to exhibit a high level of growth inhibition in
leukemia, renal, and prostate cancer cell lines with a GI
less than 10 nM. Aurilide B
(I) showed net tumor cell killing activity in the NCI's hollow fiber assay, an
model for assessing a chemical's anticancer activity.
in vivo
109
INTRODUCTION
Cyanobacteria are phenomenal producers of structurally intriguing and
biologically active secondary metabolites including such important molecules as the
curacins1
and
cryptophycins.2
In our ongoing program to explore these organisms as
sources of novel anticancer leads, we recently discovered dolabellin3'4 and
lyngbyabellins
E-14
from a Papua New Guinea collection of the marine
cyanobacterium Lyngbya ma] uscula Gomont (Oscillatoriaceae). In addition, we have
identified two new cytotoxins from this collection, aurilides B (1) and C (2), which are
closely related to aurilide (3),5 originally isolated from the sea hare Do/abe/Ia
auricularia. Herein, this paper describes the isolation, structure determination, and
biological activities of aurilides B and C.
JL
NH
0
NH
OH
Aurilide B (1)
Aurilide C (2)
O'
Aurilide (3)
110
RESULTS AND DISCUSSION
Collections of a shallow water (1-3 m) strain of L. majuscula were made in
Alotau Bay, Papua New Guinea, 2002. The alga was extracted with CH2Cl2/MeOH
(2:1) and fractionated by silica gel vacuum liquid chromatography. Preliminary
bioassay of the relatively polar EtOAc-hexanes eluted fraction showed toxicity in the
brine shrimp model (LD
- I ppm). Guided by this assay, this fraction was further
chromatographed over a Mega Bond RP18 solid-phase extraction (SPE) cartridge and
then reversed-phase HPLC to afford two new metabolites, aurilides B (1) and C (2).
The molecular formula of aurilide B (1) was established as CH75N5O10 on the
basis of HR-FABMS [mlz 834.5603 (M + H) (A -1.1 mmu)J. The 1H NMR data
showed the presence of two amide NH groups at ö 7.69 and 6.75, and three Nmethylamide groups at 63.23, 2.88, and 2.63, suggesting the peptidic nature of 1.
Detailed analysis of the 2D NMR data enabled us to assign all signals from the proton
and carbon NMR spectra (Table V. 1, and Figure V. 1), and revealed a structural
framework consisting of peptide and polyketide sections (substructures a and b,
respectively). Substructure a was composed of six amino acid residues based on
interpretation of 2D NMR data (Figure V. 2-4). Elucidation of three N-methyl amides
(N-methylglycine, N-methylalanine, and N-methylisoleucine) was started with HMBC
correlations from the N-methyl carbon to the a-proton of each amino acid residue
(Figure V. 4), and followed by interpretation of 1H-1H COSY and HSQC-TOCSY NMR
spectra to complete construction of each unit within one spin system (Figure V. 5-6).
The presence of two valine residues was confirmed by 1H-1H correlations of their
amide NH protons at 6 7.69 and 6.75 to the a-protons at 6 5.12 and 4.78,
respectively, which in turn showed 1H-1H correlations to the appropriate remaining
protons within each unit in the HSQC-TOCSY NMR spectrum (Figure V. 6). A 2hydroxyisoleucic acid residue (Hila) was determined by a 1JCH from H-26 (6 4.90) to a
downfield carbon at 78.5 ppm (C-26), while a TOCSY correlation was observed from
H-26, via the methine proton H-27, to a methyl group at C-30, and the methylene
protons H2-28, which in turn showed correlation to the methyl groups at C-29. The
sequence of these residues
(two
valines, N-methylglycine, N-methyaIanine, N-
methylisoleucine, and 2-hydroxyisoleucic acid residue) was deduced from HMBC
111
correlations between H3-4/C-5, NH(1 )/C-1 0, H-1 2/C-I 3, H3-I 9/C-20, and NH(2)/C-25
(Table V. 1) to generate substructure a.
Substructure b was elucidated as follows. COSY analysis connected proton
signals from the olefinic proton H-33, via the allyllic methylene protons H2-34 and H-
35 oxymethine signal, to the methine proton H-36, which in turn showed correlations
to the methyl group at C-43 and the oxymethine proton H-37. A second spin system
could be traced from CH3-41 to the allyllic methylene protons at H2-40 and then to the
other olefinic proton H-39. HMBC long range correlations between the resonance of
H3-44 and those of C-37, C-38 and C-39 revealed that CH3-44 was attached to C-38,
and provided the connectivity of the two COSY fragments in substructure b (Figure V.
6). In a similar manner the HMBC correlations observed between H3-42 and C-31,
C-32 and C-33 completed substructure b (Table V. 1). The geometry of both the
double bonds
A3233
and
was assigned as E on the basis of the 13C NMR
chemical shifts of CH3-42 ( 12.7) and CH3-44 ( 11.3).
7.5
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.3
Figure V. 1. 1H and 13C spectra of aurilide B (1).
3.0
2.3
2.0
1.3
0:5
112
Table V. 1. NMR Data of Aurilide B (1) and C (2) in C6D6.
Position
1
2
3
4
5
6
7
8
9
10
11
öc
170.2
58.9
13.8
36.1
172.1
54.3
31.0
20.1
17.3
169.9
51.8
12
13
14
15
16
17
18
19
36.8
170.0
58.6
33.9
27.4
20
173.1
21
54.7
31.7
22
23
24
25
26
27
28
29
30
4.40, d (18.0)
3.80, d (18.0)
3.23, s
5.24, d (10.0)
2.48, m
1.86, 1.30, m
18.1
26.1
11.8
14.9
14.1
12.7
10.2
11.3
H-6, 8,9
H-9
H-6, 8
NH(1),H-11
H-12
H-li
Aurilide C (2)
H (J in Hz)
170.2
59.6
14.0
36.8
54.4
32.0
20.4
17.5
170.11
51.9
37.1
170.14
58.7
7.74, t (9.0)
2.19, m
3.97, m
2.07, m
5.18,d(1l.2)
H-39,43,44
132.1
H-37, 40,44
H-40, 41,44
134.6
21.4
14.3
12.8
4.78, dd (8.8, 8.8)
1.98, m
0.89, d (6.0)
0.90, d (6.0)
4.90, d (6.1)
2.17, m
1.50, 1.14, m
0.83, t (7.7)
1.03, d (6.0)
5.61, t (7.7)
1.95, 1.92, m
0.89
1.95, s
0.64, d (7.0)
1.54,s
7.69 br, d (9.1)
6.75 br, d (8.8)
H-41
H-40
H-33, 34
H-36
3.08, m
1.25, d (7.1)
2.55, s
172.1
H-15, 16
H-16, 18
H-15, 18
H-16, 17
H-14, 15
H-14
H-14, 19, 21
H-23
H-21, 23
H-24
H-23
NH(2), H-26
H-27, 30
H-26, 29, 30
H-29, 30
H-28
H-27
H-26, 33, 42
H-34, 42
H-34, 42
H-35
H-43
H-37, 43
0.85, d (7.0)
2.88, s
20.2
170.3
78.5
37.2
HMBC
H-2, 3, 37
H-3, 4
H-2
H-2
H-2, 4, 6
H-8, 9
H-il, 12,14
14.8
30.7
41
42
43
44
NH (1)
NH (2)
5.12, dd (9.0, 7.4)
1.97, m
1.15, d (7.0)
1.25, d (7.0)
1.03
40
32
33
34
35
36
37
38
39
3.23, m
1.21, d (7.1)
2.63, s
12.1
169.3
128.0
145.3
30.9
71.0
41.1
82.5
131.4
134.2
21.4
31
Aurilide B (1)
H (J in Hz)
34.1
27.6
12.2
15.1
30.6
173.2
54.9
31.0
18.9
20.3
170.3
80.4
30.5
18.7
18.4
169.7
128.3
146.0
30.9
71.2
41.2
82.6
10.1
11.4
5.15, dd (9.0, 5.0)
1.98, m
1.17, d (7.0)
1.28, d (7.0)
4.39. d (18.0)
3.80, d (18.0)
3.22, s
5.26, d (10.0)
2.49, m
1.89, 1.30, m
1.03
0.86, d (7.0)
2.85, s
4.75, dd (8.6, 7.5)
1.95, m
0.88
0.90
4.54, d (7.5)
2.36, m
1.00, d (7.0)
0.88, d (7.0)
7.75, t (9.0)
2.14, m
3.98, m
2.02, m
5.17, d (11.2)
5.62, t (7.7)
1.95, 1.92, m
0.89
1.95, s
0.66, (7.0)
1.54, s
H-37,39
7.66 br, d (9.1)
6.70 br, d (8.8)
113
ppal
0 .5
II
1
.0
HI
1.
1
2.
a,
2.
3,
3.
4.
.
$.,
4.
.1
5.
:
5.
6.
6.
.7
..:
..,
,.
.0
.i.
J.0
Figure V. 2. COSY spectrum of aurilide B (1).
2.5
2.0
1.5
1.0
0.5
ppm
114
pp
10
20
30
S.
40
50
0
o
S
60
0
80
p
S
90
100
110
120
130
0
140
4
150
160
b.0
1.0
l.0
0.3
0.11
3.0
3.11
%.3
1.0
0.3
Figure V. 3. HSQC spectrum of aurilide B (1).
0.0
4.3
4.11
1.3
1.0
0.3
ppm
115
ppm
b.0
I.
7.0
6.5
6.0
5.5
5.0
4.5
4.0
3.5
3.0
Figure V. 4. HMBC spectrum of aurilide B (1).
2.5
2.0
1.5
1.0
0.5
ppm
116
p
o4 i
I
'
:
à.d'
:
C
0
I
4
0
II
o
0
o
.1
II'
'I
P1
1
1
1
1
.
1
C
8.0
7.5
7.0
6.5
6.0
5.5
5.0
4.5
I
4.0
3.5
3.0
2.5
Figure V. 5. HSQC-TOCSY spectrum of aurilide B (1).
2.0
1.5
1.0
0.5
ppm
117
substructure a
cosy
HMBC
-i----
NOE
44
42
substructure b
Figure V. 6. Key COSY, HMBC, and NOE correlations for 1.
Substructures a and b were connected on the basis of HMBC data. The aproton (H-26) of Hila showed a cross-peak to the C-31 carbonyl carbon of
substructure b, and H-37 of b correlated with the C-I carbonyl carbon of substructure
a to complete a 26-membered ring.
Diagnostic NOEs from CH3-43 to H-35 and H-37, as well as the close
similarity of 1H and 13C NMR shifts and
JH,H
values between I and the known
compound (3), were indicative of the relative stereochemistry of the polyketide portion
of I as 35S*, 36S*, 37S*. To deduce the absolute stereochemistry of C-35 through
C-37, S- and R-MTPA esters were introduced to the C-35 hydroxyl group of I,
respectively.6
The
LA5(SR)values
for H-33, 34, 36, 39, 40, 41, 42, 43, and 44 were
indicative of a 5S, 6S, 7S configuration (Figure V. 7), which is identical to that of
aurilide (3).
118
To assign the absolute configuration of the amino acid residues (substructure
a), I was acid hydrolyzed and analyzed by chiral HPLC and chiral GC-MS analysis.
The absolute configurations of the three components, Val, Melle, and isoleucic acid,
were determined to be L, allo-L, allo-D, respectively. A portion of hydrolysate of I
was used to determine the absolute configuration of MeAla by Marfey
analysis;7
this
residue was shown to be L, thus completing the structure as shown.
0
OMTPA o
002
+008
-0.02
+0.26
-0.03 -0.11
-0.01
Figure V.7. M(SR)values (ppm) of MTPA esters of aurilide B (1) obtained in CD3CN.
Aurilide C (2) was isolated by RP-HPLC from the same fraction containing
aurilide B (1). High-resolution FABMS analysis of aurilide C (2) revealed an [M+H]
ion (m/z 820.5449) consistent with a molecular formula of C43H74N5010. Aurilide C
had high structural homology to I as evidenced by nearly identical 1H and 13C NMR
chemical shifts for most of the molecule (Table V. 1, and Figure V. 8). However, it
displayed differences for the isoleucic acid unit of 1. 2D NMR and mass
spectroscopy analysis revealed that the isoleucic acid residue has been replaced by
a 2-hydroxyisovaleric acid (Hiva). Hydrolysis and stereoanalysis of the peptide
portion of 2 were undertaken as described above for aurilide B (I). The absolute
configurations of the four components, Val, MeAla, Melle, and Hiva, were determined
to be L-, L-, allo-L-, and D-, respectively. We propose that the polyketide portion is of
the same configuration as that of I because of their comparable spectroscopic
properties.
Aurilide B (I) and C (2) were tested for cytotoxicity to NCI-H460 human lung
tumor and neuro-2a mouse neuroblastoma cells. Intriguingly, aurilide B was
119
approximately 4-fold more toxic than C to these cell lines. The
LC50
for aurilide B was
0.01 pM and 0.04 pM for neuro-2a and H460 cells, respectively, and 0.05 pM and
0.13 pM, respectively, for aurilide C. Aurilide B (1) was evaluated in the NCI 60 cell
line panel and found to exhibit a high level of growth inhibition in leukemia, renal, and
prostate cancer cell lines with a Gl
less than 10 nM. Aurilide B (1) showed net
tumor cell killing activity in the NCI's hollow fiber assay, an in vivo model for
assessing a chemical's anticancer activity.8 In the microfilament disruption assay
compound I caused at 2.5 pm/mL partial loss of the microfilament network when
tested against A-I 0 cells (smooth muscle cells). However the compound might
possess a multimodal mode of action taking into account its general toxicity observed
at much lower concentration.
7.5
7.0
6.5
6.0
5.0
5.5
4.5
4.0
3.5
90
80
3.0
2.5
2.0
60
50
1.5
0.5 ppn
1.0
I......
170
160
150
140
130
120
110
100
70
Figure V. 8. 1 H and I 3C spectra of aurilide C (2).
40
30
20
pt
120
Biosynthetically, the aurilides likely derive from an assembly by nonribosomal
polypeptide synthetases (NRPS) and polyketide synthases (PKS),9 and thus, the
structural variation of the aurilides might be due to adenylation domains with relaxed
substrate specificities. Recently, another cytotoxic depsipeptid,
kulokekahilide-2,1°
was isolated from a cephalaspidean mollusk, which is also closely related to aurilide
(3) obtained from the sea hare D. auricularia. Observation that some metabolites
occur in unrelated genera of marine invertebrates, as well as the isolation of related
compounds from microbial sources, provides further evidence that many natural
products are actually produced by symbionts rather than their invertebrate
hosts.11'12
For example, the isolation here of aurilide B (1) and C (2) from the marine
cyanobacterium Lyngbya majuscula suggests that marine invertebrates obtain and
accumulate the related metabolites aurilide (3) and kulokekahilide-2 as a result of
their cyanobacterial diet.
Experimental
General Experimental Procedures.
Optical rotations were measured on a
Perkin-Elmer 141 polarimeter. IR and UV spectra were recorded on Nicolet 510 and
Beckman DU64OB spectrophotometers, respectively. NMR spectra were recorded on
Bruker Avance DPX 400 MHz and Bruker Avance 300 MHz spectrometers with the
solvent C6D6 used as an internal standard (H at 7.16, 6c at 128.4). High resolution
mass spectra were recorded on a Kratos MS-50 TC mass spectrometer. Chiral GCMS analysis was accomplished on a Hewlett-Packard gas chromatograph 5890
Series II with a Hewlett-Packard 5971 mass selective detector using an Alltech
capillary column (CHIRASIL-VAL phase 25 m x 0.25 mm). HPLC was performed
using Waters 515 HPLC pumps and a Waters 996 photodiode array detector.
Collection. The marine cyanobacterium Lyngbya majuscula (voucher
specimen available from WHG as collection number PNG5-27-02-1) was collected
from shallow waters (1-3 m) in Alotau Bay, Papua New Guinea, on May 7, 2002. The
material was stored in 2-propanol at -20°C until extraction.
121
Extraction and Isolation. Approximately 150 g (dry wt) of the alga were
extracted repeatedly with CH2Cl2/MeOH (2:1) to produce 3.30 g of crude organic
extract. The extract (3.2 g) was fractionated by silica gel vacuum liquid
chromatography using a stepwise gradient solvent system of increasing polarity
starting from 10% EtOAc in hexanes to 100% MeOH. The fraction eluting with 90%
EtOAc was found to be active at I ppm in the brine shrimp toxicity assay. This
fraction was further chromatographed on Mega Bond RP15 solid-phase extraction
(SPE) cartridges using a stepwise gradient solvent system of decreasing polarity
starting from 80% MeOH in HO to 100% MeOH. The most active fractions after SPE
(85% toxicity at 1 ppm to brine shrimp) were then purified by HPLC [Phenomenex
Sphereclone 5 p ODS (250 x 10 mm), 9:1 MeOH/H20, detection at 211 nm] giving
compounds 1 (80 mg), and 2 (5 mg).
Aurilide B (1): colorless amorphous solid;
(MeOH)
Amax
[cLJ24D
-17° (c 0.34, MeOH); UV
222 nm (log 64.65); IR (neat) 3481 (br), 3353 (br), 2964, 1740, 1687,
1646, 1251,1205 cm1; 1H and 13C NMR data, see Table 1; HR FABMS m/z[M + H]
834.5603 (calcd for CH76N5O10, 834.5592).
Aurilide C (2): colorless amorphous solid;
(MeOH)
Amax
[a]240
-19° (c 0.39, MeOH); UV
222 nm (log c 4.56); IR (neat) 3487 (br), 3350 (br), 2962, 1740, 1687,
1647, 1251, 1207 cm1; 1H and 13C NMR data, see Table 1; HR FABMS m/z EM + H]
820.5449 (calcd for C43H74N5010, 820.5436).
MTPA Esters of 1. Aurilide B (1, 0.5 mg each) was reacted with R- or SMTPACI (10 pL) in 300 pL of CH2Cl2 containing 10 mg of DMAP. The reaction
mixtures were partitioned with EtOAcIO.1 M NaHCO3, and the EtOAc layers were
washed with 0.1 M HCI and H20. The EtOAc layer was evaporated and then
separated by ODS HPLC [Phenomenex Sphereclone 5 pODS (250 x 10mm), 9:1
MeOH/H20, detection at 220 nmj to yield S- and R- MTPA esters.
S-MTPA ester: 1H NMR (CD3CN) ö 7.20 (H-33), 2.55 (H-34a), 2.19 (H-34b),
5.83 (H-35), 2.38 (H-36), 5.05 (H-37), 5.51 (H-39), 2.05 (H-40), 0.95 (H-41), 1.58 (H-
42), 0.85 (H-43), 1.61 (H-44); FABMS m/z 1050.7 (M + H).
R-MTPA ester: 1H NMR (CD3CN) ö 7.28 (H-33), 2.67 (H-34a), 2.24 (H-34b),
5.80 (H-35),
a
122
2.37 (H-36), 5.05 (H-37), 5.49 (H-39), 2.04 (H-40), 0.93 (H-41), 1.83 (H-42), 0.73 (H-
43), 1.60 (H-44); FABMS mlz 1050.7 (M + H).
Absolute stereochemistry of the peptide portion of 1. Aurilide B (1, 0.3
mg) was hydrolyzed in 6 N HCI at 110 °C for 24 h, then dried under a stream of N2
and further dried under vacuum. The residue was reconstituted with 300 pL of H20
prior to chiral HPLC analysis. Mobile phase I: 2 mM CuSO4 in MeCN/H20 (5:95),
flow rate I mL/min [Phenomenex Chirex 3126 (D), 4.6 x 250 mm; UV 254 nm
detector]. Mobile phase I elution times (tR, mm) of authentic standards: allo-L-Melle
(16.8), L-Melle (17.4), allo-D-Melle (22.2), D-MeIIe (23.0).
Mobile phase II: 2 mM CuSO4 in MeCN/H20 (15:85), flow rate I mL/min
[column, Phenomenex chirex 3126 (D), 4.6 x 50 mm; UV 254 nm detector]. Mobile
phase II elution times (tR, mm) of authentic standards: allo-L-isoleucic acid (8.2), L-
isoleucic acid (11.5), allo-D-isoleucic acid (14.2), D-isoleucic acid (18.2).
The hydrolysate was chromatographed alone and co-injected with standards
to confirm assignments of allo-L-Melle, allo-D-isoleucic acid. The presence of L-Val
(2 eq) was confirmed by chiral GCIMS using established methods.13
The configuration of MeAla was determined by Marfey's analysis.7 A portion
of the hydrolysate was evaporated to dryness and resuspended in H20 (100 p.L). A
0.1% 1-fluoro-2,4-dinitrophenyl-5-L-alaninamide solution in acetone (L-Marfey's
reagent, 20 p.L) and IN NaHCO3(10 1tL) were added to a portion of the hydrolysate,
and the mixtures were heated at 40°C for I h. The solutions were cooled to room
temperature, neutralized with 2N HCI (5 j.tL) and evaporated to dryness. The
residues were resuspended in H20 (50 jiL) and analyzed by reversed-phase HPLC
[Microsob-MV
C18,
4.6 x 250 mm, UV detection at 340 nm] using a linear gradient of
9:1 50mM triethylamine phosphate (TEAP) buffer (pH 3.1)/CH3CN to 1:1
TEAP/CH3CN over 60 mm. The derivatized MeAla residue in the hydrolysate of I
eluted at the same retention time as the derivatized standard L-MeAla (12.0 mm) but
not that of D-MeAIa (13.7 mm).
The absolute stereochemistry of aurilide C (2) was analyzed as described
above for the aurilide B (1) hydrolysate. Allo-L-Melle (16.8 mm), and D-Hiva (9.2 mm)
were assigned to aurilide C (2) by chiral HPLC; L-MeAla (12.0 mm) was determined
123
based on Marfey's analysis. The presence of L-Val (2 eq) was confirmed by chiral
GC/MS using established
methods.13
Biological Activity. Brine shrimp (Artemia sauna) toxicity was measured as
previously described.14 After a 24 h hatching period, aliquots of a 10 mg/mL stock
solution of compounds A-D were added to test wells containing 5 mL of artificial
seawater and brine shrimp to achieve a range of final concentrations from 0.1 to 100
ppm. After 24 h the live and dead shrimp were tallied.
Cytotoxicity was measured in NCI-H460 human lung tumor cells and neuro-2a
mouse neurablastoma cells using the method of Alley et
al,15
with cell viability being
determined by MU reduction.16 Cells were seeded in 96-well plates at 6000 cells/well
in 180 p1 of medium. Twenty-four hours later, the test chemicals were dissolved in
DMSO and diluted into medium without fetal bovine serum and then added at 20
pg/well DMSO was less than 0.5% of the final concentration. After 48 hr, the
medium was removed and cell viability determined.
REFERENCES
(1)
Gerwick, William H.; Proteau, Philip J.; Nagle, Dale G.; Hamel, Ernest;
Blokhin, Andrei; Slate, Doris L. J. Org. Chem. 1994, 59, 1243-5.
(2)
Trimurtulu, Golakoti; Ohtani, Ikuko; Patterson, Gregory M. L.; Moore,
Richard E.; Corbett, Thomas H.; Valeriote, Frederick A.; Demchik, Lisa. J.
Am. Chem. Soc. 1994, 116, 4729-37.
(3)
Sone, H.; Kondo, T.; Kiryu, M.; Ishiwata, H.; Ojika, M.; Yamada, K. J. Org.
Chem. 1995, 60, 4774-4781.
(4)
Han, B.; McPhail, K. L; Gross, H.; Goeger, D.; Moobery, S. L.; Gerwick, W.
H. (manuscript in preparation).
(5)
Suenaga, K.; Mutou, T.; Shibata, T.; Itoh, 1.; Fujita, 1; Takada, N.;
Hayamizu, K.; Takagi, M.; Irifune, T.; Kigoshi, H.; Yamada, K. Tetrahedron
2004, 60, 8509-8527.
124
(6)
Ohtani, Ikuko; Kusumi, Takenori; Kashman, Yoel; Kakisawa, Hiroshi. J.
Am. Chem. Soc. 1991, 113, 4092-6.
(7)
Marfey, Peter. Carlsberg Res. Commun. 1984, 49, 591-6.
(8)
Plowman, J.; Dykes, D. J.; Hollingshead, M.; Simpson-Herren, L; Alley, M.
C. In Anticancer Drug Development Guide: Preclinical Screening, Clinical
Trials, and Approval. B. Teicher, Ed.; Humana Press, Inc. Totowa, NJ.
1997, pp 101-125.
(9)
Gerwick, W. H.; Tan, L. T.; Sitachitta, N. In The Alkaloids; Cordell, G. A.,
Ed.; Academic Press: San Diego, 2001; Vol. 57, pp 75-1 84.
(10) Nakao, Yoichi; Yoshida, Wesley Y.; Takada, Yuuki; Kimura, Junji; Yang,
Liu; Mooberry, Susan L.; Scheuer, Paul J. J. Nat. Prod. 2004, 67, 13321340.
(11) Haygood, M. G.; Schmidt., E. W.; Davidson, S. K.; Faulkner, D. J. J. Mo!.
Microbiol. Biotechno!. 1999, 1, 33-43.
(12) Osinga, R.; Armstrong, E.; Burgess, J. G.; Hoffmann, F.; Reitner, J.;
Schumann-Kindel, G. Hydrobiologia 2001, 461, 55-62.
(13) Trimurtulu, G.; Ohtani, I.; Patterson, G. M. L.; Moore, R. E.; Corbett, T. H.;
Valeriote, F. A.; Demchik, L. J. Am. Chem. Soc. 1994, 116, 4729-37.
(14) Meyer, B. N.; Ferrigni, N. R.; Putnam, L. B.; Jacobsen, L. B.; Nichols, D.
E.; McLaughlin, J. L. Planta Med. 1982, 45, 31-3.
(15) Alley, M. C.; Scudiero, D. A. CancerRes. 1988, 48, 589-601.
(16) Manger, R. L.; Leja, L. S. J. AOAC. !nt. 1995, 78, 521-527.
125
CHAPTER SIX
CONCLUSIONS
As is quite evident from the findings reported in the thesis, marine
cyanobacteria are extraordinary prolific in their production of elaborate and complex
bioactive secondary metabolites. This investigation provided a good representation
of the diversity of organic molecules produced by marine cyanobacteria of the genus
Lyngbya. A total of eighteen new compounds were identified from the organic
extracts of two strains of Lyngbya from Papua New Guinea. The isolation and
purification of these molecules was achieved by different chromatographic
techniques, including VLC and HPLC, while their structures were established by
extensive ID and 2D NMR experiments and mass spectrometry. All of the secondary
metabolites found in this work are nitrogen-containing compounds with molecular
sizes ranging from 325 MW (semiplenamide C) to 1008 MW (wewakpeptin D).
Preliminary bioassays of the crude organic extract of the marine
cyanobacterium Lyngbya semip!ena collected from Papua New Guinea in 1997
showed good activity in the brine shrimp toxicity model at 10 ppm. Guided by this
assay, seven new anandamide-like fatty acid amides, semiplenamides A to G,
together with four cyclic depsipeptides, wewakpeptins A to D, were identified. Due to
the structural resemblance of the novel ethanolamide derivatives (semiplenamide AG) with anandamide, (N-arachidonoyl-ethanolamine), an endogenous agonist of
cannabinoid receptors, semiplenamide A-G were tested on the well characterized
proteins of the endocannabinoid system: 1) the "central" cannabinoid CB1 receptors;
2) the anandamide membrane transporter (AMT), which is responsible for
anandamide cellular uptake; and 3) the fatty acid amide hydrolase (FAAH), which
catalyses anandamide hydrolysis. Three showed modest potency in displacing
radiolabeled anandamide from the cannabinoid receptor (CBI), and one was a
modest inhibitor of the anandamide membrane transporter (AMT). The wewakpeptins
were tested for cytotoxicity to NCI-H460 human lung tumor and neuro-2a mouse
neuroblastoma cells. Intriguingly, wewakpeptin A and B were approximately 10-fold
more toxic than C and D to these cell lines.
126
Lyngbya majuscula has been recognized as a chemically and biologically rich
strain. Five new lyngbyabellin analogs, lyngbyabellins E-I, along with the known
compound dolabellin, originally isolated from sea hare Do/abel/a auricularia, were
identified from a 2002 Papua New Guinea collection of the marine cyanobacterium
Lyngbya majuscula. The lyngbyabellins were tested for cytotoxicity to NCI-H460
human lung tumor and neuro-2a mouse neuroblastoma cells and had
LC50
values
between 0.2 and 4.8 pM. Intriguingly, Iyngbyabellin E and H appeared to be more
active against the H460 cell line with LC values of 0.4 pM and 0.2 pM, respectively,
compared to LC values of 1.2 and 1.4 pM in the neuro-2a cell line. Lynbyabellin I
was the most toxic to neuro-2a cells (LC
0.7 pM), whereas lyngbyabellin G, was the
least cytotoxic of all compounds to either cell line. On the basis of this limited
screening, it appears that lung tumor cell toxicity is enhanced in the cyclic
representatives, and overall potency is increased in those containing an elaborated
side chain.
Additionally, two new cytotoxins, aurilides B and C, which are closely related
to aurilide, originally isolated from the sea hare Dolabella auricularia, have been
identified from the same extract where the lyngbyabellins E-1 were isolated. Aurilides
B and C were tested for cytotoxicity to NCI-H460 human lung tumor and neuro-2a
mouse neuroblastoma cells. Interestingly, aurilide B was approximately 4-fold more
toxic than C to these cell lines. The LC for Aurilide B was 0.01 pM and 0.04 pM for
neuro-2a and H460 cells, respectively, and 0.05 pM and 0.13 pM, respectively, for
aurilide C. Aurilide B (1) was evaluated in the NCI 60 cell line panel and found to
exhibit a high level of growth inhibition in leukemia, renal, and prostate cancer cell
lines with a
Gl50
less than 10 nM. Aurilide B (1) showed net tumor cell killing activity
in the NCI's hollow fiber assay, an in vivo model for assessing a chemical's
anticancer activity.
The principle biogenetic theme in the natural products chemistry of marine
cyanobacteria is the integration of nonribosomal polypeptide synthetase (NRPS) and
polyketide synthesase (PKS) pathways, in a variety of configurations, so as to
produce a great structural diversity. Study of the biosynthetic pathways employing
traditional isotope-labeled precursor incorporation studies will multiply as knowledge
of the culture requirements of these organisms increases. In turn, knowledge of
127
these unique structures and their biosynthetic pathways will intensify studies at the
molecular genetic level. At present, the mechanisms and pathways that are used to
regulate secondary metabolite expression in these creatures are wholly unknown,
and breakthroughs in this area could provide the proverbial "golden key" for unlocking
the full biosynthetic potential of these organisms.
Given the prevalent antitubulin/antiactin theme in the mechanism of action of
cyanobacterial metabolites, the discovery of new compounds of utility to the
chemotherapy of neoplastic diseases can be anticipated with considerable certainty.
However, similar to the natural products of other taxonomic groups, the biochemical
mechanisms of action of many marine cyanobacterial compounds are largely
unknown, and their continued study will provide a rich collection of tools for molecular
pharmacologists. It is likely that new biochemical targets for toxins and other
bioactive compounds will be revealed in the mechanism of action studies. Additional
insight into the mechanism of action of marine cyanobacterial toxins will also come
from continued studies of the chemical ecology of these organisms.
Debate also continues on the true source of these bioactive marine natural
products. The accumulated experimental and circumstantial evidence is in favor of
the hypothesis that numerous nonribosomal peptides from marine invertebrates are
actually produced by symbiotic microorganisms. For instance, both dolabellin and
aurilide were originally isolated from the sea hare
Dolabella
auricularia with isolated
yields of I O%. However, we have isolated both dolabellin and two aurilides from a
single collection of marine cyanobacteria Lyngbya majuscula from Papua New
Guinea. Moreover, the yield of aurilide B is over I O% of the wet weight of the
sample. Isolation and cultivation of the suspected microbial producers of bioactive
natural products either from the surrounding sea water or from the tissue of
invertebrates through careful design of special media could provide a much needed
answer to the pressing supply problem that is currently presented by many
pharmacologically interesting compounds are hampered from further development as
a drug due to limited supply problems. If these microorganisms are indeed the
producers of bioactive metabolites of interest, we believe the investigation of the
chemical ecology associated with these natural product-producing microbes is
128
essential, and represents the marine biologist's 'why' to go along with the chemist's
'what' and the bioprocess engineer's 'where' and 'how'.
Novel compounds are discovered and reported on a daily basis. The marine
environment is highly complex, however, and thus a highly interdisciplinary approach
is required in order to realize its potential. As one of the two remaining environments
(marine and extraterrestrial) which is yet to be fully explored, a greater understanding
could be achieved by a close collaboration between marine biologists,
microbiologists, geneticists, bioprocess engineers and natural product chemists.
129
BIBLIOGRAPHY
Aimi N, Odaka H, Sakai S, Fujiki H, Suganuma M, Moore RE and Patterson GM
(1990) Lyngbyatoxins B and C, two new irritants from Lyngbya majuscula. J Nat Prod
53:1593-1596.
Amador ML, Jimeno J, Paz-Ares L, Cortes-Funes H and Hidago M (2003) Ann Onco!
14:1607-1615.
Ang KKH, Holmes MJ, Higa T, Hamann MT and Kara UAK (2000) Antimicrobial
Agents Chemothera. 44:1645-1649.
Baden DG, Bikhazi G, Decker SJ, Foldes FF and Leung 1(1984) Neuromuscular
blocking action of two brevetoxins from the Florida red tide organism Ptychodiscus
brevis. Toxicon 22:75-84.
Bai R, Friedman SJ, Pettit GR and Hamel E (1992) Biochem. Pharmacol. 43:2637-45.
Bai R, Pettit GR and Hamel E (1990a) Biochem. Pharmaco!. 39:1941-9.
Bai R, Pettit GR and Hamel E (1990b) J. Biol. Chem. 265:17141-9.
Barrows LR, Radisky DC, Copp BR, Swaffar OS, Kramer RA, Warters RL and Ireland
CM (1993) Anti-Cancer Drug Design 8:333-347.
Basu A, Kozikowski AP and Lazo JS (1992) Structural requirements of lyngbyatoxin A
for activation and downregulation of protein kinase C. Biochemistry 31:3824-3830.
Bergman W and Feeney RJ (1951) J. Org. Chem. 16:981-987.
Berman FW, Gerwick WH and Murray IF (1999) Toxicon 37:1645-1648.
Blokhin AV, Yoo HO, Geralds RS, Nagle DG, Gerwick WH and Hamel E (1995) Mo!
Pharmacol 48:523-531.
Bodey GP, Freirich EJ, Monto RW and Hewlett JS (1969) Cancer Chemother. 53:5966.
Buchanan RAand Hess F (1980) Science 127:143-171.
Burja AM and Hill RT (2001) Hydrobio!ogica 461:41-47.
CardeHina JH, 2nd, Marner FJ and Moore RE (1979) Seaweed dermatitis: structure of
lyngbyatoxin A. Science 204:193-195.
Carter DC, Moore RE, Mynderse JS, Niemczura WP and Todd JS (1984) J. Org.
Chem. 49:236-241.
130
Chang Z, Flatt P, Gerwick WH, Nguyen VA, Willis CL and Sherman DH (2002) The
barbamide biosynthetic gene cluster: a novel marine cyanobacterial system of mixed
polyketide synthase (PKS)-non-ribosomal peptide synthetase (NRPS) origin involving
an unusual trichloroleucyl starter unit. Gene 296:235-247.
Chen J and Forsyth CJ (2004) Proc. Nat!. Acad. Sd. USA 101:12067-12072.
Clark WD and Crews P (1995) Tet. Lett. 36:1185-1188.
Crews P, Manes LV and Boehler M (1986) Tetrahedron Letters 27:2797-2800.
Cruz-Monserrate Z, Mullaney JT, Harran PG, Pettit GR, Mullaney JT, Harran PG and
Pettit GR (2003) Eur. J. Biochem. 270:3822-28.
De Clercq E (2000) Res. Rev. 20:323-349.
de Fiebre CM, Meyer EM, Henry JC, Muraskin SI, Kern WR and Papke RL (1995)
Mo!. Pharmacol. 47:164-171.
de Silva ED and Scheuer P (1980) Tet. Left. 21:1611-14.
Deems RA, Lombardo 0, Morgan DP, Mihelich ED and Dennis EA (1987) Biochim.
Biophys. Acta. 917:258-68.
Degnan BM, Hawkins CJ, Lavin MF, McCaffrey EJ, Parry DL, van den Brenk AL and
Watters DJ (1989) J. Med. Chem. 32:1349-1 354.
Dileo P CP, Bacci G, et al. (2002) Proceedings of the 2002 ASCO Annual Meeting.
Abstr #1628.
Dumdei EJ, Simpson JS, Garson MJ, Bryriel KA and Kennard CHL (1997) Australian
Journal of Chemistiy 50:139-144.
Edwards DJ and Gerwick WH (2004) JAm Chem Soc 126:11432-11433.
Edwards DJ, Marquez BL, Nogle LM, McPhail K, Goeger DE, Roberts MA and
Gerwick WH (2004) Structure and biosynthesis of the jarnaicamides, new mixed
polyketide-peptide neurotoxins from the marine cyanobacterium Lyngbya majuscula.
Chem Biol 11:817-833.
El Sayed KA, Kelly M, Kara UAK, Ang KKH, Katsuyarna I, Dunbar DC, Khan AA and
Hamann MT (2001) J. Am. Chem. Soc. 123:1804-1808.
Fu X, Do T, Schmitz FJ, Andrusevich V and Engel MH (1998) J. Nat. Prod. 61:15471551.
Fujiki H, Mon M, Nakayasu M, Terada M, Sugimura I and Moore RE (1981) proc.
Nat!. Acad. Sd. USA 78:3872-3876.
131
Gerwick WH, Proteau PJ, Nagle DG, Hamel E, Blokhin AV and Slate DL (1994) J.
org. Chem. 59:1243-1245.
Halstead BW (1965) Posinous and Venomous Marine Animals of the World. US
Government Printing Office, Washington DC. 1:1-155.
Hamel E (2002) Biopolymers 66:142-160.
Hamel E, Blokhin AV, Nagle DG, Yoo HD and Gerwick WH (1995) Drug Delivery
Research 34:110-120.
Harrigan GG, Goetz GH, Luesch H, Yang S and Likos J (2001) Journal of Natural
Products 64:1133-1138.
Harrigan GG, Luesch H, Yoshida WY, Moore RE, Nagle DG, Paul VJ, Mooberry SL,
Corbett TH and Valeriote FA (1998) Journal of Natural Products 61:1075-1077.
Haygood MG, Salomon CE and Faulkner DJ (2003) Ecteinascidin Family
Compounds: Compositions and Methods. International Patent EPI 360337.
Haygood MG, Schmidt EW, Davidson SK and Faulkner DJ (1999) J. Mol. Microbiol.
Biotechnol. 1:33-43.
Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J and Moore BS
(2002) Applied and Environmental Microbiology 68:4431-4440.
Hill RT, Hamann MI, Peraud 0 and Kasanah N (2003) Ref Type: Patent:41 15-4180
PCT.
In Y, Doi M, lnoue M, Ishida T, Hamada Y and Shioiri T (1994) Acta Ciystallogr. C.
50:432-434.
Ishibashi M, Iwasaki T, lmai S, Sakamoto S, Yamaguchi K, Ito A, Iwasaki T, Imai S,
Sakamoto S, Yamaguchi K and Ito A (2001) J. Nat. Prod. 64:108-110.
Jacobs RS, Culver P, Langdon R, O'Brien T and White S (1985) Tetrahedron 41:9814.
Jaspars M (1998) Advances in Drug Discovery Techniques;. Haniey, A. L., Ed.;
Wiley: New York,:65-84.
Jiménez JI and Scheuer PJ (2001) J. Nat. Prod. 64:200-203.
Kelecom A (1999) Chemistry of Marine Natural Products: Yesterday, Today and
Tomorrow. An Acad Bras Cienc 71:249-263.
Kem WR, Abbott BC and Coates RM (1971) Toxicon 9:15-22.
Kobayashi M and Kitagawa K (1999) J Nat Toxins 8:249-258.
132
Kozikowski AP, Shum PW, Basu A and Lazo JS (1991) Synthesis of structural
analogues of lyngbyatoxin A and their evaluation as activators of protein kinase C.
Journal of medicinal chemistiy 34:2420-2430.
Kraft AS (1993) J. Nat!. Cancer Inst. 85:1790-1792.
Kraft AS, Woodley S, Pettit GR, Gao F, Coil JC and Wagner F (1996) Cancer
Chemother. Pharmacol. 37:271-278.
Li WI, Berman FW, Okino T, Yokokawa F, Shioiri T, Gerwick WH and Murray TF
(2001) Proc. Nat!. Acad. Sci. USA 98:7599-7604.
Li WI, Marquez BL, Okino T, Yokokawa F, Shioiri T, Gerwick WH and Murray TF
(2004) Characterization of the preferred stereochemistry for the neuropharmacologic
actions of antillatoxin. J Nat Prod 67:559-568.
Lin Y-Y, Risk M, Ray SM, Van Engen 0, Clardy J, Golik J, James JC and Nakanishi K
(1981) J. Am. Chem. Soc. 103:6773-6775.
Look SA, Fenical W, Matsumoto GK and Clardy J (1986) J. Org. Chem. 51:51405145.
Luduena RF, Prasad V, Roach MC, Banerjee M, Yoo HD and Gerwick WH (1997)
Drug Delivery Research 40:223-229.
Luesch H, Yoshida WY, Moore RE, Paul VJ and Corbett TH (2001) Total structure
determination of apratoxin A, a potent novel cytotoxin from the marine
cyanobacterium Lyngbya majuscula. JAm Chem Soc 123:5418-5423.
Luesch H, Yoshida WY, Moore RE, Paul VJ and Mooberry SL (2000) Isolation,
structure determination, and biological activity of Lyngbyabellin A from the marine
cyanobacterium lyngbya majuscula. J Nat Prod 63:611-615.
Luibrand RT, Erdman TR, Voilmer JJ, Scheuer PJ, Finer J and Clardy J (1979)
Tetrahedron 35:609-612.
MacMillan JB, Trousdale EK and Molinski TF (2000) Org. Lett. 2:2721-2723.
Mahnir V, Lin B, Prokai-Tatrai K and Kem WR (1998) Biopharm. Drug Disposit.
19:147-151.
Marquez BL, Watts KS, Yokochi A, Roberts MA, Verdier-Pinard P, Jimenez JI, Hamel
E, Scheuer PJ and Gerwick WH (2002) Structure and absolute stereochemistry of
hectochlorin, a potent stimulator of actin assembly. J Nat Prod 65:866-871.
May WS, Sharkis SJ, Esa AH, Gebbia V, Kraft AS, Pettit GR and Sensenbrenner LL
(1987) Proc. Nat!. Acad. Sd. USA 84:8483-8487.
Mayer AMS and Hamann MT (2004) Marine Biotechnology 6:37-52.
133
McDonald LA and Ireland CM (1992) J. Nat. Prod. 55:376-379.
Milligan KE, Marquez BL, Williamson RT and Gerwick WH (2000) Lyngbyabellin B, a
toxic and antifungal secondary metabolite from the marine cyanobacterium Lyngbya
majuscula. J Nat Prod 63:1440-1443.
Mooberry SL, Leal RM, Tinley TL, Luesch H, Moore RE and Corbett TH (2003)
International Journal of Cancer 104:512-521.
Moore RE and Scheuer PJ (1971) Science 172:495-498.
Murakami Y, Oshima Y and Yasumoto T (1982) Nippon Suisan Gakkaishi 48:69-72.
Murata M, Legrand A-M, Ishibashi Y and Yasumoto 1 (1989) J. Am. Chem. Soc.
111:8927-8931.
Murata M, Nakoi H, Iwashita 1, Matsunaga S, Sasaki M, Yokoyama A and Yasumoto
T (1993) J. Am. Chem. Soc. 115:2060-2062.
Newman DJ and Cragg GM (2004) J Nat Prod 67:1216-1 238.
Norimura T, Sasaki M, Matsumori N, Miata M, Tachibana K and Yasumoto T (1996)
Angew. Chem. mt. Ed. EngI. 35:1675-1678.
Nuijen B, Bouma M and Manada C (2000) Anti-Cancer Drugs 11:793-811.
Orjala J and Gerwick WH (1996) Barbamide, a chlorinated metabolite with
molluscicidal activity from the Caribbean cyanobacterium Lyngbya majuscula. J Nat
Prod 59:427-430.
Orjala J, Nagle DG, Hsu V and Gerwick WH (1995) Antillatoxin: An Exceptionally
lchthyotoxic Cyclic Lipopeptide from the Tropical Cyanobacterium Lyngbya
ma] uscula. Journal of the American Chemical Society 117:8281-8282.
Osinga R, Armstrong E, Burgess JG, Hoffmann F, Reitner J and Schumann-Kindel G
(2001) Hydrobiologia 461:55-62.
Perez E, Beatriz A, Perez T, Velasco I, Henriquez P, Munoz M, Moss C and
McKenzie D (2004) Sequences from an Endosymbiont and their Uses. International
Patent W02004015143.
Pettit GR, Herald CL, Doubek DL, Herald DL, Arnold E and Clardy J (1982) J. Am.
Chem. Soc. 104:6846-6848.
Pettit GR, Kamano Y, Herald CL, Tuinman AA, Boettner FE, Kizu H, Schmidt JM,
Baczynkyji L, Tomer KB and Bontems RJ (1987) J. Am. Chem. Soc. 109:7581-82.
Poll MA (2002) Recent Advances in Marine Biotechnology 7:1-30.
134
Reddy MV, Rao MR, Rhodes D, Hansen MS, Rubins K, Bushman ED, Venkateswarlu
Y and Faulkner DJ (1999) J. Med. Chem. 42:1901-1907.
Rinehart K, GloerJ, Cook J, CarterJ, Mizsak Sand Scahill T (1987) JAm Chem Soc
103: 1857-1 859.
Rinehart KL, Holt TG, Fregeau NL, Stroh JG, Kiefer PA, Sun F, Li LH and Martin DG
(1990)J. Org. Chem. 55:4512-4515.
Roussis V, Wu Z, Fenical W, Strobel SA, Van Duyne GD and Clardy J (1990) J. Org.
Chem. 55:4916-4922.
Salomon CE and Faulkner DJ (2002) J. Nat.
Prod.
65:689-692.
Salomon CE, Magarvey NA and Sherman DH (2004) Nat Prod Rep 21:105-121.
Schaufelberger DE and Pettit GR (1991) J. Nat.
Prod.
54:1265-1270.
Schmidt EW, Sudek S and Haygood MG (2004) J. Nat.
Prod.
67:1341-1 345.
Simmons LT, Andrianasolo E, McPhail K, Flatt P and Gerwick WH (2005) Mo!.
Cancer Ther. 4:333-342.
Sitachitta N, Rossi J, Roberts M, Gerwick WH, Fletcher MD and Willis CL (1998)
Biosynthesis of the marine cyanobacterial metabolite barbamide. 1. Origin of the
trichloromethyl group. Journal of the American Chemical Society 120:7131-7132.
Smith CD, Zhang X, Mooberry SL, Patterson GM and Moore RE (1994) Cancer
Research 54:3779-3784.
Sone H, Kondo T, Kiryu M, Ishiwata H, Ojika M and Yamada K (1995) Dolabellin, a
Cytotoxic Bisthiazole Metabolite from the Sea Hare Dolabella auricularia: Structural
Determination and Synthesis. Journal of Organic Chemistry 60:4774-4781.
Thacker RW and Starnes S (2003) Marine Biology 142:643-648.
Trimurtulu G, Ohtani I, Patterson GM, Moore RE, Corbett TH, Valeriote FA and
Demchik L (1994) JAm Chem Soc 116:4729-4737.
Vacelet J and Donadey C (1977) J. Exp. Mar. Eco!. 30:301-314.
Vaishampayan H, Glode M, Du W, Kraft A, Hudes G, Wright J and Hussain M (2000)
Clin. Canc. Res. 6:4205.
van Kesteren C, de Vooght M, Mathot R, Schellens J, Jimeno JM and Beijnen JH
(2003) Anti-Cancer Drugs 14:487-502.
Verdier-Pinard P, Lai JY, Yoo HD, Yu J, Marquez B, Nagle DG, Nambu M, White JD,
Falck JR, Gerwick WH, Day BW and Hamel E (1998) Mo! Pharmacol 53:62-76.
135
White JD, Hanselmann Rand Wardrop DJ (1999) JAm Chem Soc 121:1106-1107.
Whitton BA and Potts M (2000) Introduction to cyanobacteria. In The ecology of
cyanobacteria. Their diversity in time and space., B.A. Whitton and M.Potts, eds.
(Dordrecht: KluwerAcademic Publisher),:1-1 I.
Williams DE, Burgoyne DL, Rettig SJ, Andersen RJ, Fathi-Afshar ZR and Allen TM
(1993) J. Nat. Prod. 56:545-551.
Wipf P, Reeves JT and Day BW (2004) Curr. Pharm. Des. 10:1417-1437.
Wright AD and Konig GM (1996) J. Nat.
Prod.
59:710-716.
Wright AE, Forleo DA, Gunawardana GP, Gunasekera SP, Koehn FE and McConnell
OJ (1990) J. Org. Chem. 55:4508-4511.
Wu M, Okino T, Nogle LM, Marquez BL, Williamson RT, Sitachitta N, Berman FW,
Murray IF, McGough K, Jocobs R, Colsen K, Asano T, Yokokawa F, Shioiri T and
Gerwick WH (2000) Structure, Synthesis, and Biological Properties of Kalkitoxin, a
Novel Neurotoxin from the Marine Cyanobacterium Lyngbya ma] uscula. Journal of the
American Chemical Society 122:12041-12042.
Wylie BL, Ernst NB, Grace KJS and Jacobs RS (1997)
Prog.
Surg. 24:146-1 52.
Yasumoto I, Oshima Y, Sugawara W, Fukuyo Y, Oguri H, Igarashi T and Fujita N
(1980) Nippon Suisan Gakkaishi 46:1405-1411.
Yokokawa F, Fujiwara H and Shioiri I (2000) Total synthesis and revision of absolute
stereochemistry of antillatoxin, an ichthyotoxic cyclic lipopeptide from marine
cyanobacterium Lyngbya majuscula. Tetrahedron 56:1759-1775.
Yokokawa F and Shioiri I (1998) J. Org. Chem. 63:8638-8639.
Zabriskie TM, Kiocke JA, Ireland CM, Marcus AH, Molinski TF, Faulkner DJ, Xu CF
and Clardy J (1986) Jaspamide, a modified peptide from a Jaspis sponge, with
insecticidal and antifungal activity. Journal of the American Chemical Society
108:3123-3124.
Zewail-Foote M and Hurley L (1999) J Med Chem 42:2493-2497.
136
APPEN DICES
137
APPENDIX A
ISOLATION AND STRUCTURE ELUCIDATION OF A NEW CYCLIC
DEPSIPEPTIDE, WEWAKMIDE A, FROM A PAPUA NEW GUINEA COLLECTION
OF THE MARINE CYANOBACTERIUM LYNGBYA SEMIPLENA
In our continuing effort to discover new biologically active and structurally
intriguing secondary metabolites from a Papua New Guinea collection of the marine
cyanobacterium Lyngbya semiplena, a new cyclic depsipeptide, wewakpeptide A (1),
was isolated from the same crude extract as the wewakpeptins had previously been
identified
wewakamide A (1)
lie
aba
R,3R)-3-amino-2-m
thylbutanoic acid
Hiv
Guineamide B (2)
Hiv
Dolastatin D (3)
(maba)
138
While the IR spectrum displayed absorption bands at 1736 and 1657 cm1 for both
ester and amide functionalities, the HRFABMS of compound I analyzed for
C53H86N8010,
indicating a molecule of peptide origin. 1H and 13C NMR spectra of I
were well dispersed in COd3 (Figure A. 1, Table A. 1), allowing for construction of
nine partial structures by 20 NMR and accounting for all atoms in the molecular
formula of I (Figures A. 2
4). Seven standard amino acids were deduced as
Proline (Pro), Valine (Val), Leucine (Leu), Phenylalanine (Phe), N-methyl-alanine (N-
MeAla), N-methyl-isoleucine (N-Melle), and N-methyl-leucine (N-MeLeu). Two
additional residues were assembled to form the a-hydroxy acid, 2-hydroxyisovaleric
acid (Hiva) and the 13-amino acid, 3-amino-2-methylbutanoic acid (Maba). Amino acid
sequencing was accomplished by long range 13C-1H correlation experiments (HMBC)
with different mixing times and a ROESY experiment. Further evidence supporting
this sequence was developed from (MS) experiments (Figure A. 5).
Figure A. I. 1H and 13C NMR spectra of wewakamide A (I).
139
IiI
'II,
e
PS
d
,01
U'
*+'
a
'I
ri
I
Si
I
U
S
U
I
S
$4
SI
a
I
U
S
a
I,..'.
Figure A. 2. COSY spectrum of wewakamide A (1).
140
Is's
IuI
5,
t
0
I
I
0
p
100
120
ppm
pp.
6
6
Figure A. 3. HSQC spectrum of wewakamide A (1).
141
d
UI
U
0
25
.
:
U.
I
S
50
mc
*0'
J
ii*
75
I
100
US
1
U
'4
125
I
150
Figure A. 4. HMBC spectrum of wewakamide A (4).
142
m/z 586 (+H)
ii
m/z 439 (+H)
m/z 340 (+1-1)
m/z 312 (+H)
m/z 227 (+H)
iva ProMe Ala Me LeuMe lie Vat Mab aLeu Phe
wewakamide A (1)
m/z 882 (+H)
m/z 783 (1-1)
m/z 684(+H)
NHPile
Hiva
Pro Me Ala MeLeuMe lie Vat Mail aLeu
-
0
wewakamide A (1)
Figure A. 5. Key fragments from collisionally induced ESI-MS/MS experiments with
wewakamide A (1).
143
Table A. 1. NMR Spectral Data for Wewakamide A (1) in
Unit
Position
Hiv
1
Pro
2
3
4
5
6
7
8
9
10
N-Me-Ala
N-Me-Leu
11
12
13
14
15
16
17
18
19
20
21
N-Me-lie
Val
22
23
24
25
26
27
28
NH
29
30
31
Maba
Leu
32
33
NH
34
35
36
37
38
NH
39
40
41
42
43
(J in Hz)
5.07, d (9.2)
2.37
0.98
1.08
CDCI3.
HMBCa
H
75.8
30.6
18.3
19.3
172.6
47.1
23.1
31.6
59.8
2, 3, 4, 5, 53
1,5
1,2,4
1,2,3
3.36, 3.57, m
1.79, 1.59
1.81, 1.61
2.86
7, 9
3.02
10, 12
10, 11, 13, 14, 15
9
6,9
7, 8, 10, 11
169.1
30.3
50.4
15.0
171.3
30.8
55.3
38.3
25.5
24.3
24.4
176.4
33.2
61.3
36.7
16.4
28.1
13.2
5.66,q(7.0)
1.32,d(6.6)
12,14
2.32
5.53
1.76, 1.69
1.48
0.98, d (6.6)
1.09, d (6.6)
12, 14, 16
14, 15, 17,21
16, 18, 19,20
3.08
5.51, d (6.0)
2.29
1.04
1.56, 1.36
0.97
21,23
21, 22, 24, 25,26,28
5.85, d (7.0)
3.66, dd (8.5, 7.5)
2.01, m
1.20, d (6.8)
28, 29, 30
28, 30, 31, 32, 33
29, 31, 32,
29, 30, 32
29, 30, 31
8.30, d (9.8)
4.04, ddq (9.8, 2.8, 7.0)
0.90, d (7.0)
2.32, dq (2.8, 7.0)
1.07, d (7.0)
29, 33, 34
35, 36, 37
34, 36
33, 34, 35, 37, 38
34, 36, 38
7.12 d (7.8)
4.73, dd (6.7, 6.1)
1.76, 1.32
1.78
1.05
38, 39
36, 40, 41
39, 41, 42, 43, 44
39
17, 18, 20
17, 18, 19,
25, 26, 28
23, 24,
24,27
24
172.9
61.8
29.8
20.0
20.2
170.4
47.9
14.2
44.7
13.8
174.3
50.5
38.7
25.8
21.5
21.3
1.10
1.01
40, 41,43
40, 41,42
144
44
Phe
45
46
47
48
49
50
a
169.30
NH
54.4
40.9
137.4
129.1
51
129.6
127.6
129.6
52
53
129.1
170.1
6.45, d (7.1)
4.66, m
3.18, 2.89, dd (6.0, 3.2)
44, 45, 53
46, 53
45, 47, 48/52, 53
7.33
7.35
7.28
7.35
7.33
46,49
47, 48
49,51
47, 52
46, 50, 51
Proton showing HMBC correlation to indicated carbon.
The absolute configuration of I was established by analysis of degradation
products. A small sample was hydrolyzed with 6N HCI to its constituent amino and
hydroxy acid units. These were analyzed by chiral HPLC as well as chiral GO-MS
and compared with the retention times of authentic standards. All of the usual amino
acids as well as the Hiva unit were shown to possess L-configuration, while the Maba
unit was not determined due to unavailability of the standards. A search of the
chemical literature indicated that the Maba unit has only been identified in two natural
products, guineamide B (2) 2 and dolastatin D (3)
Diagnostic NOEs from H-34 to H-
36 were indicative of the relative stereochemistry as 34S*, 36S*. Therefore, we
propose that the Maba unit is of the same enantiomeric series as that of dolastatin D
because of the comparable spectroscopic properties between I and 3.
Wewakamide A (1) was screened for biological activity using brine shrimp
toxicity
sodium channel modulation
and both the NCI-H460 human lung tumor
and the neuro-2a mouse neuroblastoma cell lines. This compound, however, proved
inactive in all of these assays, except for brine shrimp toxicity, with LC50 = 5 ppm.
Interestingly, four other related metabolites, wewakpeptins A - D from the same
extract of Lyngbya semiplena display potent cytotoxicity against both the NCI-H460
human lung tumor and the neuro-2a mouse neuroblastoma cell lines.
145
EXPERIMENTAL
General Experimental Procedures.
Optical rotations were measured on a
Perkin-Elmer 141 polarimeter. IR and UV spectra were recorded on Nicolet 510 and
Beckman DU64OB spectrophotometers, respectively. NMR spectra were recorded on
Bruker Avance DPX 400 MHz and Bruker Avance 300 MHz spectrometers with the
solvent CDCI3 used as an internal standard (H at 7.26, ö at 77.4). High resolution
mass spectra were recorded on a Kratos MS-50 TC mass spectrometer. Tandem
mass spectrometric data were obtained on an electrospray ionization (ESI)
quadrupole ion trap mass spectrometer (Finnigan LCQ, San Jose, CA). For ESIMS/MS analysis, samples were injected onto a Cl 8 trap column for desalting and
introduced into the mass spectrometer by isocratic elution using 50% acetonitrile
containing 0.1 % formic acid. For MS/MS investigations, the protonated molecular ion
clusters were isolated in the ion trap and collisionally activated with different collision
energies to find optimal fragmentation conditions. For the most intense fragment ions
MS3 and MS4 experiments were performed. Chiral GC-MS analysis was
accomplished on a Hewlett-Packard gas chromatograph 5890 Series II with a
Hewlett-Packard 5971 mass selective detector using an Alitech capillary column
(CHIRASIL-VAL phase 25 m x 0.25 mm). HPLC was performed using Waters 515
HPLC pumps and a Waters 996 photodiode array detector.
Collection. The marine cyanobacterium Lyngbya semiplena (voucher
specimen available from WHG as collection number PNGI2-7Dec99-3) was collected
from shallow waters (1-3 m) in Wewak Bay, Papua New Guinea, on December 7,
1999. Taxonomy was assigned by microscopic comparison with the description given
by Desikachary.6 The material was stored in 2-propanol at -20°C until extraction.
Extraction and Isolation. Approximately 138 g (dry wt) of the alga were
extracted repeatedly with CH2Cl2/MeOH (2:1)to produce 3.05 g of crude organic
extract. The extract (3.0 g) was fractionated by silica gel vacuum liquid
chromatography using a stepwise gradient solvent system of increasing polarity
starting from 10% EtOAc in hexanes to 100% MeOH. The fraction eluting with 100%
MeOH was found to be active at I ppm in the brine shrimp toxicity assay. This
fraction was further chromatographed on Mega Bond RP18 solid-phase extraction
146
(SPE) cartridges using a stepwise gradient solvent system of decreasing polarity
starting from 80% MeOH in H20 to 100% MeOH. The most active fractions after SPE
(85% toxicity at I ppm to brine shrimp) were then purified by HPLC [Phenomenex
Sphereclone 5 p ODS (250 x 10 mm), 9:1 MeOH/H20, detection at 211 nmj giving
compounds 1 (2.5 mg),
Wewakamide A (1): glassy oil; [a]
-83° (c 0.03, CHCI3); UV (MeOH)
Amax
215 nm (loge 4.6); IR (neat) 3317, 2966, 2922, 2854, 1736, 1657, 1546, 1461, 1238
cm1; 1H and 13C NMR data, see Table Al; HRFABMS m/z [M + H] 995.6544
(calculated for C53H87N8010, 995.6545).
Absolute stereochemistry of 1. Wewakamide A (1, 500 pg) was
hydrolyzed in 6 N HCI at 105 °C for 16 h, then dried under a stream of N2 and further
dried under vacuum. The residue was reconstituted with 300 pL of H20 prior to chiral
HPLC analysis [Phenomenex Chirex 3126 (D), 4.6 x 250 mm; UV 254 nm detector],
mobile phase I: 100% 2 mM CuSO4 in
mM CuSO4 in MeCN/H20 (15:85),
flow
H20, flow
rate 0.7 mL/min; mobile phase II: 2
rate 0.8 mL/min; mobile phase III: 2 mM
CuSO4 in MeCN/H20 (5:95), flow rate 1 mL/min; [column, Phenomenex chirex 3126
(D), 4.6 x 50 mm; UV 254 nm detector], mobile phase IV: 2 mM CuSO4 in MeCN/H20
(15:85), flow rate 0.8 mLlmin. Mobile phase I elution times (tR, mm) of authentic
standards: L-MeAla (16.0), D-MeAla (16.5), L-Pro (28.4), D-Pro (63.0), L-Val (38.4),
D-Val (68.5). Mobile phase II elution times (tR, mm) of authentic standards: L-MeLeu
(12.5), D-MeLeu (14.1), L-Leu (16.0), D-Leu (17.1), L-Phe (39.8), D-Phe (41.3).
Mobile phase Ill elution times (tR, mm) of authentic standards: L-Allo-Melle (17.7), L-
Melle (18.8), D-Allo-Melle (27.7), D-Melle (28.2). Mobile phase IV elution times (tR,
mm) of authentic standards: L-Hiv (9.2), D-Hiv (14.5). The hydrolysate was
chromatographed alone and co-injected with standards to confirm assignments (LMeAla, L-Pro, L-Val, L-MeLeu, L-Leu, L-Phe, L-Melle, L-Hiv).
Biological Activity. Brine shrimp (Artemia sauna) toxicity was measured as
previously described.4 After a 24 h hatching period, aliquots of a 10 mg/mL stock
solution of compound A was added to test wells containing 5 mL of artificial seawater
and brine shrimp to achieve a range of final concentrations from 0.1 to 100 ppm.
After 24 h the live and dead shrimp were tallied. Modulation of the voltage-sensative
147
sodium channel in mouse neuro-2a neuroblastoma cells was also examined as
previously described.
Cytotoxicity was measured in NCI-H460 human lung tumor cells and neuro-2a
mouse neurablastoma cells using the method of Alley et. al 7with cell viability being
determined by MIT reduction.5 Cells were seeded in 96-well plates at 6000 cells/well
in 180 p1 medium. Twenty-four hours later, the test chemical was dissolved in DMSO
and diluted into medium without fetal bovine serum and then added at 20 pg/well
DMSO was less than 0.5% of the final concentration. After 48 hr, the medium was
removed and cell viability determined.
REFERENCES
1.
Han, B.; Goeger, D.; Maier, C. S.; Gerwick, W. H. J. Org. Chem. 2005, 70,
3 133-3 139.
2.
Tan, L. T.; Sitachitta, N.; Gerwick, W. H. J. Nat. Prod. 2003, 66, 764-771.
3.
Sone, H.; Nemoto, T.; lshiwata, H.; Ojika, M.; Yamada, K. Tet. Left. 1993,
34, 8449-52.
4.
Meyer, B. N.; Ferrigni, N. R.; Putnam, L. B.; Jacobsen, L. B.; Nichols, D. E.;
McLaughlin, J. L. Planta Med. 1982, 45, 31-3.
5.
Manger, R. L.; Leja, L. S.; Lee, S. Y.; Hungerford, J. M.; Hokama, Y.; Dickey,
R. W.; Granade, H. R.; Lewis, R.; Yasumoto, T.; Wekell, M. M. J. AOAC. mt.
1995, 78, 521-527.
6.
Marfey, P. Carlsberg Res. Commun. 1984, 49, 591-596.
7.
Alley, M. C., Scudiero, D. A. Cancer Res. 1988, 48, 589-601.
148
APPENDIX B
ISOLATION AND STRUCTURE ELUCIDATION OF GUINEAMIDE G FROM A
PAPUA NEW GUINEA COLLECTION OF THE MARINE CYANOBACTERIUM
LYNGBYA MAJUSCULA
Cyanobacteria are phenomenal producers of structurally intriguing and
biologically active secondary metabolites such as curacins1 and
cryptophycins.2 In
our ongoing program to explore these organisms as sources for novel anticancer
agents, our most recent investigations have led to the findings of dolabellin,3
',
lyngbyabellins E-1 and aurilides B and C from a collection of the pantropical
marine cyanobacterium Lyngbya majuscula Gomont (Oscillatoriaceae). Herein, this
paper describes the isolation, structure determination, and a discussion of biological
activities of guineamide G (1).
An [M + H] peak observed in HR FABMS for guineamide G (1) suggested a
molecular formula of C42H55N507, accounting for 18 degrees of unsaturation. The
peptidic nature of this molecule was again established from tell-tale exchangeable NH
protons resonating at 6 8.84 and 6 5.86 in the 1H NMR data. Similarly to guineamide
F
(2),6
four singlet CH3 proton signals (6 3.02, 6 3.44, 6 1.23, and 6 1.20) were
present in the 1H NMR data (Table B. 1). The two low field singlet proton signals at 6
3.02 and 6 3.44 suggested N-methylations. Six distinct low field carbon signals, due
to ester/amide carbonyls, were observed in the 13C NMR data of guineamide G (1)
(Figure B. 1). At least 12 olefinic carbon signals, with four peaks having two
overlapping carbon signals, were present in the 128-132 ppm range. Taken together
with low field aromatic proton signals (6.8-7.4 ppm) in the 1H NMR data, two monosubstituted phenyl groups in I were suggested. Additionally, the 13C NMR spectrum
showed two distinctive carbon signals at 6 83.6 and 6 69.3,
149
N-MePhe-1
Dhoya
19
V0
15o10
Gly
Pro
26
i=O
0
N
23
0
0
N
HN
42
38
13N
N-MePhe-2'34
Guineamide F (2)
Guineamides G (1)
8.5
170
8.0
160
7.5
150
7.0
140
6.5
130
6.0
5.5
120
5.0
110
4.5
100
4.0
90
3.5
3.0
2.5
80
70
60
Figure B. 1. 1H and 13C NMR spectra of Guineamide G (1).
2.0
50
1.5
40
1.0
30
0.5
20
ppm
ppm
150
ppm
0-
'
.
i
*
10
..
:
:.
:
5P
.0
I
I.
6
7
B
9
0
1
0
3
1
Figure B. 2. COSY spectrum of Guineamide G (1).
1
U
ppm
151
ppm
10
20
rii,
0.
30
S
40
S
Co
50
0
60
S
70
S
80
90
100
110
120
130
140
150
9
8
7
6
5
4
3
Figure B. 3. HSQC spectrum of Guineamide G (1).
2
1
0
ppm
152
ppm
20
'
I
I
I
I
P
I
0
40
60
Is,
I
La
I
LaO
80
b
*
100
120
4
*
1
140
160
...
,..
9
8
7
6
5
4
3
Figure B. 4. HMBC spectrum of Guineamide G (1).
2
1
0
ppm
153
Table B. 1. NMR data for Guineamide G (1) at 400 MHz (1H) and 100 MHz (13C) in
CDCI3.
multf(Hz)
Oc
1
175.8
2
3
brt
47.1
77.1
Position
4a
4b
5a
5b
6
7
8
9
10
HMBC
2,2-dimethyl-3-hydroxy-7-octynoic acid (Dhoya)
oH
5.27
1.85
1.69
1.55
1.55
2.27
1.99
1.26
1.24
9.0
29.3
29.3
25.0
25.0
18.4
83.7
69.5
18.3
26.1
m
m
m
m
m
s
s
1,5,9, 10, 11
3,5
3,6
5,8
1,2
1,2
Gly
11
12a
12b
NH
3.24
4.72
8.85
m
dd
d
16.9, 9.6
9.4
170.8
41.3
41.3
11,13
13
N-MePhe-1
13
14
15a
15b
3.89
2.90
3.70
16
17/21
7.10
7.23
19
7.23
22(N-CH3)
18/20
169.0
10.2, 3.1
64.2
13.7, 10.2 34.3
14.5, 3.1
34.3
138.5
129.7
m
m
127.2
127.4
m
31.5
3.03s
dd
dd
dd
13, 15, 16, 22, 23
13, 14, 16, 17
15,19
14,23
Pro
23
24
25a
25b
26a
26b
27a
27b
3.41
m
0.74 m
-0.03 m
1.27 m
1.27 m
3.20 m
3.39 m
171.3
57.8
30.2
30.2
22.2
22.2
46.5
46.5
25, 26, 27
23, 26, 27
24,27
25, 26
N-MePhe-2
28
29
30a
30b
5.14
2.78
3.12
dd
dd
dd
7.00
7.23
m
m
31
32/36
33/35
168.9
10.2, 4.9 54.2
12.6, 4.9 37.9
12.6, 10.3
137.2
129.3
129.2
28, 30, 31, 37, 38
28, 29, 31, 32
30, 33, 34
31
154
7.19
37(N-CH3)
34
38
39
40
41
42
NH
4.54
1.93
0.95
0.95
5.85
m
3.44s
Val
brt
7.4
m
m
m
d
7.6
127.5
31.7
173.2
55.2
31.0
18.5
19.3
29,38
1, 38, 40, 41,
38, 39, 41
39, 40, 42
40, 41
1
consistent with a terminal acetylenic functionality. As previously observed, the carbon
at
69.3 exhibited weak HSQC correlations but showed a
a methine proton at ö 1.93 in the HMBC
JCH
spectrum.3'4
1JCK coupling
of 249 Hz to
This proton also exhibited a
HMBC correlation to the quaternary carbon at ö 83.6, confirming the presence of
an acetylene. An interesting feature of the 1H NMR spectrum of guinamide G was the
presence of a shielded diastereotopic proton resonance at ö 0.03.
Careful analyses of COSY, HSQC, and HMBC data of guineamide G (1)
revealed that it was consisted of five usual amino acid residues as guineamide B
(Figures B. 2-4), which are two units of N-MePhe, one Pro, one Val, one Gly, and
the 2,2-dimethyl-3-hydroxy-hexanoic acid (Dmhha) residue in 4 was replaced by 2,2-
dimethyl-3-hydroxy-7-octynoic acid (Dhoya) in 1. The placement of the shielded
proton signal at
0.03 was assigned as one of the methylene protons on 13-C of the
Pro unit. Such a shielding effect could arise from the influence of nearby aromatic
amino acids, such as N-MePhe-1 and N-MePhe-2 in the molecule. The final planar
structure of guineamide G (1) was determined from key proton correlations observed
in the HMBC data (Table A 2), which provided the same sequence as presented in
structure of 2. Although no HMBC correlations were detected for the N-MePhe-2-Pro
sequence, it is the remaining linkage needed to account for the degrees of
unsaturation calculated from the molecular formula.
Hydrolysis and stereoanalysis of metabolite guineamide G (1) was not
undertaken due to the limited time and the desire to preserve as much sample as
possible for biological testing.
Initially reported, some of the guineamides possess moderate cytotoxicity to a
mouse neuroblastoma cell line with lC values of 15 and 16 pM, respectively.6
155
However, guineamide G has shown stronger cytotoxicity to a mouse neuroblastoma
cell line with LC values of 2 pglmL.
REFERENCES
1.
Gerwick, W. H.; Proteau, P. J.; Nagle, D. G.; Hamel, E.; Blokhin, A.; Slate,
D. L. J. Org. Chem. 1994, 59, 1243-5.
2.
Trimurtulu, G.; Ohtani, I.; Patterson, G. M. L.; Moore, R. E.; Corbett, T. H.;
Valeriote, F. A.; Demchik, L. J. Am. Chem. Soc. 1994, 116, 4729-37.
3.
Sone, H.; Kondo, T.; Kiryu, M.; lshiwata, H.; Ojika, M.; Yamada, K. J. Org.
Chem. 1995, 60,4774-81.
4.
Han, B.; MacPhail, K.; Gross, H.; Goeger, D.; Moobery, S. L.; Gerwick, W. H.
(manuscript in preparation)
5.
Han, B.; Gross, H.; Goeger, D.; Gerwick, W. H. (manuscript in preparation)
6.
Tan, L. T.; Sitachitta, N.; Gerwick, W. H. J. Nat. Prod. 2003, 66, 764-71.