Aestuaramides, a natural library of cyanobactin cyclic peptides resulting from isoprene-derived Claisen rearrangements

Article abstract of DOI:10.1021/cb300614c

We report 12 cyanobactin cyclic peptides, the aestuaramides, from the cultivated cyanobacterium Lyngbya aestuarii. We show that aestuaramides are synthesized enzymatically as reverse O-prenylated tyrosine ethers that subsequently undergo a Claisen rearrangement to produce forward C-prenylated tyrosine. These results reveal that a nonenzymatic Claisen rearrangement dictates isoprene regiochemistry in a natural system. They also reveal one of the mechanisms that organisms use to generate structurally diverse compound libraries starting from simple ribosomal peptide pathways (RiPPs).

Full text of DOI:10.1021/cb300614c

Letters
pubs.acs.org/acschemicalbiology
Aestuaramides, a Natural Library of Cyanobactin Cyclic Peptides
Resulting from Isoprene-Derived Claisen Rearrangements
†
John A. McIntosh, Zhenjian Lin, Ma. Diarey B. Tianero, and Eric W. Schmidt*
Department of Medicinal Chemistry, College of Pharmacy, University of Utah, Salt Lake City, Utah 84112, United States
*
S Supporting Information
ABSTRACT: We report 12 cyanobactin cyclic peptides, the aestuar-
amides, from the cultivated cyanobacterium Lyngbya aestuarii. We show
that aestuaramides are synthesized enzymatically as reverse O-prenylated
tyrosine ethers that subsequently undergo a Claisen rearrangement to
produce forward C-prenylated tyrosine. These results reveal that a
nonenzymatic Claisen rearrangement dictates isoprene regiochemistry in
a natural system. They also reveal one of the mechanisms that organisms
use to generate structurally diverse compound libraries starting from
simple ribosomal peptide pathways (RiPPs).
atural products are important starting points in the
1
N
process of drug discovery. In part, this is because they
present a diversity of scaffolds and other chemical features. A
major goal of natural products research has been to understand
how enzymes dictate regioselective chemical modifications that
underlie this structural diversity. For instance, many natural
products are alkylated by isoprene equivalents, such as
dimethylallylpyrophosphate (DMAPP), which is installed by a
2
large enzyme class known as prenyltransferases. Prenyltrans-
ferases are known to carry out regioselective prenylation of
aromatic carbons, hydroxyl groups, amines, and thiols, leading
to diverse, bioactive products.
Figure 1. Prenylation in cyanobactin cyclic peptides. (A) Previously
identified cyanobactin natural products are reverse O-prenylated on
Ser/Thr or forward O-prenylated on Tyr. (B) LynF reverse O-
prenylates Tyr, and the product undergoes a spontaneous Claisen
rearrangement in vitro.
Mechanistically, prenyltransferases function by ionization of
prenyl diphosphate, followed by electrophilic alkylation of the
3
substrates. A longstanding hypothesis has been that prenyl-
transferases dictate regioselectivity by positioning substrates.
However, recent biochemical work supports the hypothesis that
the final regioselectivity observed may not always be enzymati-
cally controlled. Instead, in some cases a nonenzymatic Claisen
Purified LynF accepts many different Tyr-containing
substrates, including cyclic peptides with diverse amino acid
4
4
rearrangement may dictate the isolated chemistry. In other
sequences and simple Tyr derivatives, such as boc-Tyr. For all
cases, regioselectivity is at least partly substrate controlled,
of the substrates attempted in in vitro reactions, LynF O-
prenylates Tyr in the reverse orientation, which is an otherwise
unprecedented biochemical modification. Surprisingly, we
found that this reverse-prenylated intermediate undergoes a
spontaneous Claisen rearrangement, even in the absence of
enzyme, to yield a forward C-prenylated moiety that is very
common in natural products (Figure 1B). This result supported
earlier hypotheses that, perhaps, spontaneous Claisen rear-
rangements might underlie many of the observed prenylated
where different substrates exhibit differences in prenylation
5
,6
pattern by the same enzyme.
We have been studying this phenomenon using a family of
small, cyclic peptides known as cyanobactins, which present a
7
−10
diverse series of prenylated metabolites.
Cyanobactins
begin as small proteins synthesized on the ribosome.
Subsequently, they are posttranslationally modified in a variety
of ways. For example, cyanobactins are often prenylated. The
isoprene unit usually results from DMAPP, which is installed in
a forward or reverse orientation on the amino acids Ser, Thr,
Tyr, and possibly Trp by a series of prenyltransferases
Received: November 13, 2012
Accepted: February 15, 2013
4
exemplified by LynF (Figure 1).
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
1
1
natural products. To the best of our knowledge, this may be
the only biological relevant Claisen rearrangement outside of
resulted from the core sequence VCYMPCYP and were present
in sufficient quantity to enable purification and NMR analysis.
Aestuaramide A (1) was assigned a molecular formula of
12
the well-studied shikimate pathway.
+
A weakness of our previous study was that the actual natural
products made by LynF, from the cyanobacterium Lyngbya
aestuarii, were not known. Instead, we predicted the structures
of the natural products based upon the L. aestuarii genome
sequence and, on that basis, synthesized putative substrate
C H N O S (m/z 822.3160 [M + H] ), as determined by
40
51
7
6 3
FTICR and MS/MS analyses. This formula, in tandem with the
predicted linear sequence VCMPCYP, revealed that 1 was
likely isoprenylated. NMR spectroscopy in DMSO-d con-
6
firmed the presence of five spin systems that could be assigned
to the amino acids (Val, Met, Pro, Tyr, and Pro) and one that
could be assigned to isoprene (Figure 3 and Table 1).
Additionally, two sharp singlets (δ = 7.77 and 8.03) in the
aromatic region were consistent with the presence of two Cys-
derived thiazoles. The amino acid sequence was confirmed by a
combination of ROESY, MS/MS, and chemical degradation.
ROESY data showed that Tyr was adjacent to Pro1, while Met
was adjacent to Pro2. MS/MS confirmed that the peptide was
cyclic and matched the proposed peptide sequence precisely
(Figure 4). Further evidence was obtained by carefully
hydrolyzing related peptide 2 and analyzing it by Marfey’s
method. We optimized hydrolysis following a protocol that
4
,7
analogues. Because it is now known that the regioselectivity
5
,6
of at least some prenyltransferases is substrate controlled, it
remained possible that this Claisen rearrangement was
substrate-dependent and not naturally occurring. Therefore,
we sought to isolate these putative new compounds from their
natural source. Here, we report isolation of novel compounds,
aestuaramides A (1) and B (2), and we demonstrate that the
spontaneous Claisen rearrangement of the natural products is
important in their natural structural diversity. These are the first
novel compounds reported from L. aestuarii, a globally
13−15
abundant cyanobacterium.
The L. aestuarii genome sequence contained a cyanobactin
pathway that we predicted would build cyclic peptides using the
ribosomally derived core sequences, ACMPCYP and
16
leaves thiazole intact. All of the predicted amino acids were
present in the molecule, according to this analysis. In particular,
the thiazole-Pro and thiazole-Val connectivities were confirmed
from this result. Additionally, this method revealed that the
configurations of all amino acids were L.
7
VCMPCYP (Figure 2). We obtained the sequenced strain,
NMR spectroscopy unambiguously showed that 1 contained
a reverse O-Tyr isoprene moiety. The ROESY spectrum
revealed correlations between the methyl groups of the reverse-
1
prenyl moiety and the Tyr ε protons. In addition, the H and
13
C chemical shifts of the Tyr aromatic moiety and isoprene
were very similar to those previously reported for reverse-O-
4
prenylated boc-Tyr. Finally, MS/MS data confirmed this
assignment. We previously showed using multiple substrates
that reverse O-prenylated Tyr readily loses isoprene in the
course of mass spectrometry, while the C-prenylated Tyr does
4
not lose isoprene under most conditions. Similarly, 1 readily
lost isoprene even under mild fragmentation conditions (Figure
4
).
On the basis of FTICR-MS, aestuaramide B (2) had a
molecular formula of C H N O S (m/z 754.2153 [M +
35
43
7
6 3
+
H] ), indicating that it was likely an unprenylated analogue of
1
1
. Indeed, comparative analysis of H NMR and MS data
revealed that these compounds were nearly identical, except
that peaks corresponding to isoprene were absent in 2 (Figure
1
4
and Table 1). Although the H NMR spectra of 1 and 2 were
largely identical, the greatest chemical shift differences observed
were localized to the phenol ring of tyrosine (1, δ = 6.91 and
1
Figure 2. Ribosomal synthesis of aestuaramides. The LynE precursor
peptide is shown at top, with the sequence encoding products
underlined and bold. Compounds identified in this study are
numbered. On the basis of biochemical precedent, LynADG would
lead to production of cyclic peptides 2 and 8. Subsequently,
prenylation by LynF would lead to reverse O-prenylated Tyr products
6
.88; 2, δ = 6.83 and 6.67). Moreover, the H NMR spectrum
of 2 contained a sharp singlet consistent with a phenolic OH (δ
1
=
9.33), while this signal was clearly absent in the H spectrum
of 1. To further confirm the amino acid sequences and position
of prenylation of these peptides, we carefully examined their
MS/MS data, which were consistent with the proposed
structures (Figure 4). A comparison of MS/MS spectra for 1
and 2 was particularly telling since the fragments were similar,
but 2 lacked isoprene.
1
and 7.
L. aestuarii PCC 8106, and cultivated it over a six-month
period. This led to gradual accumulation of a cyanobacterial
broth (20 L) from continuous culture, which was rapidly
extracted with methanol to repress isolation artifacts. The crude
extract was analyzed by Fourier transform ion cyclotron
resonance (FTICR) MS/MS, leading to the identification of
In 1:1 D O/CD OD at 50 °C, 1 was underwent the Claisen
2
3
rearrangement over a 21 h period (Figure 5 and Table 2). This
rearrangement could be clearly observed by following the
chemical shifts of the aromatic residues and comparing them to
those for authentic standards of forward C-prenylated and
reverse O-prenylated Tyr that we had previously characterized.
The rate of rearrangement was slower than what we had
1
1
2 cyclic peptides, which we named aestuaramides A−L (1−
2) (Figure 2). Two of these, aestuaramides A and B (1 and 2),
B
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
Figure 3. Summary of key NMR data for 1.
previously observed in aqueous enzyme reaction buffers with
LynF, indicating either an important effect of aqueous buffers
aestuaramides D−F (4−6), corresponding to the sulfoxide
derivative of 1−3, leading to six total cyclic peptides derived
from the VCMPCYP sequence. FTICR-MS provided high-
resolution masses. Further, the MS/MS fragmentation patterns
were virtually identical to those for the nonsulfoxide analogues,
except fragments containing Met are larger by 16 Da. We have
previously observed a similar pattern of oxidized and
4
or substrate dependence in the rearrangement itself. An
identical, C-prenylated compound was also found in the freshly
harvested L. aestuarii extract, and therefore, this compound was
named aestuaramide C (3).
1
Compound 3 had a nearly identical H NMR spectrum to
17
that of 1 in 1:1 D O/CD OD (Supporting Information Figure
nonoxidized Met in other cyanobactins.
2
3
S3K). The major differences observed were in the chemical
Similarly, six related products resulting from the ACMPCYP
precursor peptide sequence were identified. These compounds
(aestuaramides G−L, 7−12) were reverse O-prenylated,
unprenylated, or forward C-prenylated, with or without Met
oxidation. The structures were assigned using a genome
sequencing-based prediction and very similar MS/MS spectra
to those of their NMR-characterized relatives, 1 and 2.
shifts of the aromatic moiety and in the isoprene unit. In
1
contrast to the p-phenol of 1, 3 had three H signals in the
aromatic region, representing a 1,3,4-trisubstituted aromatic
ring. The chemical shifts of these protons (δ = 7.02 (H-6), 6.91
(H-5), and 6.94 (H-8)) were consistent with a dialkylated
phenol. The observed coupling constants showed that an
additional alkyl group was present at H-7 (Supporting
Information Figure S3J). Chemical shifts of the isoprene
moiety had changed from those for a reverse-prenylated
compound (1) to a forward-prenylated compound. In
particular, the isoprene methyls were at δ = 1.54 in 1 and δ
Characterization of RiPP products remains a challenge in
part because low yields are sometimes obtained. Countering
this problem, an advantage of working with RiPPs is that the
peptide sequence is genomically encoded, so that if a fragment
containing the predicted sequence is found with high resolution
by mass spectrometry, then based upon the size of the genome
of the producer, it is probable (or in some cases virtually
certain) that the compound in question is encoded by the RiPP
=
1.79 and 1.82 in 3. By FTICR-MS analysis, 3 and its forward-
C-prenylated relatives exhibit strikingly different fragmentation
patterns than the reverse-O-prenylated series. While the O-
prenylated 1 readily loses C H , this loss is not observed in 3 or
1
8,19
sequence.
With aestuaramides, standard amino acids Pro,
5
8
in any other member of the C-prenylated series. In the course
of analysis, the rearrangement product 3 was spontaneously
oxidized. By LC-FTICR and MS/MS, the C-prenylated and
Met-oxidized 3 was identical with one of the compounds in the
crude extract. The rearrangement and single oxidation product
of 1 was identical by LC-FTICR-MS/MS to compound 6
Tyr, and Met were clearly observed. The main problem lies in
the assignment of posttranslational modifications. In the case of
aestuaramides, biochemical precedent indicated that the
products should be N−C cyclic (presence of LynA and
LynG) and that they should contain thiazoles (LynD and
oxidase). Indeed, in the MS spectra, the major compounds
lacked H O in comparison with the predicted linear peptides.
(
Supporting Information Figure S5 and see below). Aside from
10
3
1
these differences, the only obvious difference in the H NMR
spectra was that the thiazole-2 proton exhibited a downfield
shift from δ = 8.04 to δ = 8.08 during the conversion of 1 to 3,
while the thiazole-1 proton was invariant.
The fragmentation methods used here cannot break the
thiazole−amino acid linkage, such that Val−thiazole, Ala−
thiazole, and Pro−thiazole could be inferred. Fragments for
these predictions were present in all spectra. Moreover, when a
peptide is cyclic rather than linear, the number of possible
A limitation of the above data is that, owing to the amount of
1
3
20
material recovered, C data was only available for compound 1
fragments greatly increases, and indeed, the evidence here
1
and not 2 or 3. Assignments thus relied on H NMR data and
favors cyclization. However, these are hypotheses that require
further testing. Fortunately, in the case of 1, a complete NMR
data set confirmed the amino acid assignments and provided
additional strong support in favor of the presence of thiazoles
FTICR-MS/MS. The MS/MS spectra of these compounds
were identical except for differences directly attributable to
prenylation of tyrosine. Moreover, by NMR this prenylation
could be clearly observed to occur on tyrosine in both 1 and 3.
In addition to 1−3, nine other related products could be
observed in the L. aestuarii crude extract by MS analysis
1
and N−C circularization. Additional H NMR spectra of 2 and
3 served to further support these structures. Comparison of the
fragmentation pattern of 1 to those for 2−12 showed that they
were very similar to 1, with the differences being explained by
the minor change in peptide sequence and in prenylation
pattern, as described above. Because compounds 4−12 were
elucidated solely on the basis of MS and genome sequence,
they are tentatively assigned with a lower level of evidence.
(Figures 1 and S2, Supporting Information). Aestuaramides A−
C (1−3) resulted from posttranslational modification of the
precursor peptide sequence, VCMPCYP. In addition to these
reverse O-prenylated, unprenylated, and forward C-prenylated
compounds, we also observed three further products,
C
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
Table 1. NMR Data for 1 and 2 in DMSO-d6
that all intermediates could be directly observed. The
identification of these natural compounds confirms that this
spontaneous Claisen rearrangement is important in natural
isoprenoid chemistry. Our initial work with purified enzyme
itself suggests that it can be difficult to directly observe this
1
2
13C
1H (J in Hz)
1H (J in Hz)
Met
NH
1
7.02, d (8.6)
7.04, d (8.3)
4
ND
49.4, CH 4.78, m
32.3, CH2 1.95, m; 1.75 m 1.97, m; 1.73 m
transformation, and that the possibility should be considered
2
4.77, m
more seriously in future studies of phenol prenylation.
The presence of both forward- and reverse-prenylated cyclic
peptides in a single organism adds to the chemical complexity
of ribosomal peptide natural products, which are now referred
3
4
30.4, CH2 2.40, m
14.4, CH3 2.04, s
169.4, C
2.41, m
2.08, s
5
a
21
thiazole 1
Val
1
to as RiPPs. In the aestuaramide biosynthetic gene cluster
2
ND
(lyn), a single precursor peptide, LynE, encodes the sequences
for both the VCMPCYP and ACMPCYP series of compounds
a
3
123.8,
CH
7.76, s
9.08, s
7.77, s
(
Figure 2). A heterocyclase, LynD, was proposed to synthesize
NH
1
9.10, d (9.5)
thiazole from Cys, with the help of an oxidase found in
172.8, C
22,23
LynG.
Subsequently, LynA and LynG proteases cleave and
2
54.9, CH 4.75, m
30.6, CH 2.34, m
19.6, CH3 0.58, d (6.5)
19.2, CH3 0.78, d (6.5)
170.4, C
4.75, m
circularize the compounds, in reactions that have been
3
2.38, m
24−26
characterized in related cyanobactin systems.
On the
4
0.60, d (6.7)
0.82, d (6.7)
basis of our work with purified LynF, it is only at this late stage,
5
when peptides have already been macrocyclized, that
Pro 1
Ptyr
1
4
prenylation would take place. We previously showed that
2
61.3, CH 3.08, m
3.10, m
prenylation is perhaps the least efficient enzymatic step in
3
27.8, CH2 1.72, m; 1.50, m 1.74,m; 1.56, m
23.4, CH2 1.64, m; 1.57,m 1.65, m; 1.58, m
48.5, CH2 3.90, m; 3.11, m 3.93, m; 3.11, m
27
cyanobactin pathways. In heterologous expression experi-
4
ments in Escherichia coli, the natural cyanobactin trunkamide
was prenylated by 0, 1, or 2 DMAPP units. Our results here
show that these prenyltransferases can be quite slow even in
their native hosts. Indeed, LynF and relatives exhibit excep-
5
NH
1
7.50, d (7.0)
7.51, d (7.0)
174.5, C
52.4, CH 4.03, m
2
3.98, m
24
tionally relaxed substrate specificity, which is likely directly
3
38.5, CH2 2.98, m; 2.52, m 2.94, dd (13.0, 4);
.52, m
17
related to this slow rate. Finally, Met autoxidation leads to a
2
further six derivatives, for 12 total compounds resulting from a
single precursor peptide. Autoxidation is probably non-
enzymatic (spontaneous) since oxidation increases over time
even in extracted or purified materials.
4
133.2, C
5/6
129.6,
CH
6.91, d (8.0)
6.88, d (8.0)
6.85, d (8.0)
6.69, d (8.0)
7
/8
121.6,
CH
This represents a striking structural diversity from a single
RiPP precursor peptide. Most commonly, only a single product
results from a single precursor peptide, although there are a few
cases in which up to four products result. Twelve natural
products from a single precursor peptide may represent a
record for RiPPs. The production of numerous derivatives from
a single pathway is one of the ways in which variable tailoring
steps allow RiPP pathways to contribute to chemical diversity
in bacteria. Of note, there are two copies of the Val-containing
sequence and one of the Ala-containing sequence encoded on
LynE. As shown above, we noted a greater amount of Val-
containing products in comparison to Ala-containing products.
One possibility is that this ratio directly correlates to a different
abundance of the two species. In cyanobactin precursor
peptides, there are several cases with multiple repeats of
product-coding sequences in a single peptide, although
repetition of the same coding sequence within a peptide
9
9
1
1
1
158.1, C
-OH
0
9.33, s
79.3, C
1/12 26.4, CH3 1.36, s
3
143.9,
CH
6.07, dd (18.0,
11.4)
14
113.6,
CH2
5.12, m
a
thiazole 2
Pro 2
1
2
3
170.7, C
ND
a
124.7,
CH
8.05, s
8.05, s
1
2
3
4
5
ND
58.5, CH 5.86, brd (7.3)
28.8, CH2 2.41, m; 1.96,m
5.85, d (7.3)
2.44, m; 2.04, m
21.4, CH2 1.92, m; 1.74, m 1.92, m; 1.73, m
46.4, CH2 3.67, m; 3.52, m 3.70, m; 3.52, m
a
17
Assignments may be interchanged with one another.
appears to be relatively rare. There are also many cases in
which only a single coding sequence is found in a cyanobactin
precursor peptide. Our current working hypothesis is that
multiple coding sequences within a single peptide result from a
duplication of the entire precursor peptide, followed by
recombination to eliminate nonidentical sequences. In support
of this hypothesis, we have documented several cases in which
multiple precursor peptides (up to nine) are found in a single
Previously, we showed that LynF reverse O-prenylates Tyr in
diverse artificial substrates and that these intermediates
spontaneously rearrange to form the forward C-prenylated
4
derivatives. We proposed that this Claisen rearrangement
might possibly underlie many different types of C-prenylated
phenolic natural products. Here, we show that LynF reverse O-
prenylates its natural substrates and that these rearrange to
form forward C-prenylated products within their natural
environments. With these specific substrates, the Claisen
rearrangement is quite slow and, in fact, much slower than
17
cyanobactin operon. In cases that we have examined, it
appears that only some of these precursors lead to isolable
17
products. It seems likely that this repetition may be more
common than generally realized since it creates an assembly
challenge that can lead to artifacts in shotgun-sequenced
4
with the substrate analogues that we previously used, meaning
D
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
Figure 4. Summary of key MS data. FTICR-MS/MS shows that (A) 1 is O-prenylated, while (B) 2 is not; the MS/MS spectra are otherwise
identical. (C). MS/MS data provide evidence for the fragments shown, providing unambiguous evidence of peptide sequence (see Supporting
Information for full details).
Table 2. Key NMR Data for Spectra Shown in Figure 5
No.
1
3
5
6
7
8
7.20, d (8.0)
7.20, d (8.0)
7.11, d (8.0)
7.11, d (8.0)
1.54, s
6.94, s
7.02, d (7.6)
6.91. d (7.6)
1.82, s
1
1
1
1
1
2
3
4
1.54, s
1.79, s
6.24, dd (17.6, 10.9)
5.28, m
5.35, brs
multiple coding sequences in a single precursor peptide has also
2
9
been found in N−C cyclic peptides from plants.
L. aestuarii is a biologically important, dominant benthic
(
bottom-dwelling) component of shallow bays in oceans
13−15
around the world.
To the best of our knowledge, the
only previously reported products of L. aestuarii are widely
3
0,31
occurring UV-shielding pigments.
Therefore, aestuaramides
represent the first novel compounds from this widespread
source. It has been shown that other cyanobactins are
components of numerous types of cyanobacteria, including
7,8,10,17
species that are dominant components of the plankton.
Cyclic peptides such as aestuaramides may be exceptionally
widespread metabolites in natural ecosystems. The natural roles
of the compounds are largely unknown, but their abundance,
structural diversity, and occasionally potent biological activity in
pharmaceutical screens hints that they may be important in the
biology of cyanobacteria and other organisms.
Figure 5. Claisen rearrangement converting compound 1 to 3. Shown
is the Tyr aromatic region of the H NMR spectrum as compound 1
subjected to heating as described in the methods section; y-axis, 1,
1
time 0, RT; 2, time 0, 50 °C; 3, 0.5 h, 50 °C; 4, 2 h, 50 °C; 5, 19 h, 50
METHODS
°
C; 6, 21 h, 50 °C. At time points 1−4, the starting material 1
■
General Methods. 1H NMR, 13C NMR, and 1H− H COSY
1
dominates the spectrum, while, by time period 6, 1 is completely
absent and has been replaced by 3 (see Supporting Information Figure
S3K for more details).
1
13
spectra were recorded on a Varian INOVA 500 ( H 500 MHz and
125 MHz) NMR spectrometer with a 3 mm Nalorac MDBG probe at
C
3
6
00 K. HSQC and HMBC spectra were recorded on a Varian INOVA
1
13
00 ( H 600 MHz and C 150 MHz) NMR spectrometer equipped
1
13 15
with a 5 mm H[ C, N] triple resonance cold probe with a z-axis
gradient, utilizing residual solvent signals for referencing. High-
resolution mass spectra (HRMS) were obtained using a Bruker
APEXII FTICR mass spectrometer equipped with an actively shielded
genomes. Alternatively, we have previously documented
28
between-strain cyanobactin pathway recombination. This
indicates that the environment also serves as a reservoir for
coding sequence diversity. This is very similar to the
duplication−recombination scenario above, except that multi-
ple precursor peptides would be created in multiple strains,
rather than in a single strain. Interestingly, the phenomenon of
9
.4 T superconducting magnet, an external Bruker APOLLO ESI
source, and a Synrad 50W CO2 CW laser. Onyx monolithic C18
semipreparative columns were used for HPLC, as conducted on a
Hitachi Elite Lachrom System equipped with a Diode Array L-2455
E
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
detector. C18 flash chromatography was performed using silica gel 90
Present Address
†
(
Fluka).
John A. McIntosh: Division of Chemistry and Chemical
Engineering, California Institute of Technology, Pasadena,
California 91125, United States.
Cultivation. L. aestuarii PCC 8106 was obtained from the Pasteur
Culture Collection and maintained in MN marine medium. To
produce aestuaramides, the cyanobacteria were cultivated in MN
marine medium (∼10 mL) in plastic 100 × 15 mm Petri dishes on a
windowsill at RT. MN medium (1 L): artificial seawater to 75%
strength; MgSO ·7H O (40 mg); CaCl (20 mg); NaNO (750 mg);
Notes
The authors declare no competing financial interest.
4
2
2
3
K HPO ·3H O (20 mg); citric acid (3 mg); ferric ammonium citrate
2
4
2
ACKNOWLEDGMENTS
■
(
3 mg); EDTA disodium potassium salt (0.5 mg); NaHCO (2 mg);
3
This work was funded by National Institutes of Health
GM071425 and GM092009. We thank K. Parsawar, C. Nelson,
and J. Cox (U. Utah) for obtaining the mass spectra used in this
study.
trace metal mix (1 mL). Trace metal mix (1 L): H BO (2.86 g);
3
3
MnCl ·4H O (1.81 g); ZnSO ·7H O (222 mg); NaMoO ·2H O (390
2
2
4
2
4
2
mg); CuSO ·5H O (79 mg); Co(NO ) ·6H O (49.4 mg). Seed
4
2
3
2
2
cultures of L. aestuarii were transferred into the Petri dishes, which
were then allowed to grow for a 2-month period.
Extraction and Purification. L. aestuarii was removed from the
media by filtration to yield a cell mass (33.8 g wet weight), which was
REFERENCES
■
(
1) Newman, D. J., and Cragg, G. M. (2012) Natural products as
sources of new drugs over the 30 years from 1981 to 2010. J. Nat. Prod.
flash frozen with liquid N and ground with a mortar and pestle. This
2
material was extracted with methanol (4 × 500 mL). Wet HP-20 resin
Supelco) was added to 75% of the initial crude extract (1.5 L), and
the extract was dried onto the resin by rotary evaporation. The resin
was extensively washed with water and eluted with a methanol gradient
50%, 75%, and 100%), followed by 75% acetone in water. Fractions
7
(
5, 311−335.
2) Heide, L. (2009) Prenyl transfer to aromatic substrates: genetics
and enzymology. Curr. Opin. Chem. Biol. 13, 171−179.
3) Haug-Schifferdecker, E., Arican, D., Bruckner, R., and Heide, L.
2010) A new group of aromatic prenyltransferases in fungi, catalyzing
a 2,7-dihydroxynaphthalene 3-dimethylallyl-transferase reaction. J. Biol.
Chem. 285, 16487−16494.
4) McIntosh, J. A., Donia, M. S., Nair, S. K., and Schmidt, E. W.
2011) Enzymatic basis of ribosomal peptide prenylation in
cyanobacteria. J. Am. Chem. Soc. 133, 13698−13705.
5) Wollinsky, B., Ludwig, L., Hamacher, A., Yu, X., Kassack, M. U.,
and Li, S. M. (2012) Prenylation at the indole ring leads to a
significant increase of cytotoxicity of tryptophan-containing cyclic
dipeptides. Bioorg. Med. Chem. Lett. 22, 3866−3869.
6) Chen, J., Morita, H., Wakimoto, T., Mori, T., Noguchi, H., and
Abe, I. (2012) Prenylation of a nonaromatic carbon of indolylbute-
none by a fungal indole prenyltransferase. Org. Lett. 14, 3080−3083.
7) Donia, M. S., Ravel, J., and Schmidt, E. W. (2008) A global
assembly line for cyanobactins. Nat. Chem. Biol. 4, 341−343.
8) Leikoski, N., Fewer, D. P., and Sivonen, K. (2009) Widespread
occurrence and lateral transfer of the cyanobactin biosynthesis gene
cluster in cyanobacteria. Appl. Environ. Microbiol. 75, 853−857.
9) Schmidt, E. W., and Donia, M. S. (2009) Cyanobactin Ribosomally
Synthesized Peptides-A Case of Deep Metagenome Mining, (Hopwood, D.
A., Ed.), pp 575−596, Chapter 23, Elsevier, New York.
10) Sivonen, K., Leikoski, N., Fewer, D. P., and Jokela, J. (2010)
Cyanobactins-ribosomal cyclic peptides produced by cyanobacteria.
Appl. Microbiol. Biotechnol. 86, 1213−1225.
11) Quillinan, A. J., and Scheinmann, F. (1971) Application of the
(
(
̈
(
(
eluting in 100% methanol or 75% acetone were dried and resuspended
in methanol, to which DMSO was then added yielding 80:20 DMSO/
methanol mixtures that were purified by C18 flash chromatography.
Columns were equilibrated in 1:1 methanol/water, the applied extract
was washed with the same mixture, and the desired products were
eluted with 75% acetonitrile. Finally, C HPLC was performed with a
(
(
1
8
(
−1
flow rate of 3 mL min and a gradient from 1% to 60% acetonitrile
over 60 min, to yield 1 (0.4 mg, 1.6 × 10 % of wet weight) and 2 (0.1
mg, 3.9 × 10 % of wet weight). For FTICR of the crude extract, a
small-scale methanol extract was used with freshly flash-frozen
material.
−
3
−
4
(
LC-MS Analysis of D/L-FDLA Derivatives. Compound 2 (20 μg)
was dissolved in 6 N HCl (200 μL) and heated in sealed ampule vials
at 110 °C for 8 h. The solvent was removed in vacuo. The acid
(
hydrolysate of 2 was dissolved in H O. To a 50 μL aliquot was added 1
2
(
N NaHCO (20 μL) and 1-fluoro-2,4-dinitrophenyl-5-L-leucinamide
3
(
°
1% solution in acetone, 100 μL), and the mixture was heated to 40
C for 50 min. The solution was cooled to RT, neutralized with 1 N
HCl (20 μL), and then dried in vacuo. The residue was dissolved in
MeOH and then analyzed by LC-MS. The analysis of the L- and D-
FDLA derivatives was performed on a EclipseXDB-18 column (150 ×
(
(
4
.6 mm, 5 μm) employing a linear gradient from 20% to 80% CH CN
3
−1
in 0.01 M ammonium formate at 0.5 mL min over 50 min. The
retention times of the D- and L-FDLA derivatives, respectively, were as
(
−
follows: L-Ala, 17.90, 16.50 min, m/z 428 [M − H] ; L-Val-Tzl, 31.80,
Claisen rearrangement to the synthesis of heterocyclic bicyclo2,2,2oc-
tenones: an approach to the morellins based on new biogenetic
suggestions. J. Chem. Soc. D, 966−967.
−
2
−
2
9.50 min, m/z 493 [M − H] ; L-Pro, 24.18, 20.61 min, m/z 408 [M
−
−
H] ; L-Tyr, 17.90, 17.05 min, m/z 474 [M − H] ; L-Pro-Tzl, 26.38,
−
6.28 min, m/z 491 [M − H] .
(
12) Andrews, P. R., Smith, G. D., and Young, I. G. (1973)
Claisen Rearrangement. Compound 1 (0.4 mg) in 1:1 CD OD/
D O was heated to 50 °C in the NMR probe and followed by H
3
Transition state stabilization and enzymic catalysis. Kinetic and
molecular orbital studies of the rearrangement of chorismate to
prephenate. Biochemistry 12, 3492−3498.
1
2
NMR for 21 h.
(
13) Martínez-Alonso, M., Mir, J., Caumette, P., Gaju, N., Guerrero,
R., and Esteve, I. (2004) Distribution of phototrophic populations and
primary production in a microbial mat from the Ebro Delta, Spain. Int.
Microbiol. 7, 19−25.
ASSOCIATED CONTENT
■
*
S
Supporting Information
(
14) Paerl, H. W., Prufert, L. E., and Ambrose, W. W. (1991)
FTICR/IRMPD characterization of 1 and 2; LC-FTICR and
MS/MS characterization; NMR spectra; LC-MS analysis of D/
L-FDLA derivatives; identity of thermally produced aestuar-
amide C. This material is available free of charge via the
Internet at http://pubs.acs.org.
Contemporaneous N fixation and oxygenic photosynthesis in the
2
nonheterocystous mat-forming cyanobacterium Lyngbya aestuarii.
Appl. Environ. Microbiol. 57, 3086−3092.
(15) Rath, J., Mandal, S., and Adhikary, S. P. (2012) Salinity induced
synthesis of UV-screening compound scytonemin in the cyanobacte-
rium Lyngbya aestuarii. J. Photochem. Photobiol., B 115, 5−8.
(16) Fujii, K., Yahashi, Y., Nakano, T., Imanishi, S., Baldia, S. F., and
Harada, K. I. (2002) Simultaneous detection and determination of the
absolute configuration of thiazole-containing amino acids in a peptide.
Tetrahedron 58, 6873−6879.
AUTHOR INFORMATION
■
*
Corresponding Author
E-mail: ews1@utah.edu.
F
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX

ACS Chemical Biology
Letters
(
17) Donia, M. S., and Schmidt, E. W. (2011) Linking chemistry and
genetics in the growing cyanobactin natural products family. Chem.
Biol. 18, 508−519.
(
18) Sudek, S., Haygood, M. G., Youssef, D. T., and Schmidt, E. W.
(
2006) Structure of trichamide, a cyclic peptide from the bloom-
forming cyanobacterium Trichodesmium erythraeum, predicted from
the genome sequence. Appl. Environ. Microbiol. 72, 4382−4387.
(
19) Kersten, R. D., Yang, Y. L., Xu, Y., Cimermancic, P., Nam, S. J.,
Fenical, W., Fischbach, M. A., Moore, B. S., and Dorrestein, P. C.
2011) A mass spectrometry-guided genome mining approach for
natural product peptidogenomics. Nat. Chem. Biol. 7, 794−802.
20) Ng, J., Bandeira, N., Liu, W. T., Ghassemian, M., Simmons, T.
L., Gerwick, W. H., Linington, R., Dorrestein, P. C., and Pevzner, P. A.
2009) Dereplication and de novo sequencing of nonribosomal
peptides. Nat. Methods 6, 596−599.
21) Arnison, P. G., Bibb, M. J., Bierbaum, G., Bowers, A. A., Bugni,
(
(
(
(
T. S., Bulaj, G., Camarero, J. A., Campopiano, D. J., Challis, G. L.,
Clardy, J., Cotter, P. D., Craik, D. J., Dawson, M., Dittmann, E.,
Donadio, S., Dorrestein, P. C., Entian, K.-D., Fischbach, M. A.,
Garavelli, J. S., Goransson, U., Gruber, C. W., Haft, D. H., Hemscheidt,
T. K., Hertweck, C., Hill, C., Horswill, A. R., Jaspars, M., Kelly, W. L.,
Klinman, J. P., Kuipers, O. P., Link, A. J., Liu, W., Marahiel, M. A.,
Mitchell, D. A., Moll, G. N., Moore, B. S., Muller, R., Nair, S. K., Nes, I.
F., Norris, G. E., Olivera, B. M., Onaka, H., Patchett, M. L., Piel, J.,
Reaney, M. J. T., Rebuffat, S., Ross, R. P., Sahl, H.-G., Schmidt, E. W.,
Selsted, M. E., Severinov, K., Shen, B., Sivonen, K., Smith, L., Stein, T.,
Sussmuth, R. D., Tagg, J. R., Tang, G.-L., Truman, A. W., Vederas, J.
C., Walsh, C. T., Walton, J. D., Wenzel, S. C., Willey, J. M., and van der
Donk, W. A. (2013) Ribosomally synthesized and post-translationally
modified peptide natural products: overview and recommendations for
a universal nomenclature. Nat. Prod. Rep. 30, 108−160.
(
22) McIntosh, J. A., Donia, M. S., and Schmidt, E. W. (2010)
Insights into heterocyclization from two highly similar enzymes. J. Am.
Chem. Soc. 132, 4089−4091.
(
23) McIntosh, J. A., and Schmidt, E. W. (2010) Marine molecular
machines: Heterocyclization in cyanobactin biosynthesis. ChemBio-
Chem 11, 1413−1421.
(
24) McIntosh, J. A., Robertson, C. R., Agarwal, V., Nair, S. K., Bulaj,
G. W., and Schmidt, E. W. (2010) Circular logic: Nonribosomal
peptide-like macrocyclization with a ribosomal peptide catalyst. J. Am.
Chem. Soc. 132, 15499−15501.
(
25) Lee, J., McIntosh, J., Hathaway, B. J., and Schmidt, E. W. (2009)
Using marine natural products to discover a protease that catalyzes
peptide macrocyclization of diverse substrates. J. Am. Chem. Soc. 131,
2
(
122−2124.
26) Koehnke, J., Bent, A., Houssen, W. E., Zollman, D., Morawitz,
F., Shirran, S., Vendome, J., Nneoyiegbe, A. F., Trembleau, L., Botting,
C. H., Smith, M. C. M., Jaspars, M., and Naismith, J. H. (2012) The
mechanism of patellamide macrocyclization revealed by the character-
ization of the PatG macrocyclase domain. Nat. Struc. Mol. Biol. 19,
7
(
67−772.
27) Tianero, M. D. B., Donia, M. S., Young, T. S., Schultz, P. G., and
Schmidt, E. W. (2012) Ribosomal route to small-molecule diversity. J.
Am. Chem. Soc. 134, 418−425.
(
28) Donia, M. S., Fricke, W. F., Ravel, J., and Schmidt, E. W. (2011)
Variation in tropical reef symbiont metagenomes defined by secondary
metabolism. PloS One 6, e17897.
(
29) Condie, J. A., Nowak, G., Reed, D. W., Balsevich, J. J., Reaney,
M. J., Arnison, P. G., and Covello, P. S. (2011) The biosynthesis of
Caryophyllaceae-like cyclic peptides in Saponaria vaccaria L. from
DNA-encoded precursors. Plant J. 67, 682−690.
(
30) Abed, R. M. M., Kohls, K., Schoon, R., Scherf, A. K., Schacht,
̈
M., Palinska, K. A., Al-Hassani, H., Hamza, W., Rullkotter, J., and
Golubic, S. (2008) Lipid biomarkers, pigments and cyanobacterial
diversity of microbial mats across intertidal flats of the arid coast of the
Arabian Gulf (Abu Dhabi, UAE). FEMS Microbiol. Ecol. 65, 449−462.
(
31) Balskus, E. P., Case, R. J., and Walsh, C. T. (2011) The
biosynthesis of cyanobacterial sunscreen scytonemin in intertidal
microbial mat communities. FEMS Microbiol. Ecol. 77, 322−332.
G
dx.doi.org/10.1021/cb300614c | ACS Chem. Biol. XXXX, XXX, XXX−XXX