Efforts toward broadening the spectrum of arylomycin antibiotic activity

Article abstract of DOI:10.1016/j.bmcl.2013.08.026

New antibiotics are needed, and one source may be 'latent' antibiotics, natural products whose once broad-spectrum activity is currently limited by the evolution of resistance in nature. We have identified a potential class of latent antibiotics, the arylomycins, which are lipopeptides with a C-terminal macrocycle that target signal peptidase and whose spectrum is limited by a resistance-conferring mutation in many bacteria. Herein, we report the synthesis and evaluation of several arylomycin derivatives, and demonstrate that both C-terminal homologation with a glycyl aldehyde and addition of a positive charge to the macrocycle increase the activity and spectrum of the arylomycin scaffold.

Full text of DOI:10.1016/j.bmcl.2013.08.026

Bioorganic & Medicinal Chemistry Letters 23 (2013) 5654–5659
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Efforts toward broadening the spectrum of arylomycin antibiotic
activity
Jian Liu a, Peter A. Smith b, Danielle Barrios Steed a, Floyd Romesberg a,
⇑
a Department of Chemistry, The Scripps Research Institute, 10550 North Torrey Pines Road, La Jolla, CA 92037, USA
b RQx Pharmaceuticals, Inc., 11099 North Torrey Pines Road, La Jolla, CA 92037, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
New antibiotics are needed, and one source may be ‘latent’ antibiotics, natural products whose once
broad-spectrum activity is currently limited by the evolution of resistance in nature. We have identified
a potential class of latent antibiotics, the arylomycins, which are lipopeptides with a C-terminal macro-
cycle that target signal peptidase and whose spectrum is limited by a resistance-conferring mutation in
many bacteria. Herein, we report the synthesis and evaluation of several arylomycin derivatives, and
demonstrate that both C-terminal homologation with a glycyl aldehyde and addition of a positive charge
to the macrocycle increase the activity and spectrum of the arylomycin scaffold.
Received 27 June 2013
Revised 24 July 2013
Accepted 5 August 2013
Available online 14 August 2013
Keywords:
Signal peptidase
Protein secretion
Latent antibiotic
C-terminal electrophile
Ó 2013 Elsevier Ltd. All rights reserved.
The ever-growing threat of antibiotic resistant bacteria contin-
ues to motivate efforts to discover and develop new antibiotics for
clinical use, in particular, those acting via novel mechanisms and
possessing a broad-spectrum of activity.1,2 Natural product antibi-
otics benefit from eons of evolution and thus provide ideal scaf-
folds for optimization, and indeed, the majority of clinically used
antibiotics were derived from natural sources.3,4 However, new
classes of natural product antibiotics with broad-spectrum activity
have become difficult to identify,5 and the more commonplace,
narrow-spectrum compounds are commonly thought to be com-
promised by intrinsic limitations of their scaffold, such as poor
penetrance, or of their target, such as insufficient conservation or
essentiality. Thus, it is inferred that these scaffolds never had
broad-spectrum activity, and that it would likely be challenging
to optimize them to have sufficient spectrum for development.
However, some of these scaffolds may be narrow spectrum due
to the previous evolution of resistance during the same natural
arms race that drove their development. We predict that these for-
merly more broad-spectrum scaffolds may be more amenable to
optimization, just as antibiotics compromised during clinical use
have been successfully ‘re-optimized,’ resulting in the many
next-generation variants used in the clinic today. We refer to such
natural products as ‘latent’ antibiotics.6,7
several derivatives.8–10 Discovered as secondary metabolites of a
strain of Streptomyces in 2002, the arylomycins are five or six res-
idue peptides with a C-terminal hydroxyphenyl glycine (Hpg)-ala-
nine–tyrosine tripeptide macrocycle and an N-terminal aliphatic
chain (Fig. 1).11–13 In vitro, the arylomycins were shown to inhibit
type I bacterial signal peptidase (SPase), an essential Ser-Lys dyad
protease that is highly conserved, positioned on the outer leaflet of
the inner membrane, and essential due to its role in removing the
N-terminal leader peptide that targets proteins for secretion via
the general secretory pathway.14–18 Although only a few strains
of bacteria were tested, the arylomycins were deemed to have a
narrow spectrum of activity and were not considered to be prom-
ising for development. However, after a wider survey of activity,
we discovered not only that the arylomycins have potent activity
against Staphylococcus epidermidis,8 but also that this pathogen
evolves resistance to the arylomycins by mutation of Ser29 to
Pro in the target SPase and that most bacteria that are inherently
resistant already possess an SPase with Pro at the analogous posi-
tion (i.e., position 84 in Escherichia coli).6,19 Moreover, we found
that a wide variety of both Gram-positive and Gram-negative bac-
teria with SPases lacking an analogous Pro are sensitive to the aryl-
omycins, and that mutation of the Pro to Ser is sufficient to render
Staphylococcus aureus, Escherichia coli, and Pseudomonas aeruginosa
sensitive to the arylomycins. Finally, a broad survey of diverse bac-
teria revealed that sensitivity is most correlated with the absence
of an analogous Pro and not with phylogenetic relatedness. Based
on this data, we hypothesized that the arylomycins are latent
antibiotics and that their scaffold might be optimized for
broad-spectrum activity.
We recently reported the synthesis of several members of a no-
vel class of natural product antibiotics, the arylomycins, as well as
⇑
Corresponding author. Tel.: +1 858 784 7290; fax: +1 858 784 7472.
E-mail address: floyd@scripps.edu (F. Romesberg).
0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2013.08.026

J. Liu et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5654–5659
5655
Figure 1. The arylomycin family of natural product antibiotics.
To begin to probe the potential of the arylomycins for optimiza-
tion, we initially focused on modifications of the lipopeptide tail,9
because it is interactions with the lipopeptide tail that are dis-
rupted by the resistance-conferring mutation.6 While derivatives
were identified that gained activity against S. aureus,9 a more gen-
eral strategy for the development of peptidase inhibitors is the
introduction of a suitably positioned moiety that mimics the
acylation transition state or an electrophilic moiety that forms a
covalent bond with the catalytic nucleophile. However, while the
C-terminus of the arylomycins binds SPase in a position that is
proximal to the catalytic nucleophile, the naturally present carbox-
ylate engages in hydrogen-bonding and ionic interactions10,20 that
are predicted to be essential for binding,21 suggesting that the
C-terminus might not be amenable to modification.
In addition to target affinity, target accessibility can limit the
spectrum of an antibiotic.22 With Gram-positive bacteria, the cell
wall is comprised predominantly of peptidoglycan and negatively
charged polymers such as teichoic acids,23,24 and is the only barrier
separating SPase from potential inhibitors. While the cell wall is
not usually thought to act as a barrier to small molecules
(<ꢀ50,000 Da),25 several studies suggest that the physicochemical
properties of a small molecule can affect its rate of diffusion
through the cell wall, and thus it is possible that optimization of
an inhibitor’s physicochemical properties could broaden its
Gram-positive spectrum.26 Target access is particularly challenging
with Gram-negative bacteria, and although we previously reported
that the arylomycins penetrate the Gram-negative outer mem-
brane,6 the efficiency of this process is unclear, as is the mecha-
nism. In this work we report the first studies directed at
modifying the arylomycin C-terminus to optimize interactions
with the catalytic residues of SPase and at exploring the properties
affecting penetration and access to SPase in whole cells.
mediated oxidative cleavage. Compound 6 was similarly prepared
from the homologated arylomycin derivative, which was synthe-
sized from the protected macrocycle via Arndt–Eistert homologa-
tion, followed by global deprotection with BBr3 and oxidative
cleavage with NaIO4 (Scheme 2).
The biochemical activity (Table 1) of each derivative was evalu-
ated by determining the half maximal inhibitory concentration
(IC50) with E. coli and S. aureus SPases (see Supplementary Data
for details). The IC50 for the parental arylomycin A–C16 is 70 nM
with the E. coli enzyme and 1800 nM with the S. aureus enzyme.
Relative to the parent arylomycin, the phosphonic acid 1, a poten-
tial transition state mimic, lost ꢀ10-fold activity against both pro-
teins. Phosphonates, particularly those with electrophilic
substituents such as diaryl phosphonates, can potentially react
irreversibly with active site nucleophiles; however, all of the phos-
phonates tested, including diaryl phosphonate 4, lost virtually all
affinity for both proteins. In contrast, compound 5, which intro-
duces a glycyl aldehyde at a position analogous to that occupied
by the scissile bond of natural peptide substrate, was fivefold more
potent than arylomycin A–C16 against the E. coli protein and 40-
fold more potent against the S. aureus protein. This gain in activity
is clearly dependent on the specific positioning of the aldehyde, as
the homologated variant 6 lost virtually all activity.
Antibacterial activity (Table 2) was measured for each deriva-
tive by determining the minimal inhibitory concentration (MIC)
against wild type strains of S. epidermidis (RP62A), S. aureus (meth-
icillin sensitive S. aureus (MSSA) strain 8325 and methicillin resis-
tant S. aureus (MRSA) strain USA300) and E. coli (MG1655), as well
as against mutant strains of S. aureus 8325 and E. coli MG1655 that
were rendered sensitive to the arylomycins via mutation of the
resistance-conferring SPase Pro residue to Ser (S. aureus:
spsB(P29S)) or Leu (E. coli: lepB(P84L)). As reported previously,
arylomycin A–C16 is potently active against S. epidermidis, but not
against MSSA, MRSA, or E. coli.6 Relative to the parent arylomycin
A–C16, derivatives 1–4 lost activity against all of the strains of bac-
teria tested, although the loss was somewhat smaller with 1
against S. epidermidis, as well as against the sensitized strains of
E. coli and P. aeruginosa. In contrast, derivative 5 retained activity
against wild-type S. epidermidis (2 mg/mL), as well as both of the
sensitized strains (MICs of 8 and 4 mg/mL, respectively), and most
significantly, gained activity against MRSA (16 mg/mL, relative to
no detectable activity for the parent arylomycin A–C16).
C-terminal derivatization. To begin to explore the potential of C-
terminal modifications of the arylomycin scaffold, we derivatized
arylomycin A–C16 (Fig. 1), as it has been the most extensively char-
acterized.9 We thus synthesized compounds 1–6, where the C-ter-
minal carboxylate is replaced with a phosphonate or aldehyde
moiety (Scheme 1 and Supplementary data). Compounds 1–4 were
synthesized via coupling of the protected arylomycin A–C16 with
the corresponding aminomethylphosphonate (or the benzyl phos-
phonic ester in the case of compound 1), followed by deprotection
via Pd-catalyzed hydrogenation. Compound 5 was prepared from
protected arylomycin A–C16 by coupling with ( )-3-amino pro-
pane-1,2-diol, followed by Pd-catalyzed hydrogenation, and NaIO4
Hpg derivatization. The most interesting result from the
C-terminal modified derivatives is the significant increase in

5656
J. Liu et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5654–5659
Scheme 1. Reagents and conditions: (a) (i) 20% TFA/DCM, (ii) R’OH, DEPBT, NaHCO3, THF, (iii) Me3SnOH, DCE, 80 °C; (b) (i) (1) For 1 and 4, HOBt, EDC, TEA, H2NCH2PO(OR1)
(1: R1 = Bn; 4: R1 = Ph), DMF; (2) For 2 and 3, DEPBT, NaHCO3, H2NCH2PO(OR1) (2: R1 = Me; 3: R1 = Et), THF, (ii) 10% Pd/C, H2, AcOH/MeOH; (c) (i) iBuOCOCl, NMM,
3-aminopropane-1,2-diol, THF, (ii) Pd/C, H2, AcOH/MeOH, (iii) NaIO4, THF/H2O.
Scheme 2. Reagents and conditions: (a) (i) LiOH(aq), THF, rt, (ii) iBuOCOCl, NMM, CH2N2, THF, (iii) CF3CO2Ag, MeOH, rt; (b) (i) 20% TFA/DCM, (ii) R0OH, DEPBT, NaHCO3, THFs,
i
(iii) Me3SnOH, DCE, 80 °C; (c) (i) BuOCOCl, NMM, 3-aminopropane-1,2-diol, THF, (ii) BBr3, DCM, ꢁ78 °C to rt, (iii) NaIO4, THF/H2O.
biochemical activity of aldehyde of 5; however, while this trans-
lates into increased activity against MRSA, and activity is generally
retained against other bacteria, it does not translate into increased
activity against the other strains tested. It is possible that the re-
moval of the sole charged group of the arylomycin scaffold affects
its physicochemical properties in a way that decreases target ac-
cess. To test this hypothesis, and also to explore modifications at
a second site within the macrocycle, three pairs of compounds
with modifications to the macrocycle core Hpg were synthesized
(Schemes 3–5 and Supplementary data). Compounds 7 and 8 were
synthesized to examine the effects of an added negative charge; 9
and 10, to examine the effects of an added positive charge; and 11
and 12, to examine the effects of an added hydroxyl group. In
general, synthesis of these derivatives proceeded via phosphoryla-
tion or alkylation of the protected macrocycle, followed by
Table 1
Biochemical activity (IC50s (ng/mL)) of arylomycin derivatives against purified SPase
Compd
E. coli
820
20,000
>20,000
>20,000
15
S. aureus
1
2
3
4
5
6
7
8
9
16,000
>20,000
>20,000
>20,000
49
>20,000
1900
180
3200
950
140
230
51
84
70
9
10
11
12
320
3900
61
Table 2
Antibacterial activity (MICs (
l
g/mL)) of arylomycin derivatives
Compd
S. epidermidis
S. aureus
MRSA
E. coli
S. aureus (spsB(P29S))
E.coli (lepB(P84L))
E. coli (lptD)
E. coli (lptD lepB(P84L))
a
A–C16
1
2
3
4
5
6
7
8
0.25
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
64
>64
>64
>64
16
>64
>64
>64
32
>64
>64
64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
>64
4
>64
>64
>64
>64
8
>64
>64
32
4
32
8
8
0.5
4
64
>64
>64
4
>64
1
16
0.5
2
16
>64
>64
>64
>64
>64
>64
32
64
4
>64
16
16
0.02
0.1
2
2
8
8
8
8
>64
2
32
0.3
1
0.1
1
0.06
2
0.03
0.3
0.004
0.06
0.02
0.03
9
10
11
12
1
2
1
2
a
6
Arylomycin A–C16, included for reference.

J. Liu et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5654–5659
5657
Scheme 3. Reagents and conditions: (a) CCl4, DMAP, DIEA, (BnO)2P(O)H, ACN, rt; (b) (i) 15% TFA/DCM, 0 °C, (ii) R’OH, DEPBT, NaHCO3,THF, rt; (c) Pd/C, H2, HCO2H/MeOH, rt.
i
Scheme 4. Reagents and conditions: (a) (i) LiOH (aq), THF, (ii) NMM, BuOCOCl, allylamine; (b) CCl4, DMAP, DIEA, (BnO)2P(O)H, ACN; (c) (i) 15% TFA/DCM, 0 °C, (ii) R’OH,
DEPBT, NaHCO3,THF, rt; (d) K2OsO4ꢂH2O, K3Fe(CN)6, then NaIO4, THF/H2O; (e) Pd/C, H2, HCO2H/MeOH, rt.
Scheme 5. (a) ICH2CH2R3, base, acetone, reflux (for 9 and 10, R3 = NHCbz, base = Cs2CO3; For 11 and 12, R3 = OBn, base = K2CO3); (b) (i) 20% TFA/DCM, (ii) R’OH, DEPBT,
NaHCO3, THF, (iii) Me3SnOH, DCE, 80 °C; (c) Pd/C, H2, HCO2H/MeOH, rt; (d) (i) iBuOCOCl, N-methylmorpholine, THF, 3-aminopropane-1,2-diol, 0 °C to rt, (ii) Pd/C, H2, AcOH/
MeOH, (iii) NaIO4, THF, H2O.
coupling to the lipopeptide using DEPBT, and then global deprotec-
tion via Me3SnOH mediated hydrolysis and Pd/C catalyzed
hydrogenation. The synthesis of 7 required C-terminal benzyl
(as opposed to methyl) protection to prevent decomposition of
the phosphate group during hydrolysis. In the case of 8, a modified
route was employed that involved first installing the C-terminal
allylamine, followed by phosphorylation and lipopeptide coupling,
and then finally dihydroxylation of the olefin and NaIO4 mediated
cleavage to produce the desired aldehyde. For 10 and 12, the
aldehyde moiety was installed as described above prior to
hydrogenation.
to arylomycin A–C16, the derivatives installed on the natural
C-terminal carboxylic acid scaffold (7, 9, and 11) had little effect
on binding against either SPase (IC50s within ꢀtwofold). Similarly,
when installed on the glycyl aldehyde backbone, the hydroxy-ethyl
moiety of 12 had little effect on the affinity for either SPase.
However, the ethyl amine derivative 10 bound to both SPases with
sixfold lower affinity, and the phosphate derivative 8 bound to the
S. aureus SPase with fourfold lower affinity and to the E. coli SPase
with 15-fold lower affinity.
The antibacterial activity of compounds 7–12 was determined
using the same strains of bacteria described above (Table 2). In
addition, to more directly examine the effects on penetration
through the Gram-negative outer membrane, we measured
The biochemical activity of compounds 7–12 was again deter-
mined for both the E. coli and S. aureus SPases (Table 1). Relative

5658
J. Liu et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5654–5659
activity of arylomycin A–C16 and each derivative against
a
phosphate derivative 7 relative to arylomycin A–C16, despite
similar affinity for the S. aureus SPase. Further exploration of this
effect is warranted and should include varying the position of the
charge and the addition of multiple positively charged moieties,
for example, to the macrocycle and throughout the peptide back-
bone. Regardless, along with the previously reported lipopeptide
modifications that bestow the scaffold with Gram-positive activity,
the identification of two macrocycle derivatives with broadened
Gram-positive activity (compounds 5 and 9) is consistent with
the hypothesis that the arylomycins are latent antibiotics and are
well suited for further optimization of Gram-positive activity.
Arylomycin A–C16, as well as every derivative examined, is sig-
nificantly more active against E. coli harboring the lepB(P84L)
mutation compared to its isogenic counterpart. Interestingly, other
methods of membrane permeabilization, such as moderate
concentrations of polymyxin B nonapeptide, have little effect on
arylomycin sensitivity,6 suggesting the penetrance is complex
and bimodal, with relatively little penetration across an intact
outer-membrane and increased penetrance occurring only after
significant disruption of the lipopolysaccharide barrier. Due to
the detection limits of the assay, the precise fold-changes in sensi-
tivity are not available for all compounds, but in general, the
hyperpermeable phenotype results in 25- to 125-fold greater sus-
ceptibility. Interestingly, the relative increase in sensitivity upon
permeabilization is larger for the derivatives than for arylomycin
A–C16 itself. The relative increase is the greatest with compound
9, and because the biochemical activity and the antibacterial activ-
ity against the unpermeabilized strains of arylomycin A–C16 and 9
are similar, the difference appears to result from increased pene-
tration of 9 with the hyperpermeable strain. This again may result
from ionic interactions that localize the inhibitor to the cell wall,
possibly predisposing it to take advantage of its disruption, and/
or the Donnan effect,29–31 which drives the accumulation of posi-
tively charged compounds within the perisplasmic space. Regard-
less of the exact mechanism of arylomycin penetration, the data
emphasize that the barriers posed by the Gram-negative outer
membrane remain a clear challenge for developing arylomycins
with potent Gram-negative activity.
hyperpermeable strain of E. coli harboring an lptD mutation, which
renders the outer-membrane permeable to large molecules such as
vancomyin and rifampin.27 Also, to examine the combined effects
of outer-membrane disruption and SPase sensitization, we mea-
sured activity of each compound against a strain of E. coli that har-
bors both the lptD mutation and the lepB(P84L) mutation that
confers arylomycin sensitivity. Arylomycin A–C16 shows significant
activity against these permeabilized strains (MICs = 16 and
0.02 mg/mL, respectively), clearly demonstrating that the
resistance conferring Pro and compound exclusion by the outer-
membrane both play significant roles in the low susceptibility of
wild-type E. coli.
Compounds 7, 9, and 11 generally retained activity against the
bacteria against which the parent compound is active. The only
exception where activity was lost was with compound 7 against
sensitized S. aureus. An exception where activity was gained was
compound 9, which gained measurable activity against MRSA. In
addition, while 9 did not gain activity against either sensitized
Gram-negative strain, it gained more activity than the parent com-
pound or any other derivative against both the sensitized and
unsensitized strains upon permeabilization. As predicted from
the biochemical data, the same modifications installed within the
C-terminal aldehyde scaffold of 5 (compounds 8, 10 and 12) gener-
ally resulted in similar or decreased antibacterial activity.
To examine the optimization of the arylomycin scaffold for
broad-spectrum antibacterial activity we pursued two routes.
The first route was based on increasing the affinity of the scaffold
for its peptidase target via introduction of a phosphonate transition
state mimic or an electrophile that might form a covalent bond
with the catalytic nucleophile. The second route was based on
increasing target access via the introduction of a charged or hydro-
gen-bonding substituent to the macrocycle. In general, C-terminal
homologation with an aminomethylphosphonic acid or ester ab-
lated virtually all affinity of the arylomycin scaffold for SPase. This
suggests that the modifications either eliminate an essential inter-
action mediated by the free carboxylate, or introduce steric clashes
within the binding site. However, homologation with a glycyl alde-
hyde increased affinity by 5- and 40-fold for the E. coli and S. aureus
SPases, respectively, resulting in low nM inhibition of both en-
zymes and also imparting the scaffold with antibacterial activity
against MRSA.
It is unclear whether the glycyl aldehyde mediates its effects via
the formation of a covalent bond with SPase. However, the data
suggest that its presence at least induces a rather different mode
of binding. This conclusion is drawn from the very different effects
on SPase binding of adding a substituent to the Hpg of the natural
and glycyl aldehyde homologated scaffolds. The natural scaffold is
generally insensitive to the modifications, which is consistent with
crystal structure data that show that the Hpg phenolic hydroxyl
group is directed away from the protein and into solvent.10,28 How-
ever, both the phosphate and ethyl amine derivatizations result in
significantly reduced affinity for SPase when made within the gly-
cyl aldehyde scaffold. Regardless of the exact mechanism by which
the glycyl aldehyde influences SPase binding, it is clear that the
free C-terminal carboxylate is not required for binding, and that
other C-terminal modifications should be explored.
Acknowledgments
This work was supported by the Office of Naval Research
(Awards N000140310126 and N000140810478) and the National
Institutes of Health (AI081126).
Supplementary data
Supplementary data associated with this (synthetic procedures,
biological methods, and spectroscopic data (1H, 13C, MS, and
HRMS)) article can be found, in the online version, at http://
dx.doi.org/10.1016/j.bmcl.2013.08.026.
References and notes
1. Hancock, R. E. Mol. Syst. Biol. 2007, 3, 142.
2. Fischbach, M. A.; Walsh, C. T. Science 2009, 325, 1089.
3. von Nussbaum, F.; Brands, M.; Hinzen, B.; Weigand, S.; Habich, D. Angew.
Chem., Int. Ed. 2006, 45, 5072.
Among the Hpg modifications examined, the most promising is
the ethyl amine of 9, which despite not significantly altering the
affinity of the scaffold for S. aureus SPase, bestowed it with modest
anti-MRSA activity. This activity may result from the positively
charged primary amine introducing ionic interactions with the
negatively charged teichoic acids of the Gram-positive cell wall
and thus increasing the effective concentration of the arylomycin
proximal to its target. This is also consistent with the loss of
anti-S. aureus activity observed with the negatively charged
4. Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2007, 70, 461.
5. Payne, D. J.; Gwynn, M. N.; Holmes, D. J.; Pompliano, D. L. Nat. Rev. Drug Disc.
2007, 6, 29.
6. Smith, P. A.; Roberts, T. C.; Romesberg, F. E. Chem. Biol. 2010, 17, 1223.
7. Tan, Y. X.; Romesberg, F. E. MedChemComm 2012, 3, 916.
8. Roberts, T. C.; Smith, P. A.; Cirz, R. T.; Romesberg, F. E. J. Am. Chem. Soc. 2007,
129, 15830.
9. Roberts, T. C.; Schallenberger, M. A.; Liu, J.; Smith, P. A.; Romesberg, F. E. J. Med.
Chem. 2011, 54, 4954.
10. Liu, J.; Luo, C.; Smith, P. A.; Chin, J. K.; Page, M. G.; Paetzel, M.; Romesberg, F. E.
J. Am. Chem. Soc. 2011, 133, 17869.

J. Liu et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5654–5659
5659
11. Holtzel, A.; Schmid, D. G.; Nicholson, G. J.; Stevanovic, S.; Schimana, J.;
21. Li, T.; Froeyen, M.; Herdewijn, P. J. Mol. Graphics Modell. 2008, 26, 813.
22. Poole, K. J. Antimicrob. Chemother. 2005, 56, 20.
23. Weidenmaier, C.; Peschel, A. Nat. Rev. Microbiol. 2008, 6, 276.
24. Swoboda, J. G.; Campbell, J.; Meredith, T. C.; Walker, S. ChemBioChem 2010, 11,
35.
25. Demchick, P.; Koch, A. L. J. Bacteriol. 1996, 178, 768.
26. Herbert, S.; Bera, A.; Nerz, C.; Kraus, D.; Peschel, A.; Goerke, C.; Meehl, M.;
Cheung, A.; Gotz, F. PLoS Pathog. 2007, 3, e102.
Gebhardt, K.; Fiedler, H. P.; Jung, G. J. Antibiot. 2002, 55, 571.
12. Schimana, J.; Gebhardt, K.; Holtzel, A.; Schmid, D. G.; Sussmuth, R.; Muller, J.;
Pukall, R.; Fiedler, H. P. J. Antibiot. 2002, 55, 565.
13. Kulanthaivel, P.; Kreuzman, A. J.; Strege, M. A.; Belvo, M. D.; Smitka, T. A.;
Clemens, M.; Swartling, J. R.; Minton, K. L.; Zheng, F.; Angleton, E. L.; Mullen, D.;
Jungheim, L. N.; Klimkowski, V. J.; Nicas, T. I.; Thompson, R. C.; Peng, S. B. J. Biol.
Chem. 2004, 279, 36250.
14. Paetzel, M.; Dalbey, R. E.; Strynadka, N. C. Pharmacol. Ther. 2000, 87, 27.
15. Paetzel, M.; Karla, A.; Strynadka, N. C.; Dalbey, R. E. Chem. Rev. 2002, 102, 4549.
16. Date, T. J. Bacteriol. 1983, 154, 76.
27. Sampson, B. A.; Misra, R.; Benson, S. A. Genetics 1989, 122, 491.
28. Luo, C.; Roussel, P.; Dreier, J.; Page, M. G.; Paetzel, M. Biochemistry 2009, 48,
8976.
17. Cregg, K. M.; Wilding, I.; Black, M. T. J. Bacteriol. 1996, 178, 5712.
18. Zhang, Y. B.; Greenberg, B.; Lacks, S. A. Gene 1997, 194, 249.
19. Smith, P. A.; Powers, M. E.; Roberts, T. C.; Romesberg, F. E. Antimicrob. Agents
Chemother. 2011, 55, 1130.
29. Donnan, F. G.; Harris, A. B. J. Chem. Soc. Trans. 1911, 99, 1554.
30. Tanford, C. Physical Chemistry of Macromolecules; Wiley: New York, 1961.
31. Bolton, G. R.; Boesch, A. W.; Basha, J.; Lacasse, D. P.; Kelley, B. D.; Acharya, H.
Biotechnol. Prog. 2011, 27, 140.
20. Paetzel, M.; Goodall, J. J.; Kania, M.; Dalbey, R. E.; Page, M. G. P. J. Biol. Chem.
2004, 279, 30781.