Phylogeny-guided isolation of ethyl tumonoate a from the marine cyanobacterium cf. Oscillatoria margaritifera

Article abstract of DOI:10.1021/np200236c

The evolutionary relationships of cyanobacteria, as inferred by their SSU (16S) rRNA genes, were used as predictors of their potential to produce varied secondary metabolites. The evolutionary relatedness in geographically distant cyanobacterial specimens was then used as a guide for the detection and isolation of new variations of predicted molecules. This phylogeny-guided isolation approach for new secondary metabolites was tested in its capacity to direct the search for specific classes of new natural products from Curacao marine cyanobacteria. As a result, we discovered ethyl tumonoate A (1), a new tumonoic acid derivative with anti-inflammatory activity and inhibitory activity of calcium oscillations in neocortical neurons.

Full text of DOI:10.1021/np200236c

ARTICLE
pubs.acs.org/jnp
Phylogeny-Guided Isolation of Ethyl Tumonoate A from the Marine
Cyanobacterium cf. Oscillatoria margaritifera
†
†
†,‡
†
†
§
Niclas Engene, Hyukjae Choi, Eduardo Esquenazi, Tara Byrum, Francisco A. Villa, Zhengyu Cao,
§
†,^,||
†
,†,||
Thomas F. Murray, Pieter C. Dorrestein,
Lena Gerwick, and William H. Gerwick*
†
Center for Marine Biotechnology and Biomedicine, Scripps Institution of Oceanography, University of California, San Diego,
La Jolla, California 92093, United States
‡
Department of Biology, University of California, San Diego, La Jolla, California 92093, United States
§
Department of Pharmacology, School of Medicine, Creighton University, Omaha, Nebraska 68178, United States
^
Department of Chemistry and Biochemistry, University of California, San Diego, La Jolla, California 92093, United States
Skaggs School of Pharmacy and Pharmaceutical Sciences, University of California, San Diego, La Jolla, California 92093, United States
S Supporting Information
b
ABSTRACT: The evolutionary relationships of cyanobacteria,
as inferred by their SSU (16S) rRNA genes, were used as pre-
dictors of their potential to produce varied secondary metabo-
lites. The evolutionary relatedness in geographically distant
cyanobacterial specimens was then used as a guide for the detec-
tion and isolation of new variations of predicted molecules. This
phylogeny-guided isolation approach for new secondary meta-
bolites was tested in its capacity to direct the search for specific
classes of new natural products from Cura c- ao marine cyano-
bacteria. As a result, we discovered ethyl tumonoate A (1), a
new tumonoic acid derivative with anti-inflammatory activity and inhibitory activity of calcium oscillations in neocortical neurons.
12
yanobacteria form a monophyletic bacterial phylum extra-
strains from the environment; and (iv) distinguish cyanobac-
terial chemotypes.
1,2
14
C
ordinarily rich in bioactive secondary metabolites.
A
number of these metabolites are potent toxins associated with
harmful algal blooms. Ironically, many of these same bioactive
According to basic evolutionary principles, metabolic and
biosynthetic pathways are evolving between geographically
isolated or distant populations as an adaptation to new envi-
ronments. This concept of biogeographical diversification corres-
ponds to a major underlying rationale and approach in the search
3
molecules have simultaneously been discovered to have a variety
of potential pharmaceutical applications. In order to support
and enhance the search for novel natural products, it is valuable
to have proper understanding of the taxonomy of the secondary
2
13,14
for new natural products.
While in freshwater cyanobacteria
4
metabolite-producing cyanobacteria. The traditional system
secondary metabolites such as microcystins and cylindrosper-
mopsins have been reported from phylogenetically unrelated
genera, the production of secondary metabolites in marine
cyanobacteria as well as the distribution of their biosynthetic
pathways has been linked with the phylogenetic positions of the
of classifying cyanobacteria, which is based on phenotypical
observations, has recently shown major incongruities with phy-
logenetic classifications where evolutionarily informative house-
keeping genes are analyzed. This is partly a result of significant
morphological plasticity among cyanobacteria, even at the genus
level. Furthermore, cyanobacterial classification systems are
largely founded on a relatively limited number of morphological
5,6
15,3
secondary metabolite production. However, to date, the two
concepts of evolutionary relatedness and geographic diversifica-
tion have not been effectively combined for predictive explora-
tion of new natural products from marine cyanobacteria.
7
characters, which often appear similar due to convergent evolu-
8
tion. Thus, the recent inclusion of phylogenetic analysis in
The major focus of this study was to use evolutionary relation-
ships of marine cyanobacteria inferred by their SSU (16S) rRNA
genes as predictors of their potential to produce various second-
ary metabolites. The evolutionary relatedness of geographically
distant populations was then used as a guide for the detection and
isolation of new variations of predicted molecules. Furthermore,
classification of these microorganisms has led to a more accurate
taxonomic understanding of these cyanobacterial groups.
These developments have benefitted not only the field of
taxonomy but also natural products drug discovery efforts. For
example, phylogenetic approaches have been used to (i) accu-
rately identify natural product-producing strains; (ii) analyze
the microbial diversity in natural product-producing assem-
9,10
11
Received: March 16, 2011
Published: July 13, 2011
1
2,13
blages;
(iii) re-collect specific natural product-producing
Copyright r 2011 American Chemical Society and
American Society of Pharmacognosy
1737
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
ARTICLE
because new geographic locations likely represent unique envi-
ronments with different evolutionary pressures, such analogues
may possess different biological activities. This phylogeny-guided
isolation approach for new secondary metabolites was tested
for its capacity to direct the search for specific classes of new
natural products from Cura c- ao marine cyanobacteria. As a result,
we discovered ethyl tumonoate A (1), a new tumonoic acid
derivative with anti-inflammatory activity in murine macrophage
cells and inhibitory activity of calcium oscillations in neocortical
neurons.
Figure 1. (A) Underwater pictures of cf. Oscillatoria margaritifera
NAC8-46 from 8ꢀ10 m in Caracas baii, Cura c- ao. (B) Photomicrograph
(
400ꢁ) of filaments of cf. O. margaritifera NAC8-46.
clade. The original tumonoic acid producer, also described as a
Lyngbya, was likely also misidentified because of the lack of
21
phylogenetic inference.
On the basis of the close evolutionary relatedness with the
tumonoic acid-producing B. cantharidosmum, the cf. O. margar-
itifera strains were predicted to have a high likelihood of
possessing the genes encoding for the biosynthesis of these or
closely related molecules. Therefore, these four cf. O. margar-
itifera strains were screened for tumonoic acid-type molecules
by a combination of LC-ESI-MS and MALDI-TOF-MS. All four
cf. O. margaritifera populations displayed molecular ions corre-
sponding to the known tumonoic acids AꢀC and EꢀG, as well as
their methyl esters (methyl tumonoates A and B). The identities
of these were confirmed by comparison of retention times and
MS/MS fragmentation patterns with the authentic compounds
(see Supporting Information).
’
RESULTS AND DISCUSSION
In October 2008 the tropical depression Omar made landfall
on the Caribbean island of Cura c- ao, bringing torrential rainfall
16
and beach erosion. As a probable result of this disruptive
weather system, extensive black, mat-forming cyanobacterial
blooms emerged at several sites along the island’s leeward coast
(
Figure 1). Four geographically dispersed populations of this
cyanobacterium (NAC8-45, NAC8-46, NAC8-54, and NAC8-
5) were collected and morphologically and phylogenetically
5
compared. All four specimens fell under the criteria of Lyngbya
according to traditional morphology-based classification sys-
tems. The SSU (16S) rRNA genes, however, revealed a relatively
close phylogenetic relation with the Oscillatoria reference strain
PCC 7515 as well as other marine Oscillatoria species (Figure 2).
The genera Lyngbya and Oscillatoria are evolutionarily distant,
but share a number of morphological features and are therefore
The substantial geographic distance between the related
Pacific B. cantharidosmum PNG05-4 and the Caribbean cf.
O. margaritifera strains inspired further investigation of their
organic extracts by LC-ESI-MS for new variations of the tumonoic
acids. A major constituent with a molecular weight of m/z 368
17,18
+
often mistaken for each other.
As a consequence, the four
[M + H] was detected in all four cf. O. margaritifera populations,
specimens were reclassified as members of the genus Oscillatoria
based on their phylogenetic position in relation to the Oscillatoria
reference strain. The specimens were more specifically identified
as Oscillatoria margaritifera (K €u tzing ex Gomont, 1892) based on
a polyphasic combination of phylogenetic and morphological
characters (Figure 2; Table 1). Because tropical marine Oscillatoria
are also related to the genus Trichodesmium, the paraphyly of
Oscillatoria suggests the need for revision of tropical marine
Oscillatoria as a distinct and separate generic entity, hence the
definition cf. Oscillatoria margaritifera (Figure 2).
and database searches failed to suggest its identity. Three mid-
polar vacuum liquid chromatography (VLC) fractions of NAC8-
46 contained the metabolite, and these were subsequently
subjected to RP-HPLC to yield a pale yellow oil in a 3.0%
yield (1, 55.1 mg). HR-ESI-MS gave a peak at m/z 390.2617
+
[M + Na] , indicating a molecular formula of C H NO .
2
1
37
4
The IR spectrum showed characteristic absorption bands at
ꢀ
1
1630 and 1733 cm , indicating the presence of amide and ester
1
13
groups in the molecule, respectively. The H and C NMR
spectra of 1 showed an olefinic methine (δ 5.34, t; δ 129.0), an
H
C
Interestingly, the cf. O. margaritifera strains were phylogen-
etically most closely related (p-distance = 99.3%) to tumonoic
acid-producing Blennothrix cantharidosmum PNG05-4 from
R-proton of an amino acid (δ 4.39, dd; δ 58.7), an oxygenated
H
C
methine (δ 4.04, d; δ 79.9), and an oxygenated methylene
H
C
(δ 4.08, m; δ 61.0). The overlapped methylenes (δ 1.16ꢀ
H
C
H
19
Papua New Guinea (Figure 2). The morphologically similar
genus Blennothrix has been shown to share a close evolutionary
history with the genus Oscillatoria, indicating that taxonomic
1.25) and methyls (δ 0.78, t; 0.99, d) suggested a branched fatty
H
acid chain. On the basis of 2D NMR data analysis, all of the
protons and carbons of a proline residue, an ethyl group, and two
terminal regions of a fatty acid were assigned. HMBC correla-
19,20
delimitation of these two genera is also needed.
Other closely
0
0
0
related strains include the Panamanian venturamide-producing
Oscillatoria sp. PAB-21 and the viridamide-producing Oscillatoria
nigro-viridis strain 3LOSC from Cura c- ao. Phylogenetic inference
suggests that PAB-21 and 3LOSC strains diverged as a prior
evolutionary event into a more distant basal lineage. This basal
group also includes the microcolin-producing strain LP163,
which originally was characterized as a Lyngbya polychroa.
Herein, LP16 will be assumed to be a species of Oscillatoria
because of its clear phylogenetic nesting within the Oscillatoria
tions from H -1 to C-1 indicated that the ethyl group was
2
attached to the proline via an ester linkage. The HMBC correla-
0
tions from H-2, H -13, and H -5 to C-1 revealed that the proline
3
2
and the fatty acid were linked through an amide bond. Three
1
13
remaining methylenes, having similar H and C chemical shifts
(δ 1.18, m; δ 29.3, δ 1.17, m; δ 29.2, δ 1.16, m; δ 31.8),
H
C
H
C
H
C
were assigned as linking the two termini of the fatty acid. A 1D
NOE correlation between H -14 and H -6 indicated the E
3
2
geometry of the double bond. The H-2 pentet (δH 2.67, dq,
1
738
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
ARTICLE
Figure 2. Maximum-likelihood (PhyML) phylogenetic inference of the cf. Oscillatoria margaritifera strains based on the SSU (16S) rRNA gene. The
evolutionarily distant Gloeobacter violaceus PCC 7421 was used as outgroup. A lineage of the unicellular order Chroococcales and a lineage of the
heterocystous order Nostocales as well as several genera of the order Oscillatoriales were added to place the Oscillatoria lineage in a broader taxonomic
perspective. Representative reference strains were selected from Bergey’s Manual of Systematics and are distinguished with asterisks. The Oscillatoria
T
lineage is defined on the basis of the phylogenetic nesting with the Oscillatoria reference strain PCC 7515 and is highlighted. Specimens are designated
as taxa, strain, and acc. Nr. in brackets. The secondary metabolites produced by the Oscillatoria species are numbered next to the Oscillatoria lineage.
Corresponding secondary metabolite-producing strains are numbered with the equivalent number. The statistical supports are indicated at each node
according to the maximum-likelihood (ML), Bayesian inference (MrBayes), and maximum parsimony (MP) methods. The scale bar is equivalent to 0.04
substitution per nucleotide position.
a
Table 1. Morphological Characterization of cf. Oscillatoria margaritifera Strains from Cura c- ao
filament width
cell width
cell length
cross-wall
b
b
b
c
strain
thallus
sheaths
(μm)
(μm)
(μm)
constriction
apical cells
NAC8-45
NAC8-46
NAC8-54
NAC8-55
mats/clumps
mats/clumps
mats/clumps
mats/clumps
thin, barely visible
thin, barely visible
thin, barely visible
thin, barely visible
20.3 ( 0.7
22.8 ( 0.7
22.4 ( 0.4
38.0 ( 0.7
16.8 ( 0.8
20.2 ( 1.0
20.8 ( 0.9
35.4 ( 1.0
2.0 ( 0.1
1.9 ( 0.6
2.1 ( 0.5
3.1 ( 0.3
>0.5 (>3.0%)
>0.5 (>2.5%)
>0.5 (>2.4%)
>0.5 (1.4%)
slightly rounded with thickened cell walls
slightly rounded with thickened cell walls
slightly rounded with thickened cell walls
slightly rounded with thickened cell walls
a
10,17 b
Species assigned by morphology in accordance with current taxonomic systems.
Filament diameters are the average of three filament measurements, and cell measurements the average of 10 adjacent cells of three filaments. Amount
The size measurements presented in this table are averages in μm.
c
of cell cross-wall constriction in μm and percentage of total cell width.
J = 7.2, 7.2 Hz) in the fatty acid chain suggested that H -13 and
to confirm that these molecules were real secondary metabo-
lites and not isolation artifacts. Molecular weights of m/z 368
3
19,21,22
the 3-OH substituent were of anti configuration (Table 2).
13
0
+
+
The small difference in the C NMR chemical shifts of C-3 from
[M + H] and m/z 562.2 [M + Na] were detected, correspond-
ing to ethyl tumonoate A and methyl tumonoate B, respectively.
Methyl esters of NRPS-type metabolites have been shown to be
produced in other bacteria by an unusual S-adenosyl-L-methio-
0
C-4 (Δδ
0
0
= 4.3 ppm) indicated that the proline peptide
C3 ꢀC4
23
bond was in the trans conformation (Table 2).
The absolute configuration of proline in ethyl tumonoate A
1) was determined by acid hydrolysis followed by derivatization
2
6
(
nine (SAM)-dependent methyl-transferase. Additionally,
naturally occurring methyl esters have been found as natural
with Marfey’s reagent fluorodinitrophenyl-5-L-valine amide
1
1
(
FDVA) and subsequent LCMS analysis and revealed that it
products in other cyanobacterial collections. From the
MALDI-TOF MS experiments described above and this pre-
cedence, we conclude that the ethyl group of compound 1 and
methyl group of methyl tumonoate B are true natural products
of cf. O. margaritifera.
24
was of L configuration. The absolute configuration of the 3-OH
group was determined as S by derivatization with Mosher’s
reagent (R/S-MTPA-Cl) followed by H NMR data analysis
1
2
5
(Figure 3). Therefore the absolute configuration of the 2,4-
dimethyl-3-hydroxydodec-4-enoic acid was 2R,3S.
Live filaments from cultured specimens of cf. O. margaritifera
were screened by MALDI-TOF-MS for methyl and ethyl tumonoates
Ethyl tumonoate A (1) shares structural resemblance with a
number of other cyanobacterial secondary metabolites, such as
the viridamides and the microcolins; the latter are derived from
1
739
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
ARTICLE
Table 2. NMR Spectroscopic Data for Ethyl Tumonoate A
1
13
(
1) in CDCl at 600 MHz ( H) and 150 MHz ( C)
3
a
position
δ
C
δ
H
multi (J in Hz) COSY
HMBC
Fatty acyl group
1
2
3
4
5
6
7
8
9
174.9
41.1 2.67, dq (7.2, 7.2) 2-Me, 3 1, 3, 4, 13
79.9 4.04, d (7.2)
134.2
2
1, 2, 4, 5, 13, 14
Figure 3. Δδ
ethyl tumonoate A (1).
values around C-3 between the Mosher’s esters of
SꢀR
129.0 5.34, t (7.1)
27.6 1.92, m
29.4 1.25, m
29.3 1.18, m
29.2 1.17, m
31.8 1.16, m
22.7 1.18, m
14.1 0.78, t (7.0)
14.4 0.99, d (7.2)
11.4 1.52, s
6
3, 6, 7, 14
4, 5, 7, 8
5
25
5, 7
6
concentrations of 10 μg/mL or higher against H-460 human lung
tumor cells. The low toxicity of ethyl tumonoate A (1) corre-
sponds with the lack of toxicity previously reported for related
19,21
tumonoic acid analogues.
1
1
1
1
1
3
0
The original VLC fractions were also screened for their ability
to activate or suppress calcium flux in neocortical neurons. VLC
fractions E and F produced calcium influx at high concentrations
1
12
11
2
2
10, 11
1, 2, 3
3, 4, 5
3
(
0.05 mg/mL), while fractions E, F, and G suppressed sponta-
4
neous calcium oscillations at lower concentrations. Because these
fractions predominately contained the tumonoic acids, these
metabolites were isolated and individually tested as pure com-
pounds in the same assay. While ethyl tumonoate A at 10 μM
-OH
3.32, brs
Proline
0
1
172.2
0
0
0
0
0
0
0
2+
2
58.7 4.39, dd (8.6, 4.0)
29.1 2.10, m, 1.88 m
24.8 1.96, m, 1.90 m
47.0 3.57, m
3
1 , 3 , 4 , 5
produced a nearly complete inhibition of Ca oscillations, tumo-
noic acids A and F showed partial inhibition at this concentration,
and tumonoic acid G was completely inactive.
0
0
0
0
0
0
0
3
2 , 4
1 , 2 , 4 , 5
0
0
0 0
0
0
4
3 , 5
2 , 3 , 5
0
0
0 0
2 , 3 , 4
5
4
Ethyl group
’ CONCLUSIONS
0
0
00
0
00
1
2
61.0 4.08, m
2
1 , 2
Evolutionary relationships inferred from common house-
keeping genes have had an enormous impact on cyanobacterial
taxonomy and systematics. Moreover, phylogenetics has been
00
00
00
14.1 1.16, t (7.3)
13
1
1
a
1
10
From H to the indicated C.
increasingly informative for natural products research in the
characterization, comparison, and re-collection of natural pro-
ducts-producing organisms. In this study, we demonstrated a
direct correlation between phylogenetics and class of secondary
metabolite production in marine cyanobacteria. It should be
noted that, in freshwater cyanobacteria, secondary metabolites
such as microcystins and cylindrospermopsins have been re-
phylogenetically related cyanobacteria (Figure 2). These mol-
ecules are composed of 8ꢀ12-carbon-long fatty acid chains
connected with amino acid moieties, a structure type predicted
to be biosynthesized by mixed polyketide synthase (PKS) and
nonribosomal peptide synthetase (NRPS) clusters. Because of
the structural similarities of the molecules as well as the evolu-
tionary relatedness of the organisms, these secondary metabo-
lites are likely derived from homologous PKS/NRPS pathways.
However, marine cyanobacteria have often been found to con-
29
ported from phylogenetically unrelated genera. However, the
origin of the biosynthetic pathways encoding these toxins has not
been firmly established, and it remains uncertain whether they
were laterally transferred or evolved from an ancient ancestor
8,27
tain multiple, genetically different biosynthetic pathways.
29
followed by gene loss in nonproducing taxa. The direct
Thus, the fact that the venturamides are structurally distinct
from the aforementioned secondary metabolites is most likely
explained by their biosynthesis involving analogous genetic
pathways. The venturamides appear structurally more similar
to trichamide, a metabolite isolated from the related cyanobac-
terium Trichodesmium erythraeum IMS 101.
correlation found in marine cyanobacteria allows for the predic-
tion of orthologous biosynthetic pathways and production of
specific secondary metabolites. Herein, we applied this knowl-
edge to investigate related but geographically distant cyanobac-
terial populations for new bioactive secondary metabolites. This
phylogeny-guided isolation approach led to the discovery of ethyl
tumonoate A, a new tumonoic acid derivative with anti-inflam-
matory and ion modulation activities.
The structural resemblance of the tumonoic acids to bacterial
homoserine lactones stimulated investigations into the quorum
1
9
sensing activities of these molecules. In fact, several of the
tumonoic acids moderately inhibit cell-to-cell communication,
resulting in a reduction in bioluminescence in Vibrio harveyi.
19
’ EXPERIMENTAL SECTION
Similarly, the potent immunosuppressive activity in the structur-
General Experimental Procedures. Optical rotations were
measured on a JASCO P-2000 polarimeter and IR spectra on a
ThermoElectron Nicolet IR100 FT-IR spectrometer. NMR spectra
ally and evolutionary related microcolins A and B was a rationale
2
8
for testing ethyl tumonoate A for these activities. Ethyl
tumonoate A (1) showed in vitro anti-inflammatory activity in the
RAW264.7 murine macrophage cell-based nitric oxide assay with
an IC50 of 9.8 μM (3.6 μg/mL) with little or no cytotoxi-
city. Additionally, only a low level of toxicity was observed at
were recorded with chloroform as an internal standard (δC 77.2, δ
H
1
7.26) on a Bruker 600 MHz spectrometer (600 and 150 MHz for H
1
3
and C NMR, respectively) equipped with a 1.7 mm MicroCryo-
Probe. LR- and HR-ESI-MS were obtained on a ThermoFinnigan LCQ
1
740
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
ARTICLE
Advantage Max mass detector and Thermo Scientific LTQ-XL Orbitrap
mass spectrometer, respectively. LCMS analysis was carried out on a
Finnigan LCQ Advantage Max spectrometer with a Finnigan Surveyor
HPLC system equipped with Thermo Finnigan Surveyor PDA plus
detector. HPLC was performed using a Waters 515 pump and a Waters
most parsimonious tree. Bootstrap support was obtained from 1000
replicates.
Isolation and Structure Elucidation. Biomass of each cyano-
bacterial specimen was extracted exhaustively with CH Cl /MeOH
2 2
(2:1). The extracts were fractionated by silica gel VLC using a stepwise
gradient from 100% hexanes to 100% MeOH in nine fractions (AꢀI) of
increasing polarity. VLC fractions were purified over a 500 mg C-18 Sep-
Pak followed by RP-HPLC (Phenomenex Jupiter 10 μm C18, 300 Å,
996 photodiode array detector.
Sampling and Taxonomic Characterization. Cyanobacterial
strains were collected by scuba or snorkeling from the following four
sites along the leeward coast of Cura c- ao, Netherlands Antilles: NAC8-45
10 ꢁ 250 mm, 85% CH
3 2
OH/H O at 3 mL/min). For MALDI-TOF-
MS, each specimen (5ꢀ10 μg wet wt) was extracted with 1 μg/μL of
0
0
(
Marie Pampoen; harbor inlet; 12ꢀ05 47.22 N, 68ꢀ54 54.54 W; 2ꢀ3 m
0
0
depth), NAC8-46 (Caracas baii; coral reef; 12ꢀ04 31.48 N, 68ꢀ51 49.37
matrix solution (70 mg/mL alpha-cyano-4-hydroxycinnamic acid and
2,5-dihydroxybenzoic acid (1:1), 750 μL of CH CN, 248 μL of dH O,
2 μL of trifluoroacetic acid) in 96-well plastic plates for 20ꢀ30 s.
One microliter of matrix extract was deposited on the well of a Bruker
Microflex MSP 96 stainless steel target plate and run on a Bruker
Microflex mass spectrometer equipped with flexControl 3.0.
Ethyl Tumonoate A (1): pale yellow oil; [R]
IR (neat) νmax 1733, 1630 cm ; H, C, and 2D NMR data, see
Table 2; HRESIMS m/z [M + Na] 390.2617 (calcd for C21H37NO
4
390.2620).
Absolute Configuration of the Proline Residue in Ethyl
Tumonoate A (1) by Marfey’s Method (ref 24). Ethyl tumonoate
A (100 μg) was treated with 100 μL of 6 N HCl at 110 ꢀC for 30 min.
The reaction products obtained following lyophilization of the crude
reaction mixture were dissolved in 100 μL of H
The dried hydrolysate was dissolved in 100 μL of 1 M NaHCO
0
W; 8ꢀ10 m depth), NAC8-54 (Pierbaai reef; coral reef; 12ꢀ05 40.07 N,
3 2
0
6
8ꢀ54 47.55 W; 8ꢀ10 m depth), and NAC8-55 (Jan Theil baii; coral
0
0
reef; 12ꢀ04 33.80 N, 68ꢀ52 54.55 W; 3ꢀ4 m depth). Specimens were
cleaned from macroflora/fauna under a dissecting scope. Algal tissue (ca.
200 mg) was preserved for genetic analysis in 10 mL of RNAlater
2
5
(
Ambion), for chemical analysis in seawater/EtOH (1:1) at ꢀ20 ꢀC, and
D
ꢀ77.5 (c 1.0, CHCl
3
);
ꢀ1
1
13
in seawater filtered through 0.2 μm Acrodisc Syringe filters (PALL Life
Sciences) for culturing and morphological analysis. Morphological
characterizations were performed using an Olympus IX51 epifluorescent
microscope (100ꢁ) equipped with an Olympus U-CMAD3 camera.
Taxonomic identification of cyanobacterial specimens was performed in
+
Na,
10,17
accordance with modern phycological systems.
Polymerase Chain Reaction (PCR) and Cloning. Genomic
DNA was extracted using the Wizard Genomic DNA purification kit
O and relyophilized.
, and
2
(
Promega Inc.) following the manufacturer’s specifications. DNA con-
centration and purity were measured on a DU 800 spectrophotometer
Beckman Coulter). The 16S rRNA genes were PCR-amplified from
isolated DNA using the cyanobacteria-specific primers 106F and 1509R,
3
then 25 μL of 1% L-FDVA (1-fluoro-2,4-dinitrophenyl-5-L-valine amide)
in acetone was added. The solution was vortexed and incubated at 40 ꢀC
for 60 min. The reaction was quenched by the addition of 25 μL of 2 N
HCl and diluted with 100 μL of MeOH, and then a 10 μL aliquot was
analyzed by LC-MS using the following RP-HPLC conditions: [HP
Lichrosphere 100 RP-18 column, 5.0 μm, 4.0 ꢁ 125 mm, with a stepped
(
30
as previously described. The PCR reaction volumes were 25 μL
containing 0.5 μL (∼50 ng) of DNA, 2.5 μL of 10 ꢁ PfuUltra IV
reaction buffer, 0.5 μL (25 mM) of dNTP mix, 0.5 μL of each primer
(10 μM), 0.5 μL of PfuUltra IV fusion HS DNA polymerase, and 20.5 μL
gradient elution of 0.1% trifluoroacetic acid in H
CH
2
O (eluent A) and 100%
of dH O. The PCR reactions were performed in an Eppendorf Master-
2
3
CN (eluent B); gradient program 0ꢀ5 min, B, 30%, 5ꢀ25 min; B,
cycler gradient as follows: initial denaturation for 2 min at 95 ꢀC, 25
cycles of amplification (20 s at 95 ꢀC, 20 s at 50 ꢀC, and 1.5 min at
30ꢀ70%, flow rate, 500 μL/min]. The Marfey derivatives of authentic
D- and L-Pro eluted at 15.88 and 13.74 min, respectively. The Marfey
derivative from acid hydrolysates of 1 was eluted at 13.47 min, and co-
injection with the authentic L-Pro derivative gave a single peak.
72 ꢀC), and final elongation for 3 min at 72 ꢀC. PCR products were
purified using a MinElute PCR purification kit (Qiagen) before sub-
cloned using the Zero Blunt TOPO PCR cloning kit (Invitrogen)
following the manufacturer’s specifications. Plasmid DNA was isolated
using the QIAprep Spin Miniprep kit (Qiagen) and sequenced with M13
primers. The gene sequences are available in the DDBJ/EMBL/Gen-
Bank databases under acc. nos. GU724196, GU724197, GU724207, and
GU724208.
Absolute Configuration of the 3-Hydroxy Group of Ethyl
Tumonoate A (1) by Mosher’s Method (ref 25). Dried com-
pound 1 (1 mg) was dissolved in 200 μL of anhydrous pyridine, and a
catalytic amount of DMAP (4-dimethylaminopyridine) and an excess
amount of (R)-MTPA-Cl were added. The reaction vial was maintained
at room temperature for 24 h, and the reaction progress was monitored
by normal-phase thin-layer chromatography (NP TLC). The (S)-
MTPA ester of 1 was isolated by preparative NP TLC with a developing
solvent of hexanes/EtOAc (1:1) and an eluent of 100% EtOAc. Using
the same procedure with (S)-MTPA-Cl, the (R)-MTPA ester of 1 was
also obtained. The absolute configuration of the 3-hydroxy group was
Phylogenetic Inference. Gene sequences were aligned bidirec-
tionally using the L-INS-i algorithm in MAFFT 6.717. A total of 1378 bp
(310 parsimony informative sites) of the 16S rRNA gene were analyzed
without data exclusion. The evolutionary distant unicellular cyanobac-
T
terium Gloeobacter violaceus PCC 7421 (NC005125) was included as
T
1
an outgroup. Representative type-strains ( ) were selected from Bergey’s
determined by H NMR analysis of the (R)/(S)-MTPA esters (see
17
Manual. Phylogenetic analyses were compared using the maximum
likelihood, Bayesian inference, and maximum parsimony methods.
Appropriate nucleotide substitution models were compared and se-
lected using uncorrected/corrected Akaike information criterion (AIC/
AICc), Bayesian information criterion (BIC), and the decision-theoretic
Supporting Information).
3-(S)-MTPA ester of ethyl tumonoate A (1): pale yellow oil; H
1
3 H
NMR (CDCl , 600 MHz) δ 5.79 (1H, t, J = 7.2 Hz, H-5), 5.61 (1H, d,
0
0
0
J = 10.6 Hz, H-3), 4.15 (2H, m, H-1 ), 4.14 (1H, m, H-2 ), 3.56 (1H, m,
0
0
H-5 a), 3.47 (1H, m, H-5 b), 2.96 (1H, dq, J = 10.6, 6.9 Hz, H-2), 2.07
0
0
(
DT) in jModelTest 0.1.1. Bayesian analysis was conducted using
(2H, m, H-6), 1.94 (1H, m, H-4 a), 1.93 (1H, m, H-3 a), 1.86 (1H, m,
31
0
0
MrBayes 3.1. The AIC1, AIC2, DIC, and BIC criteria all selected
GTR+I+G as the optimum model. The Markov chains, one cold and
three heated, were run for 3 000 000 generations. The maximum like-
lihood inference was performed using PhyML v2.4.4. The analysis was
run using the GTR+I+G model with 1000 bootstrap replicates. The
maximum parsimony analysis was performed in PAUP* 4.0b10 using a
heuristic search through the branch-swapping tree-bisection-reconnec-
tion algorithm with the addition of 10 000 random replicates to find the
H-3 b), 1.77 (1H, m, H-4 b), 1.64 (3H, s, 4-Me), 1.37 (2H, dd, J = 13.3,
6.9 Hz, H-7), 1.32ꢀ1.21 (8H, m, H-8, H-9, H-10, H-11), 1.25 (3H, t,
0
0
J = 7.1 Hz, H-2 ), 1.01 (3H, d, J = 7.0 Hz, 2-Me), 0.87 (3H, t, J = 7.1 Hz,
32
+
+
3H); LR ESIMS m/z 583.94 [M + H] , 606.17 [M + Na] .
3-(R)-MTPA Ester of Ethyl Tumonoate A (1): pale yellow oil;
H NMR (CDCl , 600 MHz) δ 5.72 (1H, t, J = 6.9 Hz, H-5), 5.50 (1H,
3 H
1
0
00
d, J = 10.6 Hz, H-3), 4.36 (1H, dd, J = 7.9 Hz, H-2 ), 4.18 (2H, m, H-1 ),
0
0
3.59 (1H, m, H-5 a), 3.52 (1H, m, H-5 b), 2.95 (1H, dq, J = 10.7, 6.9 Hz,
1
741
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
H-2), 2.05 (2H, m, H-6), 1.98 (1H, m, H-3 a), 1.96 (1H, m, H-4 a), 1.91
ARTICLE
0
0
Software). The EC50 values were determined by nonlinear regression
analysis using a logistic equation.
0
0
(1H, m, H-3 b), 1.80 (1H, m, H-4 b), 1.49 (3H, s, 4-Me), 1.36 (2H, m,
H-7), 1.33ꢀ1.20 (8H, m, H-8, H-9, H-10, H-11), 1.27 (3H, t, J = 7.1 Hz,
00
H-2 ), 1.03 (3H, d, J = 6.9 Hz, 2-Me), 0.87 (3H, t, J = 6.6 Hz, 3H); LR
’
ASSOCIATED CONTENT
+
+
ESIMS m/z 583.82 [M + H] , 606.11 [M + Na] .
Nitric Oxide Assay. Cells from the mouse macrophage cell line
RAW264.7 (ATCC; Manassas, VA) were cultured in DMEM with 4 mM
L-glutamine and 4.5 g/L glucose supplemented with 10% fetal bovine
serum (FBS), penicillin, and streptomycin. Unless otherwise stated
S
Supporting Information. LR- and HR-ESIMS, one-
b
1
13
dimensional ( H and C) and two-dimensional (COSY, HSQC,
and HMBC) NMR spectra of ethyl tumonoate A (1), and H
NMR spectra of (S)- and (R)-MTPA esters of ethyl tumonoate
A. This material is available free of charge via the Internet at
http://pubs.acs.org.
1
4
RAW264.7 cells were seeded in 96-well plates (5 ꢁ 10 cells/well), and
after settling for 1 day they were stimulated with 3 μg/mL LPS in the
absence or presence of ethyltumonate A (1 to 10 μg/mL) for 24 h in
triplicate wells at 37 ꢀC with 5% CO
2
. The generation of NO was
assessed in the supernatant of cell cultures by quantification of nitrite
’
AUTHOR INFORMATION
33
using the Griess reaction. In brief, 50 μL of each supernatant were
added to 96-well plates together with 50 μL of 1% sulfanilamide in 5%
phosphoric acid plus 50 μL of 0.1% naphthylenediamine dihydrochlo-
Corresponding Author
*Tel: (858) 534-0578. Fax: (858) 534-0576. E-mail: wgerwick@
ucsd.edu.
ride in H O, and the optical density was measured at 570 nm. The
2
IC50 value, the sample concentration resulting in 50% inhibition of
NO production, was determined using nonlinear regression analysis
’ ACKNOWLEDGMENT
(
% nitrite versus concentration). Cytotoxicity was measured in NCI
We dedicate this work to J. Kom ꢀa rek for his contributions to
cyanobacterial taxonomy. We gratefully acknowledge the gov-
ernment of Cura c- ao for collection permits and the CARMABI
Research Station for assistance and research support. We also
thank J. Nunnery for assistance with the cyanobacterial collec-
tions. This research was supported by the Halliday award (SIO)
and NIH Grant NS 053398.
33
H-460 human lung carcinoma cells using the MTT assay. Cells were
4
seeded in 96-well plates at 3.3 ꢁ 10 cells/mL in 180 μL of RPMI 1640
medium with 10% FBS. After overnight recovery, the test compounds
were dissolved in DMSO and diluted into medium without FBS and
then added to the wells (tested at final concentrations of 30 and 3 μg/
mL). After 48 h, cell viability was determined by MTT staining.
Neocortical Neuron Culture. Primary cultures of neocortical
neurons were obtained from embryonic day 16 Swiss-Webster mice.
Briefly, pregnant mice were euthanized by CO
2
asphyxiation, and
’ REFERENCES
(1) Tan, L. T. Phytochemistry 2007, 68, 954–979.
(2) Gerwick, W. H.; Coates, R. C.; Engene, N.; Gerwick, L. G.;
Grindberg, R.; Jones, A.; Sorrels, C. Microbe 2008, 3, 277–284.
(3) Sharp, K.; Arthur, K. E.; Gu, L.; Ross, C.; Harrison, G.;
Gunasekera, S. P.; Meickle, T.; Matthew, S.; Luesch, H.; Thacker,
R. W.; Sherman, D. H.; Paul, V. J. Appl. Environ. Microbiol. 2009, 75,
embryos were removed under sterile conditions. Neocortices were
collected, stripped of meninges, minced by trituration with a Pasteur
pipet, and treated with trypsin for 25 min at 37 ꢀC. The cells were then
dissociated by two successive trituration and sedimentation steps in
soybean trypsin inhibitor and DNase containing isolation buffer, cen-
trifuged, and resuspended in Eagle’s minimal essential medium with
Earle’s salt (MEM) and supplemented with 1 mM L-glutamine, 10%
FBS, 10% horse serum, 100 IU/mL penicillin, and 0.10 mg/mL
streptomycin, pH 7.4. Cells were plated onto poly-L-lysine-coated 96-
well (9 mm) clear-bottomed black-well culture plates (Costar) at a
2
879–2888.
4) Gerwick, W. H. Ocean Agenda 21: Passages to the Pacific Century;
Oregon State University Sea Grant, 1989; pp 59–65.
5) Wilmotte, A.; Herdman, M. Bergey’s Manual of Systematic
Bacteriology, 2nd ed.; Springer: New York, 2001; pp 487–493.
(
(
5
density of 1.5 ꢁ 10 cells/well. Cells were then incubated at 37 ꢀC in a
(
(
(
6) Kom ꢀa rek, J. Algae 2006, 21, 349–375.
7) Sumina, E. L. Microbiology 2006, 75, 459–464.
8) Engene, N.; Choi, H.; Esquenazi, E.; Rottacker, E. C.; Ellisman,
5
% CO and 95% humidity atmosphere. Cytosine arabinoside (10 μM)
2
was added to the culture medium on day 2 after plating to prevent
proliferation of nonneuronal cells. The culture media was changed on
days 5 and 7 using a serum-free growth medium containing Neurobasal
Medium supplemented with B-27, 100 IU/mL penicillin, 0.10 mg/mL
streptomycin, and 0.2 mM L-glutamine. Neocortical cultures were used
in experiments at 8ꢀ13 days in vitro (DIV). All animal use protocols
were approved by the Institutional Animal Care and Use Committee
M. H.; Dorrestein, P. C.; Gerwick, W. H. Environ. Microbiol. 2011, 13,
1601–1610.
(9) Hoffman, L.; Kom ꢀa rek, J.; Ka ꢁs tovsk ^y ., J. Arch. Hydrobiol. 2005,
159, 95–115.
(10) Kom ꢀa rek, J.; Anagnostidis, K. In S €u sswasserflora von Mitteleuropa;
B €u del, B.; G €a rtner, G.; Krienitz, L.; Schagerl, M., Eds.; Gustav Fischer: Jena,
2005; p 19/2.
(
IACUC).
2+
Intracellular Ca Monitoring. Neocortical neurons grown in
6-well plates were used for [Ca ]
(11) Simmons, T. L.; Engene, N.; Ure ~n a, L. D.; Romero, L. I.;
2+
ꢀ
Ortega-Barria, E.; Gerwick, L.; Gerwick, W. H. J. Nat. Prod. 2008, 71,
544–1550.
12) Simmons, T. L.; Coates, R. C.; Clark, B. R.; Engene, N.;
9
i
measurements at 12ꢀ13 DIV.
1
Briefly, the growth medium was removed and replaced with dye loading
buffer (50 μL/well) containing 4 μM Fluo-3 and 0.04% pluronic acid
F-127 in Locke’s buffer (8.6 mM Hepes, 5.6 mM KCl, 154 mM NaCl,
(
Gonzalez, D.; Esquenazi, E.; Dorrestein, P. C.; Gerwick, W. H. Proc.
Natl. Acad. Sci. U. S. A. 2008, 105, 4587–4594.
5.6 mM glucose, 1.0 mM MgCl
2 2
, 2.3 mM CaCl , 0.0001 mM glycine, pH
(
13) Taniguchi, M.; Nunnery, J. K.; Engene, N.; Esquenazi, E.;
Byrum, T.; Dorrestein, P. C.; Gerwick, W. H. J. Nat. Prod. 2009, 73,
93–398.
14) Tidgewell, K.; Engene, N.; Byrum, T.; Media, J.; Doi, T.;
Valeriote, F. A.; Gerwick, W. H. ChemBioChem 2010, 11, 1458–1466.
15) Thacker, R. W.; Paul, V. J. Appl. Environ. Microbiol. 2004, 70,
7.4). After 1 h incubation in dye loading buffer, the neurons were washed
four times in fresh Locke’s buffer (200 μL/well) using an automated cell
washer (BioTek instrument) and transferred to a FLEXstation II
3
(
(Molecular Devices). The final volume of Locke’s buffer in each well
2+
was 150 μL. Cells were excited at 485 nm, and Ca -bound Fluo-3
emission was detected at 535 nm.
(
3305–3312.
Data Analysis. Timeꢀresponse and concentrationꢀresponse
(16) Meterological service of Netherlands Antilles and Aruba
(http://www.weather.an/reports/documents/EffectsofOmar.pdf).
graphs were generated using Graphpad Prism software (Graphpad
1
742
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743

Journal of Natural Products
ARTICLE
(
17) Castenholz, R. W. Bergey’s Manual of Systematic Bacteriology,
nd ed.; Springer: New York, 2001; pp 474–487.
18) Engene, N.; Coates, R. C.; Gerwick, W. H. J. Phycol. 2010, 46,
91–601.
19) Clark, B. R.; Engene, N.; Teasdale, M. E.; Rowley, D. C.;
Matainaho, T.; Valeriote, F. A.; Gerwick, W. H. J. Nat. Prod. 2008, 71,
530–1537.
20) Abed, R. M.; Palinska, K. A.; Camoin, G.; Golubic, S. FEMS
Microbiol. Lett. 2006, 260, 171–177.
21) Harrigan, G. G.; Luesch, H.; Yoshida, W. Y.; Moore, R. E.;
Nagle, D. G.; Biggs, J.; Park, P. U.; Paul, V. J. J. Nat. Prod. 1999, 62,
64–467.
22) Paik, S.; Carmeli, S.; Cullingham, J.; Moore, R. E.; Patterson,
G. M. L.; Tius, M. A. J. Am. Chem. Soc. 1994, 116, 8116–8125.
23) Dorman, D. E.; Borvey, F. A. J. Org. Chem. 1973, 38, 1719–
722.
2
5
(
(
1
(
(
4
(
(
1
(
(
24) Marfey, P. Carlsberg Res. Commun. 1984, 49, 591–596.
25) Dale, J. A.; Dull, D. L.; Mosher, H. S. J. Org. Chem. 1969, 34,
2
543–2549.
26) Weinig, S.; Hecht, H.-J.; Mahmud, T.; M €u ller, R. Chem. Biol.
003, 10, 939–952.
27) Jones, A. C.; Monroe, E. A.; Podell, S.; Hess, W.; Klages, S.;
(
2
(
Esquenazi, E.; Niessen, S.; Hoover, H.; Rothmann, M.; Lasken, R.; Yates
III, J. R.; Reinhardt, R.; Kube, M.; Burkart, M.; Allen, E. E.; Dorrestein,
P. C.; Gerwick, W. H.; Gerwick, L. Proc. Natl. Acad. Sci. U. S. A. 2011,
1
7
1
08, 8815–8820.
28) Zhang, L.-H.; Longley, R. E.; Koehn, F. E. Life Sci. 1997, 60,
51–762.
(
(29) St €u ken, A.; Jakobsen, K. S. Microbiology (Reading, U.K.) 2010,
56, 2438–2451.
(
30) N €u bel, U.; Garcia-Pichel, F.; Muyzer, G. Appl. Environ. Micro-
biol. 1997, 63, 3327–3332.
31) Ronquist, F.; Huelsenbeck, J. P. Bioinformatics 2003, 12,
572–1574.
(
1
(32) Guindon, S.; Gascuel, O. Syst. Biol. 2003, 52, 696–704.
(33) Green, L. C.; Wagner, D. A.; Glogowski, J.; Skipper, P. L.;
Wishnok, J. S.; Tannenbaum, S. R. Anal. Biochem. 1982, 126, 131–138.
(34) Mosman, T. J. Immunol. Methods 1983, 65, 55–63.
1
743
dx.doi.org/10.1021/np200236c |J. Nat. Prod. 2011, 74, 1737–1743