Catalytic Promiscuity in the Biosynthesis of Cyclic
Peptide Secondary Metabolites in Planktonic Marine
Cyanobacteria
The MIT Faculty has made this article openly available. Please share
how this access benefits you. Your story matters.
Citation
Li, B. et al. “Catalytic Promiscuity in the Biosynthesis of Cyclic
Peptide Secondary Metabolites in Planktonic Marine
Cyanobacteria.” Proceedings of the National Academy of
Sciences 107.23 (2010): 10430–10435. CrossRef. Web.
As Published
http://dx.doi.org/10.1073/pnas.0913677107
Publisher
National Academy of Sciences (U.S.)
Version
Final published version
Accessed
Fri May 27 01:21:29 EDT 2016
Citable Link
http://hdl.handle.net/1721.1/77585
Terms of Use
Article is made available in accordance with the publisher's policy
and may be subject to US copyright law. Please refer to the
publisher's site for terms of use.
Detailed Terms
Catalytic promiscuity in the biosynthesis of
cyclic peptide secondary metabolites in
planktonic marine cyanobacteria
Bo Lia,1, Daniel Sherb,1, Libusha Kellyb, Yanxiang Shic, Katherine Huangb, Patrick J. Knerrc,
Ike Joewonod, Doug Rusche, Sallie W. Chisholmb,2, and Wilfred A. van der Donka,c,d,2
a
Department of Biochemistry, cDepartment of Chemistry, and dthe Howard Hughes Medical Institute, University of Illinois at Urbana-Champaign, 600 S.
Mathews Avenue, Urbana, IL 61801; bDepartment of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139;
and eJ. Craig Venter Institute, Rockville, MD 20850
Edited by Craig A. Townsend, Johns Hopkins University, Baltimore, MD, and accepted by the Editorial Board April 17, 2010 (received for review
December 7, 2009)
Our understanding of secondary metabolite production in bacteria
has been shaped primarily by studies of attached varieties such as
symbionts, pathogens, and soil bacteria. Here we show that a
strain of the single-celled, planktonic marine cyanobacterium Prochlorococcus—which conducts a sizable fraction of photosynthesis
in the oceans—produces many cyclic, lanthionine-containing
peptides (lantipeptides). Remarkably, in Prochlorococcus MIT9313
a single promiscuous enzyme transforms up to 29 different linear
ribosomally synthesized peptides into a library of polycyclic,
conformationally constrained products with highly diverse ring
topologies. Genes encoding this system are found in variable abundances across the oceans—with a hot spot in a Galapagos hypersaline lagoon—suggesting they play a habitat- and/or communityspecific role. The extraordinarily efficient pathway for generating
structural diversity enables these cyanobacteria to produce as
many secondary metabolites as model antibiotic-producing bacteria, but with much smaller genomes.
lantibiotic ∣ Synechococcus ∣ combinatorial biosynthesis ∣ Global Ocean
Survey metagenome
S
econdary metabolites are among the most functionally and
structurally diverse molecules in nature and play critical roles
in a wide variety of processes such as metal transport, cell-cell
communication, and chemical defense (1). Two major pathways
exist for the production of peptide secondary metabolites—
ribosomal and nonribosomal. In the former case, the resulting
ribosomally produced peptide is often modified by cognate
enzymes to produce the mature metabolite. Recent genomeenabled studies have shown that many classes of peptide natural
products initially suspected to be of nonribosomal origin are in
fact gene-encoded (e.g., refs. 2–6) and that structural motifs
previously thought to be found only in nonribosomal secondary
metabolites are also found in ribosomally synthesized compounds
(7). Clearly, synthesis of secondary metabolites via the ribosome
turns out to be much more widespread than originally anticipated.
Lanthionine-containing peptides are a family of ribosomally
synthesized microbial secondary metabolites characterized by intramolecular thioether cross-links formed between a dehydrated
Ser or Thr residue (dehydroalanine or dehydrobutyrine, respectively) and a Cys residue (Fig. 1A). Most of the known lanthionine-containing peptides have antimicrobial activity and are
referred to as lantibiotics (8). Family members with other, often
unknown, functions are termed lantipeptides (9). Their precursors are encoded by short genes, generically termed lanAs.
The precursor LanA peptides translated from these genes contain an N-terminal leader peptide that is removed in the final step
of biosynthesis (10); a C-terminal core peptide is transformed
into the mature natural product through a dehydration and cyclization process (Fig. 1A), which is catalyzed for class II lantibiotics
by a bifunctional synthetase generically termed LanM (11).
10430–10435 ∣ PNAS ∣ June 8, 2010 ∣ vol. 107 ∣ no. 23
In most lantibiotic-producing bacteria, the LanM lanthionine
synthetase modifies only a single LanA precursor peptide. However, LanM enzymes reveal low substrate specificity under
laboratory conditions, a property that has been exploited for
the generation of variants of the natural products (12–14). Wondering whether organisms in nature have taken advantage of this
substrate promiscuity to generate natural product libraries with a
single streamlined biosynthetic toolset, we conducted a search of
the sequenced bacterial genomes for cases where a single bacterium can produce multiple lanthionine-containing peptides using
the same LanM enzyme. We discovered that the genomes of several strains of marine Prochlorococcus and Synechococcus contain
multiple lanA-like genes but only a single lanM-like gene. This
finding is of interest not only because these two cyanobacteria
account for as much as half of the chlorophyll—i.e., photosynthetic capacity—in the tropical and subtropical oligotrophic
oceans (15), but also because their “lifestyle” is very different
from that of microbes known to produce these types of compounds; they are single-celled and free-floating and live in a very
dilute habitat where the function of secondary metabolites is not
readily apparent. Using a cultured strain of Prochlorococcus and
available metagenomic databases from the oceans, we asked
whether Prochlorococcus can indeed utilize this system to produce a diverse array of lantipeptides, whether there is evidence
for this capability in ocean metagenomic databases, and whether
it is found more commonly in specific oceanic regions or habitats
than in others.
Results
Putative Lantibiotic Gene Clusters in Cyanobacteria. By using the
mersacidin synthetase gene mrsM from Bacillus sp. strain HIL
Y-85 as a query, a single lanM gene was found in the genomes
of two closely related strains of Prochlorococcus (MIT9313 and
MIT9303) and in one distantly related (16) strain of marine
Synechococcus (RS9916) (noted also by ref. 17). The genomes
of these strains contained 29 (Fig. 1B), 15, and 10 (Fig. S1)
putative lanA genes, respectively. We focused our work on
Prochlorococcus MIT9313, because this strain contains the largest
Author contributions: B.L., D.S., S.W.C., and W.A.v.d.D. designed research; B.L., D.S., Y.S.,
P.J.K., and I.J. performed research; K.H. and D.R. contributed new reagents/analytic tools;
B.L., D.S., L.K., Y.S., P.J.K., S.W.C., and W.A.v.d.D. analyzed data; and B.L., D.S., L.K., S.W.C.,
and W.A.v.d.D. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission. C.A.T. is a guest editor invited by the
Editorial Board.
1
B.L. and D.S. contributed equally to this work.
2
To whom correspondence may be addressed. E-mail: chisholm@mit.edu or vddonk@
illinois.edu.
This article contains supporting information online at www.pnas.org/lookup/suppl/
doi:10.1073/pnas.0913677107/-/DCSupplemental.
www.pnas.org/cgi/doi/10.1073/pnas.0913677107
BIOCHEMISTRY
Fig. 1. Enzymatic transformations and biosynthetic genes for lantipeptide synthesis in Prochlorococcus MIT9313. (A) Dehydration and cyclization reactions in
lantibiotic/lantipeptide biosynthesis. (B) Genomic clusters in MIT9313 encoding ProcM and multiple ProcA peptides, their nomenclature, and locations on the
genome. (C) High level of conservation in the N-terminal leader sequence and hypervariability of the C-terminal core peptide of ProcAs. The GG/GA cleavage
site is marked by an arrow.
number of putative lanA genes. Its genome contains one gene
encoding a lanM homolog, designated procM, which is located
in a cluster with seven genes that encode putative substrate peptides. Twenty-two additional procA genes are found elsewhere in
the genome, 20 of which are clustered together in three other
regions on the genome (Fig. 1B). We named the lanA-like genes
procAs, followed by the cluster number on the genome and the
order of the specific procA gene within this cluster (e.g., procA1.
1-procA1.7; see Fig. 1B). The products of the procA genes display
very high sequence identity in their putative leader peptides but
strikingly high diversity in their putative core peptides, the part of
the molecule that will form the mature product (Fig. 1C). The
leader and core peptides are separated by a GG or GA cleavage
Li et al.
motif, used by ABC transporters with an N-terminal cysteine protease domain to remove leader peptides during class II lantibiotic
secretion (18). The genome of MIT9313 encodes one homolog of
these transporters (Fig. S1). Intriguingly, not a single pair of core
peptides shows detectable sequence identity, and their length
varies greatly (from 13 to 32 amino acids). Furthermore, collectively the core peptides contain serines and threonines, the amino
acids that are dehydrated in lantibiotic substrates, at every position from the 1st residue to the 24th residue following the leader
peptide (Fig. 1C). Similarly, cysteine residues are found in every
position from residue 1 of the core peptide to residue 23. For a
single ProcM protein to process all 29 peptides, it would have to
be extraordinarily promiscuous.
PNAS ∣
June 8, 2010 ∣
vol. 107 ∣
no. 23 ∣
10431
In Vitro Dehydration and Cyclization Activity. To test whether one
ProcM enzyme could possibly process these 29 highly diverse core
peptides, seven procA genes located in the vicinity of procM, ten
procA genes located distally on the genome, and procM itself were
cloned and heterologously expressed in Escherichia coli as
N-terminally His6 -tagged fusion proteins. Remarkably, incubation of each of the 17 purified peptides with ProcM in the presence of ATP and Mg2þ resulted in efficient dehydration (Fig. 2
and Figs. S2 and S3) as determined by MALDI-TOF MS after
proteolytic removal of most of the leader peptide (for detailed
experimental procedures, see the SI Appendix). For many substrates, the number of dehydrations equaled the number of
Ser/Thr present in the core peptide (Figs. 1C and 2). For some
substrates such as ProcA1.2, one of the Ser/Thr in the core peptide escapes dehydration, which is not uncommon in lanthioninecontaining peptides (18). Analysis by tandem electrospray ionization MS (ESI-MSMS) showed regions in the ProcA products that
are resistant to fragmentation (Fig. 3A), suggestive of thioether
cross-links (11). To provide further evidence that these rings consist of lanthionine linkages, larger scale generation of the product
of ProcA2.8 was carried out, most of the leader peptide was
removed by using protease GluC, and the C-terminal proteolytic
fragment corresponding to the core peptide was purified by
HPLC. Complete hydrolysis of the peptide and derivatization
of the resulting amino acids as described previously (19), followed
by GC-MS analysis and comparison with a meso-lanthionine standard confirmed the lanthionine linkages (Figs. S4–S6). It is likely,
given the high sequence conservation in the leader peptides, that
all 29 ProcA peptides are substrates for ProcM. We propose the
name prochlorosin (Pcn) for the resulting family of compounds
produced by different Prochlorococcus strains, the nomenclature
following that of the procA genes described above, with the
addition of the strain name—e.g., prochlorosin 1.1 (MIT9313).
Determination of Ring Topologies. A subset of ProcA products
modified by ProcM was further investigated to establish the topology of thioether cross-links. After removal of most of the
leader peptide by using the commercial proteases LysC, trypsin,
or GluC, the modified ProcAs were analyzed by ESI-MSMS. For
the products of ProcA1.1, ProcA2.8, and ProcA4.3 modified by
ProcM (Fig. 3A and Figs. S7 and S8), the observed fragmentation
pattern demonstrates ring systems that do not overlap. In contrast, the constellation of Cys and Ser/Thr residues in ProcA1.7
requires overlapping rings, and indeed no fragmentation was observed between residues 8 and 22 (Fig. 3A). To determine the
connectivity of the thioether rings, a series of mutant peptides
of ProcA1.7 was generated to disrupt each individual ring by mutation of a Ser or Thr to Ala (Fig. S9). ESI-MSMS analysis of
these ProcA1.7 mutants after modification by ProcM support
the ring structure shown in Fig. 3B. Overlapping rings were also
found in prochlorosin 1.5, 2.11 and 3.3 (Fig. 3B and Figs. S10–
S12). We cannot rule out the presence of minor products with
alternative ring topologies for which the intensities of the fragment
ions would fall below our detection limit.
Production of Prochlorosins in Vivo. We next investigated whether
Prochlorococcus MIT9313 makes prochlorosins under laboratory
conditions, by using several approaches. procM and several
procA genes are transcribed by exponentially growing MIT9313
cultures (Fig. 4A). Analysis of previously published whole genome expression studies shows that their transcription is downregulated under nitrogen starvation (Fig. S13) but is not responsive to phosphate (20) or iron (21) starvation. These findings
show that these genes are integrated into the MIT9313 transcriptional circuitry, can respond to changes in environmental conditions, and hence are likely functional. Furthermore, several
prochlorosins were detected in spent media of late-exponential
stage MIT9313 cells, showing that they are produced. ESI-MSMS
Fig. 2. In vitro dehydration of ProcAs by ProcM. MALDI-TOF mass spectra of representative ProcAs modified by ProcM in vitro and subsequently proteolytically
cleaved (LysC for ProcA1.2, GluC for all others) to remove part of the leader peptide (solid spectra). ProcM treatment was omitted in the dashed spectra. The
sequence of each ProcA after protease cleavage is shown, and the number of dehydrations is determined by the mass difference between ProcM-treated and
control samples (each dehydration ¼ −18 Da). The control spectrum for ProcA1.7 does not include the parent peak because GluC cleavage of the unmodified
core peptide results in two smaller peptides outside of the window shown.
10432 ∣
www.pnas.org/cgi/doi/10.1073/pnas.0913677107
Li et al.
Fig. 3. In vitro cyclization of ProcAs by ProcM. (A) ESI-MS/MS analysis of ProcA1.1 G-1E (Upper) and ProcA1.7 (Lower) treated successively with ProcM and
GluC. The labeled fragment ions support two nonoverlapping thioether rings
for prochlorosin 1.1 (MIT9313) and show that, in prochlorosin 1.7 (MIT9313),
Dhb1 and Dhb5 are not involved in rings. (B) Proposed structures of Pcn1.1,
1.7, 2.8, 3.3, 4.3, and 2.11 (MIT9313); see SI Appendix. Arrows illustrate the
start of the putative core peptide. Asterisks indicate prochlorosins containing
residues from the leader peptide (underlined).
analysis of in vivo produced prochlorosin 2.1 (MIT9313) showed
two key fragment ions (y′′17 and y′′18; for nomenclature, see
Scheme S1) that agree with the fragmentation observed for
the compound produced in vitro (Fig. 4B). Similar fragmentation
patterns for the in vivo and in vitro produced compounds were
also observed for prochlorosin 2.11 and 3.2 (MIT9313)
(Fig. S14). Thus, these three prochlorosins are produced in vivo,
the leader cleavage site is located at the anticipated motif, and the
in vitro prepared compounds have the same ring pattern as the
natural products.
Distribution of Prochlorosin Gene Clusters in the Global Ocean Metagenomic Survey. The recent global ocean metagenomic survey
(GOS), in which Prochlorococcus and Synechococcus genes
are relatively abundant (22), offers an opportunity to assess
Li et al.
the prevalence of lanthionine biosynthetic genes in these and
other microbes and examine their geographic distributions. To
this end, the GOS database was searched by using tBLASTn, with
the MIT9313 procM sequence as a query. We also analyzed the
occurrence of seven single-copy, reference genes from Prochlorococcus and Synechococcus at each station (Table S1), as a metric
for the number of genome equivalents from these groups captured in the sample (23). We found 21 distinct procM-like
sequences predicted to originate from either Prochlorococcus
or Synechococcus (the resolution of the phylogenetic tree does
not enable us to determine from which of these two genera
the sequences originate; Fig. S15). We also found eight additional
procM-like sequences, some of which are similar to those found in
the genomes of proteobacteria, firmicutes, and actinobacteria
(17). By using the conserved leader peptides of the procA genes
from MIT9313 as a query, we found 152 distinct procA like sequences in GOS (Table S2). As many as three different procA
genes were found in tandem on the same GOS read. Together
with the high ratio of procA to procM genes detected, this suggests
that, among the Prochlorococcus and Synechococcus cells
sampled by GOS, the genetic capability for promiscuous biosynthesis of lantipeptides is the norm rather than an exception.
We assembled the sequencing reads from one GOS site (GOS33,
see below) into contigs, two of which (10–14 kbp) encode, in
addition to procA and/or procM, putative lantipeptide ABC
PNAS ∣
June 8, 2010 ∣
vol. 107 ∣
no. 23 ∣
10433
BIOCHEMISTRY
Fig. 4. In vivo production of prochlorosins. (A) Transcription of procM and
procA genes by a log-phase culture of MIT9313. RT-reverse-transcriptase
added (+) or omitted (−). rnpB is a control housekeeping gene. (B) Production
of prochlorosin 2.1 (MIT9313) in vivo. The upper panel depicts ESI-MSMS
analysis of ProcA2.1 modified in vitro by ProcM and digested with GluC;
the lower panel shows ESI-MSMS analysis for prochlorosin 2.1 (MIT9313) from
the spent media of a late-exponential stage culture of MIT9313. Fragment
ions are indicated. The b2 and y′′28-y′′31 ions in the upper panel are derived
from fragmentation in a stretch of residues (underlined) remaining from the
leader peptide after GluC cleavage. Therefore, they are not observed in
the product isolated from spent medium. Site directed mutagenesis studies
suggest that Ser7 of Pcn2.1 (MIT9313) escapes dehydration (Fig. S18).
Fig. 5. Prochlorococcus and Synechococcus lantipeptide genes in the Global
Ocean Survey database. (A) Lantipeptide biosynthetic clusters in assembled
metagenomic reads from the hypersaline lagoon site, GOS33. Blue—procM,
red—procA, green—ABC transporters (red *—peptidase domain), light green
—tolC, yellow—response regulator. (B) Distribution of procM- and procA-like
genes in the GOS sites. Upper: Average number of core gene copies (normalized for gene length to 1,000 aa and to the database size at each GOS site in
Mb), representing the relative abundance of Prochlorococcus and Synechococcus cells at each site. Lower: Ratio of procM and procA gene copies to core
genes at each site. Wedges denote the location of the sampling sites in the
bar graph. The sites of isolation of the Prochlorococcus MIT9313 and MIT9303
and Synechococcus RS9916 strains that contain lantipeptide genes are shown
as yellow triangles.
transporters, suggesting that at least some products of this system
are secreted (Fig. 5A).
If we assume that procM is found in only one copy per genome,
we roughly estimate that the capacity to produce lantipeptides is
shared by 0.5–5% of the Prochlorococcus and Synechococcus genomes sampled by the GOS database (SI Appendix). This percentage is likely an underestimate of the fraction of cells at these
locations that contain the genes, because GOS samples were
collected from surface waters, which are dominated by highlight-adapted Prochlorococcus (15); the only strains from our culture collection that contained these genes, MIT9313 and 9303,
are low-light-adapted, and this ecotype is abundant only in deeper waters (15). procM and procA-like genes were detected in the
Eastern Tropical Pacific, around the Galapagos Islands, and in
the Eastern Indian Ocean (Fig. 5B). However, the highest frequency, compared to reference genes, was found at a hypersaline
lagoon in the Galapagos (GOS33), where up to 25% of the total
Prochlorococcus and Synechococcus cells sampled carried the
procM gene.
Discussion
Several different forms of natural combinatorial biosynthesis
have been recognized as contributing to the diversity in secondary
metabolites. For instance, nonribosomal peptide synthetases and
polyketide synthases create secondary metabolites from multienzyme “assembly lines” composed of a series of functional modules
acting sequentially, with new domain combinations leading to the
generation of unique products (24). In addition, enzymes such as
the isoprenoid cyclases can produce multiple products from a
10434 ∣
www.pnas.org/cgi/doi/10.1073/pnas.0913677107
single substrate to generate a group of diverse compounds (25,
26). Alternatively, a spectrum of structurally related secondary
products produced by different members of a microbial consortium can contribute to the defensive capability of the consortium
as a whole [e.g., cyanobactins (27)]. The lantipeptide producing
machinery described here for prochlorosin biosynthesis by
Prochlorococcus MIT9313 is unique in that it shows that a single
organism can produce as many as 29 different secondary metabolites from distinct gene-derived precursors by using one promiscuous biosynthetic enzyme.
Remarkably, Prochlorococcus MIT9313 has the genetic
capacity to produce as many secondary metabolites as the model
antibiotic-producing actinomycetes Streptomyces coelicolor and
Streptomyces avermitilis (28, 29), but with a genome less than
one-third in size (2.4 MB compared with 8.6–9.0 MB) (30).
Rather than maintaining a large (∼20–60 KB) gene cluster for
the synthesis of each secondary metabolite, Prochlorococcus
and Synechococcus maintain a single biosynthetic enzyme gene
and a series of short, ∼300 bp, precursor genes to generate a
topologically diverse set of conformationally constrained peptides in a remarkably effective two-step reaction sequence. When
taking into account all of the cyanobacterial genomes analyzed
here and in the GOS dataset, we have detected 150 unique
putative lantipeptide sequences illustrating the remarkable diversity of products that can be generated by this pathway (see the
SI Appendix). For instance, the ring patterns determined for
all six lantipeptides analyzed in this study (Fig. 3) are different
from the ∼20 previously documented lantibiotic/lantipeptide ring
topologies, which originate from a multitude of different organisms (Fig. S16).
How can the structural diversity manifested in prochlorosins
be introduced into a series of linear peptides by a single enzyme?
The number and position of Ser/Thr and Cys residues varies
greatly in the 17 investigated core peptides, and yet all of these
peptides are dehydrated and cyclized. As depicted in Fig. 3B, the
rings have very different sizes and are formed from Cys residues
located both C- and N-terminal to the dehydro amino acid
partners with which they react. It is difficult to envision an active
site geometry of ProcM that would actively catalyze each of these
cyclizations, and we hypothesize that only a subset of rings are
generated by the enzyme, preorganizing the resulting product
for further regioselective, nonenzymatic cyclization; such nonenzymatic cyclization has been demonstrated with synthetic
peptides (31). Notably, the modification process is likely guided
by the conserved leader peptide, because ProcM did not modify
truncated ProcA substrates lacking the leader peptides (Fig. S17).
The promiscuity demonstrated here may be rivaled by the relaxed
substrate specificity of P450 enzymes that decorate terpenoid
products in plants (32). However, for those systems, the molecular scaffolds have already been put in place by terpenoid
cyclases, whereas ProcM generates the topologically diverse ring
structures itself.
The promiscuous nature of the posttranslational modification
machinery is poised to allow rapid evolution of diversity, because
any nonsynonymous mutation in the gene encoding a peptidederived product potentially results in a new cyclic product. Such
evolvability stands in contrast to systems where the diversity is
generated by an enzyme, in which only a limited subset of nonsynonymous mutations (those that affect the enzymatic activity
without perturbing the enzyme structure) result in the production
of a new compound. Combined with the ease at which short precursor peptides are thought to duplicate within a genome (3, 33),
rapid diversification is facilitated as has been shown for the conopeptide family (33), and this may have been the mechanism
which resulted in 29 different procA genes being present in
MIT9313. It is noteworthy that the genomic regions in MIT9313
that contain the procA genes also contain phage integrases, transposases, and fragments of the latter (Fig. S1); genomic cluster 3
Li et al.
1. Challis GL, Hopwood DA (2003) Synergy and contingency as driving forces for the
evolution of multiple secondary metabolite production by Streptomyces species.
Proc Natl Acad Sci USA 100(Suppl 2):14555–14561.
2. Schmidt EW, et al. (2005) Patellamide A and C biosynthesis by a microcin-like pathway
in Prochloron didemni, the cyanobacterial symbiont of Lissoclinum patella. Proc Natl
Acad Sci USA 102:7315–7320.
3. Hallen HE, Luo H, Scott-Craig JS, Walton JD (2007) Gene family encoding the major
toxins of lethal Amanita mushrooms. Proc Natl Acad Sci USA 104:19097–19101.
4. Lee SW, et al. (2008) Discovery of a widely distributed toxin biosynthetic gene cluster.
Proc Natl Acad Sci USA 105:5879–5884.
5. Wieland Brown LC, Acker MG, Clardy J, Walsh CT, Fischbach MA (2009) Thirteen posttranslational modifications convert a 14-residue peptide into the antibiotic thiocillin.
Proc Natl Acad Sci USA 106:2549–2553.
6. Kelly WL, Pan L, Li C (2009) Thiostrepton biosynthesis: Prototype for a new family of
bacteriocins. J Am Chem Soc 131:4327–4334.
7. McIntosh JA, Donia MS, Schmidt EW (2009) Ribosomal peptide natural products:
Bridging the ribosomal and nonribosomal worlds. Nat Prod Rep 26:537–559.
8. Schnell N, et al. (1988) Prepeptide sequence of epidermin, a ribosomally synthesized
antibiotic with four sulphide-rings. Nature 333:276–278.
9. Goto Y, et al. (2010) Discovery of novel lantibiotic synthetases provides new mechanistic and evolutionary insights. PLoS Biol 8:e1000339 doi: 1000310.1001371/journal.
pbio.1000339.
10. Oman TJ, van der Donk WA (2010) Follow the leader: The use of leader peptides to
guide natural product biosynthesis. Nat Chem Biol 6:9–18.
11. Xie L, et al. (2004) Lacticin 481: In vitro reconstitution of lantibiotic synthetase activity.
Science 303:679–681.
12. Willey JM, van der Donk WA (2007) Lantibiotics: Peptides of diverse structure and
function. Annu Rev Microbiol 61:477–501.
13. Levengood MR, Knerr PJ, Oman TJ, van der Donk WA (2009) In vitro mutasynthesis of
lantibiotic analogues containing nonproteinogenic amino acids. J Am Chem Soc
131:12024–12025.
14. Cotter PD, et al. (2006) Complete alanine scanning of the two-component lantibiotic
lacticin 3147: Generating a blueprint for rational drug design. Mol Microbiol
62:735–747.
15. Johnson ZI, et al. (2006) Niche partitioning among Prochlorococcus ecotypes along
ocean-scale environmental gradients. Science 311:1737–1740.
16. Fuller NJ, et al. (2003) Clade-specific 16S ribosomal DNA oligonucleotides reveal the
predominance of a single marine Synechococcus clade throughout a stratified water
column in the Red Sea. Appl Environ Microbiol 69:2430–2443.
17. Begley M, Cotter PD, Hill C, Ross RP (2009) Identification of a novel two-peptide
lantibiotic, lichenicidin, following rational genome mining for LanM proteins. Appl
Environ Microbiol 75:5451–5460.
Li et al.
Materials and Methods
General materials and protocols used for molecular biology, protein purification, MS, and bioinformatics are provided in the SI Appendix. procM was
cloned into pET28b, and procA genes were cloned into pET-15b. Mutants
of procA were generated by QuikChange (Stratagene) or overlapping PCR.
His6 -ProcAs were overexpressed in insoluble form in E. coli and purified as
previously described for other LanA peptides (37). Purification of His6 -ProcM
was performed by cobalt affinity chromatography resulting in 5–10 mg∕L of
cell culture. Activity assays were carried out in 50 mM Hepes (pH 7.5), 10 mM
MgCl2 , 2.5 mM ATP, and 0.5 mM tris(2-carboxyethyl)phosphine with 25 μM
ProcA and 0.5 μM ProcM. The assay mixture was incubated at 25 °C for
15–20 h and analyzed by MALDI-TOF MS (Voyager) or ESI-MS with a Synapt
ESI quadrupole TOF System (Waters).
For transcription assays of procM and procA, an axenic culture was grown
to late-exponential stage, the cells harvested, and RNA isolated by using the
Mirvana RNA isolation kit (Ambion). RNA was reverse-transcribed by using
Superscript II (Invitrogen), DNA removed by using Ambion Turbo DNA-free,
and PCR performed by using Platinum Taq DNA polymerase (Invitrogen).
For detection of prochlorosins, a 20 L late-exponential axenic culture was
harvested, the supernatant from the cell culture was filtered (Supor
0.22 μm filter), and the cell-free spent media absorbed with 100 g of Amberlite XAD-16 resin (Sigma). The column was eluted with a stepwise gradient of
aqueous isopropanol containing 0.1% trifluoroacetic acid. The elution fractions were concentrated and purified by using a C18 solid phase extraction
column eluting with a stepwise gradient of aqueous acetonitrile containing
0.1% formic acid.
ACKNOWLEDGMENTS. This study was supported in part by grants from the
Gordon and Betty Moore Foundation, the National Science Foundation, and
the United States Department of Energy Genomics:GTL Program (S.W.C)
and the National Institutes of Health [GM58822 (to W.A.v.d.D.)]. D.S. is
supported by postdoctoral fellowships from the Fullbright Foundation and
the United States–Israel Binational Agricultural Research and Development
Fund (Vaadia-BARD Postdoctoral Fellowship Award FI-399-2007). I.J. was
supported by the Howard Hughes Medical Institute.
18. Chatterjee C, Paul M, Xie L, van der Donk WA (2005) Biosynthesis and mode of action
of lantibiotics. Chem Rev 105:633–684.
19. Ross AC, Liu H, Pattabiraman VR, Vederas JC (2010) Synthesis of the lantibiotic lactocin
S using peptide cyclizations on solid phase. J Am Chem Soc 132:462–463.
20. Martiny AC, Coleman ML, Chisholm SW (2006) Phosphate acquisition genes in Prochlorococcus ecotypes: Evidence for genome-wide adaptation. Proc Natl Acad Sci
USA 103:12552–12557.
21. Thompson AW (2009) Iron and prochlorococcus. PhD thesis (Massachusetts Institute of
Technology, Cambridge).
22. Rusch DB, et al. (2007) The Sorcerer II Global Ocean Sampling expedition: Northwest
Atlantic through eastern tropical Pacific. PLoS Biol 5:e77.
23. Howard EC, Sun S, Biers EJ, Moran MA (2008) Abundant and diverse bacteria involved
in DMSP degradation in marine surface waters. Environ Microbiol 10:2397–2410.
24. Firn RD, Jones CG (2000) The evolution of secondary metabolism—A unifying model.
Mol Microbiol 37:989–994.
25. McCaskill D, Croteau R (1997) Prospects for the bioengineering of isoprenoid biosynthesis. Adv Biochem Eng Biotech 55:107–146.
26. Abe I (2007) Enzymatic synthesis of cyclic triterpenes. Nat Prod Rep 24:1311–1331.
27. Donia MS, et al. (2006) Natural combinatorial peptide libraries in cyanobacterial
symbionts of marine ascidians. Nat Chem Biol 2:729–735.
28. Bentley SD, et al. (2002) Complete genome sequence of the model actinomycete
Streptomyces coelicolor A3(2). Nature 417:141–147.
29. Ikeda H, et al. (2003) Complete genome sequence and comparative analysis of the
industrial microorganism Streptomyces avermitilis. Nat Biotechnol 21:526–531.
30. Rocap G, et al. (2003) Genome divergence in two Prochlorococcus ecotypes reflects
oceanic niche differentiation. Nature 424:1042–1047.
31. Zhu Y, Gieselman M, Zhou H, Averin O, van der Donk WA (2003) Biomimetic studies
on the mechanism of stereoselective lanthionine formation. Org Biomol Chem
1:3304–3315.
32. Fischbach MA, Clardy J (2007) One pathway, many products. Nat Chem Biol 3:353–355.
33. Olivera BM (2002) Conus venom peptides: Reflections from the biology of clades and
species. Annu Rev Ecol Syst 33:25–47.
34. Sullivan MB, et al. (2009) The genome and structural proteome of an ocean siphovirus:
A new window into the cyanobacterial ‘mobilome’. Environ Microbiol 11:2935–2951.
35. Kuipers OP, Beerthuyzen MM, de Ruyter PG, Luesink EJ, de Vos WM (1995) Autoregulation of nisin biosynthesis in Lactococcus lactis by signal transduction. J Biol Chem
270:27299–27304.
36. Willey JM, Willems A, Kodani S, Nodwell JR (2006) Morphogenetic surfactants and
their role in the formation of aerial hyphae in Streptomyces coelicolor. Mol Microbiol
59:731–742.
37. Li B, Cooper LE, van der Donk WA (2009) In vitro studies of lantibiotic biosynthesis.
Methods Enzymol 458:533–558.
PNAS ∣
June 8, 2010 ∣
vol. 107 ∣
no. 23 ∣
10435
BIOCHEMISTRY
also contains a putative siphovirus integration site [(34)—the
peptides annotated as nif11-like are procA3.1-procA3.5]. Together, these observations suggest a mechanism for transfer of
these genes within and between genomes.
At present, we can only speculate as to the biological function
of prochlorosins. Whereas the vast majority of known lanthionine-containing peptides are bacteriocidal (12), lanthioninecontaining peptides can also act as signaling molecules (35) or
morphogenetic peptides (36). We have not observed any bacteriocidal activity in tests of four recombinant prochlorosins (Pcn1.1,
Pcn1.2, Pcn1.3, and Pcn1.5) against Lactococcus lactis 117 and
Bacillus subtilis 6633 (SI Appendix). However, these strains and
assay conditions do not even remotely represent those encountered by Prochlorococcus cells in the wild. Further tests have
been hampered by very low yields of prochlorosins from Prochlorococcus cultures (<10 μg from 20 L of cell culture) and by the
need to remove the leader peptide from recombinantly expressed
and modified peptides (SI Appendix)—obstacles that we will
attempt to surmount in future studies.
Although procM/procA genes found in the GOS database were
disproportionately abundant in a hypersaline lagoon near the
Galapagos where microbial densities are relatively high and
community diversity low (22), 42% of procM/procA genes detected were from nutrient-poor regions of the oceans, where total
microbial cell densities are relatively low (∼106 bacteria mL−1
seawater). In these habitats, the average distance between individual Prochlorococcus cells and their nearest microbial neighbor
is approximately 200 Prochlorococcus cell lengths. It is difficult to
imagine how prochlorosins could mediate intercellular functions
at these distances. Perhaps they fulfill intracellular roles, and if
so, one wonders what function would utilize such a diversity of
secondary compounds.