Enzymatic basis of ribosomal peptide prenylation in cyanobacteria

Article abstract of DOI:10.1021/ja205458h

The enzymatic basis of ribosomal peptide natural product prenylation has not been reported. Here, we characterize a prenyltransferase, LynF, from the TruF enzyme family. LynF is the first characterized representative of the TruF protein family, which is responsible for both reverse- and forward-O-prenylation of tyrosine, serine, and threonine in cyclic peptides known as cyanobactins. We show that LynF reverse O-prenylates tyrosine in macrocyclic peptides. Based upon these results, we propose that the TruF family prenylates mature cyclic peptides, from which the leader sequence and other enzyme recognition elements have been excised. This differs from the common model of ribosomal peptide biosynthesis, in which a leader sequence is required to direct post-translational modifications. In addition, we find that reverse O-prenylated tyrosine derivatives undergo a facile Claisen rearrangement at 'physiological' temperature in aqueous buffers, leading to forward C-prenylated products. Although the Claisen rearrangement route to natural products has been chemically anticipated for at least 40 years, it has not been demonstrated as a route to prenylated natural products. Here, we show that the Claisen rearrangement drives phenolic C-prenylation in at least one case, suggesting that this route should be reconsidered as a mechanism for the biosynthesis of prenylated phenolic compounds.

Full text of DOI:10.1021/ja205458h

ARTICLE
pubs.acs.org/JACS
Enzymatic Basis of Ribosomal Peptide Prenylation in Cyanobacteria
John A. McIntosh,† Mohamed S. Donia,†,§ Satish K. Nair,‡ and Eric W. Schmidt*,†
†Department of Medicinal Chemistry, University of Utah, Salt Lake City, Utah 84112, United States
‡Department of Biochemistry, University of Illinois at UrbanaÀChampaign, Urbana, Illinois 61801, United States
S Supporting Information
b
ABSTRACT: The enzymatic basis of ribosomal peptide natural
product prenylation has not been reported. Here, we character-
ize a prenyltransferase, LynF, from the TruF enzyme family.
LynF is the first characterized representative of the TruF
protein family, which is responsible for both reverse- and
forward-O-prenylation of tyrosine, serine, and threonine in
cyclic peptides known as cyanobactins. We show that LynF
reverse O-prenylates tyrosine in macrocyclic peptides. Based upon these results, we propose that the TruF family prenylates mature
cyclic peptides, from which the leader sequence and other enzyme recognition elements have been excised. This differs from the
common model of ribosomal peptide biosynthesis, in which a leader sequence is required to direct post-translational modifications.
In addition, we find that reverse O-prenylated tyrosine derivatives undergo a facile Claisen rearrangement at ‘physiological’
temperature in aqueous buffers, leading to forward C-prenylated products. Although the Claisen rearrangement route to natural
products has been chemically anticipated for at least 40 years, it has not been demonstrated as a route to prenylated natural products.
Here, we show that the Claisen rearrangement drives phenolic C-prenylation in at least one case, suggesting that this route should be
reconsidered as a mechanism for the biosynthesis of prenylated phenolic compounds.
’ INTRODUCTION
pathways discovered by metagenome sequencing, we proposed
that the TruF family of proteins might be PTs.11 Given that the
TruF family lacks the typical sequence hallmarks of PTs, we
sought to obtain biochemical evidence for this proposal. Conse-
quently, we explored the set of sequenced gene clusters to find a
soluble TruF relative for biochemical analysis.15 Among TruF
relatives, one protein, LynF from Lyngbya aestuarii,16,17 could be
solubly expressed in Escherichia coli. Although the other steps in
cyanobactin biosynthesis have been characterized,18À21 the PT
substrate was unknown prior to this study. Consequently,
substrate analogues for each possible biochemical step were
synthesized, and their products upon reaction with LynF were
analyzed (Figure 2). We show here that the LynF/TruF family
represents a remarkably broad substrate family of O-prenyltrans-
ferases, with LynF prenylating a wide variety of tyrosyl peptides
and phenol derivatives. Among biosynthetically relevant sub-
strates, only cyclic peptides are prenylated.
Prenylation is a common biochemical modification that has
been studied in detail in numerous systems.1,2 For example,
proteins are often farnesylated or geranylgeranylated on cysteine
residues, and numerous peptide natural products are known to
be prenylated at diverse positions. Prenylation is key to the
biological activity of these molecules.3,4 Several enzyme families
have been described that catalyze prenyl transfer, and indeed
whole new prenyltransferase (PT) families continue to be
discovered.5À9 For example, the ABBA family has been shown
to catalyze aromatic C-prenylation on a variety of substrates,10
especially ortho to phenolic oxygen. Several ribosomal peptide
natural products, such as ComX and the cyanobactins, are
prenylated in unique ways that greatly increase the chemical
diversity of the resulting compounds.11À13 PTs for this growing
ribosomal peptide natural product group have yet to be enzy-
matically characterized.14
Unexpectedly, reactions catalyzed by LynF led to carbon-pre-
nylated phenolic products. Generally speaking, PTs catalyze elec-
trophilic alkylation of their substrates. In the case of aromatic
substrates, reactions are thought to proceed via electrophilic aro-
matic substitution.10 However, several alternative mechanisms exist.
For one, it has been proposed since at least the early 1970s that
C-prenylated phenols might also arise via reverse O-prenylation of
phenols followed by a Claisen rearrangement (Figure 1).22 This
chemical proposal was exploited in several elegant ‘biomimetic’ total
Cyanobactins are a broadly distributed group of ribosomally
derived, macrocyclic peptides whose biosynthetic genes are
homologous. These compounds are often highly post-transla-
tionally modified. Indeed, many cyanobactins are prenylated by
dimethylallyl pyrophosphate (DMAPP) on the oxygen atom of
serine, threonine, or tyrosine (Figure 1), all of which are
biochemically unprecedented reactions in the context of riboso-
mal peptides. Among cyanobactins, prenylation is known to
occur in the “reverse” position (DMAPP 3-carbon) with serine
and threonine and in the “forward” position (DMAPP 1-carbon)
with tyrosine. However, no obvious candidate PTs exist in
cyanobactin gene clusters.11 By comparing several cyanobactin
Received: June 13, 2011
Published: July 18, 2011
r
2011 American Chemical Society
13698
dx.doi.org/10.1021/ja205458h J. Am. Chem. Soc. 2011, 133, 13698–13705
|

Journal of the American Chemical Society
ARTICLE
syntheses of complex natural products.23,24 Despite a resur-
gence of biochemical studies of aromatic PTs in recent
years,6,8,10,25À30 the Claisen rearrangement proposal has not
been revisited. Here, in addition to characterizing a novel PT
family, we show that enzymatically synthesized aromatic C-pre-
nylated phenols can indeed arise via the Claisen rearrangement
pathway.
’ RESULTS
Expression of LynF and Synthesis of Substrates. We found
that several of the TruF-group PTs were difficult to overexpress.
Fortuitously, in our search for candidate PTs, LynF (44% amino
acid identity with TruF1), from the lyn pathway of the cyano-
bacterium L. aestuarii, was readily expressed in E. coli in soluble
form. We predicted that LynF should prenylate phenols, but
there are no known natural products of the lyn pathway.16,17
Therefore, analogues of possible substrates were synthesized,
including linear peptides representing putative pathway inter-
mediates as well as cyclo[APMPPYP] (6), which is similar to the
predicted lyn pathway product (Figure 2, Table 1). Additionally,
reactions with several wholly unnatural phenol derivatives and
linear peptides containing tyrosine were attempted. The purity
and the identity of the synthetic substrates were assessed
spectroscopically.
LynF is a Tyrosine PT. When incubated with DMAPP and
MgCl2, LynF catalyzed the prenylation of peptides and a subset
of phenols related to tyrosine (Table 1). Mg2+ was added because
many PTs require it for function.10 Indeed, LynF does not
function in the absence of either Mg2+ or Mn2+. Tyrosine
prenylation was demonstrated by Fourier-transform ion cyclo-
tron resonance (FT-ICR) and MS/MS fragmentation, which
revealed the presence of the expected ions (Table S2, Supporting
Information) and also localized the prenylation to tyrosine
(Figure S1, Supporting Information).
LynF prenylated several tyrosine-containing substrates and showed
a strong preference for reaction with cyclic over linear peptides
(Table 1). In contrast to typical ribosomal peptide biosynthesis,31
LynF did not act on pathway intermediates (3À5) that still contained
a leader sequence or enzyme “recognition elements” (Figures 2 and
S1 and S2, Supporting Information). Of biosynthetically relevant
Figure 1. Prenylation in natural products. (A) Representative cyano-
bactin peptide natural products, showing groups derived from DMAPP
in blue. (B) Two possible enzymatic mechanisms of phenol ortho-C-
prenylation. First, DMAPP is dephosphorylated to yield a cation that can
react either at oxygen (pathway I) or at carbon (pathyway II). In
principle, a Claisen rearrangement from pathway I could then yield the
C-prenylated product. Only pathway II has been previously linked to
enzymatic modification.
Figure 2. Defining the biosynthetic route to prenylated cyanobactins. Proposed biosynthetic scheme for Lyn pathway showing modification of
precursor peptide by heterocyclization, proteolysis, macrocyclization, and prenylation. The first steps in this route are supported by previous
enzymological studies, but the timing of prenylation was not known. Enzyme recognition elements are highlighted. Analogues (3À6) were used to assess
prenylation of each possible biosynthetic intermediate and are shown in analogous colors beneath each proposed biosynthetic intermediate. Pro was
substituted as an approximate isostere for thiazole in later analogues. No reaction was observed with any intermediate except the final cyclic peptide,
showing that prenylation occurs at a late step, after all enzyme recognition elements have been excised.
13699
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
Table 1. Substrates Assayeda
no.
substrate
yield
3
TruLy1 precursor peptide
TruLy1 w/3 heterocycles
APMPPYPSYDDAE
cyclo[APMPPYP]
APMPPYP
NR
NR
NR
48%
12%
10%
47%
43%
37%
94%
1%
4
5
6
7
8
N-acetyl APMPPYP
cyclo[APMPPPAPMPPYP]
cyclo[APMPPYPAPMPPYP]
cyclo[KKPYILP]
cyclo[KPYILP]
KPYILP
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
boc-L-Tyr
71%
66%
NR
NR
NR
NR
NR
3%
boc-D-Tyr
boc-4-cyano-L-Phe
boc-O-allyl-L-Tyr
boc-4-iodo-L-Phe
boc-4-methoxy-L-Phe
L-Phe
N-acetyl-L-Tyr
L-Tyr
NR
NR
NR
NR
NR
NR
dopamine
Figure 3. LynF is ortho-C-prenylating. The aromatic/olefin region of
the heteronuclear quantum coherence NMR spectra is shown for (A)
boc-tyrosine (14) and (B) its purified enzymatic product with LynF,
clearly indicating a single, forward C-prenylation event. Similar spectra
were also observed for reactions containing 6. NMR and MS character-
ization of compounds is presented in Figures S1, S3, and S5, Supporting
Information.
phenol
L-Trp
cyclo[QGGRGDWP]
QGGRGDWPAYDGE
a All yield quantitations are based on HPLC (for 6À8 and 14À25) or MS
(3À5 and 9À13) analyses of 24 h reactions and do not represent isolated
yields. NR (“no reaction”) denotes no detectable prenylation of tyrosine or
derivatives. All substrates were run at 100 μM except 3, 4, and 9, 10, 11, and
12 (14, 14, 70, 30, 20, and 20 μM, respectively). Both full (100 μM) and
reduced concentrations (28, 56 μM) were employed with substrates 6 and
14 in order to compare data for substrates tested at reduced concentration.
Scheme 1. NMR Characterized Products
substrates, LynF acted only on the mature cyclic peptides, such as 6.
Overall, LynF was capable of prenylating a broad variety of cyclic
peptides, including substrates (9À12) containing 6-, 7-, 13-, and
14-amino acid residues. Surprisingly, LynF also prenylated several
tyrosine-containing substrates that are not relevant to the natural
biosynthetic pathway. Tyrosine itself was not a substrate, but boc-
L-tyrosine and other N-terminally blocked tyrosine derivatives were
readily prenylated (14, 15 and 21). Derivatives lacking a free phenolic
ÀOH were not substrates (16À20). Overall, LynF appears to require
a blocked N-terminus but otherwise exhibits relaxed substrate
specificity.
LynF Products are ortho-C-Prenylated. Based upon the
known products of cyanobactin pathways, we expected that
LynF would catalyze forward O-prenylation of phenol.11,32 We
performed large-scale enzymatic reactions with long incubation
times to generate sufficient quantities of products for NMR
analysis. Products of two different LynF substrates, 6 and 14,
were isolated by HPLC and characterized by NMR and high-
resolution MS. Surprisingly, by comparison to previously de-
scribed compounds,33 we established that both 6 and 14 were
forward C-prenylated, ortho to the phenolic hydroxyl group
(Figures 3 and S3, Supporting Information).
substitution at the ortho position in a manner identical with that
reported for ABBA PTs. With an eye toward constructing a
Hammett plot of reactivity, we assayed the LynF catalyzed
prenylation of a series of 4-substituted Tyr derivatives
(16À20) (Table 1). However, all of these reactions failed. If
LynF-catalyzed prenylation were to occur via electrophilic aro-
matic substitution, then one would expect a broader scope of
reactivity, as was found for dimethylallyl tryptophan synthase.34
An alternative mechanism for the formation of ortho-C-pre-
nylated phenols involves reverse O-prenylation followed by
Claisen rearrangement of the resulting O-allyl intermediate.22 In
C-Prenylation is the Result of a Claisen Rearrangement.
Initially, we assumed that LynF catalyzed electrophilic aromatic
13700
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
Figure 5. Time course of boc-L-Tyr (14) reaction followed by HPLC.
The reaction was performed in quadruplicate, with variation indicated
by error bars (A) and (B). Initial product of the reaction is almost
exclusively O-prenylated as shown at 1, 2, and 4 h time points (B). After
4 h, the level of O-prenylated intermediate reaches steady-state and its levels
are constant through 24 h, accompanied by steady increase in the concentra-
tion of C-prenylated final product 29. This allowed kinetic constants for the
Claisen rearrangement to be directly determined, since at steady state the
concentration of the O-prenyl intermediate can be assumed to be a constant.
Figure 4. LynF catalyzes reverse O-prenylation of tyrosine. The aro-
matic/olefin region of 1H NMR spectra are shown (A) for boc-tyrosine
(14), (B) the HPLC-purified intermediate LynF product (31), and
(C) the final reaction product (30). These spectra clearly indicate that
the first product of the LynF reaction is reverse-O-prenyl tyrosine, which
subsequently rearranges to give the C-prenylated product.
O-prenylated products. Fortuitously, we were able to isolate
one of these compounds, resulting from reaction of 15 (Figures 4
and S5, Supporting Information). NMR analysis of the purified
material confirmed that the product (31) was O-prenylated and
conclusively demonstrated that O-prenylation occurred in the
reverse orientation. Purified 31 was then added to aqueous buffer
at 37 °C (for buffer composition see Materials and Methods
Section), and it rapidly and spontaneously rearranged to form the
forward C-prenylated product 30 which was identical to the
previously NMR-characterized product, 29.
Reverse O-prenylated phenols are known to undergo the
Claisen rearrangement to yield forward ortho-C-prenylated
products.22,24,38,39 Thus, we realized that if the sole enzymatic
reaction catalyzed by LynF were reverse O-prenyltransfer on Tyr,
then thiswouldleadtothe mixture ofproducts wehadconsistently
observed with all substrates. Alternatively, we supposed that LynF
might carry out reverse O-prenyltransfer in addition to direct
electrophilic aromatic substitution in the forward direction on
carbon. To distinguish between these two possibilities, a kinetic
analysis for reactions containing 14 was performed in which C-
and O-prenyl products were followed over 24 h in reactions
performed in quadruplicate (Figure 5). The O-prenyl product
that vein, careful examination of LC-FT-ICR analyses of reaction
mixtures showed that upon reaction with LynF, all substrates gave
rise to two products (Figure S1, Supporting Information). These
products were isobaric and prenylated on tyrosine. However, one
product was prenylated on carbon and the other on oxygen.
Forward carbon prenylation had been established using 28
(Scheme 1), which was purified and characterized by NMR as
described above. Of the two products of LynF upon reaction with
6, purified 28 was found to be identical to the early eluting product
(FigureS4, Supporting Information). Moreover, nofragmentation
of the C-prenyl moiety on 28 could be observed in MS-MS
experiments. Similarly, for all LynF products, we observed that the
early eluting compound was prenylated on tyrosine but did not
lose isoprene in MS-MS, indicating C-prenylation.
In contrast, all late-eluting products evinced prominent loss of
isoprene (C5H8) in their MS-MS spectra. Loss of C5H8 from
prenylated phenols is diagnostic of O-prenylation, as shown in
previous studies.35À37 This reaction was more difficult to char-
acterize by NMR, owing to the apparent instability of the
13701
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
have observed reverse O-prenylated phenols to rearrange is
unsurprising in light of the known aqueous acceleration of the
Claisen rearrangement.40À45 Hereweshow by experimentthatthe
rearrangement goes to completion at 37 °C in aqueous buffers.
Consequently, our conditions may provide a particularly mild
reaction condition for the Claisen rearrangement for use in future
synthetic studies.
To further examine the rapid rearrangement observed with 31,
we purchased O-allyl-boc-L-tyrosine and examined it to see if its
rearrangement might be accelerated with enzyme or the buffer
conditions employed with 31. In contrast to the prenylated
substrates, this compound did not undergo the Claisen rearrange-
ment under the aforementioned conditions. Based upon these
results, it seems that the geminal methyl groups adjacent to the
phenolic oxygen are required to promote the Claisen rearrange-
ment at relatively low temperatures. This effect can be rationalized
as an example of the gem substituent effect, which is believed to
accelerate the Claisen rearrangement.46 Indeed, it has been shown
that the presence of bulky substituents R- to oxygen can accelerate
the Claisen rearrangement and similar reactions,47,48 though to the
best of our knowledge this is the first direct comparison of these
substrates in the aromatic Claisen rearrangement.
Phylogenetic analysis of LynF and homologues nicely ratio-
nalizes the observed pattern of reactivity, where LynF is most
closely related to TruF1 and PagF (Figure 6). In light of the
above biochemical evidence, TruF1 can be assigned the role of a
reverse O-prenyltransferase acting on Ser and Thr, while PagF
can be assigned the role of a forward O-prenyltransferase acting
on Tyr. Thus, that LynF would carry out reverse O-prenylation of
Tyr is unsurprising in light of its phylogenetic profile.
Figure 6. Phylogenetic analysis of the TruF/LynF family. (A) Tru
F-group proteins cluster according to whether natural products are
prenylated (top) or nonprenylated (bottom). Chemical products are
shown for each TruF-like protein, where they exist. No TruF-like
sequence relatives can be identified outside of cyanobactin gene clusters
using either BLAST searching or sequence alignments with other PT
family proteins, indicating that this is a novel group of PTs. (B) Actual
product structures of prenylagaramide (left) and trunkamide (right)
pathways; predicted enzymatic product of lyn pathway (middle) prior to
Claisen rearrangement.
Kinetic Measurements. LynF catalyzed reaction rates were
measured in triplicate using two different substrates: cyclo-
[APMPPYP] (6), and an unnatural substrate, boc-L-tyrosine
(14) (Figure S7, Supporting Information). Using HPLC analysis,
the turnover numbers for 6 and 14 were similar (14 and 63 hÀ1
,
respectively), as were Km values (4 and 14 mM, respectively).
These rates are slower than those typically reported for prenyl-
transferases.27,49 However, the apparent slowness of reactions
catalyzed by LynF is not unusual when compared with rates
observed with other cyanobactin biosynthetic enzymes. Some of
these reactions are quite slow, a fact that has been attributed to
their extremely broad substrate tolerance and by extension their
relatively low affinities for any given substrate.18À21
appeared first, with a delayed onset of 29. After 4 h of reaction
time, a steady state was reached in which the rate of prenylation
was equal to the rate of the Claisen rearrangement. Interpretation
of the kinetic data, which assumed a unimolecular mechanism and
steady-state levels of O-prenyl intermediate, yielded a rate of
rearrangement of 8.3 μM/h and a rate constant for the Claisen
rearrangement (kwhererateofClaisen=k*[O-prenyl intermediate])
of 0.23 hÀ1. Taken together with the absence of reactivity observed
with analogues lacking a free phenolic ÀOH, these data show that the
initial enzymatic reaction is reverse O-prenylation, followed by slower
conversion to a forward C-prenylated phenol
Having shown that the O-prenylated compound is the product of
initial prenyltransfer, we sought to determine whether the rearran-
gement to the forward C-prenylated compound was enzyme
catalyzed. To do so, purified O-prenylated 31 was added to enzyme,
buffer, or boiled enzyme. Under all three conditions, 31 was
efficiently converted into C-prenylated 30 without any appreciable
enzymatic acceleration (Figure S6, Supporting Information).
This spontaneous Claisen rearrangement might seemsurprising
given that in the synthetic literature, reverse prenylated phenols
require elevated temperatures for rearrangement.24,38 However,
these synthetic reactions take place in organic solvents, while we
have employed aqueous solvents. Indeed, the speed with which we
’ DISCUSSION
To the best of our knowledge, LynF represents the first
enzymatically characterized PT leading to the synthesis of riboso-
mal peptide natural products. Further, serineand threonine O-PTs
have not been previously described nor have tyrosine O-PTs
acting on ribosomal peptides.14 Although these post-translational
modifications are currently known only in the cyanobactin family
of natural products, cyanobactins are present in perhaps ∼30% of
all cyanobacteria onEarth and thereforeconstitute a major fraction
of bioactive natural products globally.50 Another salient feature of
this enzyme group is that it clearly acts on polypeptide products,
while most other natural product DMAPP transferases act on
starting amino acids or on small dipeptides. For example, forward
O-prenylated tyrosine has recently been characterized in sirodes-
min diketopiperazine biosynthesis.49,51 However, in this case
tyrosine itself is the substrate for prenylation, and the product
does not result from ribosomal synthesis.
13702
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
Scheme 2. Claisen Rearrangement Pathway
Phenols themselves are C- or O-prenylated in many small
molecule natural products.10 In all cases that have been character-
ized so far, it is thought that C-prenylated phenols arise from direct
electrophilic aromatic substitution.10 Although the Claisen rear-
rangement was long-predicted from ‘biomimetic’ chemistry, a
biochemical demonstration of its relevance as a route to C-pre-
nylated phenols was lacking. Here, we show that the Claisen
rearrangement route can indeed occur to afford C-prenylated
products (Scheme 2). This route could easily be missed, since the
O-prenylated intermediates are short-lived and not easily detected
by commonly used analytical methods. For example, the inter-
mediates have extremely weak absorption at λ = 280 nm, and their
fluorescence spectra are different than for unsubstituted or C-pre-
nylated phenols. The rearrangement is relatively rapid and con-
tinues even after enzymes have been denatured or inactivated.
Given aqueous acceleration of the Claisen rearrangement and the
acceleration provided by the geminal methyl groups of the reverse
prenylation, the ease with which the Claisen rearrangement might
occur in a cellular context has perhaps been underestimated.
Overall, forward prenylation via electrophilic aromatic substitution
and reverse O-prenylation followed by the Claisen rearrangement
will be indistinguishable under many conditions.
clusters, it remains possible that they are structurally related to
known PTs. However, no putative ABBA-like required residues
are present in the correct places in these proteins (Figure S8,
Supporting Information). Since PTs are often deeply divergent
and sequence similarity is completely lacking for this protein class,
a final comparison will await structural study. It is also unknown
why nonprenylating cyanobactin clusters usually contain (and
even require) LynF-like proteins.16,50 In these nonprenylating
cases, all enzymatic roles have been assigned, so that LynF
homologues serve no obvious enzymatic function.20,21 However,
removal of the LynF homologue from heterologous expression of
the nonprenylating pat pathway in E. coli abolishes compound
production.16 Possible roles include a chaperone function or
perhaps interaction with the leader sequence.
In conclusion, we show that the TruF/LynF group of proteins
represents a new family of PTs that catalyze unprecedented
enzymatic reactions and that are quite distinct from previously
characterized proteins. LynF represents the first ribosomal pep-
tide natural product prenyltransferase to be characterized, open-
ing the door to the study of prenylated ribosomal peptide natural
products.
We initially expected that the Claisen rearrangement might be
enzymatically accelerated. In synthetic chemistry, several guani-
dinium-based synthetic catalysts of the Claisen rearrangement
have been reported.38,52 Additionally, in the premier example of a
biological Claisen rearrangement, chorismate mutase has been
calculated to provide rate enhancements of >106.53 However, it is
clear from our results using purified reverse-O-prenylated 31 that
in this case the reaction is spontaneous. Given the spontaneous
nature of this transformation and the similarity to other reported
Claisen rearrangements, the simplest hypothesis is that 31
proceeds to 30 via the Claisen rearrangement and not via some
other, more complicated mechanism.
We have previously shown that proteins in this group were
involved in pathways to very sequence-diverse prenylated natural
products in vivo,16 and here we show that purified LynF accepts many
different cyclic substrates. This broad specificity is especially remark-
able in that LynF substrates share no common sequence features that
would provide robust enzyme recognition elements. In ribosomal
peptide natural product synthesis, enzymes commonly recognize
conserved motifs in a leader peptide, which is subsequently cleaved
and discarded, allowing the enzymes to modify diverse sequences.31
However, in this case the reaction proceeds after the leader sequence
and the recognition elements have already been removed.
’ MATERIALS AND METHODS
For detailed methods, see Materials and Methods in the Supporting
Information.
Substrates. Dimethylallyl pyrophosphate (DMAPP) was synthe-
sized following previously established procedures.54À56 Excepting 11,
12, and 26 whose synthesis and characterization has been reported
elsewhere,15,19 peptide substrates were synthesized at the University of
Utah DNA/peptide synthesis core facility. Synthesis of boc-protected
4-iodo-L-phenylalanine and 4-methoxy-L-phenylalanine was performed
according to previously established procedures.57 Boc-L-tyrosine, so-
dium hydrogen pyrophosphate, dimethylallyl bromide, tetrabutylam-
monium hydroxide, and dopamine HCl were purchased from Sigma.
N-acetyl-L-tyrosine and phenol were purchased from Fisher Scientific.
All other Tyr and Phe derivatives were purchased from ChemImpex.
Genes and Cloning. A codon-optimized version of lynF was
synthesized and cloned into pET28 in frame with the N-terminal his-
tag sequence using NdeI and EcoRI (Genscript). TruLy1 was cloned via
modification of a previously described vector,15 which was subsequently
cloned into pET28b using NdeI and BamHI.
Protein Expression and Purification. LynF was expressed in
BL21(DE3) cells, purified initially by Ni-NTA chromatography, which
was followed by size-exclusion chromatography to yield homogeneous
protein. Purification of TruLy1 was likewise performed by Ni-NTA
chromatography, with the main difference being that rather than attempt-
ing to isolate soluble protein, TruLy1 was strongly overexpressed with the
intent of driving the protein into inclusion bodies, after which time
purification under denaturing conditions was performed.
The structure and catalytic mechanism of this new family of PTs
remains to be determined. Although the proteins bear no homol-
ogy to any other characterized protein outside of cyanobactin gene
13703
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
Enzyme Assays. Enzyme reactions typically contained enzyme
(3.8 μM) and variable substrate concentration (100 μM for most
substrates; higher concentrations, i.e., 1 mM, were occasionally em-
ployed with boc-protected amino acid derivatives. Exceptions include
substrates 11 and 12, which were used at 20 μM final concentration as
well as substrates 9 and 10, which were used at 70 and 30 μM,
respectively). Several additives (1 M of NaCl, 40 mM of glycylglycine
pH of 9.0, 12 mM of MgCl2, 3 mM of tris(2-carboxyethyl) phosphine
(TCEP), and 1 mM of DMAPP) were added to all reactions. Reactions
were incubated at 37 °C for 24 h in a DNA Engine Peltier thermocycler
(Bio-Rad). Enzyme reactions with full-length precursor peptide con-
tained TruLy1 (28 μM), ATP (0.8 mM), with or without heterocyclase
enzyme TruD (90 nM), and additives as above. Controls were run to
ensure that LynF was active in the presence of TruD and TruLy1 and
vice versa. Products were characterized by MS or diode array (λ = 220
and 280 nm) and fluorescence (λ = 271 nm excitation and 303 nm
emission) HPLC. Reactions assessing the rate of rearrangement of
purified 31 were performed at 37 °C with time points taken at 0 and 8 h
and included the standard additives described above. For descriptions of
specific assays, see Materials and Methods in the Supporting Information.
Phylogenetic Tree Construction. The amino acid sequences of
LynF homologues from the functionally characterized cyanobactin
pathways were aligned using CLUSTALX. Maximum likelihood analysis
with molecular clock PROMLK (PHYLIP) using the bootstrap test
method (1000 replicates) was performed to assess the phylogenetic
relationship between the different homologues. The same tree branches
were also supported using other phylogenetic experiments such as
maximum parsimony (MEGA 4.0) using 1000 bootstrap replicates.
General Methods. ESI-MS and FT-ICR analyses were performed
at the University of Utah Mass Spectrometry and Proteomics core
facility. MALDI-MS analyses were performed on a Micromass MALDI
micro MX instrument (Waters). HPLC separations were performed on
a LaChrom Elite system (Hitachi). NMR spectra were collected on
either 400 or 500 MHz spectrometers (Varian). CD spectra were
collected on a Jasco J-815 spectrometer, and data were plotted in Excel.
’ REFERENCES
(1) Brandt, W.; Br€auer, L.; G€unnewich, N.; Kufka, J.; Rausch, F.;
Schulze, D.; Schulze, E.; Weber, R.; Zakharova, S.; Wessjohann, L.
Phytochemistry 2009, 70, 1758–1775.
(2) Nguyen, U. T.; Goody, R. S.; Alexandrov, K. ChemBioChem
2010, 11, 1194–1201.
(3) Gibbs, R. A. Nat. Chem. Biol. 2005, 1, 7–8.
(4) Casey, P. J.; Seabra, M. C. J. Biol. Chem. 1996, 271, 5289–5292.
(5) Sacchettini, J. C.; Poulter, C. D. Science 1997, 277, 1788–1789.
(6) Edwards, D. J.; Gerwick, W. H. J. Am. Chem. Soc. 2004, 126,
11432–11433.
(7) Reiss, Y.; Goldstein, J. L.; Seabra, M. C.; Casey, P. J.; Brown, M. S.
Cell 1990, 62, 81–88.
(8) Kuzuyama, T.; Noel, J. P.; Richard, S. B. Nature 2005, 435,
983–987.
(9) Gebler, J. C.; Poulter, C. D. Arch. Biochem. Biophys. 1992, 296,
308–313.
(10) Saleh, O.; Haagen, Y.; Seeger, K.; Heide, L. Phytochemistry 2009,
70, 1728–1738.
(11) Donia, M. S.; Ravel, J.; Schmidt, E. W. Nat. Chem. Biol. 2008, 4,
341–343.
(12) Carroll, A. R.; Coll, J. C.; Bourne, D. J.; MacLeod, J. K.; Zabriskie,
T.; Ireland, C. M.; Bowden, B. F. Aust. J. Chem. 1996, 49, 659–667.
(13) Okada, M.; Sato, I.; Cho, S. J.; Iwata, H.; Nishio, T.; Dubnau,
D.; Sakagami, Y. Nat. Chem. Biol. 2005, 1, 23–24.
(14) McIntosh, J. A.; Donia, M. S.; Schmidt, E. W. Nat. Prod. Rep.
2009, 26, 537–559.
(15) Donia, M. S.; Schmidt, E. W. Chem. Biol. 2011, 18, 508–519.
(16) Donia, M. S.; Hathaway, B. J.; Sudek, S.; Haygood, M. G.;
Rosovitz, M. J.; Ravel, J.; Schmidt, E. W. Nat. Chem. Biol. 2006, 2, 729–735.
(17) Leikowski, N.; Fewer, D. P.; Sivonen, K. Appl. Environ. Micro-
biol. 2009, 75, 853–857.
(18) McIntosh, J. A.; Schmidt, E. W. ChemBioChem 2010, 11, 1413–1421.
(19) McIntosh, J. A.; Robertson, C. R.; Vinayak, A.; Satish, N. K.;
Bulaj, G. W.; Schmidt, E. W. J. Am. Chem. Soc. 2010, 132, 15499–15501.
(20) McIntosh, J. A.; Donia, M. S.; Schmidt, E. W. J. Am. Chem. Soc.
2010, 132, 4089–4091.
(21) Lee, J.; McIntosh, J. A.; Hathaway, B. J.; Schmidt, E. W. J. Am.
Chem. Soc. 2009, 131, 2122–2214.
’ ASSOCIATED CONTENT
(22) Quillinan, A. J.; Scheinmann, F. J. Chem. Soc. D 1971, 966–967.
(23) Tisdale, E. J.; Slobodov, I.; Theodorakis, E. A. Org. Biomol.
Chem. 2003, 1, 4418–4422.
S
Supporting Information. Additional mass spectrometry,
b
NMR, kinetic, protein purification, sequence, and substrate data,
and full methods. This material is available free of charge via the
Internet at http://pubs.acs.org.
(24) Nicolaou, K. C.; Li, J. Angew. Chem., Int. Ed. 2001, 40, 4264–4268.
(25) Jost, M.; Zocher, G.; Tarcz, S.; Matuscheck, M.; Xie, X.; Li,
S. M.; Stehle, T. J. Am. Chem. Soc. 2010, 132, 17849–17858.
(26) Haagen, Y.; Uns€old, I.; Westrich, L.; Gust, B.; Richard, S. B.;
Noel, J. P.; Heide, L. FEBS Lett. 2007, 581, 2889–2893.
(27) Ding, Y.; de Wet, J. R.; Cavalcoli, J.; Li, S.; Greshock, T. J.;
Miller, K. A.; Finefield, J. M.; Sunderhaus, J. D.; McAfoos, T. J.;
Tsukamoto, S.; Williams, R. M.; Sherman, D. H. J. Am. Chem. Soc.
2010, 132, 12733–12740.
’ AUTHOR INFORMATION
Corresponding Author
ews1@utah.edu
Present Addresses
§Department of Bioengineering and Therapeutic Sciences and
California Institute for Quantitative Biosciences, University of
California, San Francisco, San Francisco, CA 94158
(28) Schultz, A. W.; Lewis, C. A.; Luzung, M. R.; Baran, P. S.; Moore,
B. S. J. Nat. Prod. 2010, 73, 373–377.
(29) Balibar, C. J.; Howard-Jones, A. R.; Walsh, C. T. Nat. Chem.
Biol. 2007, 3, 584–592.
(30) Pojer, F.; Wemakor, E.; Kammerer, B.; Chen, H.; Walsh,
C. T.; Li, S. M.; Heide, L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
2316–2321.
(31) Oman, T. J.; van der Donk, W. A. Nat. Chem. Biol. 2010,
6, 9–18.
(32) Murakami, M.; Itou, Y.; Ishida, K.; Shin, H. J. J. Nat. Prod. 1999,
62, 725–755.
(33) Ahmed, F.; Ohtsuki, T.; Aida, W.; Ishibashi, M. J. Nat. Prod.
2008, 71, 1963–1966.
’ ACKNOWLEDGMENT
This work was supported by NIH GM071425. We thank C. Dale
Poulter, Jeffrey Rudolph, Gary E. Keck, and John Heemstra for
helpful discussions; Chad Nelson, Krishna Parsawar, and Jim Muller
for mass spectrometry assistance; Scott Endicott and Robert
Schackmann for peptide synthesis; Seth Lilavivat for circular
dichroism assistance; and Jack Skalicky, Jay Olsen, Dai Tianero,
and Zhenjian Lin for NMR assistance. We dedicate this paper to the
late Prof. D. John Faulkner for his pioneering work on the Claisen
rearrangement and marine natural products.
(34) Gebler, J. C.; Woodside, A. B.; Poulter, C. D. J. Am. Chem. Soc.
1992, 114, 7354–7360.
(35) Donia, M. S.; Schmidt, E. W. Chem. Biol. 2011, 18, 508–519.
13704
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705

Journal of the American Chemical Society
ARTICLE
(36) Sobolev, V. S.; Potter, T. L.; Horn, B. W. Phytochem. Anal. 2006,
17, 312–322.
(37) Miranda, C. L.; Stevens, J. F.; Ivanov, V.; McCall, M.; Frei, B.;
Deinzer, M. L.; Buhler, D. R. J. Agric. Food Chem. 2000, 48, 3876–3884.
(38) Uyeda, C.; Jacobsen, E. N. J. Am. Chem. Soc. 2008, 130, 9228–9229.
(39) Hoarau, C.; Pettus, R. R. T. Synlett 2003, 127–137.
(40) Bagnell, L.; Cablewski, T.; Strauss, C. R.; Trainor, R. W. J. Org.
Chem. 1996, 61, 7355–7359.
(41) Narayan, S.; Muldoon, J.; Finn, M. G.; Fokin, V. V.; Kolb, H. C.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2005, 44, 3275–3279.
(42) Majumdar, K. C.; Alam, S.; Chattopadhyay, B. Tetrahedron
2008, 64, 597–643.
(43) Knowles, J. R.; Copley, S. D. J. Am. Chem. Soc. 1987, 109,
5008–5013.
(44) Grieco, P. A. Aldrichimica Acta 1991, 24, 59–66.
(45) White, W. N.; Wolfarth, E. U. J. Org. Chem. 1970, 35,
2196–2199.
(46) Jung, M. E.; Piizzi, G. Chem. Rev. 2005, 105, 1735–1766.
(47) Harfenist, M.; Thom, E. J. Org. Chem. 1972, 37, 841–848.
(48) Goering, H. L.; Jacobson, R. R. J. Am. Chem. Soc. 1958, 80,
3277–3285.
(49) Zou, H. X.; Xie, X.; Zheng, X. D.; Li, S. M. Appl. Microbiol.
Biotechnol. 2011, 89, 1443–1451.
(50) Donia, M. S.; Schmidt, E. W.Comprehensive Natural Products
Chemistry II; Mander, L., Liu, H.-W., Eds.; Elsevier: Oxford, U.K., 2008;
Vol. 2, pp 539558.
(51) Kremer, A.; Li, S. M. Microbiology 2010, 156, 278–286.
(52) Uyeda, C.; R€otheli, A. R.; Jacobsen, E. N. Angew. Chem., Int. Ed.
2010, 49, 9753–9756.
(53) Andrews, P. R.; Smith, G. D.; Young, I. G. Biochemistry 1973,
12, 3492–3498.
(54) Woodside, A. B.; Huang, Z.; Poulter, C. D. Org. Synth. 1988,
66, 211–215.
(55) Woodside, A. B.; Huang, Z.; Poulter, C. D. Org. Synth. 1993,
8, 616–620.
(56) Davisson, V. J.; Woodside, A. B.; Neal, T. R.; Stremler, K. E.;
Muehlbacher, M.; Poulter, C. D. J. Org. Chem. 1986, 51, 4768–4779.
(57) Stankova, I. G.; Videnov, G. I.;Golovinsky, E. V.; Jung, G.J. Peptide
Sci. 1999, 5, 392–398.
13705
dx.doi.org/10.1021/ja205458h |J. Am. Chem. Soc. 2011, 133, 13698–13705