Effects of Cyclic Lipodepsipeptide Structural Modulation on Stability, Antibacterial Activity, and Human Cell Toxicity

Article abstract of DOI:10.1002/cmdc.201200016

Bacterial infections are becoming increasingly difficult to treat due to the development and spread of antibiotic resistance. Therefore, identifying novel antibacterial targets and new antibacterial agents capable of treating infections by drug-resistant bacteria is of vital importance. The structurally simple yet potent fusaricidin or LI-F class of natural products represents a particularly attractive source of candidates for the development of new antibacterial agents. We synthesized 18 fusaricidin/LI-F analogues and investigated the effects of structure modification on their conformation, serum stability, antibacterial activity, and toxicity toward human cells. Our findings show that substitution of an ester bond in depsipeptides with an amide bond may afford equally potent analogues with improved stability and greatly decreased cytotoxicity. The lower overall hydrophobicity/amphiphilicity of amide analogues in comparison with their parent depsipeptides, as indicated by HPLC retention times, may explain the dissociation of antibacterial activity and human cell cytotoxicity. These results indicate that amide analogues may have significant advantages over fusaricidin/LI-F natural products and their depsipeptide analogues as lead structures for the development of new antibacterial agents.

Full text of DOI:10.1002/cmdc.201200016

MED
DOI: 10.1002/cmdc.201200016
Effects of Cyclic Lipodepsipeptide Structural Modulation
on Stability, Antibacterial Activity, and Human Cell Toxicity
Nina Bionda,[a, b] Maciej Stawikowski,[a] Roma Stawikowska,[a] Marꢀ Cudic,[a] Fabian Lꢁpez-
Vallejo,[a] Daniela Treitl,[b] Josꢀ Medina-Franco,[a] and Predrag Cudic*[a]
Bacterial infections are becoming increasingly difficult to treat
due to the development and spread of antibiotic resistance.
Therefore, identifying novel antibacterial targets and new anti-
bacterial agents capable of treating infections by drug-resistant
bacteria is of vital importance. The structurally simple yet
potent fusaricidin or LI-F class of natural products represents
a particularly attractive source of candidates for the develop-
ment of new antibacterial agents. We synthesized 18 fusarici-
din/LI-F analogues and investigated the effects of structure
modification on their conformation, serum stability, antibacteri-
al activity, and toxicity toward human cells. Our findings show
that substitution of an ester bond in depsipeptides with an
amide bond may afford equally potent analogues with im-
proved stability and greatly decreased cytotoxicity. The lower
overall hydrophobicity/amphiphilicity of amide analogues in
comparison with their parent depsipeptides, as indicated by
HPLC retention times, may explain the dissociation of antibac-
terial activity and human cell cytotoxicity. These results indi-
cate that amide analogues may have significant advantages
over fusaricidin/LI-F natural products and their depsipeptide
analogues as lead structures for the development of new anti-
bacterial agents.
Introduction
Antibiotic resistance is increasing at a faster rate than the de-
velopment of new antibiotics,[1–4] and the prevalence of multi-
drug-resistant (MDR) bacteria has become a global public
health problem.[4] The majority of life-threatening infections
worldwide are caused by the ESKAPE pathogens (Enterococcus
faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acineto-
bacter baumannii, Pseudomonas aeruginosa, and Enterobac-
ter spp.).[5,6] This group of bacteria is encountered in more than
40% of hospital-acquired infections, and is resistant to the ma-
jority of commonly used antibiotics. Therefore, the identifica-
tion of novel antibacterial targets and new antibacterial agents
capable of treating infections by drug-resistant bacteria is cru-
cial.
entering phase III clinical trials for the treatment of Clostridium
difficile-associated diarrhea.
Despite progress in the development of new antibacterial
agents, it is inevitable that resistant strains of bacteria will
emerge in response to widespread use of a particular antibiotic
and limit its usefulness. Structurally simple yet potent fusarici-
dins or the LI-F family of natural products (Figure 1) represent
particularly attractive candidates for the development of new
antibacterial agents capable of reverting infections caused by
MDR bacteria.[7,17–20]
Fusaricidins/LI-Fs are cyclic lipodepsipeptide antifungal anti-
biotics isolated from Paenibacillus sp. Their common structural
feature is the macrocyclic ring consisting of six amino acid resi-
dues, three of which, Thr1, d-allo-Thr4 (d-aThr4), and d-Ala6, are
conserved throughout the family, as well as the 15-guanidino-
3-hydroxypentadecanoic acid tail attached to the N-terminal
Thr1 residue by an amide bond. Fusaricidins/LI-Fs are cyclized
by a lactone bridge between the N-terminal Thr1 hydroxy
group and the C-terminal d-Ala6. Among isolated fusaricidin/LI-
F antibiotics, fusaricidin A or LI-F04a (Figure 1)[20] have shown
the most promising antimicrobial activity against a variety of
Naturally occurring cyclic lipodepsipeptides that contain one
or more ester bonds along with the amide bonds have
emerged as promising lead compounds for the discovery of
new antibiotics.[7–10] The biosynthesis of these peptides pro-
ceeds non-ribosomally and is catalyzed by a complex of multi-
functional enzymes termed non-ribosomal peptide synthases
(NRPSs). NRPSs have a unique modular structure in which each
module contains the requisite domains for the recognition and
activation of a single amino acid, generating a wide structural
and functional diversity of non-ribosomal peptides.[11] Within
this class of natural products, the cyclic lipodepsipeptide dap-
tomycin (Cubicin, Cubist Pharmaceuticals Inc.)[7,12,13] has already
been approved for use in the USA, European Union, and
Canada for the treatment of infections caused by MDR bacteri-
al strains. Ramoplanin (Nanotherapeutics Inc.)[7,14–16] is another
example of a cyclic lipodepsipeptide with the potential for re-
verting bacterial multidrug resistance. Ramoplanin is currently
[a] N. Bionda, Dr. M. Stawikowski, R. Stawikowska, Dr. M. Cudic,
Dr. F. Lꢀpez-Vallejo, Dr. J. Medina-Franco, Dr. P. Cudic
Torrey Pines Institute for Molecular Studies
11350 SW Village Parkway, Port St. Lucie, FL 34987 (USA)
E-mail: pcudic@tpims.org
[b] N. Bionda, D. Treitl
Department of Chemistry and Biochemistry
Florida Atlantic University, 777 Glades Road, Boca Raton, FL 33431 (USA)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201200016.
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
871

P. Cudic et al.
MED
for antimicrobial activity, improved serum stability, and separa-
tion of antibacterial activity from toxicity toward human cells.
Results
Solid-phase synthesis
The amino acid sequences and lipid tails of synthesized fusari-
cidin A/LI-F04a analogues 1–18 are shown in Figure 2. Ana-
logues 1–12 are depsipeptides that contain an ester bond be-
tween Thr1 and d-Ala6 or Gly6 residues. In analogue 13, natural-
ly occurring amino acid residues Thr1 and d-Ala6 are replaced
with Lys, thereby substituting the ester moiety as well as two
chiral centers, while keeping the same number of the atoms in
the ring. The remaining four cyclic analogues 14–17 have an
ester bond replaced with an amide bond by substituting Thr1
for Dap1. Analogue 18 is a linear version of 6 and was pre-
pared as a control. The synthesis of depsipeptide analogues 1–
12 is shown in Scheme 1.[24] In the first step, the C-terminal
amino acid Fmoc-d-Asp-OAllyl was attached to a PEG–PS-
based amide resin (TentaGel S RAM) via the side chain by using
a standard HBTU/HOBt/NMM synthetic protocol. Standard
Figure 1. Structures of the fusaricidin/LI-F family of compounds.
fungi, including the clinically important Candida albicans and
Cryptococcus neoformans, and against Gram-positive bacteria
such as S. aureus (MIC value range: 0.78–3.12 mgmLꢀ1). Fusarici-
dins/LI-Fs have not, however, shown activity against Gram-neg-
ative bacteria.[18,19] The mode of action of fusaricidins/LI-Fs is
still unknown. The total synthesis of the fusaricidin A/LI-F04a
natural product using a combination of solid- and solution-
phase synthetic approaches was recently reported by Co-
chrane et al.[21] Two research groups, Jensen and colleagues[22]
and Park and co-workers,[23] reported the identification and iso-
lation of the putative fusaricidin/LI-F synthetase gene, fusA,
from Paenibacillus polymyxa, opening the possibility for the de-
velopment of biosynthetic approaches toward this family of
naturally occurring cyclic lipodepsipeptides and their ana-
logues.
We previously reported a complete Fmoc solid-phase syn-
thesis of a fusaricidin A/LI-F04a analogue containing 12-guani-
dinododecanoic acid instead of the naturally occurring 15-gua-
nidino-3-hydroxypentadecanoic acid (Figure 1).[24] Total synthe-
sis of fusaricidin/LI-F antifungal antibiotics, and particularly un-
limited access to their synthetic analogues, represent impor-
tant initial steps toward full exploitation of their antimicrobial
potential. Herein we describe our efforts to determine the
structural aspects of fusaricidin/LI-F peptides that are required
Figure 2. Sequences of synthesized fusaricidin A/LI-F04a analogues. [a] Dif-
ferences among the sequences of naturally occurring fusaricidin A/LI-F04a
and synthetic analogues are highlighted in bold. [b] Rt: retention time ob-
tained by analytical RP HPLC; method: 2% solvent B for 0.5 min followed by
linear gradient 2!98% solvent B over 30 min, for which solvent A is 0.1%
TFA in H2O, and B is 0.08% TFA in CH3CN.
872
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882

Structural Modulation of Cyclic Lipodepsipeptides
MED
Scheme 1. Synthesis of fusaricidin A/LI-F04a analogues 1–17. Reagents and conditions: a) Fmoc-d-Asp-OAllyl and Fmoc-AA-OH, standard Fmoc SPPS deprotec-
tion and coupling protocols; b) Ac-Thr-OH, standard Fmoc SPPS deprotection and coupling protocols; c) Fmoc-Thr-OH, standard Fmoc SPPS deprotection and
coupling protocols; d) Fmoc-ADA-OH, standard Fmoc SPPS deprotection and coupling protocols; e) Fmoc-Lys-OH, standard Fmoc SPPS deprotection and cou-
pling protocols; f) Fmoc-Dap(Mtt)-OH, standard Fmoc SPPS deprotection and coupling protocols; g) Alloc-d-Ala-OH or Alloc-Gly-OH (4 equiv), DIC (4 equiv),
DMAP (0.2 equiv), CH2Cl2, RT, 18 h; h) TFA/CH2Cl2 (1% v/v), RT, 30 min; i) Pd(Ph3P)4 (0.1 equiv), HN(CH3)2·BH3 (4 equiv), CH2Cl2, RT, 2ꢃ10 min; j) PyBOP/HOBt/
DIEA (2:2:6 equiv), DMF, RT, 18 h; k) piperidine/DMF (20% v/v), RT, 25 min, N,N-bis(tert-butoxycarbonyl)thiourea (3 equiv), 2-chloro-1-methylpyridinium iodide
(3 equiv), TEA (4 equiv), DMF, RT, 18 h; l) piperidine/DMF (20% v/v), RT, 25 min; m) TFA/TIA/H2O (95:2.5:2.5 v/v/v), RT, 3 h.
Fmoc SPPS strategies were used throughout. The ester bond
was formed between the Thr1 side chain hydroxy group and
Alloc-d-Ala6-OH or Alloc-Gly6-OH carboxyl groups using DIC/
DMAP coupling conditions in CH2Cl2. To avoid potential epime-
rization during Alloc-d-Ala6-OH coupling, a catalytic amount of
DMAP (0.2 equiv) was used.[25,26] Under applied experimental
conditions no epimerization was observed as indicated by ana-
lytical RP HPLC (data not shown). The lipid tail, Fmoc-aminodo-
decanoic acid (Fmoc-ADA-OH), was incorporated into the
linear peptide precursor prior to d-Ala6/Gly6 coupling via ester
bond and on-resin cyclization in order to avoid an undesired
O!N acyl shift known to occur under basic conditions re-
quired for Fmoc removal.[24,27,28]
Amide analogues 13–17 were prepared by replacing the
Thr1 residue with Lys (compound 13) or Dap (14–17;
Scheme 1). For this purpose, an identical Fmoc SPPS strategy
was employed. In the case of analogue 13, upon synthesis of
the linear precursor, Allyl and Alloc protecting groups were re-
moved, and the peptide was cyclized through the Lys1 e-amino
group and the d-Asn5 a-carboxyl group. In the case of ana-
logues 14–17, selective removal of the Mtt protecting group
from Dap1 with 2% TFA in CH2Cl2 allowed coupling of Alloc-d-
Ala-OH through standard coupling protocols. The final synthet-
ic steps were performed as described above. Linear analogue
18 was prepared on 2-chlorotrityl chloride resin (Cl-TrtCl)[29]
using standard Fmoc chemistry, starting with the attachment
of Fmoc-d-Ala-OH via the carboxylic group. Loading of the
resin was determined to be 0.5 mmolgꢀ1.[30] The use of Cl-TrtCl
resin in this case afforded a linear peptide with a C-terminal
carboxylic group, identical to the hydrolysis product of 6. All
analogues were purified by preparative RP HPLC, and purity (ꢁ
95%) was confirmed with analytical RP HPLC and MALDI-TOF
MS.
After selective removal of Alloc and Allyl protecting groups
by treatment with Pd(Ph3P)4 and non-basic borane dimethyla-
mine complex as scavenger,[13] the linear peptide was cyclized
through an amide bond between d-Ala6 or Gly6 and d-Asn5
residues using PyBOP. The conversion of the lipid tail amino
group into the desired guanidino function was achieved by re-
moval of the Fmoc protecting group using a standard piperi-
dine deprotection protocol and subsequent treatment of the
peptidyl resin with N,N-bis(tert-butoxycarbonyl)thiourea fol-
lowed by Mukaiyama’s reagent, 2-chloro-1-methylpyridinium
iodide.[14] Final deprotection and cleavage from the resins was
carried out with a cleavage cocktail of TFA/TIA/H2O (95:2.5:2.5
v/v/v).
Conformational studies
The structural features of representative fusaricidin analogues
6 and 14 as well as the linear peptide 18 were monitored by
circular dichroism (CD) spectroscopy (Figure 3). CD spectra
were recorded in aqueous media as well as in less polar tri-
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
873

P. Cudic et al.
MED
Importantly, d-amino acid residues predominate the sequences
of depsipeptide 6, amide 14, and linear peptide 18, causing in-
version of their CD spectra (Figure 3). The CD spectrum of dep-
sipeptide 6 in 0.5% AcOH exhibits a minimum at ~197 nm and
a weak maximum ~220 nm, reminiscent of an inverted antipar-
allel b-sheet structure (Figure 3a). Water replacement with the
less polar 50% TFE/H2O mixture and 100% TFE resulted in
a drastic change in the CD spectra of depsipeptide 6. The
spectrum in 100% TFE shows a double minimum at 192 and
201 nm and a maximum at 225 nm. In addition to these
changes, a marked decrease in the intensity of the CD spec-
trum in TFE was also observed. All these spectral changes are
characteristic of overall b-sheet/b-turn structures.[39–41] b-
Sheet-/b-turn-containing small cyclic peptides are not unusual,
and examples can be found in gramicidin S and its ana-
logues.[40,42,43] CD spectra of amide analogue 14 differ markedly
from those of parent depsipeptide 6 in both aqueous and TFE
solutions, indicating significant differences in structural flexibili-
ty and conformations induced by ester-to-amide substitution.
The CD spectrum of 14 in 0.5% AcOH is characterized by
a maximum at 205 nm and minimum at 188 nm, whereas in
TFE the maximum at 205 nm is completely lost, and the inten-
sity of the spectral minimum at 188 nm is decreased and
slightly shifted toward shorter wavelengths (Figure 3b).
As expected, due to the lack of structural constraints, linear
peptide 18 shows different CD spectra in aqueous medium
and TFE. In 0.5% AcOH, peptide 18 yields characteristics of an
inverted unordered structure with
a weak maximum at
195 nm, whereas in TFE the weak minimum at 200 nm indi-
cates the presence of an inverted type I b-turn (Figure 3c).[39,41]
Besides conformational changes, ester-to-amide substitution in
cyclic peptides may also alter overall hydrophobicity and am-
phiphilicity, structural characteristics that are important for the
biological activities of peptides.[45–49] Typically, the RP HPLC re-
tention times (Rt) of peptides are used as indications of their
overall hydrophobicity.[43,44] Rt values of synthesized fusaricidi-
n A/LI-F04a analogues, listed in Figure 2, show that amide ana-
logues are less hydrophobic than their parent depsipeptides,
and this is expected due to the loss of a Thr1 methyl group.
The hydrophobicity of synthetic analogues was further altered
by replacing the Val3 residue with Ala, Tyr, and Phe, based on
the amino acid sequences found in fusaricidin/LI-F natural
products. The observed change in overall hydrophobicity in
both depsipeptide and amide analogues, as expressed by Rt, is
in the order: Ala < Tyr < Val < Phe.
Figure 3. CD spectra of fusaricidin A/LI-F04 analogues a) 6, b) 14, and c) 18
in 0.5% AcOH (g), 50% TFE (a), or 100% TFE (c).
fluoroethanol (TFE), a membrane-mimicking solvent system.[31]
Besides its membrane mimicking properties, TFE is also known
to induce formation of stable conformations for peptides that
are otherwise unstructured in aqueous solutions.[32] To increase
the solubility of peptides in aqueous media and to inhibit po-
tential peptide aggregation at the concentrations required for
CD experiments, all analogues studied were dissolved in 0.5%
aqueous AcOH.[33,34]
To gain better insight into the conformational changes in-
duced by ester-to-amide substitution in fusaricidin A/LI-F04a
analogues, we conducted molecular dynamics (MD) simula-
tions. Low-energy conformations were generated by conforma-
tional analysis in Molecular Operating Environment (MOE)[46]
using a Monte Carlo search with the Generalized Born solva-
tion model implemented in MOE.[47] The amphiphilic moment
descriptor (m), which is an established measure of balance be-
tween hydrophilic and lipophilic moieties, was computed in
MOE for the lowest-energy conformations.[48]
Because small cyclic peptides still have considerable mobility
in their backbones,[35–38] it is reasonable to expect that the
ester-to-amide substitution in our case would also lead to sig-
nificant conformational changes. Therefore, we expected differ-
ent CD spectra for depsipeptide 6 and amide analogue 14 in
water and less polar TFE, due to differences in conformational
flexibility and the ability of TFE to promote intramolecular hy-
drogen bonds and to stabilize a preferential conformation.[32]
As shown in Figure 4, the results of MD studies are in quali-
tative agreement with the experimental CD data, indicating
874
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882

Structural Modulation of Cyclic Lipodepsipeptides
MED
Table 1. Antimicrobial activity of fusaricidin A/LI-F04a analogues.
Compd
MIC [mgmLꢀ1
]
S. aureus
S. aureus
S. aureus
S. epidermidis S. pyogenes
ATCC 29213 ATCC 33591 ATCC 700699 ATCC 27626 ATCC 19615
1
2
3
4
5
ND[a]
ND
ND
ND
16
>128
>128
>128
>128
16
>128
>128
>128
>128
16
ND
ND
ND
ND
16
ND
ND
ND
ND
32
6
8
8
16
8
16
7
8
ND
16
64
16
64
16
32
8
64
16
9
ND
ND
8
>128
ND
8
ND
8
ND
>128
32
64
8
64
64
16
ND
64
16
64
8
16
64
16
64
16
64
64
8
64
64
16
>64
16
64
10
11
12
13
14
15
16
17
18
>128
64
Figure 4. Representative structures of low-energy conformers of depsipep-
tide 6 and amide 14 in H2O and TFE (as indicated) along with their calculat-
ed amphiphilic moments (m; see Experimental Section for details).
16
64
8
32
16
ND
ND
ND
ND
64
>128
>128
>128
that depsipeptide 6 can adopt different conformations in polar
and nonpolar environments, whereas conformational differen-
ces for amide analogue 14 under the same conditions are less
pronounced, suggesting a more rigid peptide backbone struc-
ture. Furthermore, calculated amphiphilic moments for depsi-
peptide 6 and amide 14 are markedly different, with the depsi-
peptide being more amphiphilic in both solvent systems
(Figure 4).
[a] ND: not determined.
four (d-Val2, Val3, d-aThr4, and d-Ala6) out of six amino acids.
The MIC values obtained (Table 1) show that substitution of
Val3 with Ala (analogue 8) preserves the antibacterial activity of
the parent depsipeptide 6, whereas substitutions of d-aThr4
and d-Val2 with d-Ala (analogues 7 and 9, respectively) results
in a significant decrease in antimicrobial activity. In contrast, re-
placement of either d-Ala6 with Gly (analogue 10) or both Thr1
and d-Ala6 with eLys1 (analogue 13) leads to a complete loss of
antibacterial activity. No significant decrease in antimicrobial
activity was observed for the case in which d-aThr4 was re-
placed by d-Thr4: analogue 5. Quite interestingly, despite sig-
nificant changes in hydrophobicity, amphiphilicity, and confor-
mation caused by ester-to-amide substitution, the antibacterial
activities of amide analogues 14–17 parallel those of the
parent depsipeptides 6, 8, 11, and 12. As mentioned earlier, fu-
saricidins/LI-F are a family of naturally occurring cyclic lipodep-
sipeptide antibiotics consisting of 12 compounds in total, with
conserved Thr1, d-aThr4, and d-Ala6 residues.[7] Aliphatic nonpo-
lar amino acid residues are present at position 2 (d-Val, d-Ile,
and d-aIle), and polar amino acids d-Asn and d-Gln are present
at position 6 of the peptide sequence. Most diverse changes
occur at position 3, where Tyr, Val, Ile, Phe, and d-Ile are found.
Taking this into consideration, we synthesized fusaricidin/LI-F
analogues possessing Phe (analogues 11 and 16) and Tyr (ana-
logues 12 and 17) residues at position 3 in the sequence. Re-
placement of Val3 with polar Tyr resulted in a decrease in anti-
microbial activity, whereas substitution with nonpolar hydro-
phobic Phe did not affect antibacterial activity (Table 1). In ad-
dition, MIC data obtained for depsipeptides 6, 8, 11, and 12
show that an increase in overall hydrophobicity, as indicated
by HPLC retention times (Figure 2) does not correlate with
their antibacterial activities. Analogues 6, 8, and 11 showed
nearly identical antibacterial activities (within experimental
error) for tested bacterial strains, whereas analogue 12 showed
Structure–activity studies
To identify structural determinants for antibacterial activity and
cytotoxicity, a total of 18 fusaricidin A/LI-F04a analogues were
prepared (Figure 2). Structural variations of the natural product
include modification of the lipid tail, substitution of amino acid
residues including an alanine scan (Ala-scan), ester-to-amide
substitution, and a linear control peptide for comparison. The
antibacterial in vitro activity of synthesized fusaricidin A/LI-
F04a analogues 1–18, expressed as the minimum inhibitory
concentration (MIC, mgmLꢀ1), was determined for Gram-posi-
tive and Gram-negative bacteria using a standard microdilution
broth method in 96-well plates (Table 1).[49,50] Similar to the nat-
ural products, fusaricidin/LI-F analogues are active against
Gram-positive bacteria, including antibiotic-resistant strains,
and do not show activity against Gram-negative bacteria (data
not shown). To probe the role of the lipid tail in antibacterial
activity, depsipeptides containing an acetyl group (1 and 2),
12-ADA (3 and 4), or 12-GDA (5–12) attached to Thr1 were syn-
thesized. No antibacterial activity was observed for the ana-
logues containing an acetyl group or 12-ADA, even at the
highest concentration tested. In contrast, depsipeptide ana-
logue 6, containing 12-GDA, showed potent antibacterial activ-
ity (MIC=8 mgmLꢀ1) against Gram-positive bacteria (Table 1).
Ala-scan and several other substitutions at key amino acid resi-
dues were performed to reveal the role of each individual
amino acid in the antimicrobial activity of synthetic fusaricidin
analogues. However, the requirements of our solid-phase syn-
thetic approach[24] (Scheme 1) permitted the replacement of
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
875

P. Cudic et al.
MED
lower activity. In contrast, similar comparison of amide ana-
logues 14–16 showed different results. Amide analogues with
hydrophobic residues at position 3 of the peptide sequence,
such as Val and Phe (analogues 14 and 16), exhibited greater
antibacterial activity than analogues with more polar Ala and
Tyr residues (analogues 15 and 17), in this case indicating a cor-
relation between the peptides’ overall hydrophobicity and an-
tibacterial activity. Despite a much higher degree of conforma-
tional flexibility and complete sequence homology with bio-
logically active depsipeptide 6, the linear control peptide 18
did not show any biological activities in the described assays.
Notably, in this assay DMSO (up to 5% v/v final concentra-
tion) was added to increase the solubility of the tested fusarici-
din A/LI-F04a analogues. Therefore, potential aggregation due
to lower solubility of amide analogues at higher concentra-
tions may contribute to the observed hemolytic activity.[51–53]
Although DMSO is hemolytic,[54,55] our control experiments
showed that under the conditions used, DMSO does not cause
hemolysis.
To further assess the therapeutic potential of synthetic fusar-
icidin A/LI-F04a analogues, we tested the in vitro toxicity of an-
alogues 1, 4, 6, and 14 on the human liver embryonic cell line
WRL 68 and the HepG2 cancer cell line. The known cytotoxic
drug doxorubicin (Adriamycin)[56] was used as a positive con-
trol in these assays. Depsipeptide analogue 1, lacking a lipid
tail and inactive in the antimicrobial assays, did not show ap-
preciable cytotoxicity toward WRL 68 or HepG2 cells within the
tested concentration range. Interestingly, lipidated depsipep-
tide analogues 4 and 6, regardless of their antibacterial or he-
molytic activities, were found to be cytotoxic at higher concen-
trations toward both cell lines tested. Analogue 4 was highly
cytotoxic at 150 mgmLꢀ1, whereas the more active depsipep-
tide analogue 6 showed a roughly similar trend in cytotoxicity
Cytotoxicity
Hemolytic activity was determined against human erythrocytes
(0.5% in PBS), as described in the Experimental Section below.
PBS and 0.5% Triton X-100 were used as references for 0 and
100% hemolysis, respectively. Peptides inactive in the antimi-
crobial assays did not give detectable hemolytic activity. None
of the active depsipeptides 6, 8, and 11 and their amide coun-
terparts, peptides 14–16, showed hemolytic activity at MICs.
However, at higher concentrations, depsipeptides 6, 8, and 11
showed considerable hemolysis relative to reference Triton X-
100. The degree of hemolysis appears to correlate with the in-
crease in depsipeptide hydrophobicity. For example, at
64 mgmLꢀ1 (8ꢃMIC) the least hydrophobic depsipeptide 8
with Ala3 exhibited the lowest hemolytic activity (~30%) fol-
lowed by analogue 6 with Val3, causing ~45% hemolysis,
whereas the most hydrophobic analogue 11 with Phe3 showed
the highest level of hemolysis at ~65%. Interestingly, none of
the amide analogues 14–16 exhibited appreciable hemolytic
activity at this concentration. At much higher concentrations,
128 and 256 mgmLꢀ1 (16ꢃ and 32ꢃMIC), depsipeptide ana-
logues 6 and 11 showed hemolysis similar to that of Triton X-
100, whereas amide analogues 14 and 16 at 256 mgmLꢀ1
reached 15 and 50% hemolysis, respectively (Figure 5). Control
peptide 18 did not exhibit appreciable hemolytic activity even
at the highest tested concentration of 64 mgmLꢀ1.
to that of doxorubicin and was highly cytotoxic at 90 mgmLꢀ1
.
In comparison, the cytotoxicity of the less hydrophobic and
amphiphilic amide analogue 14 was much lower at the same
concentrations (data not shown).
Serum stability
To investigate the effect of ester-to-amide substitution on
cyclic peptide proteolytic stability, the disappearance of the
intact peptide incubated in 50% human serum for 24 h at
378C was followed by RP HPLC[57] (Figure 6). Depsipeptide 6
and amide 14 were used in this study as representative exam-
ples of each group of synthetic fusaricidin A/LI-F04a analogues.
An approximate 35% change in the concentration of depsi-
peptide 6 was observed within 1 h incubation in 50% human
serum. However, prolonged incubation of 6 did not result in
complete depsipeptide degradation; after 24 h, ~35%
of 6 was still detected in the serum. The stability of 6
in EMEM containing 10% FBS used for cytotoxicity
assays was significantly higher. In this case a ~23%
loss in concentration of 6 was observed after 24 h (see
Supporting Information). Analysis of depsipeptide 6
degradation using RP HPLC and MALDI-TOF mass
spectrometry revealed that the main degradation
product is a result of ring opening via ester bond hy-
drolysis (depsipeptide 6: Rt =15.8 min, [M+H]+ m/z=
825.9709; hydrolysis product: Rt =15.04 min, [M+H]+
m/z=843.9988). Additional confirmation of this finding
was obtained by identical MALDI-TOF m/z and RP
HPLC Rt values for the degradation product of depsi-
peptide 6 and the control peptide 18 (Rt =15.05 min,
[M+H]+ found m/z=844.1160, calculated m/z=
843.0228). In contrast, no degradation was observed
for amide analogue 14 under the same experimental
Figure 5. Hemolytic activity of fusaricidin A/LI-F04a analogues (see Supporting Informa-
tion).
conditions (Figure 6).
876
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882

Structural Modulation of Cyclic Lipodepsipeptides
MED
ing depsipeptides to basic conditions leads to the opposite
O!N acyl shift. Incorporation of the lipid tail or acetyl group
into the linear peptide precursor prior to d-Ala6 coupling via
ester bond and on-resin cyclization completely suppressed the
O!N acyl shift, resulting in the desired cyclic lipodepsipeptide
(Scheme 1). In addition, we found that the efficacy of ester
bond formation strongly depends on the nature of the solid
support and solvent. Optimal results were obtained with PEG-
based resins, a DIC/DMAP coupling method, and CH2Cl2.[24] The
superiority of PEG–PS over PS-based resins in ester bond for-
mation could be explained by a better solvation of peptide–
PEG–PS resins,[66,67] rapid DIC activation of the carboxylic
group,[68] and significant suppression of N-acylurea by-product
formation[69] in a nonpolar solvent such as CH2Cl2. On the other
hand, synthesis of amide analogues 13–17 using the same ap-
proach was straightforward (Scheme 1). The desired cyclic lipo-
peptides were obtained in satisfactory yields.
Figure 6. Stability of depsipeptide 6 and amide 14 in 50% human serum.
Discussion
Due to their structural diversity and high potency, naturally oc-
curring cyclic depsipeptides have attracted a great deal of at-
tention for the discovery of new antibiotics with superior activ-
ity against MDR bacteria.[7,58] However, full exploitation of the
antimicrobial potential of this class of natural products strongly
depends on synthetic access to their structures, in particular
their analogues, and an understanding of the details important
for their mode of action. It has been reported that the antibac-
terial and hemolytic activities of antimicrobial peptides depend
on various structural components, such as charge, size, secon-
dary structure, hydrophobicity, and amphipathicity.[42,59–61]
Among naturally occurring antibacterial peptides, the fusarici-
dins/LI-F family of cyclic lipodepsipeptides represents an at-
tractive series of lead compounds for the development of new
antibiotics because of their unique structure in which peptide
sequences are composed of six mostly hydrophobic amino
acids and a positive charge located at the end of the lipid
tail.[7] In addition, there is substantial evidence to show that
substitution of the ester bond in depsipeptides with an amide
bond may afford derivatives with similar activities and marked-
ly increased serum stabilities.[62–64] With this in mind, we syn-
thesized 18 fusaricidin A/LI-F04a analogues and investigated
the effects of structural modification on their overall hydropho-
bicity/amphiphilicity, conformation, stability, and biological ac-
tivity. All analogues were synthesized on a solid support using
Fmoc methodology. However, solid-phase synthesis of depsi-
peptides 1–13 turned out to be particularly challenging. Our
strategy for the solid-phase synthesis of cyclic lipodepsipepti-
des involved attachment of d-Asp5 via side chain stepwise
Fmoc synthesis of a linear peptide precursor and on-resin
head-to-tail cyclization (Scheme 1). The key step in the synthe-
sis of cyclic depsipeptides is ring closure. Considering the
greater reactivity of the amino group, and therefore minimal
possibility of side reactions, we chose macrolactamization for
depsipeptide ring closure, as opposed to macrolactonization.
Although this synthetic strategy appears to be a better choice
for solid-phase depsipeptide synthesis, an undesired intramo-
lecular O!N acyl shift may occur under basic conditions.[27,65]
Reversible intramolecular O!N or N!O acyl shifts are well de-
scribed side reactions in peptide chemistry.[27,65] Typically, pep-
tides containing Ser or Thr residues undergo N!O acyl shift
under acidic conditions, whereas the exposure of correspond-
Several studies have highlighted the importance of amide-
to-ester substitution on peptide conformation.[35,70–72] In gener-
al, replacement of a hydrogen-donating NH group with an
oxygen atom in depsipeptides results in removal of the back-
bone hydrogen bond, thus affecting the conformation of pep-
tides. As long as the side chains are unaltered, intramolecular
hydrogen bonds affect the equilibrium distribution of peptide
conformers. The effect of ester-to-amide substitution on the
conformation of fusaricidin A/LI-F04a analogues was assessed
by CD spectroscopy (Figure 3). The dramatic differences ob-
served in CD spectra of depsipeptide 6 in aqueous medium
and less polar TFE, and the absence of such spectral changes
for amide analogue 14, indicate a higher degree of flexibility
of the ester bond as opposed to the amide bond. These differ-
ences also suggest the greater ability of synthesized depsipep-
tide analogues to alternate conformations in different cellular
environments. However, complete interpretation of both depsi-
peptide 6 and amide 14 CD spectra in aqueous medium and
TFE is rather difficult, as the spectra obtained cannot be attrib-
uted to a single conformation. MD calculations were per-
formed to further explore the conformational differences
caused by ester-to-amide substitution in fusaricidin A/LI-F04a
analogues and to assess the possibility of forming more organ-
ized structures under conditions that mimic the membrane en-
vironment. The results are shown in Figure 4. Our experimental
CD data fully support the MD simulations, indicating higher
conformational flexibility of depsipeptide analogues. The MD
calculations showed no significant conformational differences
between depsipeptide 6 and amide analogue 14 in water,
whereas the differences become markedly more pronounced
in a less polar environment such as TFE.
Besides conformational changes, ester-to-amide substitution
altered the biochemical properties of the fusaricidin analogues.
Assessments of the cytotoxicity and stability of synthesized fu-
saricidin A/LI-F-04a analogues in human serum are important
secondary screening assays, mainly because these assays elimi-
nate cytotoxic analogues and those with short half-lives.[73] Our
experimental data suggest that the low serum stability of dep-
sipeptide 6 can be attributed mainly to the hydrolysis of its
ester bond. This finding is in accordance with published data
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
877

P. Cudic et al.
MED
showing that cyclic depsipeptides can be degraded through
competition between protease and esterase activities. Howev-
er, the depsipeptides were more easily subjected to hydrolysis
by esterase-type enzymes.[74–76] The cyclic structure, the pres-
ence of d-amino acids, and the greater stability of an amide
bond altogether contribute to the enhanced proteolytic stabili-
ty of the amide analogues.[77–84] Quite importantly, ester-to-
amide substitution did not affect the antibacterial potency of
these peptides. A positive charge positioned at the end of the
lipid tail, hydrophobicity, and amphiphilicity appeared to be
crucial for the antibacterial activities of fusaricidin A/LI-F04a an-
alogues as well as separation of their antibacterial activity and
toxicity toward human cells. Antibacterial assays showed that
only analogues with a guanidylated lipid tail are effective
against Gram-positive bacteria, indicating the importance of
the lipid tail and positively charged guanidino functionality for
their antimicrobial activity (Table 1). Based on the role of the
lipid tail in other lipopeptide antibiotics[85–88] and our experi-
mental data showing that nonlipidated analogue 2 is inactive
against tested bacterial strains, we can assume that the lipid
moiety helps target fusaricidin analogues to the bacterial
membrane. The antibacterial activity of peptides with a guani-
dylated lipid moiety could be explained by the strong ability
of the guanidinium group to bind the anionic phosphate of
the bacterial phospholipid membrane through a combination
of hydrogen bonding and charge–charge interactions. In gen-
eral, a high pKa value, diffused charge density, and the geome-
try of the guanidinium group which allows better alignment of
hydrogen bonds, all contribute to better interactions with the
phosphate anion relative to the ammonium group.[89]
the enhanced hemolytic activities (Figure 5). More hydrophobic
depsipeptides 6 and 11 were also more hemolytic. Similarly,
depsipeptides 6, 8, 11, and 12 exhibited higher hemolytic ac-
tivities than their less hydrophobic amide counterparts 14–17.
At concentrations 8ꢃMIC (64 mgmLꢀ1) none of the amide ana-
logues 14–17 were found to be hemolytic. The fact that linear
peptide 18 is not hemolytic (Supporting Information) suggests
that hydrolysis of the ester bond which leads to opening of
the depsipeptide ring is not a cause of hemolysis.
Liver toxicity is one of the most critical issues in drug devel-
opment, often leading to the delay or failure of drug candi-
dates.[90–92] Therefore, peptides that show no in vitro liver cell
toxicity may prove to be better candidates for further preclini-
cal or clinical studies. HepG2 and WRL 68 cell lines were used
to assess the potential toxicity of fusaricidin A/LI-F04a ana-
logues. In both cell lines, depsipeptide 6 exhibited much
higher in vitro cytotoxicity than amide counterpart 14. Howev-
er, the absence of cytotoxicity toward human cells observed
for amide analogue 14 cannot be explained solely by lower hy-
drophobicity. Conformational differences among these fusarici-
din/LI-F analogues, as illustrated by the MD calculations and
the differences in CD spectra between analogues 6 and 14
(Figures 3 and 4), should be taken into consideration as well.
Whereas hemolytic depsipeptide 6 has the ability to form
more ordered structures in a membrane-mimicking environ-
ment, non-hemolytic amide analogue 14 fails to undergo any
significant conformational change, regardless of the solvent
system. Conformational changes and hence changes in amphi-
pathicity have been ascribed to the low hemolytic activity of
some synthetic and naturally occurring cyclic cationic antimi-
crobial peptides.[42,59,61,93] Typically, for these cyclic peptides,
the presence of cationic amino acids in the sequence, a defined
secondary structure (b-sheet and hairpin loop), and amphiphil-
ic character are common structural characteristics that deter-
mine their biological activities.[59,93] In the case of fusaricidin A/
LI-F04a and its analogues, a positive charge is positioned at
the terminus of the 12-carbon-atom lipid tail, and peptide se-
quences are composed mostly of hydrophobic amino acids
(Figure 2). Nevertheless, experimental CD data and molecular
modeling studies showed that the peptide ring in both depsi-
peptide 6 and amide 14 analogues can adopt different amphi-
philic conformations depending on the environment, with
depsipeptides being more amphiphilic. Relatively lower amphi-
philicity of the amide analogues may contribute to the dissoci-
ation of antibacterial activity from human cell cytotoxicity.
The results of the antibacterial assays of the Ala-scan ana-
logues revealed that residues d-Val2, d-aThr4, and d-Ala6 are
important for antibacterial activity. Replacement of any of
these amino acids with corresponding Ala, analogues 7, 9, and
10, resulted in complete loss of depsipeptide antibacterial ac-
tivity (Table 1). On the other hand, position 3 in the depsipep-
tide sequence is more tolerable to changes. Depsipeptide ana-
logues containing neutral Ala3 (8) and bulky hydrophobic Val3
(6) or Phe3 (11) exhibited nearly identical antibacterial activi-
ties, whereas analogue 12 with polar Tyr3 was inactive. Equally
potent in vitro antibacterial activity was observed for their
amide counterparts, analogues 14 and 16, possessing hydro-
phobic Val3 and Phe3, respectively. However, differences in anti-
bacterial activity were observed between depsipeptide 8 and
its amide counterpart 15, both containing Ala at position 3.
Depsipeptide 8 exhibited antibacterial activity similar to the
most potent analogues, whereas amide 15 showed significant
decrease in activity (Table 1). Because the relative overall hy-
drophobicity ranking order of Ala < Val < Phe is the same for
both amide and depsipeptide analogues, the greater change
in amphiphilicity upon substitution of lipophilic Val3 with neu-
tral Ala3 in structurally constrained amide 15 may explain the
loss of its antibacterial activity.
Conclusions
Structure–activity relationship studies of fusaricidin A/LIF-04a
depsipeptide analogues reported herein reveal key structural
requirements for antibacterial activity and decreasing cytotox-
icity. The positively charged guanidinium group at the end of
the 12-carbon-atom lipid tail and the presence of hydrophobic
amino acids in the depsipeptide sequence are crucial for anti-
bacterial activity. Ala-scan results and comparison of the fusari-
cidin/LI-F natural product sequences suggest that position 3 in
the depsipeptide sequence is more tolerable to amino acid
As mentioned earlier, comparisons of the overall hydropho-
bicity of depsipeptides 6, 8, 11, and 12 (Figure 2) show no cor-
relation with their antibacterial activities. In contrast, an in-
crease in depsipeptide hydrophobicity can be associated with
878
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882

Structural Modulation of Cyclic Lipodepsipeptides
MED
American Type Culture Collection (ATCC, Manassas, VA, USA). Dehy-
drated culture media and agar, and polystyrene plates used for an-
timicrobial assays were purchased from BD (Franklin Lakes, NJ,
USA). Control antibiotics were purchased from Sigma–Aldrich. Anti-
microbial activity assays were carried out in standard sterile 96-well
plates, and MIC values were determined by measuring turbidity at
600 nm using a Stat Fax 2100 Microplate reader (Awareness Tech-
nology Inc., Palm City, FL, USA). Human red blood cells (hRBCs)
were purchased from Innovative Research (Novi, MI, USA). Human
serum was purchased from Sigma–Aldrich.
change. By introduction of neutral and hydrophobic amino
acids into this position, we were able to manipulate the depsi-
peptide’s overall hydrophobicity and amphiphilicity without
loss of antibacterial potency. However, structural changes lead-
ing to an increase in the depsipeptide’s overall hydrophobicity
and amphiphilicity resulted in an increase in cytotoxicity. On
the other hand, substitution of an ester bond in depsipeptides
by an amide bond gave more stable cyclic lipopeptide ana-
logues with preserved in vitro antibacterial activities, yet great-
ly improved serum stabilities and minimized human cell toxici-
ty. Lower overall hydrophobicity/amphiphilicity of amide ana-
logues in relative to their parent depsipeptides may explain
the dissociation of antibacterial activity from human cell cyto-
toxicity. More stable and less cytotoxic amide analogues may
have significant advantages over naturally occurring fusaricidi-
n A/LI-F04a and its depsipeptide analogues as lead structures
for the development of new antibacterial agents. In addition,
amide analogues are synthetically more accessible than the
parent depsipeptides, allowing for further structural optimiza-
tion using a combinatorial chemistry approach. Synthesis of
a focused combinatorial library based on fusaricidin A/LI-F04a
amide analogues and elucidation of the mode of action of
both the depsipeptide and amide analogues are currently un-
derway.
General procedure for peptide synthesis and purification
Linear peptidyl-resin precursors for cyclic lipopeptides 1–17 were
synthesized on amide TentaGel S RAM resin (substitution
0.26 mmolgꢀ1, 0.25 mmol scale) using an automated peptide syn-
thesizer. The solid-phase synthesis of cyclic peptides 1–17 was
started by attaching C-terminal Fmoc-d-Asp-OAllyl via side chain
to the resin using HBTU/HOBt/NMM protocol. The same coupling
protocol was used throughout, including coupling of the lipid tail
(Fmoc-ADA-OH, 1.5 equiv). In the case of depsipeptide analogues
1–12, Alloc-d-Ala-OH (4 equiv) or Alloc-Gly (4 equiv) was coupled
manually via ester bond to the hydroxy group of Fmoc-Thr using
DIC (4 equiv) and DMAP (1 equiv) coupling reagents in CH2Cl2.
Amide analogues 13–17 were prepared by coupling Fmoc-Dap-
(Mtt) instead of Fmoc-Thr-OH using the same coupling protocol as
above. Selective removal of Allyl and Alloc protecting groups was
performed by treatment of peptidyl-resin precursors with borane
dimethylamine complex (4 equiv), followed by addition of
Pd(PPh3)4 (0.1 equiv) in CH2Cl2 under argon.[94] Mtt was selectively
removed under mild acidic conditions (1% TFA in CH2Cl2, 30 min).
Solid-phase cyclization of linear precursors was carried out in
Experimental Section
Chemicals and instrumentation
TentaGel S RAM resin was obtained from Advanced ChemTech
(Louisville, KY, USA). 2-Chlorotrityl chloride was obtained from No-
vabiochem (Gibbstown, NJ, USA). Fmoc-protected amino acids and
coupling reagents (HOBt, HBTU, PyBOP) were purchased from
Chem-Impex (Wood Dale, IL, USA) or Novabiochem. DIC was pur-
chased from Acros Organics (Thermo Fisher Scientific, Waltham,
MA, USA). DMAP was purchased from Sigma–Aldrich (St. Louis,
MO, USA). All solvents were purchased from Fisher Scientific (Atlan-
ta, GA, USA) or Sigma–Aldrich, and were analytical reagent grade
or better. Linear peptidyl-resin precursors were synthesized on
a PS3 automated peptide synthesizer (Protein Technologies Inc.,
Tucson, AZ, USA). Mass spectrometry was performed on MALDI-
TOF Voyager-DE STR (Applied Biosystems, Foster City, CA, USA) in
reflector mode using a-cyano-4-hydroxycinnamic acid as a matrix
and positive mode. Analytical RP HPLC analyses and peptide purifi-
cations were performed on 1260 Infinity (Agilent Technologies,
Santa Clara, CA, USA) liquid chromatography systems equipped
with a UV/Vis detector. For analytical RP HPLC analysis, a C18 mono-
meric column (Grace Vydac, 250ꢃ4.6 mm, 5 mm, 120 ꢄ), 1 mLminꢀ1
flow rate, and elution method with a linear gradient of 2!98% B
over 30 min, where A is 0.1% TFA in H2O, and B is 0.08% TFA in
CH3CN was used. For peptide purification, a preparative C18 mono-
meric column (Grace Vydac, 250ꢃ22 mm, 10 mm, 120 ꢄ) was used.
Elution method was identical to the analytical method except for
the flow rate, which was 19 mLminꢀ1. CD spectra were recorded
on a JASCO 810 spectropolarimeter (Easton, MD, USA) using
a quartz cell of 0.1 mm optical path length. Spectra were measured
over a wavelength range of 180–250 nm with an instrument scan-
ning speed 200 nmminꢀ1 and a response time of 1 s. The concen-
trations of peptides were 0.1–0.2 mm. Cytotoxicity assays were ana-
lyzed on a Synergy H4 microplate reader (BioTek, Winooski, VT,
USA). Microbial strains and human cells were purchased from
a
manual reaction vessel overnight using PyBOP/HOBt/DIEA
(2:2:6 equiv) in DMF. The conversion of the lipid tail amino group
into the desired guanidino group was achieved by removal of the
Fmoc protecting group using standard piperidine deprotection
protocol and treatment of the peptidyl-resin with N,N-bis(tert-bu-
toxycarbonyl)thiourea (3 equiv) followed by Mukaiyama’s reagent
2-chloro-1-methylpyridinium iodide/TEA (3:4 equiv) in DMF.[95]
Control peptide 18 was synthesized on 2-chlorotrityl chloride resin
(substitution 1.3 mmolgꢀ1). The synthesis started by attaching
Fmoc-d-Ala-OH (4 equiv) to the resin using an equimolar amount
of DIEA in CH2Cl2 followed by resin end-capping with MeOH,[64]
and chain elongation using standard Fmoc chemistry. Quantitative
Fmoc substitution of the resin (0.5 mmolgꢀ1) was determined by
Fmoc cleavage and absorption measurement at 304 nm.[30] In all
cases, the reaction progress was monitored by RP HPLC, MALDI-
TOF MS, and where applicable, ninhydrin colorimetric test.[96]
Peptides were removed from the resin using TFA/TIA/H2O
(95:2.5:2.5 v/v/v) for 3 h. The crude peptides were precipitated
with cold methyl tert-butyl ether, and purified using preparative RP
HPLC. HPLC fractions were analyzed for purity, combined, and
lyophilized to give a white powder. The final purity of synthesized
peptides was confirmed by analytical RP HPLC, and was ꢁ95% in
all cases.
Concentrations of peptides in all experiments were determined
using RP HPLC and calibration curve based on analogue 6. The
peptide content of 6 was determined by quantitative amino acid
analysis to be 62.32%.
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
879

P. Cudic et al.
MED
~0.001 (based on calculation). Concentrations of analogues 1–18,
as well as control antibiotics, were in the range 1–128 mgmLꢀ1. All
samples were loaded in duplicate, and the average OD value was
taken for calculating MIC. Each assay was repeated twice. Stock sol-
utions of synthesized analogues 1–18, as well as control antibiotics
were prepared in 5–10% DMSO/H2O (v/v) solvent mixture, depend-
ing on the analogue solubility. After dilutions with MHB, final con-
centration of DMSO in wells was 0.5–1%. Plates were loaded with
90 mL bacterial suspension (with initial OD600 of 0.001) of the tested
microorganism, and 10 mL aliquots of twofold serial dilutions of the
analogues 1–18 or control antibiotics. Plates were then incubated
at 378C overnight with gentle shaking. In the case of S. pyogenes,
plates were incubated under 5% CO2 atmosphere. Inhibition of
bacterial growth was determined by measuring OD600; a decrease
in OD600 indicates inhibition of bacterial growth.
Circular dichroism
All CD spectra were recorded on JASCO 810 spectropolarimeter at
258C using a 0.1 cm path length cell. The spectra were acquired in
the range 180–250 nm, 1 nm bandwidth, four accumulations and
200 nmminꢀ1 scanning speed. All spectra were obtained using
0.1–0.2 mm concentrations in 0.5% or 1% AcOH, 25–100% TFE/
H2O (v/v) solution. Spectra at 0.5 and 1% AcOH were virtually iden-
tical for each peptide. Each experiment was repeated at least once
and at various concentrations. No concentration-dependent CD
spectral changes were observed.
Molecular modeling
First low-energy conformations were generated by conformational
analysis in Molecular Operating Environment (MOE)[46] using
a Monte Carlo search with the Generalized Born solvation model
implemented in MOE.[47] The dielectric constant was set to 26.14 or
80 to simulate the search in TFE or water, respectively. The search
was conducted with the MMFF94x force field using default param-
eters. The lowest-energy conformers found for each compound in
the Monte Carlo search were the starting points of 40 ns MD simu-
lation using the MacroModel 9.9 module from Maestro software.[97]
In brief, the initial structures were equilibrated by 2 ps. The “Bonds
to Hydrogens” option from the Shake procedure was selected in
order to use 2 fs as time step. The simulation temperature was set
at 300.0 K as default. The optimized potential for liquid simulation
(OPLS)-2005 force field[98] was used to calculate the potential
energy. Conformations were sampled every 20 ps and optimized
using the same force field. The dielectric constant was set to 26.14
to simulate the search in TFE. The GB/SA solvation model imple-
mented in MacroModel[99] was used for water simulation.
Hemolytic activity
Human red blood cells (hRBCs) were diluted with PBS to 1%. De-
pending on solubility, analogues 1–18 were dissolved in 5–10%
DMSO/H2O (v/v) solvent mixture to concentrations of 16–
512 mgmLꢀ1. Into each well of the clear, flat-bottom 96-well plate,
50 mL of the hRBCs were placed followed by addition of 50 mL ana-
logue solution to final peptide concentrations of 8–256 mgmLꢀ1
.
Assays were performed in triplicate, and each experiment was re-
peated twice. To determine the potential effect of DMSO on hemo-
lytic activity, controls containing 5–10% DMSO in H2O (v/v) were
added to the assay setup. As a positive control, 50 mL Triton X-100
in H2O was used at a final concentration of 0.5% (v/v). As a nega-
tive control, 50 mL H2O and PBS was used. Plates were incubated
for 1 h at 378C. To each well 100 mL of PBS was added, and the
plates were centrifuged for 10 min at 1000 g. Supernatants
(150 mL) were transferred into a new plate, and absorbance at
405 nm was measured. Within the tested concentration range, the
effect of DMSO on hemolysis was minimal, and was subtracted to
obtain solely hemolytic activity of the peptides. The degree of he-
molysis was expressed in percent relative to total hemolysis effect-
ed by Triton-X.
The amphiphilic moment descriptor, m, which is an established
measure of the balance between hydrophilic and lipophilic molec-
ular properties,[26] was computed in MOE for the lowest-energy
conformations obtained in the dynamics within 5 kcalmolꢀ1 of the
lowest minimum.
Antibacterial activity
Cytotoxicity
A total of six Gram-positive and two Gram-negative bacterial
strains were used, including MDR bacterial strains: Staphylococcus
aureus ATCC 29213, Staphylococcus aureus (MRSA) ATCC 33591,
Staphylococcus aureus Mu50 (VRSA) ATCC 700699, Staphylococcus
epidermidis (MRSE) ATCC 27626, Streptococcus pyogenes ATCC
19615, Escherichia coli K-12 ATCC 29181, and Klebsiella pneumoniae
K6 ATCC 700603. Quantification of the antibacterial activity of syn-
thesized analogues 1–18 was performed in sterile 96-well flat-bot-
tomed polystyrene plates by the standard microdilution broth
method.[49,50] Tests were performed using Mꢅller–Hinton broth
(MHB) without dilution. Controls on each plate were media without
bacteria, bacterial inoculum without antimicrobials added, and bac-
terial inoculum containing methicillin or vancomycin. Assay setup:
Stocks of microorganisms maintained at ꢀ808C in 30% glycerol
were thawed and grown in media recommended by ATCC proto-
cols for each particular microorganism. The following day, an ali-
quot of the bacterial suspension (100 mL) was transferred into
10 mL fresh media and incubated for 4–5 h, until OD600 of the sus-
pension was 0.3–0.4. Bacterial suspension (100 mL) was then trans-
ferred into sterile tubes (Eppendorf) and centrifuged for 5 min at
1300 g. The supernatant was discarded, and cell pellet resuspend-
ed in MHB. Upon measuring OD600 of the suspension, appropriate
dilution was made so that the final OD of the suspension was
Cytotoxicity of analogues 1–18 was determined using the MTT col-
orimetric assay. Assays were set up in flat-bottom polystyrene 96-
well plates with 10000 cells per well grown in EMEM containing
10% FBS, 5% penicillin/streptomycin, and 5% l-glutamine (v/v).
After an overnight incubation at 378C under a humidified atmos-
phere with 5% CO2, media were removed, and fresh media con-
taining analogues 1–18 in a concentration range of 1–250 mm
were added. Plates were again incubated at 378C under a humidi-
fied atmosphere with 5% CO2. As a control, doxorubicin was used
in the same concentration range. After incubation for 24 or 48 h,
media were removed, and 100 mL MTT dissolved in serum-free
medium (1 mgmLꢀ1) was added to each well. Plates were again in-
cubated for 3 h under the same conditions. Media containing MTT
were removed, and 100 mL DMSO was added to each well. Plates
were shaken for 5 min before reading absorbance at 540 nm.
Stability assays
The stabilities of selected fusaricidin A/LI-F04a analogues in 50%
human serum, as well as in EMEM (used in cell toxicity assays)
were determined. For stability in 50% human serum, peptides 6
880
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882

Structural Modulation of Cyclic Lipodepsipeptides
MED
[12] N. Woodford, Expert Opin. Invest. Drugs 2003, 12, 117–137.
[13] W. V. Kern, Int. J. Clin. Pract. 2006, 60, 370–378.
[14] D. G. McCafferty, P. Cudic, B. A. Frankel, S. Barkallah, R. G. Kruger, W. Li,
Biopolymers 2002, 66, 261–284.
[15] D. Shah, M.-D. Dang, R. Hasbun, H. L. Koo, Z.-D. Jiang, H. L. DuPont,
K. W. Garey, Expert Rev. Anti-Infect. Ther. 2010, 8, 555–564.
[16] R. Eckert, Future Microbiol. 2011, 6, 635–651.
[17] K. Kurusu, K. Ohba, J. Antibiot. 1987, 40, 1506–1514.
[18] Y. Kajimura, M. Kaneda, J. Antibiot. 1996, 49, 129–135.
[19] Y. Kajimura, M. Kaneda, J. Antibiot. 1997, 50, 220–228.
[20] J. Kuroda, T. Fukai, M. Konishi, J. Uno, K. Kurusu, T. Nomura, Heterocycles
2000, 53, 1533–1549.
and 14 (0.5 mg each) were dissolved in 200 mL DMSO, to which
800 mL H2O and 1 mL human serum were added. The solution was
incubated at 378C. After 0 min, 45 min, 2, 4, 8, and 24 h, three sam-
ples (3ꢃ100 mL) of each peptide were taken and precipitated by
the addition of 20 mL 15% aqueous TCA. Samples were quickly vor-
texed and then centrifuged at 9300 g for 10 min. The supernatant
was analyzed by analytical RP HPLC and MALDI-TOF MS. For the
stability of depsipeptide 6 in EMEM containing 10% FBS, 0.25 mg
of the peptide was dissolved in 10 mL DMSO, followed by the addi-
tion of 990 mL media and incubation at 378C. Peptide samples
were treated and analyzed in the same way as described above.
[21] J. R. Cochrane, C. S. P. McErlean, K. A. Jolliffe, Org. Lett. 2010, 12, 3394–
3397.
[22] J. Li, P. K. Beatty, S. Shah, S. E. Jensen, Appl. Environ. Microbiol. 2007, 73,
3480–3489.
Abbreviations
AcOH, acetic acid; Alloc, allyloxycarbonyl; 12-ADA, 12-aminodode-
canoic acid; CD, circular dichroism; Dap, diaminopropionic acid;
DIC, diisopropylcarbodiimide; DIEA, diisopropylethyl amine; DMAP,
4-dimethylaminopyridine; DMF, N,N-dimethylformamide; DMSO, di-
methyl sulfoxide; EMEM, Eagle’s minimal essential medium; FBS,
fetal bovine serum; Fmoc, fluorenylmethyloxycarbonyl; 12-GDA,
12-guanidinododecanoic acid; HOBt, N-hydroxybenzotriazole;
HBTU, 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexa-
fluorophosphate; hRBCs, human red blood cells; MALDI-TOF MS,
matrix-assisted laser desorption/ionization time-of-flight mass spec-
trometry; MHB, Mꢅller–Hinton broth; MIC, minimum inhibitory
concentration; MRSA, methicillin-resistant S. aureus; MTT, 3-(4,5-di-
methylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; Mtt, 4-methyl-
trityl chloride; NMM, N-methylmorpholine; PBS, phosphate-buf-
fered saline; PyBOP, benzotriazol-1-yl-oxytripyrrolidinophosphoni-
um hexafluorophosphate; RP HPLC, reversed-phase high-perfor-
mance liquid chromatography; SPPS, solid-phase peptide synthe-
sis; TIA, thioanisole; TCA, trichloroacetic acid; TFA, trifluoroacetic
acid; TFE, trifluoroethanol; VRSA, vancomycin-resistant S. aureus.
[23] S.-K. Choi, S.-Y. Park, R. Kim, C.-H. Lee, J. F. Kim, S.-H. Park, Biochem. Bio-
phys. Res. Commun. 2008, 365, 89–95.
[24] M. Stawikowski, P. Cudic, Tetrahedron Lett. 2006, 47, 8587–8590.
[25] N. Bionda, M. Cudic, L. Barisic, M. Stawikowski, R. Stawikowska, D. Binet-
ti, P. Cudic, Amino Acids 2012, 42, 285–293.
[26] E. Atherton, N. L. Benoiton, E. Brown, R. C. Sheppard, B. J. Williams, J.
Chem. Soc. Chem. Commun. 1981, 336–337.
[27] L. Mouls, G. Subra, C. Enjalbal, J. Martinez, J. L. Aubagnac, Tetrahedron
Lett. 2004, 45, 1173–1178.
[28] M. Skwarczynski, Y. Kiso, Curr. Med. Chem. 2007, 14, 2813–2823.
[29] K. Barlos, Tetrahedron Lett. 1989, 30, 3943–3946.
[30] W. C. Chan, P. D. White in Fmoc Solid-Phase Peptide Synthesis, Vol. 222
(Ed. B. D. Hames), Oxford University Press, New York, 2003.
[31] F. D. Sonnichsen, J. E. Van Eyk, R. S. Hodges, B. D. Sykes, Biochemistry
1992, 31, 8790–8798.
[32] D. Roccatano, G. Colombo, M. Fioroni, A. E. Mark, Proc. Natl. Acad. Sci.
USA 2002, 99, 12179–12184.
[33] C. Fields, D. Mickelson, S. Drake, J. McCarthy, G. Fields, J. Biol. Chem.
1993, 268, 14153–14160.
[34] T. Sivaraman, T. K. S. Kumar, C. C. Huang, C. Yu, Biochem. Mol. Biol. Int.
1998, 44, 29–39.
[35] T. Cupido, J. Spengler, J. Ruiz-Rodriguez, J. Adan, F. Mitjans, J. Piulats, F.
Albericio, Angew. Chem. 2010, 122, 2792–2797; Angew. Chem. Int. Ed.
2010, 49, 2732–2737.
Acknowledgements
[36] R. M. Freidinger, D. S. Perlow, W. C. Randall, R. Saperstein, B. H. Arison,
D. F. Veber, Int. J. Pept. Prot. Res. 1984, 23, 142–150.
[37] R. Haubner, R. Gratias, B. Diefenbach, S. L. Goodman, A. Jonczyk, H.
Kessler, J. Am. Chem. Soc. 1996, 118, 7461–7472.
[38] H. Kessler, G. Gemmecker, A. Haupt, M. Klein, K. Wagner, M. Will, Tetrahe-
dron 1988, 44, 745–759.
[39] L. Otvos in Neuropeptide Protocols, Methods in Molecular Biology, Vol. 73
(Eds.: G. B. Irvine, C. H. Williams), Humana Press, Totowa, 1997, pp. 153–
161.
We acknowledge support of the cyclic lipopeptide work described
herein by the NIH (grant 1S06-GM073621-01) and AHA (grant
0630175N) to P.C. We also thank our colleague Dr. Laszlo Ot-
vos, Jr. for helpful comments, and Ms. Karen Gottwald for editing
the text.
Keywords: antibiotics
substitution · peptides · structure–activity relationships
· drug resistance · ester-to-amide
[40] M. Jelokhani-Niaraki, L. H. Kondejewski, S. W. Farmer, R. E. W. Hancock,
C. M. Kay, R. S. Hodges, Biochem. J. 2000, 349, 747–755.
[41] R. W. Woody in Circular Dichroism, Vol. E22b (Eds.: M. Goodman, A. Felix,
L. Moroder, C. Toniolo), Thieme, Stuttgart, 2004, pp. 739–765.
[42] L. H. Kondejewski, M. Jelokhani-Niaraki, S. W. Farmer, B. Lix, C. M. Kay,
B. D. Sykes, R. E. W. Hancock, R. S. Hodges, J. Biol. Chem. 1999, 274,
13181–13192.
[43] D. L. Lee, J. P. Powers, K. Pflegerl, M. L. Vasil, R. E. Hancock, R. S. Hodges,
J. Pept. Res. 2004, 63, 69–84.
[44] Y. Chen, M. T. Guarnieri, A. I. Vasil, M. L. Vasil, C. T. Mant, R. S. Hodges, An-
timicrob. Agents Chemother. 2007, 51, 1398–1406.
[45] C. T. Mant, J. M. Kovacs, H.-M. Kim, D. D. Pollock, R. S. Hodges, Biopoly-
mers 2009, 92, 573–595.
[46] Molecular Operating Environment (MOE) version 2008, Chemical Com-
puting Group Inc., Montreal, QC (Canada) http://www.chemcomp.com
(accessed September 15, 2011).
[47] M. Wojciechowski, B. Lesyng, J. Phys. Chem. B 2004, 108, 18368–18376.
[48] G. Cruciani, P. Crivori, P. A. Carrupt, B. Testa, J. Mol. Struct. THEOCHEM
2000, 503, 17–30.
[49] L. Otvos, M. Cudic in Broth Microdilution Antibacterial Assay of Peptides
(Ed.: G. B. Fields), Humana Press, Totowa, 2007, pp. 309–320.
[1] D. J. Payne, M. N. Gwynn, D. J. Holmes, D. L. Pompliano, Nat. Rev. Drug
Discovery 2007, 6, 29–40.
[2] New antimicrobial agents approved by the US Food and Drug Adminis-
tration from 2007 to 2009 and new indications for previously approved
agents: USFDA, Antimicrob. Agents Chemother. 2010, 54, 4033–4035.
[3] M. A. Fischbach, C. T. Walsh, Science 2009, 325, 1089–1093.
[4] B. Spellberg, R. Guidos, D. Gilbert, J. Bradley, H. W. Boucher, W. M.
Scheld, J. G. Bartlett, J. Edwards, Clin. Infect. Dis. 2008, 46, 155–164.
[5] L. B. Rice, J. Infect. Dis. 2008, 197, 1079–1081.
[6] H. W. Boucher, G. H. Talbot, J. S. Bradley, J. E. Edwards, D. Gilbert, L. B.
Rice, M. Scheld, B. Spellberg, J. Bartlett, Clin. Infec. Dis. 2009, 48, 1–12.
[7] N. Bionda, P. Cudic, Croat. Chem. Acta 2011, 84, 315–329.
[8] N. Roongsawang, K. Washio, M. Morikawa, Int. J. Molec. Sci. 2010, 12,
141–172.
[9] M. A. Marahiel, J. Pept. Sci. 2009, 15, 799–807.
[10] S. A. Sieber, M. A. Marahiel, J. Bacteriol. 2003, 185, 7036–7043.
[11] J. Grunewald, M. A. Marahiel, Microbiol. Mol. Biol. Rev. 2006, 70, 121–
146.
ChemMedChem 2012, 7, 871 – 882
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
881

P. Cudic et al.
MED
[50] CLSI in Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria
That Grow Aerobically, Approved Standard M7-A8, Clinical and Laborato-
ry Standards Institute, Wayne, PA (USA), 2011.
[77] M. Sela, E. Zisman, FASEB J. 1997, 11, 449–456.
[78] E. M. Molhoek, A. van Dijk, E. J. A. Veldhuizen, H. P. Haagsman, F. J.
Bikker, Peptides 2011, 32, 875–880.
[51] C. Subbalakshmi, V. Krishnakumari, R. Nagaraj, N. Sitaram, FEBS Lett.
1996, 395, 48–52.
[79] R. Halai, B. Callaghan, N. L. Daly, R. J. Clark, D. J. Adams, D. J. Craik, J.
Med. Chem. 2011, 54, 6984–6992.
[52] E. John, F. Jahning, Biophys. J. 1992, 63, 1536–1543.
[53] R. Feder, A. Dagan, A. Mor, J. Biol. Chem. 2000, 275, 4230–4238.
[54] H. C. Ansel, G. E. Cabre, J. Pharm. Sci. 1970, 59, 478–481.
[55] N. C. Santos, J. Figuera-Coelho, J. Martins-Silva, C. Saldanha, Biochem.
Pharmacol. 2003, 65, 1035–1041.
[80] J. S. Davies, J. Pept. Sci. 2003, 9, 471–501.
[81] P. Li, P. P. Roller, Curr. Top. Med. Chem. 2002, 2, 325–341.
[82] C. Blackburn, S. A. Kates in Solid-Phase Synthesis of Cyclic Homodetic
Peptides, Vol. 289 (Ed.: B. F. Gregg), Academic Press, 1997, pp. 175–198.
[83] S. A. Kates, N. A. Sole, F. Albericio, G. Barany in Solid-Phase Synthesis of
Cyclic Peptides (Eds.: C. Basava, G. N. Anantharamaiah), Brikhauser,
Boston, 1994, pp. 39–59.
[84] R. Tugyi, K. Uray, D. Ivꢆn, E. Fellinger, A. Perkins, F. Hudecz, Proc. Natl.
Acad. Sci. USA 2005, 102, 413–418.
[85] R. H. Baltz, V. Miao, S. K. Wrigley, Nat. Prod. Rep. 2005, 22, 717–741.
[86] F. Kopp, J. Grꢅnewald, C. Mahlert, M. A. Marahiel, Biochemistry 2006, 45,
10474–10481.
[56] R. B. Weiss, Semin. Oncol. 1992, 19, 670–686.
[57] H. Jenssen, S. I. Aspmo in Peptide-Based Drug Design Serum Stability of
Peptides, Vol. 494, Humana Press, Totowa, 2008, pp. 177–186.
[58] A. Giuliani, G. P. Pirri, S. F. Nicoletto, Cent. Eur. J. Biol. 2007, 2, 1–33.
[59] L. H. Kondejewski, S. W. Farmer, D. S. Wishart, C. M. Kay, R. E. W. Han-
cock, R. S. Hodges, J. Biol. Chem. 1996, 271, 25261–25268.
[60] C. McInnes, L. H. Kondejewski, R. S. Hodges, B. D. Sykes, J. Biol. Chem.
2000, 275, 14287–14294.
[87] F. T. Counter, N. E. Allen, D. S. Fukuda, J. N. Hobbs, J. Ott, P. W. Ensming-
er, J. S. Mynderse, D. A. Preston, C. Y. E. Wu, J. Antibiot. 1990, 43, 616–
622.
[61] C. Qin, X. Zhong, X. Bu, N. L. J. Ng, Z. Guo, J. Med. Chem. 2003, 46,
4830–4833.
[62] J. Nam, D. Shin, Y. Rew, D. L. Boger, J. Am. Chem. Soc. 2007, 129, 8747–
8755.
[63] Y. Rew, D. Shin, I. Hwang, D. L. Boger, J. Am. Chem. Soc. 2004, 126,
1041–1043.
[64] I. Izzo, G. A. Acosta, J. Tulla-Puche, T. Cupido, M. J. Martin-Lopez, C.
Cuevas, F. Albericio, Eur. J. Org. Chem. 2010, 2010, 2536–2543.
[65] A. B. Mauger, O. A. Stuart, Int. J. Pept. Prot. Res. 1987, 30, 481–488.
[66] M. Meldal in Properties of Solid Supports, Vol. 289 (Ed.: G. B. Fields), Aca-
demic Press, New York, 1997, pp. 83–104.
[67] G. B. Fields, C. g. Fields, J. Am. Chem. Soc. 1991, 113, 4202–4207.
[68] D. Hudson, J. Org. Chem. 1988, 53, 617–624.
[69] J. Podlech in 3.5 Carbodiimides (Eds.: M. Goodman, A. Felix, L. Moroder,
C. Toniolo), Stuttgart, 2004, pp. 517–533.
[70] E. T. Powers, S. Deechongkit, J. W. Kelly in Backbone–Backbone H-Bonds
Make Context-Dependent Contributions to Protein Folding Kinetics and
Thermodynamics: Lessons from Amide-to-Ester Mutations, Vol. 72 (Eds.:
L. B. Robert, B. David), Academic Press, 2005, pp. 39–78.
[71] T. Kato, H. Mizuno, S. Lee, H. Aoyagi, H. Kodama, N. Go, T. Kato, Int. J.
Pept. Prot. Res. 1992, 39, 485–492.
[72] J. A. Scheike, C. Baldauf, J. Spengler, F. Albericio, M. T. Pisabarro, B.
Koksch, Angew. Chem. 2007, 119, 7912–7916; Angew. Chem. Int. Ed.
2007, 46, 7766–7769.
[88] M. Debono, B. J. Abbot, M. R. Molloy, D. S. Fukuda, A. H. Hunt, V. M.
Daupert, F. T. Counter, J. L. Ott, C. B. Carrell, L. C. Howard, L. V. D. Boeck,
R. L. Hamill, J. Antibiot. 1988, 41, 1093–1105.
[89] R. J. T. Houk, S. L. Tobey, E. V. Anslyn, Top. Curr. Chem. 2005, 255, 199–
229.
[90] M. Fung, A. Thornton, K. Mybeck, J. H.-H. Wu, K. Hornbuckle, E. Muniz,
Drug Inf. J. 2001, 35, 293–317.
[91] N. Kaplowitz, Drug Safety 2001, 24, 483–490.
[92] D. Schuster, C. Laggner, T. Langer, Curr. Pharm. Des. 2005, 11, 3545–
3559.
[93] V. Frecer, B. Ho, J. L. Ding, Antimicrob. Agents Chemother. 2004, 48,
3349–3357.
[94] P. Gomez-Martinez, M. Dessolin, F. Guibe, F. Albericio, J. Chem. Soc.
Perkin Trans. 1 1999, 2871–2874.
[95] Y. F. Yong, J. A. Kowalski, M. A. Lipton, J. Org. Chem. 1997, 62, 1540–
1542.
[96] E. Kaiser, R. L. Colescott, C. D. Bossinger, P. I. Cook, Anal. Biochem. 1970,
34, 595–598.
[97] Maestro version 9.2, 2011, Schrçdinger LLC, New York, NY (USA) http://
www.schrodinger.com (accessed February 14, 2012).
[98] G. A. Kaminski, R. A. Friesner, J. Tirado-Rives, W. L. Jorgensen, J. Phys.
Chem. B 2001, 105, 6474–6487.
[73] M. Powell, T. Stewart, L. Otvos, L. Urge, F. Gaeta, A. Sette, T. Arrhenius,
D. Thomson, K. Soda, S. Colon, Pharm. Res. 1993, 10, 1268–1273.
[74] H. Shirahama, K. Umemoto, H. Yasuda, J. Biomater. Sci. Polym. Ed. 1999,
10, 621–639.
[99] C. W. Still, A. Tempczyk, R. C. Hawley, T. Hendrickson, J. Am. Chem. Soc.
1990, 112, 6127–6129.
Received: January 9, 2012
Revised: February 9, 2012
[75] T. Ouchi, T. Nozaki, A. Ishikawa, I. Fujimoto, Y. Ohya, J. Polym. Sci. Part A
1997, 35, 377–383.
[76] A. Schaeffer, D. S. Auld, Biochemistry 1986, 25, 2476–2479.
Published online on March 5, 2012
882
www.chemmedchem.org
ꢂ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2012, 7, 871 – 882