Structure of hierridin c, synthesis of hierridins b and c, and evidence for prevalent alkylresorcinol biosynthesis in picocyanobacteria

Article abstract of DOI:10.1021/acs.jnatprod.8b01038

Small, single-celled planktonic cyanobacteria are ubiquitous in the world's oceans yet tend not to be perceived as secondary metabolite-rich organisms. Here we report the isolation and structure elucidation of hierridin C, a minor metabolite obtained from

Full text of DOI:10.1021/acs.jnatprod.8b01038

Article
Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX
pubs.acs.org/jnp
Structure of Hierridin C, Synthesis of Hierridins B and C, and
Evidence for Prevalent Alkylresorcinol Biosynthesis in
Picocyanobacteria
Margarida Costa,†,# Ivo E. Sampaio-Dias,
‡,#
Raquel Castelo-Branco, Hugo Scharfenstein,†
†,#
§
§
§
§
Roberta Rezende de Castro, Artur Silva, Maria Paula C. Schneider, Maria Joao
̃
Arauj
́
o,
Rosar
Vitor M. Vasconcelos, and Pedro N. Leao
́
io Martins,† Valentina F. Domingues, Fat
,⊥
∥
́
ima Nogueira, Vera Camo
▽
̃
es,
▽
†
,○
*,†
̃
†Interdisciplinary Centre of Marine and Environmental Research (CIIMAR/CIMAR), University of Porto, Avenida General Norton
de Matos, s/n, 4450-208 Matosinhos, Portugal
‡
LAQV-REQUIMTE, Department of Chemistry and Biochemistry, Faculty of Sciences, University of Porto, Rua do Campo Alegre
6
§
87, 4169-007 Porto, Portugal
Institute of Biological Sciences, Center of Genomic and System Biology, Federal University of Para
Brazil
(UFPA), Belem, PA66075-110,
́ ́
⊥
Health and Environment Research Centre (CISA), School of Health, Polytechnic Institute of Porto, Rua Dr. Anton
de Almeida, 400, 4200-072, Porto, Portugal
́
io Bernardino
io Bernardino de Almeida,
́
ica, Instituto de Higiene e
∥
REQUIMTE/LAQV, Instituto Superior de Engenharia, Instituto Politec
31, 4200-072 Porto, Portugal
́
nico do Porto, Rua Dr. Anton
de Parasitologia Med
́
4
▽
Global Health and Tropical Medicine, GHTM, Unidade de Ensino e Investigaca
̧
o
̃
Medicina Tropical, IHMT, Universidade Nova de Lisboa, UNL, Rua da Junqueira no. 100, 1349-008 Lisboa, Portugal
○
Department of Biology, Sciences Faculty, University of Porto, Rua do Campo Alegre, 4169-007 Porto, Portugal
S Supporting Information
*
ABSTRACT: Small, single-celled planktonic cyanobacteria
are ubiquitous in the world’s oceans yet tend not to be
perceived as secondary metabolite-rich organisms. Here we
report the isolation and structure elucidation of hierridin C, a
minor metabolite obtained from the cultured picocyanobacte-
rium Cyanobium sp. LEGE 06113. We describe a simple,
straightforward synthetic route to the scarcely produced
hierridins that relies on a key regioselective halogenation step.
In addition, we show that these compounds originate from a
type III PKS pathway and that similar biosynthetic gene
clusters are found in a variety of bacterial genomes, most
notably those of the globally distributed picocyanobacteria
genera Prochlorococcus, Cyanobium and Synechococcus.
yanobacteria produce a wide array of natural products
Prochlorococcus, are widespread throughout coastal and open
1
C
with varied structural features. Many of these com-
pounds, in particular those of polyketide and/or nonribosomal
peptide nature, show strong potency against pharmacologically
waters around the globe and contribute strongly to primary
5
productivity in the oceans. These so-called picocyanobacteria
have smaller genomes and do not appear to be as rich in
secondary metabolite biosynthesis genes, particularly regarding
2
relevant targets, such as the toxins microcystins or the
3
4
promising anticancer compound largazole. The majority of
nonribosomal peptides and polyketides. We have recently
cyanobacterial secondary metabolites have been isolated from
reported that the marine picocyanobacterium Cyanobium sp.
LEGE 06113 produces the alkylresorcinol hierridin B (1,
Figure 1A), an O-methylated monoalkylresorcinol that
field collections or from laboratory cultures of so-called
1
6
complex cyanobacteria. Such cyanobacteria display elaborate
morphological features (e.g. differentiated cells) and have large
features a long (C15) aliphatic chain. This metabolite had
genomes that harbor multiple secondary metabolite biosyn-
4
thetic gene clusters (BGCs). Picoplanktonic cyanobacteria,
Received: December 7, 2018
such as those from the genera Synechococcus/Cyanobium and
©
XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
COSY) NMR data for 2 in DMSO-d (Table 1) revealed the
6
architecture of the aromatic moiety as a 2,3,4,6-substituted
Table 1. NMR Spectroscopic Data ( H 600 MHz, 13C 150
1
MHz, DMSO-d ) for Hierridin C (2)
6
a
no.
δC, type
138.3, C
δH (J in Hz)
HMBC
1, 2, 6
COSY
1
1
2
3
4
5
6
7
8
1
2
3
1
1
-OH
8.30, s
128.0, C
112.8, C
147.6, C
96.6, CH
146.3, C
56.4, CH3
6.66, s
1, 3, 4, 6
Figure 1. (A) Structures of hierridins B (1) and C (2). (B) Key
HMBC (arrows) and COSY (thick bonds) correlations establishing
the partial structures of the aromatic moiety and aliphatic chain
terminus of compound 2.
3.77, s
3.81, s
2.66, t (7.6)
1.42, m
1.23−1.29, m
1.23, m
4
6
55.9, CH3
′
′
27.0, CH2
1, 2, 3, 2′
1′, 3′−12′
3′−12′
14′
13′, 15′
13′, 14′
2′
1′, 3′
2′
28.2, CH2
28.7−29.0, CH2
31.3, CH2
22.1, CH2
13.9, CH3
′-12′
3′
4′
been previously reported from the filamentous cyanobacterium
Leptolyngbya ectocarpii SAG 60.90 and it was active as a
mixture (with its C17 alkyl analogue hierridin A) against the
1.24, m
0.85, t (7.0)
15′
14′
7
15′
malaria parasite Plasmodium falciparum. The only other
a
From proton to indicated carbon.
known cyanobacterial monoalkylresorcinol is cylindrofridin
A, a chlorinated intermediate in the biosynthesis of the
8
cylindrocyclophanes originating from a type III PKS pathway.
phenol with an aliphatic chain and two methoxy groups as
three of the substituents, confirming its relatedness to
metabolite 1 (Figure 1B). Nevertheless, establishing the full
structure of 2 required extensive correlational analysis, mostly
owing to the determination of substituent arrangement in the
aromatic moiety. This was facilitated by HMBC correlations
from the aromatic proton (δ 6.66, H-5) to four deshielded
carbons at δ 147.6, 146.3, 138.3, and 112.8 (C-4, C-6, C-1,
and C-3, respectively), suggesting that the carbon resonating at
δC 128.0 (C-2) should be para to H-5. The benzylic protons
As part of an effort to reisolate 1, we identified a related
chlorinated monoalkylresorcinol, hierridin C (2, Figure 1A),
the isolation, structure elucidation, antiplasmodial activity, and
9
a straightforward synthesis of which are reported herein. To
understand the biosynthesis of the hierridins in cyanobacteria,
we sequenced and mined the genome of Cyanobium sp. LEGE
H
0
6113 to find the hierridin BGC, hid, of a type III PKS nature.
C
We show that this gene cluster is present in a large number of
cyanobacterial genomes, most notably those of picocyanobac-
terial genera abundant in the world’s oceans.
(
δH 2.66, H -1′) resonance showed correlations to three
2
aromatic carbons at δ 138.3, 128.0, and 112.8 (C-1, C-2, and
C
RESULTS AND DISCUSSION
■
C-3, respectively). This allowed assigning the carbons
resonating at δ 147.6 (C-6) and 146.3 (C-4) as ortho to H-
5. In addition, each of these carbons was found to have a
methoxy substituent, on the basis of HMBC correlations from
the methoxy protons at δ 3.81 (H -8) and 3.77 (H -7) to C-6
Isolation and Structure Elucidation of Hierridin C (2).
With an initial purpose of reisolating hierridin B (1) for
biological activity studies, early stationary-phase cultures of
Cyanobium sp. LEGE 06113 were harvested and the biomass
C
H
3
3
6
extracted with CH Cl :MeOH (2:1) as previously described.
and C-4, respectively. The final substituent arrangement was
2
2
The resulting extract was fractionated by a modified version of
established from HMBC correlations between the exchange-
the normal-phase vacuum liquid chromatography (VLC)
able phenol proton (OH-1) and three carbons at δ 146.3,
C
6
protocol employed to isolate hierridin B (1). The
138.3, and 128.0 (C-6, C-1, and C-2, respectively), leading to
modification consisted of an initial gentle gradient elution
step with the purpose of obtaining less-complex hierridin B-
containing fractions and facilitate the isolation of this
metabolite. This was effectively achieved, as revealed by
the assignment of the hydroxy group to the carbon at δ 138.3
(C-1) and of its position as adjacent to both C-6 and C-2. This
and the previously assumed constraints resulted in the
C
allocation of the alkyl substituent to the carbon at δ 128.0
C
spectroscopic inspection of the fractions obtained by VLC.
(C-2) and subsequent assignment of the carbons resonating at
δC 112.8 (C-3) and 147.6 (C-4) as ortho and meta to C-2,
respectively. The partial structure was finalized by the
straightforward assignment, on the basis of COSY and
HMBC correlations, of the methylene protons resonating at
δ 1.42 (H -2′) as being in an intermediate position between
1
The H NMR spectrum (CDCl ) of fraction A1, eluting with
3
an 87:13 mixture of n-hexane and EtOAc, exhibited the
distinctive methoxy singlets of 1, resonating at δH 3.85 and
3
.76 (Figure S1). This fraction was selected for further
purification, which was carried out using consecutive semi-
preparative and analytical-scale RP-HPLC separations, result-
ing in the isolation of 1, as expected, but also of the new
metabolite 2, both in submilligram amounts (0.5 and 0.3 mg,
H
2
the benzylic methylene and the CH envelope. A second partial
2
structure (Figure 1B) was extracted from the NMR data,
corresponding to the terminus of an aliphatic chain, and was
COSY- and HMBC-correlated to the methylene envelope.
From the NMR data, the substituent at C-3 was unknown
0
.003% and 0.002% dry wt, respectively).
1
6
The H NMR spectrum of 2 in CDCl resembled that of 1.
3
Still, a single aromatic proton resonance (δH 6.44) was
observed for 2, suggesting a pentasubstituted aryl moiety. A
combination of 1D ( H, C) and 2D (HSQC, HMBC,
and the precise number of CH groups in the aliphatic chain
difficult to estimate through the integration of the H NMR
data of 2. These uncertainties were clarified by HRESIMS
2
1
1
13
B
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
analysis. Despite the low ionization under ESI conditions,
using direct infusion we were able to observe an [M − H] ion
Scheme 1. (A) Synthesis of Hierridin B (1); (B) Synthesis
of Hierridin C (2)
−
a
with an m/z value of 397.2530 that exhibited a characteristic
isotopic pattern for one chlorine atom (Figure S8),
corresponding to a molecular formula for the parent
compound of C H ClO and thereby establishing the
2
3
39
3
proposed structure for 2. Because fragmentation of this
compound under ESI conditions was limited and uninforma-
tive, we resorted to gas chromatography−mass spectrometry
(
GC-MS/MS) analysis of 2 for structural corroboration. A
single-chlorine atom isotope pattern for the aromatic “head”
following benzylic fragmentation was observed (Figure S9),
supporting the proposed structure.
Metabolite 2 is structurally related to a small number of
previously reported chlorinated monoalkylresorcinols from the
eukaryotic realm. The halogenated 5-methylresorcinol deriva-
tive 4-chloroorcinol was isolated from the phytopathogenic
fungus Colletotrichium higgisianum and induced chlorosis in
plant leaves, inhibited seed germination, and was also toxic to
10
the crustacean Artemia salina. The monochasiols are a series
of long-chain alkylresorcinols, with a single chlorinated carbon
in between the resorcinol hydroxy groups, that have been
isolated from the slime mold Dictyostelium and have mild
1
1
immunosuppressive activity. Other examples include a series
of dimerized chlorinated short-chain alkylresorcinolsthe
spiromastolsthat have been reported recently from a deep-
a
Conditions: (i) tetradecylmagnesium chloride, THF, 2 h, rt; (ii)
H SO , MeOH, Pd/C 10% (w/w), H (1 atm), 80 °C; (iii) NCS,
2
4
2
HCl 37%, 15 min, rt; (iv) tetradecylmagnesium chloride, THF, 2 h, rt;
12
sea fungus and have antibacterial activity.
(
v) Wittig reaction: myristyltriphenylphosphonium bromide, n-BuLi,
Biological Activity of Hierridin C (2). Taking into
account the bioactivity profile of 1, with negligible toxicity to
human cell lines (showing only a mild selective activity toward
HT-29 cancer cells) and low micromolar antiplasmodial
activity, we decided to evaluate the cancer cell line
THF, rt; (vi) Pd/C 10% (w/w), H (1 atm), EtOAc, 0 °C.
2
6
1
3
counterpart (o-vanillin derivative). Applying the same
protocol for derivative 4 using MeOH as solvent (due to
solubility issues) afforded 1 in moderate yields, in contrast with
the results described in the literature for the monomethoxy-
lated parent. We found that this protocol led to the
formation of an ether at the α-benzylic position as a side
product (5), which can be explained by methanol trapping of
the intermediate carbocation in acidic medium. An additional
methoxy group seems to influence the outcome of the reaction,
as no formation of ether side products was reported by the
authors.
7
cytotoxicity and antimalarial activity of 2. Upon exposure of
human cancer cell lines to up to 30 μg mL− (75 μM) of 2, no
cytotoxicity was observed (Figure S10). Both 1 and 2 were
tested for in vitro activity against Plasmodium falciparum (3D7
strain and the chloroquine-resistant Dd2 strain). While all IC50
values obtained were in the low-micromolar range (Figure
S11), compound 2 was slightly more potent than 1 against the
1
13
3
D7 strain (IC50 values of 1.5 ± 0.1 and 2.1 ± 0.1 μM,
respectively). The two metabolites showed equal potency
toward the Dd2 strain (IC50 = 2.3 ± 0.7 μM for 2 and 2.3 ±
We predicted that monochlorination at carbon 3 of the
aromatic ring in 1 would be a bottleneck step for the
preparation of 2, because it is highly activated by the presence
of electron-donating groups, which might end up competing
with 3,5-dichlorinated and/or 3- and 5-monochlorinated
products. Our attempts at this reaction were unfruitful using
standard procedures commonly used for monohalogenation of
0
.1 μM for 1). Thus, the presence of the chlorine atom seems
to have a marginal influence on the antiplasmodial activity.
Total Synthesis of Hierridin C (2). The low yields of 1
and 2 obtained from Cyanobium sp. LEGE 06113 cultures and
their relatively simple and planar structures motivated our
efforts toward synthesizing these compounds in an expeditious
and convenient fashion, as a means to enable future
biosynthetic and biological activity studies. We envisioned
that the most straightforward synthetic route to the
preparation of 2 would be from aromatic electrophilic
chlorination of 1. A previous synthetic procedure for obtaining
16
phenol derivatives. Using NaClO/NaOH no reaction took
17
place, and NCS-assisted chlorination yielded an uncharac-
terized polychlorinated adduct. Thus, we considered mono-
chlorination of the salicylaldehyde derivative 3 prior to the
Grignard addition protocol. An analogous strategy was used,
13
11
a longer-chain analogue of 1 was available in the literature
for example, in the syntheses of certain monochasiols.
but did not work in our hands. Therefore, we designed a
convenient synthetic route to its preparation (Scheme 1, A)
using the commercially available suitable aromatic ring moiety
Following a standard chlorination protocol using 1 equiv of
NCS in the presence of 0.1 equiv of p-toluenesulfonic acid
17
(TsOH), no traces of 6 were detected after 6 h. Surprisingly,
replacement of TsOH by concentrated HCl afforded 6 in very
good yield (>90%) in a short reaction time (15 min) (Scheme
1, B). Despite the high aromatic ring activation in compound
3, the electrophilic chlorination revealed total regioselectivity
(see NOESY experiment, Figure S23). Still, we were unable to
obtain compound 2 following α-hydroxy elimination of the
2
-hydroxy-3,5-dimethoxybenzaldehyde (3), which can also be
14,15
easily prepared from 3,5-dimethoxybenzaldehyde.
Grignard addition to 3 using commercial tetradecylmagne-
sium chloride afforded the corresponding carbinol 4 smoothly
in high yields (86%). Reduction of the α-hydroxy group is
already described in the literature for the monomethoxylated
C
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 2. Hierridin biosynthesis in Cyanobium sp. LEGE 06113. (A) Overview of the hierridin biosynthetic gene cluster (hid) organization and
proposed biosynthesis of hierridins (Hal, putative halogenase). (B) Schematic representation of the experimental approach to test whether HidA
methylates an alkylresorcinol substrate in hierridin biosynthesis. (C) HPLC-PDA analysis of E. coli expressing the hidA gene, with two new peaks
eluting at 23.8 and 30.2 min. (D) LC-HRESIMS analysis of the two chromatographic peaks identified from HPLC-PDA analysis, indicating that an
m/z value consistent with compound 10 is observed for one of the peaks. (E) LC-HRESIMS analysis of the collected peak eluting at 30.2 min,
showing the same retention time as a standard of compound 11 obtained by synthesis and an m/z value consistent with the protonated species ([M
+
+
H] ) of 11.
20
21
Grignard addition product 7 (Scheme 1, B); see Supporting
Information for details).
To overcome this hurdle, we considered a Wittig olefination
for the introduction of the alkyl chain and subsequent
saturation of the olefin by catalytic hydrogenation (Scheme
monomers or of other prokaryotic alkylresorcinols. In a
previous effort using degenerate primers designed to target
then-available type III PKS genes from cyanobacteria, we failed
to PCR-amplify a homologue from Cyanobium sp. LEGE
6
06113.
1
, B). The appropriate Wittig reagent was prepared by
We envisioned that a genome-mining strategy could yield
better results and sequenced the genome of this cyanobacte-
rium, having obtained 2.6 Mbp distributed over 30 contigs.
The genome data were searched for homologues of the type III
PKS cylI from Cylindrospermum licheniforme UTEX B 2014
refluxing 1-bromotetradecane with triphenylphosphine, afford-
ing the corresponding phosphonium salt, which was recrystal-
lized in Et O. The Wittig reagent was then treated with n-BuLi
in tetrahydrofuran (THF) followed by addition of salicylalde-
hyde 6, affording the (Z)-alkene 8 in good yield (88%).
Hydrogenation of 8 using catalytic amounts of Pd/C 10% in
EtOAc at 0 °C proceeded smoothly, yielding hierridin C (2)
quantitatively in 1 h. Compound 2 can thus be efficiently
obtained in just three steps (3 → 6 → 8 → 2) in high total
yield (80%).
2
20
(involved in cylindrocyclophane biosynthesis) using BlastP.
A single homologue (27% identity, 42% similarity) was found,
located in a ∼387 kb contig. Annotation of the genomic
context of this type III PKS homologue (Table S1) revealed
genes with predicted functions consistent with hierridin
structural features (one methyltransferase and one hydrox-
ylase). Based on the analysis of predicted open reading frames,
this putative hierridin BGC (hid) is proposed to comprise at
least hidA, encoding a SAM-dependent methyltransferase;
hidB, coding for a FixC superfamily FAD-dependent
oxidoreductase and the type III PKS-encoding hidC (Figure
2A). A PAP-fibrillin is encoded downstream of this proposed
minimal hid gene cluster, but its relation to the hid genes is
unclear at this stage.
Discovery of the Hierridin Biosynthetic Gene Cluster
(
hid). Despite a number of reports on the biological activity of
extracts from members of the picocyanobacteria (Synechococ-
1
8,19
cus/Cyanobium/Procholorococcus),
congruent group is notoriously lacking in PKS or NRPS
biosynthetic genes. We had previously proposed that a type
III PKS could give origin to the alkylresorcinol moiety, based
this phylogenetically
4
6
on the biosynthesis of cyanobacterial cylindrocyclophane
D
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Figure 3. Distribution of hid-like gene clusters among bacteria. FastTree 2 (approximately maximum-likelihood) phylogeny of bacterial
homologues of HidC, highlighting clades dominated by type III PKS enzymes that are encoded in clusters with the same general architecture of the
hid gene cluster, i.e., those putatively coding for an additional SAM-dependent methyltransferase (MT) and a FAD-dependent oxidoreductase
(
OxRed). In the highlighted clades, the organization of clusters lacking either the MT or the OxRed are also shown. Dotted branches correspond to
cyanobacterial type III PKSs that do not fall within the highlighted clades.
To experimentally confirm the involvement of the hid gene
cluster in the biosynthesis of the hierridins, we initially
attempted to clone and express the entire cluster in E. coli
BL21 cells, but were unable to observe any protein
overexpression by gel electrophoresis or any differential
metabolite profile by HPLC analysis. Because the hidA gene
encodes a methyltransferase that does not contain an ICMT
explain an analogous halogenation in the biosynthesis of
25
bartoloside D. Because Cyanobium sp. LEGE 06113 is not
axenic, we repeated these searches using the whole set of
contigs (cyanobacterial and non-cyanobacterial) obtained from
the genome (in fact a minimetagenome) sequencing data.
Although ChlA homologues were not found in these data,
three RebH/KtzR homologues were present in alphaproteo-
bacterial contigs (see Table S2 for details). This raises the
possibility for the noncyanobacterial conversion of 1 into 2.
Still, exposure of compound 1 to a lysate of a Cyanobium sp.
LEGE 06113 culture did not lead to any observable conversion
to metabolite 2 (not shown); therefore the nature of the
halogenation event remains unclear.
(
isoprenylcysteine carboxyl methyltransferase) domain, we
envisaged that HidA was not required for decarboxylation of
22
the aromatized product of HidC and that its natural substrate
could be the alkylresorcinol 9 (Figure 2B). The latter
compound is commercially available, which presented an
opportunity to test the ability of HidA to methylate this
substrate and confirm the involvement of the hid BGC in
hierridin production. In fact, E. coli BL21 cells carrying an
Distribution of the hid Gene Cluster among
Cyanobacterial Genomes. The occurrence of hierridin B
in phylogenetically distant cyanobacteria from the Cyanobium
expression vector with an N-His -HidA construct and grown in
6
6
the presence of 9 produced at least two new compounds
compared to an empty vector control (Figure 2C). One of
these compounds presented an HRESIMS-derived m/z value
consistent with the monomethylated alkylresorcinol 10 (Figure
and Leptolyngbya genera suggests that the hid gene cluster
could be widely distributed in the cyanobacterial tree of life.
BlastP searches, using the amino acid sequence of HidC as
query against the NCBI database, revealed an elevated number
of close homologues, found mostly among Cyanobacteria, but
also in Planctomycetes and Proteobacteria. Many of these
homologues were part of hid-like clusters, i.e., containing a
FAD-dependent oxidoreductase and a SAM-dependent meth-
yltransferase. To obtain a clearer picture of the distribution of
these clusters in the Cyanobacteria phylum we retrieved and
aligned the sequences of the top HidC homologues identified
from BlastP searches (we retrieved a total of 149 sequences of
top Blast hits from searches limited and not limited to
cyanobacterial sequences), together with the amino acid
sequences of the alkylresorcinol-generating SrsA, ArsB, and
ArsC type III PKSs. A phylogenetic analysis of the resulting
alignment revealed that several well-supported groups encoded
2
D). The other compound, eluting later in reversed-phase
conditions (Figure 2C,D), was found to correspond to the
dimethylated alkylresorcinol 11 by comparison of its LC-
HRESIMS elution profile with that of synthetic 11 (Figure
2
E), obtained by methylation of 9 using dimethylsulfate.
Curiously, the hid gene cluster does not contain a
halogenase nor could one be found among the genome data
of this cyanobacterium through Blast searches using the
sequences of the flavin-dependent halogenases ChlA (that acts
2
3
on phenolic rings), RebH or KtzR (that chlorinate
2
4
tryptophan), or those of members of other halogenase
2
4
classes as queries. Interestingly, no halogenase was found in
the genome of Synechocystis salina LEGE 06155 that could
E
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
homologues of HidC. A majority of these homologues were
encompassed in a major phylogenetic clade composed of
picocyanobacterial as well as other bacterial type III PKSs that
were part of hid-like gene clusters (Figure 3 and Figure S13).
Within this larger clade, HidC and picocyanobacterial HidC
homologues from strains belonging to the genera Prochlor-
ococcus, Synechococcus, Cyanobium, Vulcanococcus, and Aphano-
thece formed a well-supported independent subclade. In fact,
the picocyanobacterial HidC homologues were encoded within
hidABC gene clusters, while in the sister group of HidC
homologues from proteobacteria and planctomycetes, these
were encoded in gene clusters with either hidABC or hidBC
architectures. The genomes of filamentous cyanobacteria
closely related to the hierridin-producing Leptolyngbya ectocarpi
SAG 60.90 encoded HidC homologues that constituted a
separate clade in the resulting phylogenetic tree; these were
mostly found within hidCAB or hidCA gene cluster
architectures. Overall, hid-like gene clusters seem to be
prevalent within certain cyanobacterial phylogenetic groups,
most notably picocyanobacteria, which contrasts with the small
number of alkylresorcinols reported from these organisms.
which the HRESIMS spectrum was retrieved from LC-HRESIMS
analysis; see Heterologous Expression of hidA in E. coli and
Metabolite Profiling section for details), but the ESI capillary voltage
was −3100 V (capillary temperature was 275 °C), sheath gas was set
at 5 units, and the capillary voltage was set at −40 V with a tube lens
voltage set at −157 V). Both semipreparative and analytical-scale RP-
HPLC for the isolation of hierridin C were performed using a system
composed of a 1525 pump module and a 2485 UV−vis detector
(Waters) and using HPLC-grade solvents. All other solvents used
were ACS grade. GC-MS analysis was performed as reported
6
previously for 1, using a TRACE GC Ultra (Thermo Scientific)
gas chromatograph coupled with a Polaris Q ion trap mass
spectrometer. The system included an AS-3000 autosampler. A ZB-
XLB capillary column (30 m × 0.25 mm × 0.25 μm) from
Phenomenex was selected for chromatographic separation. Ultrapure
−
1
grade helium was used as a carrier gas (flow 1.0 mL min ). For the
synthesis of hierridins, all chemicals were of reagent grade and were
used without further purification: tetradecylmagnesium chloride
solution 1 M in THF (CAS: 110220-87-6, Aldrich) and 2-hydroxy-
3,5-dimethoxybenzaldehyde (97% purity, CAS: 65162-29-0, Fluo-
rochem). All reactions were carried out under an argon atmosphere.
Analytical TLC was carried out on precoated silica gel plates (Merck
6
0 F254, 0.25 mm) using UV light and an ethanolic solution of
phosphomolybdic acid (followed by gentle heating) for visualization.
Biological Material. The cyanobacterium Cyanobium sp. LEGE
CONCLUSIONS
30
■
06113 was obtained from the LEGEcc3. The strain was cultured in 6
1
−1
Our investigations of hierridin-like alkylresorcinol biosynthesis
in cyanobacteria suggest that these metabolites are prevalent in
the world’s oceans. Because many of the strains that harbor hid
gene clusters are picoplanktonic unicellular cyanobacteria with
a major primary production impact in the world’s oceans, the
natural role of these compounds and associated ecological
implications deserve further study. Phenolic lipids structurally
L flasks each with 4 L of Z8 medium supplemented with 20 g L
NaCl, under constant sterile aeration and a light (photon irradiance
−2
−1
∼
30 μmol m s )/dark cycle of 14:10 h. Temperature was
maintained at 26 °C. When cultures reached late exponential phase (6
to 8 weeks postinoculation), cyanobacterial cells were harvested by
centrifugation and rinsed with deionized H O. The biomass was then
2
freeze-dried and kept at −20 °C until extraction.
Isolation of Hierridin C (2). Freeze-dried cells from Cyanobium
sp. LEGE 06113 (16.0 g, dry wt) were extracted repeatedly with a
warm (<40 °C) mixture of CH Cl /MeOH (2:1). The resulting
2
6
related to hierridins interact with biological membranes with
21
known implications in encystment and antibacterial resist-
2
2
2
7
ance, and it is possible that hierridins play similar role(s) in
cyanobacteria, despite their low abundance in Cyanobium sp.
LEGE 06113. On the other hand, also noteworthy is the
similarity of hierridins to the chlorinated hexaphenones DIF
slurry was filtered through Whatman grade 1 qualitative filter paper
(GE Healthcare Life Sciences) and concentrated in a rotary
evaporator to yield 2.58 g of organic extract. This was fractionated
by vacuum liquid chromatography (VLC) using silica gel 60 (0.015−
0.040 mm, Merck KGaA) as stationary phase. Four fractions (A, A1,
A2, and A3) were initially collected using 9:1, 87:13, 17:3, and 41:9
mixtures (n-hexane/EtOAc), respectively, as mobile phases. A steeper
mobile-phase gradient was then applied to the column, from 4:1 (n-
hexane/EtOAc) to 100% EtOAc and then to 100% MeOH, resulting
in eight additional fractions. Fraction A1 (20.4 mg) was further
fractionated by semipreparative RP-HPLC with a Luna C18 column
(10 μm, 250 × 10 mm, Phenomenex) under isocratic conditions
2
8
(
Differentiation-Inducing Factor)-1, DIF-2, and DIF-3 and
29
to MPBD (4-methyl-5-pentylbenzene-1,3-diol), all of which
are associated with developmental signaling in the slime mold
Dictyostelum. Therefore, it is also conceivable that hierridins
have evolved to carry out a signaling or interference role in the
aquatic medium. In any case, our straightforward synthesis of
the hierridins will facilitate future experimental explorations of
these scarce metabolites, in particular of their biological
function and antimalarial properties.
−1
(97% aqueous MeCN, 3 mL min ). A chromatographic peak eluting
at t = 27.1 min was collected and found to contain 1 as a major
R
constituent. The sample was reinjected into the HPLC system fitted
with the above-mentioned column and further separated isocratically,
using 90% MeCN(aq) as mobile phase. Pure 1 (0.5 mg) eluted at
EXPERIMENTAL SECTION
■
General Experimental Procedures. Melting points were
determined using a Stuart Scientific Bibby apparatus and are
uncorrected. UV spectra were acquired on a Spectramax spectropho-
tometer (Molecular Devices). NMR spectra were recorded on a 400
MHz spectrometer (Avance III, Bruker) and on a 600 MHz
spectrometer equipped with a 5 mm cyroprobe (Avance III HD,
Bruker). CDCl or DMSO-d was used as solvent, and the chemical
41.5 min, and an additional peak eluting at 50.7 min was collected.
1
The H NMR (CDCl , 400 MHz) of this latter peak showed a major
3
constituent with a set of resonances that indicated relatedness to
compound 1. An additional round of analytical-scale HPLC with a
Synergi Fusion-RP column (4 μm, 250 × 4.6 mm, Phenomenex) and
using isocratic elution with 85% MeCN(aq) afforded 2 (tR = 38.9
min) as a white, amorphous solid (0.3 mg).
3
6
shifts are reported relative to residual CHCl (δH 7.26, δ 77.16) or
Hierridin C (2): white, amorphous solid; UV/vis (MeOH) λ (log
3
C
max
1
13
DMSO (δH 2.50, δ 39.52), respectively. HRESIMS data of pure 2
ε) 211 (3.0), 249 (2.3), 271 (2.2), 281 (2.2) nm; H and C NMR,
C
−
−
were acquired on an LTQ Orbitrap XL spectrometer, controlled by
LTQ Tune Plus 2.5.5 and Xcalibur 2.1 (Thermo Scientific). The
capillary voltage of the ESI was set to −3000 V. The capillary
temperature was 275 °C. The sheath gas flow rate (nitrogen) was set
to 40 units. The capillary voltage was −35 V, and the tube lens voltage
200 V. Samples were injected at a concentration of approximately
0 μg/mL. The same instrument and analysis were used for synthetic
compound characterization (with the exception of compound 11, for
Table 1; HRMS m/z 397.2523 [M − H] (calcd for C H ClO ,
2
3
38
3
3
97.2515).
Chemical Synthesis Procedures and Characterization Data.
E x p e r i m e n t a l P r o c e d u r e s . 4 , 6 - D i m e t h o x y - 2 - ( 1 -
hydroxypentadecyl)phenol (4). In a round-bottom flask compound
3 (100.5 mg, 0.5516 mmol) was dissolved in anhydrous THF (20
mL) followed by addition of Grignard reagent (tetradecylmagnesium
chloride 1.0 M in THF) (1.1 mL, 1.1 mmol), and the resulting
−
5
F
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
solution was left under stirring for 2 h. The solution was then
concentrated in vacuo, and to the residue was added a saturated
solution of NH Cl (20 mL) and Et O (20 mL). The phases were then
compound 6 (95.0 mg, 0.439 mmol) was reacted with tetradecyl-
magnesium chloride 1.0 M in THF (0.9 mL, 0.9 mmol), and the
resulting solution was left under stirring for 2 h. After the typical
workup the desired compound was then induced to precipitate from
the crude oil using n-hexane. The solid was filtered and washed with
fresh portions of n-hexane, affording 7 (143.8 mg, 79%) as a yellow-
4
2
separated in a separatory funnel, and the organic phase was washed
with deionized H O (3 × 20 mL). The organic extract was dried over
2
anhydrous Na SO , filtered, and concentrated in vacuo. The desired
2
4
1
compound was then induced to precipitate from the crude oil using n-
hexane. The solid was filtered and washed with fresh portions of n-
orange solid. Mp = 73−75 °C; H NMR (CDCl , 400 MHz) δ 7.49
3
H
(br s, OH), 6.46 (s, 1H, H-5), 5.30−5.22 (m, 1H, H-1′), 3.88 (s, 3H,
OCH ), 3.83 (s, 3H, OCH ), 3.47 (br s, 1H, H-1), 1.96−1.82 (m,
hexane, affording 4 (180.5 mg, 86%) as a yellowish solid: mp = 77−78
3
3
1
°
C; H NMR (CDCl , 400 MHz) δ 6.40 (d, 1H, J = 2.8 Hz, H-5),
6
1H, H-2′), 1.81−1.70 (m, 1H, H-2′), 1.58−1.47 (m, 1H, H-3′),
1.40−1.20 (m, 21H, H-3′−H-14′), 0.87 (t, 3H, J = 6.8 Hz, H-15′);
3
H
.37 (d, 1H, J = 2.8 Hz, H-3), 5.89 (s, 1H, OH), 4.85 (dt, 1H, J = 7.6
1
3
Hz, 5.6 Hz, H-1′), 3.85 (s, 3H, OCH ), 3.76 (s, 3H, OCH ), 2.57 (d,
C NMR (CDCl , 101 MHz) δ 148.5 (C, C-4), 146.4 (C, C-6),
3 C
3
3
1
H, J = 5.4 Hz, OH), 1.88−1.71 (m, 2H, H-2′), 1.53−1.21 (m, 24H,
139.4 (C, C-1), 127.5 (C, C-2), 112.3 (C, C-3), 97.5 (CH, C-5), 73.0
1
3
H-3′−H-14′), 0.88 (t, 3H, J = 6.8 Hz, H-15′); C NMR (CDCl , 101
(CH, C-1′), 57.3 (CH , OCH ), 56.5 (CH , OCH ), 36.1 (CH , C-
3
3
3
3
3
2
MHz) δ 153.2 (C, C-4), 147.4 (C, C-6), 137.2 (C, C-1), 130.0 (C,
2′), 32.1 (CH , C-3′), [29.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.7
C
2
2
2
2
C-2), 102.5 (CH, C-3), 98.4 (CH, C-5), 72.0 (CH, C-1′), 56.2 (CH ,
(CH ), 29.5 (CH ), 29.5 (CH ), 25.8 (CH ), C-4′−C-13′], 22.8
3
2
2
2
2
OCH ), 55.9 (CH , OCH ), 37.4 (CH , C-2′), 32.1 (CH , C-3′),
(CH , C-14′), 14.2 (CH , C-15′); HRESIMS m/z 413.2488 [M −
3
3
3
2
2
2
3
−
−
[
(
29.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.7 (CH ), 29.7 (CH ), 29.5
H] (calcd for C H ClO , 413.2464).
2
2
2
2
2
23 38 4
CH ), C-4′−C-13′], 22.8 (CH , C-14′), 14.2 (CH , C-15′);
(Z)-3-Chloro-4,6-dimethoxy-2-(pentadec-1-en-1-yl)phenol (8). In
a round-bottom flask previously flushed with argon, the Wittig reagent
myristyltriphenylphosphonium bromide (227.9 mg, 0.4224 mmol)
2
2
3
−
−
HRESIMS m/z 379.2885 [M − H] (calcd for C H O ,
3
2
3
39
4
79.2854).
n
4
,6-Dimethoxy-2-pentadecylphenol (Hierridin B, 1). In a round-
was dissolved in anhydrous THF (20 mL) and treated with BuLi
bottom flask compound 4 (108.0 mg, 0.2838 mmol) was dissolved in
MeOH (20 mL) and hydrogenated via balloon for 3 h at room
temperature (rt) in the presence of Pd/C 10% (w/w) and H SO 1 M
(0.26 mL, 0.42 mmol) for 15 min. After that, compound 6 (76.2 mg,
0.352 mmol) was added, and the reaction was left stirring overnight.
The resulting solution was concentrated in vacuo, then dissolved in
CH Cl (20 mL), and transferred into a separatory funnel followed by
2
4
(0.2 mL). The suspension was filtered over a Celite pad, and the
2
2
filtrate was concentrated in vacuo. The residue was then chromato-
graphed using CH Cl , affording 1 (39.3 mg, 38%) and 5 (35.8 mg,
the addition of H O. The phases were separated, the organic phase
2
2
2
was washed with H O (3 × 20 mL), dried over anhydrous Na SO ,
2
2
4
3
2%) as white, amorphous solids. Hierridin B (1): mp = 70−72 °C;
and filtered, and the filtrate was concentrated in vacuo. The resulting
residue was chromatographed using hexane/EtOAc (3:1) to afford
1
H NMR (CDCl , 400 MHz) δ 6.36 (d, 1H, J = 2.8 Hz, H-5), 6.29
3
H
1
(
d, 1H, J = 2.8 Hz, H-3), 5.26 (br s, 1H, OH), 3.85 (s, 3H, OCH3),
alkene 8 (112.3 mg, 88%) as a white solid. Mp = 58−60 °C; H NMR
3
7
.76 (s, 3H, OCH ), 2.68−2.53 (m, 2H, H-1′), 1.61 (dt, 2H, J = 15.4,
(CDCl , 400 MHz) δ 6.65−6.34 (m, 3H, H-5 + H-1′ + H-2′), 5.65
3
3
H
.5 Hz, H-2′), 1.38−1.25 (m, 24H, H-3′−H-14′), 0.89 (t, 3H, J = 6.9
(s, 1H, OH), 3.90 (s, 3H, OCH ), 3.85 (s, 3H, OCH ), 2.28 (dd, 2H,
3
3
13
Hz, H-15′); C NMR (CDCl , 101 MHz) δ 152.9 (C, C-4), 146.8
J = 13.1, 7.1 Hz, H-3′), 1.59−1.15 (m, 22H, H-4′−H-14′), 0.88 (t,
13
3
C
(
(
1
C, C-6), 137.7 (C, C-1), 128.9 (C, C-2), 105.9 (CH, C-3), 96.8
3H, J = 6.8 Hz, H-15′); C NMR (CDCl , 101 MHz) δ 148.7 (C,
3
C
CH, C-5), 56.1 (CH , OCH ), 55.9 (CH , OCH ), 32.1 (CH , C-
C-4), 145.3 (C, C-6), 139.3 (C, C-2′), 138.7 (C, C-1), 123.5 (C, C-
3
3
3
3
2
′), 30.2 (CH , C-2′), [30.0 (CH ), 29.8 (CH ), 29.8 (CH ), 29.7
2), 122.4 (CH, C-1′), 114.5 (CH, C-3), 96.6 (CH, C-5), 57.6 (CH ,
2
2
2
2
3
(
CH ), 29.5 (CH ), 22.8 (CH ), C-3′−C-14′], 14.3 (CH , C-15′).
OCH ), 56.6 (CH , OCH ), 34.3 (CH , C-3′), [32.1 (CH , C-2′),
2
2
2
3
3
3
3
2
2
4
,6-Dimethoxy-2-(1-methoxypentadecyl)phenol (5). Mp = 51−
29.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.5
2 2 2 2 2
1
5
5
3 °C; H NMR (CDCl , 400 MHz) δ 6.42 (d, J = 2.8 Hz, 1H, H-
), 6.34 (d, J = 2.8 Hz, 1H, H-3), 6.29 (br s, 1H, OH), 4.42 (dd, 1H, J
(CH ), 29.4 (CH ), C-4′−C-13′], 22.8 (CH , C-14′), 14.3 (CH , C-
− −
3
H
2
2
2
3
15′); HRESIMS m/z 395.2374 [M − H] (calcd for C H ClO ,
2
3
36
3
=
7.7, 5.7 Hz, H-1′), 3.86 (s, 3H, CH ), 3.76 (s, 3H, CH ), 3.30 (s,
395.2359).
3
3
3
H, CH ), 1.87−1.74 (m, 1H, H-2′), 1.73−1.61 (m, 1H, H-2′),
3-Chloro-4,6-dimethoxy-2-(pentadecyl)phenol (Hierridin C, 2).
In a round-bottom flask compound 8 (82.0 mg, 0.207 mmol) was
dissolved in EtOAc (20 mL) followed by addition of catalytic
amounts of Pd/C 10% (w/w), and the resulting suspension was
3
1
.49−1.17 (m, 24H, H-3′−H-14′), 0.87 (t, 3H, J = 6.9 Hz, H-15′);
13
C NMR (CDCl , 101 MHz) δ 153.2 (C, C-4), 147.8 (C, C-6),
3
C
1
8
38.1 (C, C-1), 127.2 (C, C-2), 104.5 (CH, C-3), 98.7 (CH, C-5),
1.2 (CH, C-1′), 57.2 (CH , C-1′-OCH ), 56.1 (CH , OCH ), 55.8
cooled to 0 °C using an ice bath. The system was purged with H via
2
3
3
3
3
(
(
(
CH , OCH ), 36.5 (CH , C-2′), [32.1 (CH ), 29.8 (CH ), 29.8
balloon, and the reaction was left stirring for 1 h until consumption of
the starting material. After that, the suspension was filtered over a
Celite pad and washed with EtOAc. The filtrate was concentrated in
vacuo, and the resulting residue was chromatographed using EtOAc/
hexanes (1:3) to afford hierridin C (82.3 mg, quantitative) as a white
3
3
2
2
2
CH ), 29.8 (CH ), 29.7 (CH ), 29.7 (CH ), 29.5 (CH ), 25.9
2
2
2
2
2
CH ), 22.8 (CH ), C-3′−C-14′], 14.2 (CH , C-15′); HRESIMS m/
2
2
3
−
−
z 393.3036 [M − H] (calcd for C H O , 393.3010).
2
4
41
4
2
-Chloro-6-hydroxy-3,5-dimethoxybenzaldehyde (6). In a round-
1
bottom flask compound 3 (100.1 mg, 0.5494 mmol) was dissolved in
anhydrous THF (10 mL) followed by addition of NCS (73.0 mg,
solid. Mp = 42−44 °C; H NMR (DMSO-d , 600 MHz) δ 8.28 (s,
6
H
1H, OH), 6.66 (s, 1H, H-5), 3.81 (s, 3H, H-8), 3.77 (s, 3H, H-7),
0
(
.550 mmol). To the resulting solution was added concentrated HCl
1 mL), and the reaction developed a strong yellow color. The
reaction was left stirring for 15 min, and after that cold deionized H O
2.72−2.60 (m, 2H, H1′), 1.47−1.38 (m, 2H, H-2′), 1.34−1.16 (m,
1
3
24H, H-3′−H-14′), 0.85 (t, 3H, J = 6.8 Hz, H-15′); C NMR
(DMSO-d , 150 MHz) δ 147.6 (C, C-4), 146.3 (C, C-6), 138.4 (C,
2
6
C
(
50 mL) was added to the reaction and the flask was placed in an ice−
C-1), 128.0 (C, C-2), 112.8 (C, C-3), 96.6 (CH, C-5), 56.6 (CH3,
OCH ), 56.1 (CH , OCH ), [31.2 (CH ), 29.0 (CH ), 29.0 (CH ),
water bath to induce precipitation. The precipitate was collected by
3
3
3
2
2
2
filtration and washed with cold H O. After drying, compound 6 was
29.0 (CH ), 28.9 (CH ), 28.8 (CH ), 28.6 (CH ), 28.2 (CH ), C-
2
2 2 2 2 2
obtained (107.1 mg, 90%) as a yellow solid, with no need of further
2′−C-13′] 27.0 (CH , C-1′), 22.0 (CH , C-14′), 13.9 (CH , C-15′).
2
2
3
1
purifications. Mp = 100−105 °C; H NMR (CDCl , 400 MHz) δ
1,3-Dimethoxy-5-pentadecylbenzene (11). A solution of 5-
pentadecylresorcinol (9, 20.1 mg, 0.0627 mmol) in acetone (4 mL)
was prepared in a round-bottom flask followed by the addition of
K CO (19.0 mg, 0.138 mmol) and dimethyl sulfate (13.0 μL, 0.138
3
H
1
1.75 (s, 1H, OH), 10.44 (s, CHO), 6.84 (s, H-4), 3.92 (s, 3H,
1
3
OCH ), 3.89 (s, 3H, OCH ); C NMR (CDCl , 101 MHz) δ 196.2
3
3
3
C
(
(
5
C, CHO), 148.6 (C, C-3), 147.9 (C, C-6), 147.8 (C, C-5), 116.6
2
3
CH, C-1), 115.8 (CH, C-2), 106.9 (CH, C-4), 58.0 (CH , OCH ),
mmol). The reaction was heated to 50 °C and left under stirring for
24 h. The solvent was then removed under reduced pressure, and the
crude mixture was chromatographed using hexane/EtOAc (3:1) as
eluent, affording compound 11 (16.5 mg, 75%) as a pale yellow solid.
3
3
−
6.8 (CH , OCH ); HRESIMS m/z 215.0120 [M − H] (calcd for
3
3
−
C H ClO , 215.0117).
9
8
4
3
-Chloro-4,6-dimethoxy-2-(1-hydroxypentadecyl)phenol (7).
1
Following the same protocol described for the preparation of 4,
Mp = 36−38 °C; H NMR (CDCl , 400 MHz) δ 6.35 (d, 2H, J =
3
H
G
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
2
.3 Hz, H4 + H6), 6.30 (t, 1H, J = 2.3 Hz, H2), 3.78 (s, 6H, 2 ×
cyanobacterial origin. Annotated nucleotide data have been deposited
in GenBank under accession numbers MH824152 (hid cluster) and
MK268689−MK268691 (genes encoding RebH/KtzR homologues).
Heterologous Expression of hidA in E. coli and Metabolite
Profiling. Cloning of hidA. Forward (NdeI-hidA-v2, 5′ GAG ATC
TCA TAT GGG TGG GCG CTC GGC 3′) and reverse (HindIII-
hidA-NHis, 5′ GAT CCA AGC TTT CAT CCA GCC GGA GCC
3′) primers containing recognition sequences for NdeI and HindIII,
respectively, were used to amplify the hidA gene from the gDNA of
Cyanobium sp. LEGE 06113. The hid gene cluster has a high GC
content, and successful amplification by PCR required an optimized
two-step protocol that used the Phusion High Fidelity DNA
Polymerase (New England Biolabs) with GC buffer and DMSO.
The PCR reactions were prepared in a volume of 20 μL containing 1×
GC buffer, 200 μM dNTPs, 0.75 μM of each primer, 3% DMSO, 0.4U
Phusion polymerase, and 1 μL of template gDNA. The thermal
cycling protocol consisted of an initial denaturation step at 98 °C for
45 s, followed by 35 cycles of a 10 s denaturation step (98 °C) and a
30 s annealing/extension step at 72 °C, before a final extension at 72
°C for 7 min. The resulting PCR mixture was loaded onto a 1%
agarose gel stained with Sybr Safe and visualized under a blue light
transilluminator, before excision of a slice containing the single band
(∼800 bp). The DNA was purified from the agarose using the Illustra
GFX PCR DNA and gel band purification kit (GE Healthcare) and
ligated to a pET-28a vector (Novagen), using T4 DNA ligase
(Takara) after double digestions with NdeI and HindIII (NZYTech).
OCH ), 2.60−2.48 (m, 2H, H1′), 1.66−1.55 (m, 2H, H2′), 1.36−
3
1
3
1
.22 (m, H3′−H14′), 0.89 (t, 3H, J = 6.9 Hz, H15′); C NMR
(
(
(
2
CDCl3, 101 MHz) δ 160.8 (2C, C1 + C3), 145.6 (C, C5), 106.6
2CH, C4 + C6), 97.7 (CH, C2), 55.4 (2CH3, 2 × OCH ), 36.5
CH , C1′), 32.1 (CH , C2′), [31.4 (CH ), 29.9 (CH ), 29.9 (CH ),
C
3
2
2
2
2
2
9.8 (CH ), 29.8 (CH ), 29.8 (CH ), 29.7 (CH ), 29.5 (CH ), C3′−
2
2
2
2
2
C13′], 22.8 (CH , C14′), 14.3 (CH , C15′); HRESIMS m/z
2
3
+
+
349.3100 [M + H] (calcd for C H O , 349.3101).
23 41 2
Biological Assays. Human Cell Line Cytotoxicity Assays. The
following cell lines were exposed to metabolite 2: hepatocellular
carcinoma HepG2, colon adenocarcinoma HT-29, neuroblastoma
SH-SY5Y, breast carcinoma T47D, and normal prostate cell line
PNT2 (Sigma-Aldrich), breast adenocarcinoma and RKO colon
carcinoma (ATCC), and MG-63 osteosarcoma (ATCC). Cells were
6
cultured as previously described. Cells were seeded in 96-well culture
4
−2
plates at a concentration of 10 cells cm . After 24 h of adhesion,
cells were exposed for 24 or 48 h to 99 μL of fresh medium
supplemented with 1 μL of 2 in DMSO to final concentrations of 3
−
1
and 30 μg mL . After incubation, cells were exposed to 10 μL of 0.5
mg mL− MTT. Following exposure, purple-colored formazan salts
were dissolved in 100 μL of DMSO and the absorbance was measured
at 550 nm in a microplate reader (Synergy HT, Biotek). All tests were
run in triplicate.
1
32
Antiplasmodial Assays. Laboratory-adapted P. falciparum 3D7 and
33
Dd2 strains were continuously cultured as previously described,
with minor modifications. Briefly, parasites were cultivated on human
erythrocytes suspended in RPMI 1640 medium supplemented with 25
mM HEPES, 6.8 mM hypoxanthine, and 10% AlbuMAX II (Life
Technologies), at pH 7.2. Cultures were maintained at 37 °C under
an atmosphere of 5% O and 3−5% CO and N and synchronized by
The resulting pET-28a-hidA construct encoded a NHis -tagged
version of HidA and was cloned into E. coli BL21 DE3 chemically
competent cells (NZYTech).
6
Heterologous Overexpression of HidA. The pET-28a-hidA
construct led to overexpression of a ∼35 kDa protein in E. coli
2
2
2
34
sorbitol treatment prior to the assays. Staging and parasitemia were
determined by light microscopy of Giemsa-stained thin blood smears.
The antimalarial activities of 1 and 2 were determined using the SYBR
BL21 DE3 cells in LB medium (expected MW for N-His -HidA is 31
6
kDa) that was most evident at 37 °C without IPTG addition (Figure
S12); therefore such conditions were used in the in vivo assays. These
were carried out using E. coli DE3 cells carrying either the pET-28a-
hidA or the empty pET-28a plasmid that were cultured at 37 °C in a
shaker (200 rpm) until reaching an OD600 of 0.3. At that point, 30 μL
of a 10 mM solution of 5-pentadecylresorcinol (9) in DMSO (or 30
μL of DMSO, as a control) was added to each culture. Following
overnight growth, cell pellets were extracted with EtOAc and MeOH
(pooled and repeated three times for each solvent). After solvent
removal in a rotary evaporator, the resulting extracts were dried and
kept at −20 °C until being used for HPLC separation or LC-
HRESIMS analysis.
35
Green I (Life Technologies) assay as previously described with
modifications. Briefly, early ring stage parasites were tested in
triplicate in a 96-well plate and incubated with drugs for 48 h (37
°
C, 5% CO ), and parasite growth was assessed with SYBR Green I.
2
Each compound was tested in a concentration ranging from 10 to
.0097 μM (0.2% DMSO). Fluorescence intensity was measured with
0
a multimode microplate reader (Triad, Dynex) with excitation and
emission wavelengths of 485 and 535 nm, respectively. The
fluorescence data of parasites treated with 1 or 2 were expressed as
percent survival relative to parasites not exposed to the compounds
and analyzed by nonlinear regression using GraphPad Prism 5
HPLC Analysis of E. coli Extracts, Peak Collection, and LC-
HRESIMS Analysis. The E. coli extracts resulting from the feeding
experiments with substrate 9 were resuspended in MeOH (1 mg
(GraphPad Software). Data were averaged from two independent
experiments for 3D7 and three for Dd2, each done in triplicate (six
and nine determinations in total, respectively). IC50 values and
standard deviations were determined using the same software
package.
−
1
mL ) and analyzed by HPLC-PDA. Separation was carried out in a
Synergi 4u Fusion-RP column (250 × 10 mm, Phenomenex) using a
linear gradient from 20% MeCN(aq) to 100% MeCN over 20 min,
held for 25 additional minutes before returning to the initial
Genome Sequencing and Mining. Genomic DNA was isolated
from a fresh pellet of a Cyanobium sp. LEGE 06113 culture, using a
−
1
conditions. The flow rate was 3 mL min and the injection volume
was 40 μL. Two peaks that were only observed for the extract of E. coli
cells carrying the pET-28a-hidA plasmid and supplemented with
36
CTAB-CHCl /isoamyl alcohol-based protocol. The gDNA was
3
fragmented using the AB Library Builder System (Applied
Biosystems) and subsequently submitted for sequencing on the Ion
Torrent PGM (ThermoFisher Scientific) platform. After sequencing,
compound 9 were collected individually (t = 23.8 and 30.2 min) and
R
pooled from multiple injections (scaled-up to 250 μL). After
evaporation of the eluents in a rotary evaporator, the two collected
37
the reads were evaluated for quality using FastQC25 followed by
38
assembling with SPAdes version 3.7.0.26. Using the Geneious 8.0
software package (Biomatters), the generated contigs were used to set
up a local database. The cyanobacterial type III PKS CylI (GenBank:
AFV96143) amino acid sequence was used to perform a tblastn search
against the local database of Cyanobium sp. LEGE 06113 genome
data. A single hit was found (e-value 2.4e-29), within a 387 kb contig.
The genomic context of these genes was annotated manually, by using
the products of the predicted ORFs as queries in BlastP searches
against the NCBI protein database (Table S1). Because Cyanobium
sp. LEGE 06113 is not axenic, and in spite of the close homology of
the hid cluster genes to cyanobacterial nucleotide data in GenBank,
additional Blastn searches were made in different regions of the
contig, which invariably revealed a top Blast hit sequence of
peak samples were resuspended in a 1:10:4 solution of CHCl /
3
−
1
MeOH/MeCN to a concentration of 1 mg mL and filtered through
0.2 μm syringe filters. The samples, together with a 500 μM standard
of compound 9 prepared in MeOH were then used for LC-HRESIMS
analysis on a system composed of a Dionex Ultimate 3000 HPLC
system coupled to a qExactive Focus Orbitrap mass spectrometer,
controlled by XCalibur 4.1 software (Thermo Fisher Scientific).
Separation was carried out in a Luna C18 column (3 μm, 100 × 3
−
1
mm, Phenomenex) under a flow rate of 0.4 mL min , with a gradient
from 60% solution A (H O + 0.1% formic acid)/40% solution B
2
(MeOH + 0.1% formic acid) to 100% solution B for 7 min and then
held at 100% solution B for 8 additional minutes before returning to
the initial conditions. The UV absorbance of the eluate was
H
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
monitored at 280 nm, and a full MS acquisition was carried out in
positive mode (spray voltage +2600 V, capillary temperature 320 °C,
sheath gas parameter set to 35). To clarify the identity of the peak
to IESD. In addition, this work was also partially supported by
the Associate Laboratory for Green Chemistry- LAQV, which
is financed by national funds from FCT/MCTES (UID/QUI/
collected at t = 30.2 and compare its LC-HRESIMS profile with a
R
50006/2013) and cofinanced by the ERDF under the PT2020
500 μM standard of compound 11 in MeOH, the same instrument
Partnership Agreement (POCI-01-0145-FEDER-007265).
and analysis conditions were used, except for the elution program,
which was run at 0.3 mL min− and used a gradient from 60%
1
solution A (H O + 0.1% formic acid)/40% solution B (MeCN + 0.1%
2
REFERENCES
■
formic acid) to 100% solution B for 7 min and was then held at 100%
solution B for 13 additional minutes before returning to the initial
conditions.
(
1) Tidgewell, K.; Clark, B. R.; Gerwick, W. H. In Comprehensive
Natural Products II Chemistry and Biology; Mander, L., Lui, H.-W.,
Eds.; Elsevier: Oxford, 2010; Vol. 2, pp 141−188.
(2) Yoshizawa, S.; Matsushima, R.; Watanabe, M. F.; Harada, K.;
Ichihara, A.; Carmichael, W. W.; Fujiki, H. J. Cancer Res. Clin. Oncol.
1990, 116, 609−614.
(3) Hong, J. Y.; Luesch, H. Nat. Prod. Rep. 2012, 29, 449−456.
(4) Shih, P. M.; Wu, D.; Latifi, A.; Axen, S. D.; Fewer, D. P.; Talla,
E.; Calteau, A.; Cai, F.; Tandeau de Marsac, N.; Rippka, R.; Herdman,
M.; Sivonen, K.; Coursin, T.; Laurent, T.; Goodwin, L.; Nolan, M.;
Davenport, K. W.; Han, C. S.; Rubin, E. M.; Eisen, J. A.; Woyke, T.;
Gugger, M.; Kerfeld, C. A. Proc. Natl. Acad. Sci. U. S. A. 2013, 110,
Phylogenetic Analysis of the hid Gene Cluster. The
sequences of HidC homologues were retrieved from the GenBank
following a BlastP search and were aligned using the MUSCLE
package in the Geneious software. The resulting alignment was used
for phylogenetic analysis. An approximately maximum-likelihood tree
39
was inferred using FastTree 2 (from within the Geneious software
package). In addition, the genomic context of the corresponding hidC
genes was explored and annotated (using BlastP for the predicted
amino acid sequences of the surrounding open reading frames).
ASSOCIATED CONTENT
Supporting Information
1
053−1058.
■
*
S
(5) Scanlan, D. J.; Ostrowski, M.; Mazard, S.; Dufresne, A.;
Garczarek, L.; Hess, W. R.; Post, A. F.; Hagemann, M.; Paulsen, I.;
Partensky, F. Microbiol. Mol. Biol. Rev. 2009, 73, 249−299.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.8b01038.
(
6) Leao, P. N.; Costa, M.; Ramos, V.; Pereira, A. R.; Fernandes, V.
̃
C.; Domingues, V. F.; Gerwick, W. H.; Vasconcelos, V. M.; Martins,
NMR spectra, HRESIMS and GC-MS data, bioactivity
data, detailed annotation of the hid gene cluster and its
genomic context, annotation of putative halogenases, a
version of the HidC-based phylogenetic tree featuring
taxa/accession numbers, and additional synthesis details
R. PLoS One 2013, 8, No. e69562.
(
4
7) Papendorf, O.; Ko
9, 2383−2386.
(8) Preisitsch, M.; Niedermeyer, T. H. J.; Heiden, S. E.; Neidhardt,
I.; Kumpfmuller, J.; Wurster, M.; Harmrolfs, K.; Wiesner, C.; Enke,
H.; Muller, R.; Mundt, S. J. Nat. Prod. 2016, 79, 106−115.
9) Costa, M.; Sampaio-Dias, I. E.; Castelo-Branco, R.;
Scharfenstein, H.; Rezende de Castro, R.; Silva, A.; Schneider, M.
P. C.; Araujo, M. J.; Martins, R.; Domingues, V. F.; Nogueira, F.;
Vasconcelos, V. M.; Leao, P. N. ChemRxiv 2018.
10) Masi, M.; Cimmino, A.; Boari, A.; Tuzi, A.; Zonno, M. C.;
Baroncelli, R.; Vurro, M.; Evidente, A. J. Agric. Food Chem. 2017, 65,
124−1130.
11) Kikuchi, H.; Ito, I.; Takahashi, K.; Ishigaki, H.; Iizumi, K.;
Kubohara, Y.; Oshima, Y. J. Nat. Prod. 2017, 80, 2716−2722.
12) Niu, S.; Liu, D.; Proksch, P.; Shao, Z.; Lin, W. Mar. Drugs 2015,
̈
nig, G. M.; Wright, A. D. Phytochemistry 1998,
̈
(PDF)
̈
(
AUTHOR INFORMATION
Corresponding Author
E-mail: pleao@ciimar.up.pt.
■
́
̃
*
(
ORCID
1
(
Margarida Costa: 0000-0001-9147-8621
Ivo E. Sampaio-Dias: 0000-0002-9071-6197
Valentina F. Domingues: 0000-0003-3472-849X
(
1
Pedro N. Leao: 0000-0001-5064-9164
̃
3, 2526−2540.
Author Contributions
(13) Gonzalez, A. G.; Bermejo Barrera, J.; M Rodriguez Perez, E.
#
M. Costa, I. E. Sampaio-Dias, and R. Castelo-Branco
Phytochemistry 1992, 31, 1436−1439.
(
(
14) Ellis, J. E.; Lenger, S. R. Synth. Commun. 1998, 28, 1517−1524.
contributed equally to this work.
Notes
The authors declare the following competing financial
interest(s): The antiplasmodial activity of hierridin C has
formed the basis for a patent application (WO2016207869) in
which some co-authors are inventors.
15) Liao, Y.-R.; Kuo, P.-C.; Huang, S.-C.; Liang, J.-W.; Wu, T.-S.
Tetrahedron Lett. 2012, 53, 6202−6204.
16) Hopkins, C. Y.; Chisholm, M. J. Can. J. Res. 1946, 24b, 208−
10.
17) Mahajan, T.; Kumar, L.; Dwivedi, K.; Agarwal, D. D. Ind. Eng.
Chem. Res. 2012, 51, 3881−3886.
18) Costa, M. S.; Costa, M.; Ramos, V.; Leao
Vasconcelos, V.; Martins, R. J. Toxicol. Environ. Health, Part A 2015,
8, 432−442.
(
2
(
(
̃
, P. N.; Barreiro, A.;
ACKNOWLEDGMENTS
■
7
(
We thank CEMUP for HRESIMS and NMR analyses. This
research was supported by the Structured Program of R&D&I
INNOVMAR - Innovation and Sustainability in the Manage-
ment and Exploitation of Marine Resources (reference
NORTE-01-0145-FEDER-000035, Research Line NOVEL-
MAR), funded by the Northern Regional Operational Program
́
19) Sliwins
Latała, A. Toxins 2018, 10, 48.
20) Nakamura, H.; Hamer, H. A.; Sirasani, G.; Balskus, E. P. J. Am.
Chem. Soc. 2012, 134, 18518−18521.
21) Funa, N.; Ozawa, H.; Hirata, A.; Horinouchi, S. Proc. Natl.
Acad. Sci. U. S. A. 2006, 103, 6356−6361.
22) Nakano, C.; Funa, N.; Ohnishi, Y.; Horinouchi, S. J. Bacteriol.
012, 194, 1544−1551.
23) Neumann, C. S.; Walsh, C. T.; Kay, R. R. Proc. Natl. Acad. Sci.
U. S. A. 2010, 107, 5798−5803.
(24) Agarwal, V.; Miles, Z. D.; Winter, J. M.; Eustaq
́
ka-Wilczewska, S.; Maculewicz, J.; Barreiro Felpeto, A.;
(
(
(
NORTE2020) through the European Regional Development
Fund (ERDF). This work was also funded by Fundaca para a
Ciencia e Tecnologia (FCT), under the framework of projects
(
2
(
̧ o
̃
̂
UID/Multi/04423/2013, GHTM - UID/Multi/04413/2013
and PTDC/MAR-BIO/2818/2013, IFCT contract IF/01358/
́
uio, A. S.; El
2014 to P.N.L., and Ph.D. scholarship SFRH/BD/93632/2013
Gamal, A. A.; Moore, B. S. Chem. Rev. 2017, 117, 5619−5674.
I
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
(
25) Leao, P. N.; Nakamura, H.; Costa, M.; Pereira, A. R.; Martins,
̃
R.; Vasconcelos, V.; Gerwick, W. H.; Balskus, E. P. Angew. Chem., Int.
Ed. 2015, 54, 11063−11067.
(
(
26) Stasiuk, M.; Kozubek, A. Cell. Mol. Life Sci. 2010, 67, 841−860.
27) Funabashi, M.; Funa, N.; Horinouchi, S. J. Biol. Chem. 2008,
2
83, 13983−13991.
28) Kay, R. R.; Berks, M.; Traynor, D. Development 1989, 107, 81−
0.
29) Saito, T.; Taylor, G. W.; Yang, J.-C.; Neuhaus, D.; Stetsenko,
D.; Kato, A.; Kay, R. R. Biochim. Biophys. Acta, Gen. Subj. 2006, 1760,
(
9
(
7
(
54−761.
30) LEGE Culture Collection Database.
(
31) Kotai, J. Nor. Inst. Water Res. Blindern Oslo 1972, 11/69, 5.
(32) Alley, M. C.; Scudiero, D. A.; Monks, A.; Hursey, M. L.;
Czerwinski, M. J.; Fine, D. L.; Abbott, B. J.; Mayo, J. G.; Shoemaker,
R. H.; Boyd, M. R. Cancer Res. 1988, 48, 589−601.
(33) Trager, W.; Jensen, J. B. Science 1976, 193, 673−675.
(34) Lambros, C.; Vanderberg, J. P. J. Parasitol. 1979, 65, 418−420.
(35) Machado, M.; Murtinheira, F.; Lobo, E.; Nogueira, F. Ann. Clin.
Med. Microbiol. 2016, 2, 1010.
36) Singh, S. P.; Rastogi, R. P.; Had
Microbiol. Biotechnol. 2011, 27, 1225−1230.
37) Andrews, S. 2010, available online, http://www.bioinformatics.
babraham.ac.uk/projects/fastqc.
38) Bankevich, A.; Nurk, S.; Antipov, D.; Gurevich, A. A.; Dvorkin,
(
̈
er, D.-P.; Sinha, R. P. World J.
(
(
M.; Kulikov, A. S.; Lesin, V. M.; Nikolenko, S. I.; Pham, S.; Prjibelski,
A. D.; Pyshkin, A. V.; Sirotkin, A. V.; Vyahhi, N.; Tesler, G.;
Alekseyev, M. A.; Pevzner, P. A. J. Comput. Biol. 2012, 19, 455−477.
(39) Price, M. N.; Dehal, P. S.; Arkin, A. P. PLoS One 2010, 5,
No. e9490.
J
DOI: 10.1021/acs.jnatprod.8b01038
J. Nat. Prod. XXXX, XXX, XXX−XXX