Development of Tyrocidine A analogues with improved antibacterial activity

Article abstract of DOI:10.1016/j.bmc.2007.08.007

The development of new antibacterial therapeutic agents capable of halting microbial resistance is a chief pursuit in clinical medicine. Classes of antibiotics that target and destroy bacterial membranes are attractive due to the decreased likelihood that bacteria will be able to generate resistance to this mechanism. The amphipathic cyclic decapeptide, Tyrocidine A, is a model for this class of antibiotics. Tyrocidine A is composed of a hydrophobic and a hydrophilic face, allowing for insertion into bacterial membranes, creating porous channels and destroying membrane integrity. We have used a combination of molecular modeling and solid phase synthesis to prepare Tyrocidine A and analogues 1-8. The minimum inhibitory concentrations (MICs) of these compounds were determined for a host of gram positive species and E. coli as a representative gram negative bacterium. Analogues 2 and 5 demonstrated moderate 2- to 8-fold increases in antibacterial activity over the parent Tyrocidine A for a variety of pathogenic microbes (best MICs for E. coli 32 μg/mL and 2 μg/mL for most gram positives). Examination of the structure- activity relationship between the analogues demonstrated a preference for increased amphipathicity but did not show a clear preference for increasing hydrophilicity versus hydrophobicity in improving antibacterial activity. Of note, movement of positively charged lysine residues or neutral pentafluorophenyl residues to different positions within the cyclopeptide ring system demonstrated improvements in antibacterial activity.

Full text of DOI:10.1016/j.bmc.2007.08.007

Bioorganic & Medicinal Chemistry 15 (2007) 6667–6677
Development of Tyrocidine A analogues with improved
antibacterial activity
Michael A. Marques,a Diane M. Citronb and Clay C. Wanga,*
aDepartments of Pharmacology and Chemistry, University of Southern California, 1985 Zonal Avenue, Los Angeles, CA 90089, USA
bMicrobial Research Laboratory, Los Angeles County, University of Southern California Medical Center,
1801 East Marengo Street 2G-24, Los Angeles, CA 90033, USA
Received 17 January 2007; revised 5 July 2007; accepted 7 August 2007
Available online 11 August 2007
Abstract—The development of new antibacterial therapeutic agents capable of halting microbial resistance is a chief pursuit in clin-
ical medicine. Classes of antibiotics that target and destroy bacterial membranes are attractive due to the decreased likelihood that
bacteria will be able to generate resistance to this mechanism. The amphipathic cyclic decapeptide, Tyrocidine A, is a model for this
class of antibiotics. Tyrocidine A is composed of a hydrophobic and a hydrophilic face, allowing for insertion into bacterial mem-
branes, creating porous channels and destroying membrane integrity. We have used a combination of molecular modeling and solid
phase synthesis to prepare Tyrocidine A and analogues 1–8. The minimum inhibitory concentrations (MICs) of these compounds
were determined for a host of gram positive species and E. coli as a representative gram negative bacterium. Analogues 2 and 5
demonstrated moderate 2- to 8-fold increases in antibacterial activity over the parent Tyrocidine A for a variety of pathogenic
microbes (best MICs for E. coli 32 lg/mL and 2 lg/mL for most gram positives). Examination of the structure– activity relationship
between the analogues demonstrated a preference for increased amphipathicity but did not show a clear preference for increasing
hydrophilicity versus hydrophobicity in improving antibacterial activity. Of note, movement of positively charged lysine residues or
neutral pentafluorophenyl residues to different positions within the cyclopeptide ring system demonstrated improvements in antibac-
terial activity.
Published by Elsevier Ltd.
1. Introduction
Antimicrobial agents fall into one of several classes:
b-lactams, b-lactamase inhibitors, aminoglycosides,
tetracyclines, rifamycins, macrolides, lincosamides,
glycopeptides, streptogramins, sulfonamides, oxazolidi-
nones, quinolones, gramicidins and others.7 These com-
pounds target a variety of bacterial systems including
DNA replication, transcription, folic acid metabolism,
protein synthesis, and cell wall synthesis/integrity. Each
class has its own specificities or coverage for a variety of
bacteria. The means by which bacteria acquire resistance
to such an expansive array of agents is highly variable
and complex. It has been proposed that antimicrobial
agents that target individual enzymes are the most likely
to induce resistance, whereas therapeutics that target
several structures irreversibly generate resistance at a
slower rate.9 The clearest mechanisms of genetic resis-
tance include: inactivation of the drug, modification of
the target active site, modified permeability of the cell
wall, upregulation of the target enzyme, or the complete
bypass of inhibited steps.10,11
The market share for antibiotics is greater than $25 bil-
lion per year. Unfortunately, the emergence of highly
resistant microbe strains is increasingly limiting the
effectiveness of our current arsenal of therapeutics, mak-
ing drugs such as vancomycin, daptomycin, the strep-
togramins, and linezolid, which were previously
reserved for highly resistant bacterial strains, our last
and only line of defense. Even now, resistance to these
highly potent drugs is being seen clinically.1–5 Making
matters worse, there is less motivation on the part of
pharmaceutical companies to spend billions on research
and development of a novel antibiotic when bacterial
resistance is likely to occur rapidly.6–8
Keywords: Tyrocidine; Gramicidin; Cyclopeptide; Antibiotics; Amphi-
pathic peptides; Sulfamylbutyryl resin; Solid phase synthesis.
*
With so many different paths to resistance, the goal of
generating successful new therapeutics to combat infec-
Corresponding author. Tel.: +1 323 442 1670; fax: +1 323 442
1365; e-mail: clayw@usc.edu
0968-0896/$ - see front matter Published by Elsevier Ltd.
doi:10.1016/j.bmc.2007.08.007

6668
M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
tion is also challenging. There are several common
methodologies for accomplishing this task.12,13 With
the advancements in genomic sequencing technologies,
there has been a hope that novel targets that are less
prone to cross resistance can be developed.14,15 Unfortu-
nately, deciphering bacterial genomes has not yet led to
a wealth of new targets or therapeutics. Further, as men-
tioned above, even new classes of antimicrobials such as
the oxazolidinones have met with resistant bacterial
strains.4,16–18
streptogramins (quinupristin/daflopristin), glycopep-
tides (vancomycin, teicoplanin), and lipopeptides (dap-
tomycin) have been fundamental in controlling severe
bacterial infections.
Recently, Gramicidin S and Tyrocidine A have re-
ceived a great deal of attention due to their highly
desirable mechanism of action. Both compounds are
decapeptide natural products isolated from Bacillus
subtilis.26,27 Structurally, the compounds differ only
slightly with respect to their primary sequence. Both
compounds are known to form a b-type secondary
structure leading to an amphipathic molecule.28–33
More specifically, one face of the molecule is hydro-
phobic, while the other is hydrophilic. The b-type
structure, and associated amphipathicity, imparts
Gramicidin S and Tyrocidine A with their specific
antibacterial mechanism. These cyclic amphipathic
peptides are known to associate with bacterial mem-
branes, causing them to destabilize, lose structural
integrity and ultimately fragment, resulting in bacterial
cell death.34–38 It seems that the hydrophilic residues
of the cyclopeptide associate with the negatively
charged phosphate groups of the bacterial lipids, while
the hydrophobic groups form a nonselective porous
channel.
More commonly, entirely new generations of drugs
within a class have been produced by making structural
changes to existing scaffolds, perhaps the most notable
being the penicillins, cephalosporins, quinolones, tetra-
cyclines, and macrolides.6,18–22 In light of this method,
we have selected Tyrocidine A, a natural product related
to Gramicidin S, as a platform for study (Fig. 1). Poly-
peptide antimicrobials are not new to the field of micro-
biology. Nature has long used linear and cyclic
polypeptides as natural defenses against competing or
pathogenic bacteria.23–25 Natural peptide antibiotics
such as the gramicidins and polymyxins have been used
extensively for topical therapy with excellent results.
Furthermore, cyclic peptide derivatives such as the
The amphipathic cyclopeptides demonstrate a mixed
selectivity for bacterial membranes, mildly preferring
to associate with the highly negative phosphates found
on bacterial walls, as opposed to the zwitterionic phos-
pholipids found in eukaryotes.25 Furthermore, the cho-
lesterol content of eukaryotic membranes helps to
stabilize and protect against binding of these cyclopep-
tides.39 It is noteworthy that bacteria have not yet dem-
onstrated appreciable resistance to this class of
antibiotics, making them attractive for further
study.23,25 Due to the fact that the target of these cyclo-
peptides is the bacterial membrane, pathogens would
have to alter their membrane composition and organiza-
tion to achieve resistance, a much more complicated and
costly process than simply the mutation of an enzymatic
binding site.
D-Phe10
Pro9
Val8
Orn7
O
NH2
Leu1
N
O
NH
O
HN
N
H
O
O
O
O
HN
NH
O
Leu6
HN
NH
O
HN
O
N
H2N
Orn2
Val3
Pro4
D-Phe5
Gramicidin S
With the favorable attributes of the cyclopeptides rec-
ognized, we have set out to improve the collective
understanding of their structure– activity relationship
(SAR) using Tyrocidine A as our point of reference.
The question arises, how does altering the overall
hydrophobicity, hydrophilicity, and amphipathicity
of the cyclopeptide beta sheet affect the drugs’ ability
to inhibit bacterial growth. Using molecular model-
ing, we rationalized generating a small set of Tyroci-
dine A analogues with specific point amino acid
substitutions that would affect either the polar or
non-polar face of the amphipathic cyclopeptide. A so-
lid-phase synthetic methodology was used to prepare
the parent compound, Tyrocidine A, as well as eight
other closely related analogues 1–8 (Figs. 2 and 3).
These novel analogues were then tested against a host
of clinically important gram positive and select gram
negative bacterial strains to further correlate structure
to function.
Phe3
DPhe1
Pro2
DPhe4
O
Leu10
N
O
NH
HN
NH
O
O
O
O
HN
H
N
O
Asn5
O
O
HN
NH
HN
O
H2N
O
NH
Gln6
O
Orn9
H2N
OH
H2N
Val8
Tyr7
Tyrocidine A
Figure 1. Structures of amphipathic cyclopeptides Gramicidin S and
Tyrocidine A.

M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6669
DPhe1
O
O
O
Pro2
S
Leu10
Orn9-NHBoc
Val8
N
H
R
1
O
R
Leu10
2
N
O
NH
HN
NH
R4
O
i)
O
O
O
Linker Activation
R3
HN
3
ii)
Deprotection
Cyclative Cleavage
Asn5
R2
NH
iii)
O
Asn5
O
O
HN
NH
HN
O
H2N
R1
Pro2
DPhe1
NHBoc
NH
R
O
R
Orn9
4
H2N
Val8
Tyrocidine A: R1 = Phe, R2 = DPhe, R3 = Gln, R4 = Tyr
1: R1 = Phe, R2 = DLys, R3 = Gln, R4 = Tyr
2: R1 = Phe, R2 = DLys, R3 = Lys, R4 = Tyr
5: R1 = Phe, R2 = DPhe, R3 = Lys, R4 = Tyr
6: R1 = Phe, R2 = DPhe, R3 = Gln, R4 = PFPhe
7: R1 = PFPhe, R2 = DPhe, R3 = Lys, R4 = Tyr
8: R1 = PFPhe, R2 = DLys, R3 = Lys, R4 = Tyr
3: R1 = PFPhe, R2 = DPhe, R3 = Gln, R4 = Tyr
4: R1 = PFPhe, R2 = DPhe, R3 = Gln, R4 = PFPhe
Figure 2. Solid phase synthesis of Tyrocidine A and analogues 1–8. (i) Linker activation with 25 equiv IACN, 11 equiv DIEA, in NMP; (ii)
deprotection with TFA/Phenol/TIPS/H2O (88:5:5:2), rt 2 h; (iii) cyclative cleavage with 20% DIEA in THF, rt 24 h. R1 = Phe or Pentafluorophe-
nylalanine (PFPhe), R2 = DLys or DPhe, R3 = Gln or Lys, R4 = Tyr or PFPhe.
2. Results
0.5 lg/mL for the gram positive bacteria. As previously
reported, daptomycin is a membrane disrupting cyclic
lipopeptide with broad therapeutic activity, however,
its in vitro potency has a dependency on calcium con-
centration.40,41 To determine if Tyrocidine A and the
novel analogues shared the same dependency on calcium
concentration, we screened activity at the standard cal-
cium concentration of 25 mg/L, as well as the reported
maximal value for daptomycin of 50 mg/L. Broadly,
increasing the calcium concentrations from 25 to
50 mg/L for Tyrocidine A and analogues moderately de-
creased antibacterial activity. In contrast, daptomycin
showed a marked increase in potency with increasing
calcium concentration as previously reported.40
The retention time of each compound was measured by
reverse phased LCMS, with retention time correlating
with overall predicted hydrophobicity/hydrophilicity.
Molecular modeling clearly depicted both the hydro-
phobic and hydrophilic faces of the Tyrocidine A scaf-
fold (Fig. 4). Hydrophobicity/hydrophilicity was
modulated by the placement of positive charges or the
incorporation of pentafluorophenyl residues. The order
of increasing hydrophobicity for Tyrocidine A and ana-
logues 1–8 was as follows: 2 > 8 > 1 > 7 > 5 > 6 > Tyro-
cidine A > 3 > 4 (Table 1).
The parent compound, Tyrocidine A, exhibited only
moderate antibacterial activity ranging from greater
than 128 lg/mL for E. coli to 16 lg/mL for the remain-
ing gram positive bacteria (Table 2). Analogue 1 showed
a similar profile as Tyrocidine A. Compounds 2 and 3
showed a marked improvement in antibacterial potency,
with an MIC improving from 2- to 8-fold over the par-
ent Tyrocidine A. Of note, compound 2 was the only
compound exhibiting activity against the gram negative
E. coli. Compound 4 displayed the least antibacterial
activity with an MIC range from 32 lg/mL to greater
than 256 lg/mL. Compound 5 demonstrated the most
potent and broadest antibacterial properties of all the
compounds tested, with MICs increasing from 2- to 8-
fold over the parent Tyrocidine A. Compounds 6–7
demonstrated similar antibacterial profiles with moder-
ate improvement in activity against gram positive patho-
3. Discussion
Several groups have been studying the structure–activity
relationships of streptogramin cyclopeptides Gramicidin
S34,35,42–48 and Tryocidine A (Fig. 1).49–53 Initial work
with Gramicidin S analogues has yielded a handful of
useful facts concerning the structure–activity relation-
ships of these cyclopeptides. Studies have found that
bacteriocidal and hemolytic properties are influenced
by several properties: rings size/rigidity, hydrophobicity,
and amphipathicity.44–47,54 A collection of similar stud-
ies have been conducted for the cyclopeptide Tyrocidine
A. Guo et al. have demonstrated that by using an ala-
nine substitution screen, minor changes in the peptide
composition of Tyrocidine A can yield improvements
in antibacterial potency.52 More specifically, by making
point substitutions that alter the amphipathicity of the
Tyrocidine analogue the antibacterial activity may be
systematically improved. Other recent studies by Kohli
et al. have demonstrated the utility of using nonriboso-
mal polypeptide synthetase (NRPS) technology for gen-
erating a privileged library of Tyrocidine A analogues
gens compared to Tyrocidine A. Compound
8
demonstrated improved activity against both gram posi-
tive pathogens, and the selected gram negative pathogen
E. coli.
Daptomycin exhibited broad activity and high potency
ranging from greater than 128 lg/mL for E. coli to 2–

6670
M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
Phe3
Phe3
DPhe1
DPhe1
Phe3
DPhe1
DLys4
Pro2
Pro2
DLys4
Pro2
NH
2
NH
2
DPhe4
O
O
O
Leu10
Leu10
N
N
Leu10
N
O
O
O
NH
HN
NH
HN
N
O
HN
H
NH
NH
O
O
NH
O
O
O
O
O
O
O
O
O
O
N
N
H
HN
HN
NH
NH
NH
O
N
Asn5
O
N
Asn5
O
N
O
H
O
O
HN
O
N
Asn5
HN
O
NH
HN
NH
HN
O
O
H
N
2
H N
2
NH
HN
O
H
N
2
NH
NH
NH
O
O
O
O
O
Orn9
H
2
Orn9
H
NH
2
Lys6
H
2
Gln6
Gln6
Orn9
H
N
2
2
H
2
Val8
Val8
Val8
OH
OH
OH
Tyr7
Tyr7
Tyr7
Tyrocidine A
1
2
F
F
PFPhe3
F
F
PFPhe3
Phe3
DPhe1
DPhe1
DPhe1
F
F
F
F
Pro2
Pro2
Pro2
DPhe4
F
O
DPhe4
F
O
DPhe4
O
Leu10
N
N
Leu10
N
Leu10
O
O
O
NH
O
HN
NH
HN
NH
HN
NH
NH
O
NH
O
O
O
O
O
O
O
O
O
O
O
HN
HN
HN
NH
NH
NH
Asn5
Asn5
O
O
Asn5
O
N
HN
O
O
HN
O
O
N
O
HN
NH
HN
O
H
N
O
HN
H
N
2
O
NH
HN
H
N
2
O
H N
2
NH
NH
NH
O
F
O
O
O
Gln6
Orn9
Orn9
H N
2
H
2
Orn9
H
2
NH
2
F
F
Lys6
H
N
2
Gln6
H N
2
Val8
Val8
Val8
OH
OH
F
Tyr7
Tyr7
F
PFPhe7
5
3
4
PFPhe3
F
F
PFPhe3
F
F
F
Phe3
DPhe1
DPhe1
DPhe1
F
Pro2
F
F
F
DLys4
Pro2
Pro2
DPhe4
DPhe4
F
O
NH
2
O
O
Leu10
N
N
Leu10
Leu10
N
O
O
O
NH
O
HN
NH
HN
NH
HN
NH
O
NH
O
O
O
NH
O
O
O
O
O
O
O
O
HN
HN
HN
NH
NH
NH
Asn5
O
HN
O
O
N
Asn5
O
O
N
HN
O
O
N
Asn5
O
HN
NH
HN
O
H
N
2
N
HN
H
O
N
HN
H
H N
2
O
H N
2
NH
NH
NH
O
F
O
O
O
Gln6
Orn9
H
2
Orn9
Lys6
H
2
NH
2
F
F
Orn9
Lys6
H N
2
H
2
NH
2
Val8
Val8
Val8
OH
OH
F
Tyr7
Tyr7
F
PFPhe7
7
6
8
Figure 3. Structure of Tyrocidine A and analogues 1–8. Residue substitutions on the hydrophilic and hydrophobic faces shown in bold above.
that exhibit moderate changes in antibacterial proper-
ties.50 The previous success of these groups at modulat-
ing antibacterial properties by using simple alanine
screens and point amino-acid substitutions is highly
encouraging and sets the stage for our work.
groups removed, and the resin subjected to cyclative
cleavage by treatment with diisopropylethylamine in
THF (Fig. 2). A recent study by Tariq et al. showed that
both biosynthetic and chemical mediated cyclizations
were successful for generating streptogramin–tyrocidine
derivatives, but that the chemical cyclative cleavage was
superior for generating quantities necessary for bioas-
say.55 Using similar solid phase synthesis techniques,
we were able to prepare the natural product Tyrocidine
A and eight closely related analogues 1–8. Of note,
yields were substantially improved by using a combina-
tion of HOAt in place of HOBt, NMP in place of DMF,
and heating the coupling reactions at 38 ꢀC.51,52 In con-
trast to biosynthetic methods, the synergistic benefits of
easily modulating the chemical structure without regard
for affecting enzyme efficiency, coupled with simplified
While the use of NRPS enzymes is an elegant approach,
the ability to complete a total synthesis on resin with a
single purification step is a clearly attractive option,
which we chose to utilize. A further advantage to the
complete solid phase synthesis approach was being able
to introduce point mutation without concern for affect-
ing enzymatic activity. This methodology makes use of a
commercially available sulfamylbutyryl safety catch
resin. A linear peptide may be loaded onto the resin
by standard solid phase techniques, the protecting

M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6671
Table 1. LCMS retention times for Tyrocidine A and analogues 1–8
Compound
Chargea
Rt (min)b
Tyrocidine A
+1
+2
+3
+1
+1
+2
+1
+2
+3
18.8
11.1
9.7
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
19.4
27.2c
15.3
18.7
12.4
9.8
a Charge based on number of primary amines.
b Retention time by reversed phase LCMS.
c Retention time correlates to hydrophilicity: increasing hydrophilicity
leads to shorter retention time. Addition of pentafluorophenyl
(PFPhe) residues decreases hydrophilicity.
of the cyclopeptide in question.50,52 Discussion is pre-
dominantly limited to antibacterial and hemolytic assay
results. Granted, there has yet to be definitive structural
evidence of how these antibiotics associate with the cel-
lular membrane, and molecular modeling must be trea-
ted with mild skepticism due to the variable degree of
possible conformations these cyclopeptides may take.
However, the structure of the cyclopeptide backbone
for Gramicidin S and Tyrocidine A is generally agreed
upon, and the assignment of individual residues within
the structure to either the hydrophobic or hydrophilic
face by modeling seems reasonable.56–59
The overall beta sheet conformation of these cyclopep-
tides is stabilized by intramolecular hydrogen bonding
between the carbonyl oxygen and the distal amides on
the opposing side of the cyclopeptide.60–62 It is likely
that this structural organization is what allows these lin-
ear peptides to self-cyclize using a pure solid phase
chemistry approach, as well as proving to be highly effi-
cient substrates for NRPS enzymes.49,51,60 Further, this
beta sheet conformation plays a key role in the cyclo-
peptides’ ability to disrupt bacterial membranes.34–38
The covalent linkage of the cyclopeptide locks in the
beta sheet conformation of Tyrocidine A allowing for
topographic organization and presentation of given sub-
stituents to either the hydrophobic or hydrophilic face
of the molecule. We can take advantage of the locked
conformation provided by Tyrocidine by choosing to
manipulate the substituents presented on either the
hydrophobic or hydrophilic face of the cyclopeptide.
Figure 4. Hydrophobic and hydrophilic faces of Tyrocidine A. (a)
Residues on hydrophobic face: Phe3, Asn5, Tyr7, Val8, Leu10. (b)
Chemical structure of Tyrocidine A. (c) Residues on hydrophilic face:
DPhe1, DPhe4, Gln6, Orn9.
loading and purification steps that provide milligram
quantities of product for assay, validate the protocol.
Figure 4 shows the putative hydrophobic and hydro-
philic faces of Tyrocidine A. We used this model to se-
lect positions Phe3 and Tyr7 on the hydrophobic face,
along with positions D-Phe4 and Gln6 on the hydrophilic
face for amino acid substitution. Compound 1 with a
substitution of a polar charged D-Lys4 in place of a
hydrophobic D-Phe4 on the hydrophilic face of the
cyclopeptide showed decreased antibacterial activity in
comparison to Tyrocidine A. This result is mildly sur-
prising in light of the fact that adding an additional po-
sitive charge positioned on the hydrophilic face of the
amphipathic cyclopeptide would arguably increase the
analogues’ affinity for the negatively charged bacterial
membrane. Of note, further increasing the charge of
Previously, a substantial body of work by Kondejewski
et al. on the the Gramicidin S scaffold examined how
specific changes in hydrophobicity, hydrophilicity,
amphipathicity, affinity for negative phospholipid sub-
strate, and cycle size all have intricate interlocking rela-
tionships.44–47 Such analysis remains to be accomplished
for the Tyrocidine A scaffold. Thus far, the majority of
works done to understand the intricate interplay
between structure and function of Tyrocidine A have in-
volved alanine screens or a small positional library with-
out much discussion of how these point substitutions
may affect the overall geometric or electronic properties

6672
M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
Table 2. Minimum inhibitory concentrations for Tyrocidine A analogues (lg/mL)a
Organismb
Strainc
TyrA
(1)
64
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Daptomycin (+Cad)
MRSA
MSSA
ATCC 33592
ATCC 29213
ATCC 29212
ATCC25922
MRL19010
MRL 17001
MRL 18734
MRL 18731
32
32
16
16
32
32
4
16
8
>256
>256
32
4
16
16
2
8
16
4
8
8
1
0.25
0.25
1
32
64
4
2
1
E. faecalis
E. colie
MRSE
16
4
>256
16
32
4
1
>128
>128
16
16
128
32
>256
256
32
128
2
>256
>256
>128
8
8
4
4
2
2
8
8
4
4
8
8
1
2
1
0.5
1
VRE
B. subtilis
B. cereus
32
16
8
2
4
16
16
2
16
128
256
2
2
0.25
0.25
64
2
4
0.5
a Inhibitory concentration required to completely inhibit bacterial growth.
b All gram positive bacteria with the exception of E. coli; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible
Staphylococcus aureus; E. faecalis, Enterococcus faecalis; E. coli, Escherichia coli; MRSE, methicillin-resistant Staphylococcus epidermidis; VRE,
vancomycin-resistant Enterococcus; B. subtilis, Bacillus subtilis and B. cereus, Bacillus cereus.
c MRL clinical isolates from USC-LAC Hospital.
d Calcium supplemented to 50 mg/L (normal 25 mg/L). Tyrocidine analogues did not show the same calcium dependence.
e E. coli is a gram negative bacterium.
compound
1
by adding another lysine residue
compound 2, compound 8 demonstrated a modest in-
crease in activity against the gram negative pathogen
E. coli, further supporting that these compounds may
eventually be tuned to target gram negative pathogens.
(Gln6 ! Lys6 compound 2) markedly improves the anti-
bacterial properties, providing as much as an 8-fold
improvement over the parent Tyrocidine A (Table 2).
The increased charges of compounds 2 and 8 provide
them with an improved activity toward the gram nega-
tive E. coli, indicating that future efforts may be focused
on tuning these compounds to target both gram negative
and gram positive pathogens.
With respect to delineating the importance of hydrophi-
licity versus hydrophobicity in improving antibacterial
activity, this study has shown that there is not a clear
preference for a single parameter in determining activity.
Of note, simply adding positive charge to the hydro-
philic face of Tyrocidine A does not result in improved
antibacterial activity, demonstrated by 1. Further, most
analogues showed significantly increased antibacterial
properties over the parent Tyrocidine A, while having
substantially different electronic profiles. For example,
while compounds 2 and 8 had the greatest positive
charge and hydrophilicity, compounds 3 and 6 had a
markedly increased hydrophobic nature and only a sin-
gle positive charge on the hydrophilic face (Table 1).
Such a similarity between the activities of these two
groups, respectively, seems to indicate that increasing
amphipathicity, either by increasing hydrophobicity or
hydrophilicity on an individual face, results in a moder-
ate increase in antibacterial activity. This is in contrast
to previous studies with Gramicidin S analogues assayed
against gram negative bacteria, which demonstrated a
negative correlation with increasing amphipathicity.46
Compound 4 demonstrates that there may be a limit
to how much we can increase the overall amphipathicity
of Tyrocidine A before activity is lost.
Compound 5, similar to compound 1, has a single lysine
addition, however, the substitution is at a different posi-
tion in the cyclopeptide. Interestingly, unlike compound
1, which showed poor antimicrobial activity when a sin-
gle lysine substitution was made (D-Phe4 ! D-Lys4),
there is a substantial increase in antimicrobial activity
when the lysine is positioned at a different point in the
cyclopeptide (Gln6 ! Lys6) as in compound 5. Such a
comparison supports our reasoning that simply increas-
ing charge does not directly lead to an increase in anti-
microbial properties, and that it is in fact a more
complicated interplay of structure vs. activity exists.
Substitution of the Phe3 and Tyr7 residues on the hydro-
phobic face of Tyrocidine A with pentafluorophenyl
(PFPhe) derivatives provided mixed results. The PFPhe
residue is particularly interesting due to its similar surface
area but dramatically modified electronic and hydropho-
bic nature when compared to Phe. Analogue 3, with a sin-
gle PFPhe substitution (Phe3 ! PFPhe3), showed
increased gram positive antibacterial activity similar to
that of compound 2 but failed to show gram negative
activity (Table 2). Movement of the PFPhe substitution
across the cyclopeptide to a different position, as is dem-
onstrated for compound 6 (Tyr7 ! PFPhe7), maintains
a similar antibacterial potency overall, but shows moder-
ate improvements in certain gram positive strains. Inter-
estingly, addition of a second PFPhe substitution, 4
(Phe3 ! PFPhe3) (Tyr7 ! PFPhe7), virtually abolished
antibacterial activity across the spectrum of strains tested.
The substantially improved antibacterial profile of com-
pound 5 versus 1 seems to indicate that the actual ring po-
sition of charged residues, and not just the placement of
the charge itself on the hydrophilic face of the compound,
may be of importance. The notion that residue position
and not simply facial substitution may be a factor in mod-
erating activity is supported, albeit to a lesser extent, by
the improvement of antibacterial activity of compound
6 when compared to 3 for strains of MRSE, VRE, and
E. faecalis.
Combining lysine and pentafluorophenyl substitutions
within the same molecule proved successful in increasing
the overall antibacterial activity of compounds 7 and 8
when compared to the parent Tyrocidine A. Similar to
It is noteworthy that Tyrocidine A and analogues, while
sharing the same mechanism of action as daptomycin,
do not share the same dependence on calcium.40 Fur-

M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6673
ther, the addition of calcium actually decreases the mea-
sured in vitro activity for Tyrocidine A and its ana-
logues. Whether this difference in calcium dependence
would have an implication for differences in molecular
mechanism seems unclear.
(DMF), 1-methyl-2-pyrrolidinone (NMP), 1,3-diisopro-
pylcarbodiimide (DIC), piperidine, diisopropylethyl-
amine (DIEA), 1-methylimidazole (NMI), activated
aluminum oxide, iodoacetonitrile, d6-dimethylsulfoxide,
and Sigmacoat were purchased from Aldrich St. Louis,
MO. 4-Sulfamylbutyryl AM resin (1.1 mmol/g) (01-64-
0152), Fmoc and Boc amino acids, and all other solid
phase coupling reagents were purchased from Novabio-
chem, San Diego, CA. Fluorinated amino acid deriva-
tives were purchased from PepTech Corp, Burlington,
MA.
4. Conclusion
With antibacterial resistance increasing at an alarming
rate, the development of new antibacterial therapeutic
agents capable of staving off microbial resistance is a
chief concern in clinical medicine.7,18,63 One class of
antibiotics, those that target and destroy bacterial mem-
branes, is attractive due to the decreased likelihood that
bacteria will be able to generate resistance to this mech-
anism.23–25 The amphipathic cyclic decapeptide, Tyroci-
dine A, is a model system for this class of antibiotics.26,27
Analytical and preparatory HPLC was acquired using a
Waters 600 controller equipped with a 2487 dual wave-
length detector and RCM 8 · 10 semi-prep C18 column.
Mass spectra were acquired using a Thermo/Finnigan
LCMS equipped with an Altech 10 mm · 2.1 mm C18
column (43871), autosampler, and triple wavelength
detector. UV spectra were acquired using a Shimadzu
UV-2401 spectrophotometer. NMR spectra were ac-
quired using a Varian 400 Hz instrument.
We have been able to use molecular modeling and solid
phase synthetic techniques to design and efficiently syn-
thesize Tyrocidine A and multiple novel analogues with
improved antibacterial activity. Further, we have shown
that the relationship between increasing hydrophilicity,
by the substitution of positive charge on the polar face
of the amphipathic structure, does not necessarily pro-
vide an increase in antibacterial activity, and that
charge/residue placement may be of increasing impor-
tance. Overall, we have demonstrated that an increase
in amphipathicity within a finite range improves anti-
bacterial activity. A substantial increase in the hydro-
phobic nature of Tyrocidine A by substitution with
multiple pentafluorophenyl residues demonstrates a
marked decrease in antibacterial activity, indicating a
narrower window for electrochemical modification than
originally envisioned. Finally, we demonstrated Tyroci-
dine A analogues do not have the same in vitro reliance
on calcium concentration as demonstrated by our com-
parator daptomycin.
5.2. Molecular modeling
A molecular model of Tyrocidine A was constructed
with a semi-empirical energy minimization (PM3) using
the Spartan Essential Software Package (2000). Model-
ing allowed visualization of amino acid side chains re-
ported to be on either the hydrophobic or hydrophilic
face of the amphipathic cyclopeptide (Fig. 4).50,52,57,59
Positions Phe3 and Tyr7 on the hydrophobic face, along
with positions DPhe4 and Gln6 on the hydrophilic face,
were selected for amino acid substitution. To ensure that
an observed change in antibacterial activity could be
attributed to an individual structural modification, only
a single progressive amino acid substitution was made
between Tyrocidine A and each respective analogue
(Fig. 3).
5.3. Solid phase synthesis (resin loading)
Moving forward there are a variety of other questions
that we would like to address. For example, we chose
to use lysine as our means of introducing positive charge
but would the antimicrobial activities be moderated dif-
ferently by delivery of positive charge through a differ-
ent functionality or linker length? We were able to
demonstrate an increase in gram negative antibacterial
activity by substituting the analogues with multiple po-
sitive charges and the ability to tune these compounds
for both gram negative and gram positive pathogens
with some sort of selectivity is certainly of interest. Test-
ing these compounds against a larger panel of clinically
important gram negative pathogens is in order.
Tyrocidine A and subsequent analogues were synthe-
sized using manual solid phase synthesis Fmoc and
Boc protocols. A glass resin reaction vessel was pre-
treated with Sigmacoat before use to prevent resin
from sticking to the glass surface during shaking.
Loading of the sulfamylbutyryl AM resin was accom-
plished as follows. Sulfamylbutyryl AM resin (1 g,
1.1 mmol), DCM (4 mL), and DMF (1 mL) were
added to the glass reaction vessel and shaken at room
temperature for 20 min. The resin was then drained. A
mixture of Fmoc-Leu-OH (1.94 g, 5.5 mmol), 1-meth-
ylimidazole (452 mg, 436 lL, 5.5 mmol), 1,3-diisopro-
pylcarbodiimide (DIC) (694 mg, 851 lL, 5.5 mmol),
DCM (4 mL), and DMF (1 mL) were added to the re-
sin and the mixture was shaken for 24 h at room tem-
perature. After 24 h, the resin was washed and the
loading procedure was repeated a second time as de-
scribed above. Resin loading was near quantitative
as determined by UV assay.64 Following resin loading,
the resin was washed with DMF, DCM, MeOH, and
Et2O, dried under vacuum, and stored at 4 ꢀC for la-
ter use.
5. Materials and methods
5.1. General
Methanol, diethyl ether, acetonitrile, and dichlorometh-
ane (all HPLC grade) were purchased from VWR, San
Diego, CA, and used without further purification.
Anhydrous tetrahydrofuran (THF), dimethylformamide

6674
M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
5.4. Manual solid phase synthesis
5.8. Resin self-cyclization and cleavage
To a solid phase reaction vessel charged with 0.2 g
of resin pre-loaded with Fmoc-Leu was added
DMF (2 mL) and allowed to swell at room tempera-
ture for 20 min. The resin was then drained and
washed with 20% piperdine in DMF (continuous
flow for 30 s), followed by the addition of more
20% piperdine in DMF (4 mL) and shaking at room
temperature for 30 min to accomplish Fmoc depro-
tection. Following deprotection, the resin was washed
thoroughly with DMF in preparation for the cou-
pling step. In a separate 20 mL glass vial, a mixture
of Fmoc-Orn(Boc)-OH (500 mg, 1.1 mmol), HOBt
(149 mg, 1.1 mmol), DIC (139 mg, 171 lL, 1.1 mmol),
and DMF (2 mL) was agitated until homogeneous
and allowed to stand for 10 min. This mixture was
then added to the solid phase reaction vessel contain-
ing the deprotected H2N-Leu-R resin and shaken for
3–4 h at room temperature. Following coupling, the
resin was washed thoroughly with DMF in prepara-
tion for the next round of deprotection and coupling.
If the resin was to be coupled at a later time, it was
washed with a series of DMF, DCM, MeOH, and
Et2O, dried under vacuum, and stored for later
use. The coupling of all subsequent amino acids to
synthesize resin bound linear derivatives of Tyroci-
dine A and 1–4 was accomplished as described
above.
The resin was washed with THF as described above.
Following washing, a 20% solution of DIEA in THF
(2.5 mL) was added and the mixture was shaken at room
temperature for 24 h. After 24 h the resin was filtered
and washed with THF (2 · 2 mL) and all filtrates were
combined in a round bottom flask. The solvent was
evaporated to provide a viscous solution or solid,
depending on the compound. In all cases a small
amount of methanol (200–400 lL) was added to the
round bottom to dissolve the solid or dilute the solution.
The mixture was then transferred to a 1.5 mL Eppen-
dorf tube. To the Eppendorf tube was added cold
Et2O to provide a white precipitate. The Eppendorf tube
was centrifuged at 14,000 rpm at 4 ꢀC for 5 min to pro-
vide a solid white pellet. The Eppendorf tube was dec-
anted, the pellet taken up in 200 lL of methanol and
subjected to a second round precipitation. The pellet
was then taken up in 500 lL of acetonitrile, flashed fro-
zen with liquid nitrogen, and subjected to lyophilization
to provide the final cyclized products (Tyrocidine A and
analogues 1–8) as fine white solids. It is noteworthy that
with different structural variants, such as the fluorinated
derivatives that show some solubility in the MeOH/Et2O
precipitation mixture, the use of 0.1% TFA in place of
Et2O is more effective.
5.9. Purification and characterization
5.5. Modified coupling conditions
In most cases, precipitated products were shown to be
sufficiently pure by HPLC and H NMR, and reported
yields are based on recovery following precipitation
and lyophilization. Compounds that were not pure by
precipitation were subjected to reverse phase prepara-
tory HPLC purification. The HPLC purification proto-
col is as follows: no more than 15 mg of crude
precipitate was dissolved in MeOH (1 mL) and loaded
onto the HPLC equipped with a 5 mL injection loop
and Waters 8 · 10 semi-prep C18 colum. The solvent
system was a 0.1% TFA (solvent A) and acetonitrile
(solvent B) mixture. The method consisted of a ramp
from 0 to 10% B over 5 min, followed by 10–80% B over
50 min, followed by 80–100% B over 5 min, with a flow
rate of 3 mL/min. The column was then washed at 100%
B for 5 min, followed by re-equilibration to 100% A for
the next run. HPLC spectra were recorded at 220 nm.
HPLC fractions were analyzed by LCMS. The LCMS
was equipped with an Altech 10 mm · 2.1 mm C18 col-
umn and was run using a solvent mixture of 1% phos-
phoric acid (solvent A) and acetonitrile (solvent B).
Briefly, 25 microliters of a dilute sample was autoinject-
ed to the LCMS and the standard method consisted of
the following: the column was equilibrated at 10% B
running at 125 lL/min followed by a ramp from 10 to
15% B over 5 min, followed by 15 to 80% B over
30 min, followed by 80–100% B over 5 min. The column
was then washed for 2 min at 100% B, followed by re-
equilibration. Tyrocidine A and analogue 1–8 retention
times were recorded and used as an indicator of their
hydrophobicity. Pooling of the appropriate fractions
followed by lyophilization provided pure cyclized
compounds.
It is noteworthy that changing the above conditions by
replacing HOBt with HOAt, utilizing NMP in place of
DMF, and heating the reactions to 38 ꢀC provided com-
pounds 5–8 in significantly greater quantities.
5.6. Safety catch linker activation
Following attachment of the terminal amino acid,
the resin was washed with NMP in preparation
for resin activation. To a 10 mL syringe equipped
with a 2 lm filter tip were added activated basic
alumina (200 mg), NMP (3.5 mL), DIEA (180 lL),
and iodoacetonitrile (280 lL). The mixture was then
filtered into the reaction vessel containing the unac-
tivated resin. The vessel was then wrapped in alu-
minum foil to exclude light and shaken at room
temperature for 24 h. Following 24 h the mixture
was filtered and the resin washed with DMF and
DCM in preparation for orthogonal Boc and O-
tBu deblocking.
5.7. Orthogonal deprotection
A mixture of TFA/Phenol/TIPS/H2O (88:5:5:1) was
freshly prepared and stored at 4 ꢀC for subsequent use.
To the activated resin prepared as described above was
added 3 mL of the TFA/Phenol/TIPS/H2O deblocking
mixture. The mixture was shaken at room temperature
for 2 h. Following deprotection, the resin was washed
with DCM and THF in preparation for on resin self-
cyclization of the linear peptide.

M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6675
5.10. Tyrocidine A
9.28 (d, 1H, J = 9.6 Hz), 9.11 (s, 1H), 9.07 (d, 1H,
J = 7.2 Hz), 9.00 (d, 1H, J = 9.2 Hz), 8.67 (s, 1H),
8.63 (d, 1H, J = 10 Hz), 8.04 (s, 1H), 7.84 (d, 1H,
J = 8.4 Hz), 7.64 (d, 1H, J = 8 Hz), 7.41 (d, 1H,
J = 10 Hz), 7.35 (d, 1H, J = 12.4 Hz), 7.25–7.09 (m,
11H), 6.88 (s, 1H), 5.49 (m, 1H), 5.13 (m, 1H),
4.78 (m, 1H), 4.52 (m, 3H), 4.21 (m, 1H), 4.17 (m,
1H), 3.84 (m, 1H), 3.23–2.68 (m, 8H), 2.03 (m,
2H), 1.91 (m, 4H), 1.5–1.23 (m, 6H), 0.93–0.76 (m,
12H), 0.47 (br, 2H); ES-MS m/z 1434.63 (M+H
1434.57 calcd C66H78F10N13O12). Retention time
LCMS 19.4 min.
Isolated as a fine white powder (7.4 mg 2.7% Yield) after
HPLC purification of the appropriate fractions. 1H
NMR (400 MHz, d6-DMSO, 25 ꢀC); 9.27 (s, 1H), 9.16
(s, 1H), 9.02 (m, 2H), 8.91 (d, 1H, J = 8.8 Hz), 8.70 (d,
1H, J = 4.4 Hz), 8.43 (d, 1H, J = 10 Hz), 8.01 (s, 1H),
7.91 (d, 1H, J = 7.6 Hz), 7.45 (s, 1H), 7.39 (d, 2H,
J = 8.4 Hz), 7.23–7.01 (m, 16H), 6.92 (d, 2H,
J = 8.4 Hz), 6.61 (d, 2H, J = 8.4 Hz), 5.57 (m, 1H),
5.26 (m, 1H), 4.55–4.42 (m, 4H), 4.35–4.25 (m, 2H),
4.06 (d, 1H, J = 8.4), 3.80 (m, 1H), 3.09–2.70 (m, 8H),
2.19 (m, 2H), 1.99 (m, 4H), 1.62 (m, 4H), 1.43 (m,
2H), 1.21 (m, 2H), 1.00–0.85 (m, 12H), 0.36 (br, 2H);
Compound 5: Isolated as a fine white powder after pre-
cipitation purification (45.2 mg 16.1% Yield). H NMR
1
ES-MS
m/z
1270.88
(M+H
1270.66
C66H88N13O13). Retention time LCMS = 18.8 min.
calcd
(400 MHz, d6-DMSO, 25 ꢀC); 9.07 (d, 1H, J = 7.2 Hz),
8.95 (d, 1H, J = 8.8 Hz), 8.81 (d, 1H, J = 8.4 Hz), 8.34
(d, 1H, J = 9.6 Hz), 8.00 (m, 4H), 7.36 (m, 2H), 7.26–
7.04 (m, 20H), 6.93 (d, 2H, J = 8.4 Hz), 6.63 (d, 2H,
J = 8.4 Hz), 5.62 (m, 1H), 5.21 (m, 1H), 4.54–4.47 (m,
4H), 4.31–4.25 (m, 2H), 4.05 (d, 1H, J = 8.0 Hz), 3.73
(s, 1H), 3.34–2.53 (m, 10H), 2.44–2.39 (m, 3H), 2.33
(m, 1H), 2.19–2.00 (m, 4H), 1.60–1.17 (m, 20H), 0.93–
0.89 (m, 12H), 0.32 (br, 2H); ES-MS m/z 1270.70
(M+H 1270.70 calcd C67H92N13O12). Retention time
LCMS 15.25 min.
Compound 1: Isolated as a fine white powder (7.2 mg
3.5% Yield) after HPLC purification of the appropri-
1
ate fractions. H NMR (400 MHz, d6-DMSO, 25 ꢀC);
9.26 (s, 1H), 9.17 (s, 1H), 8.99 (d, 1H, J = 9.2 Hz),
8.86 (s, 1H), 8.73 (d, 1H, J = 9.2 Hz), 8.66 (s, 1H),
8.31 (d, 1H, J = 10 Hz), 8.23 (m, 1H), 7.93 (s, 1H),
7.38–7.12 (m, 11H), 6.91 (d, 2H, J = 8.4 Hz), 6.61
(d, 2H, J = 8.4 Hz), 5.10 (m, 1H), 4.74 (m, 1H),
4.58 (m, 1H), 4.50 (m, 1H), 4.45 (m, 1H), 4.37 (m,
1H), 4.27 (m, 1H), 4.14 (m, 1H), 3.76 (m, 1H),
3.0–2.6 (m, 6H), 1.64 (m, 3H), 1.47 (m, 3H), 1.34
(m, 2H), 1.23 (m, 2H), 0.92–0.78 (m, 12), 0.40 (br,
2H); ES-MS m/z 1251.69 (M+H 1251.69 calcd
C63H91N14O13). Retention time LCMS 11.1 min.
Compound 6: Isolated as a fine white powder after precip-
itation purification (59 mg 20% Yield). 1H NMR
(400 MHz, d6-DMSO, 25 ꢀC); 9.28 (s, 1H), 9.06 (m,
2H), 8.96 (d, 1H, J = 10 Hz), 8.65 (m, 1H) 8.03 (s, 1H),
7.88 (d, 1H, J = 8.4 Hz), 7.60 (d, 1H, J = 8 Hz), 7.32–
7.03 (m, 15H), 6.87 (s, 1H), 5.57 (m, 1H), 5.29 (m, 1H),
4.56–4.42 (m, 4H), 4.28 (m, 2H), 4.05 (d, 1H,
J = 8.0 Hz), 3.84 (m, 1H), 3.15–2.65 (m, 8H), 2.43–1.98
(m, 6H), 1.68 (m, 4H), 1.61–1.21 (m, 7H), 0.93–0.88 (m,
12H), 0.34 (br, 2H); ES-MS m/z 1344.77 (M+H 1344.62
calcd C66H83F5N13O12). Retention time LCMS
18.65 min.
Compound 2: Isolated as a fine white powder after pre-
cipitation purification (15 mg 5.4% Yield) 1H NMR
(400 MHz, d6-DMSO, 25 ꢀC); 9.27 (s, 1H), 9.00 (d,
1H, J = 9.2 Hz), 8.86 (s, 1H), 8.75 (d, 1H, J = 9.2 Hz),
8.24 (d, 2H, J = 9.2), 7.99 (d, 1H, J = 4 Hz), 7.93 (s,
1H), 7.36–7.12 (m, 14H), 6.92 (d, 2H, J = 8.8 Hz), 6.63
(d, 2H, J = 8.4), 5.10, (s, 1H), 4.73 (m, 1H), 4.57 (m,
1H), 4.43 (m, 1H), 4.37 (m, 1H), 4.27 (m, 1H), 4.14 (d,
1H, J = 8.4 Hz), 4.09 (s, 1H), 3.70 (m, 1H), 2.98–2.54
(m, 6H), 1.64 (s, 3H), 1.48 (m, 6H), 1.34 (m, 5H), 1.24
(m, 2H), 1.18–1.0 (m, 3H), 0.94–0.70 (m, 12H), 0.38
(br, 2H); ES-MS m/z 1251.88 (M+H 1251.73 calcd
C64H95N14O12). Retention time LCMS 9.7 min.
Compound 7: Isolated as a fine white powder after
precipitation purification (40.4 mg 13.5% Yield). 1H
NMR (400 MHz, d6-DMSO, 25 ꢀC); 9.19 (s, 1H),
9.00 (m, 2H), 8.87 (d, 1H, J = 8 Hz), 8.28 (d, 1H,
J = 9.2) 7.96 (m, 1H), 7.86 (m, 1H), 7.41 (m, 2H),
7.22–7.07 (m, 10H), 6.92 (d, 2H, J = 8.8 Hz), 6.62
(d, 2H, J = 8.8 Hz) 5.45 (m, 1H), 5.05 (m, 1H), 4.73
(m, 1H), 4.47 (m, 2H), 4.27–4.21 (m, 4H), 4.13 (d,
1H, J = 8.0 Hz), 3.73 (m, 1H), 3.10–2.64 (m, 8H),
1.70–1.21 (m, 12H), 0.94–0.81 (m, 12H); ES-MS m/z
1360.81 (M+H 1360.65 calcd C67H87F5N13O12). Reten-
tion time LCMS 12.39 min.
Compound 3: Isolated as a fine white powder after pre-
cipitation purification (21 mg 7.0% Yield) 1H NMR
(400 MHz, d6-DMSO, 25 ꢀC); 9.27 (d, 1H, J = 9.6 Hz),
9.18 (s, 1H), 9.10, (s, 1H), 9.03 (d, 1H, J = 6 Hz), 8.96
(d, 1H, J = 9.2 Hz), 8.69 (s, 1H), 8.40 (d, 1H,
J = 9.6 Hz), 8.03 (s, 1H), 7.87 (d, 1H, J = 8.4 Hz), 7.44
(m, 6H), 7.24–7.09 (m, 12H), 6.94 (d, 2H, J = 8.4 Hz),
6.63 (d, 2H, J = 8.8 Hz), 5.50 (m, 1H), 5.11 (m, 1H),
4.53 (m, 3H), 4.31 (m, 1H), 4.22 (m, 1H), 4.17 (m,
1H), 3.81 (m, 1H), 3.01–2.67 (m, 12H), 2.00 (m, 2H),
1.66 (m, 7H), 1.25 (m, 9H), 0.92 (m, 12), 0.81 (br, 2H);
Compound 8: Isolated as a fine white powder after pre-
cipitation purification (26.1 mg 8.8% Yield). H NMR
1
(400 MHz, d6-DMSO, 25 ꢀC); 8.91 (s, 1H), 8.81 (s,
1H), 8.19 (m, 2H), 7.97–7.84 (m, 4H), 7.43–7.24 (m,
6H), 6.94 (d, 2H, J = 7.6 Hz), 6.63 (d, 2H, J = 8 Hz)
4.92 (m, 1H), 4.80 (m, 1H), 4.46 (m, 2H), 4.34–4.20
(m, 4H), 4.02 (m, 1H), 3.88 (m, 1H), 3.10–2.55 (m,
6H), 1.50–1.11 (m, 24H), 0.88–0.83 (m, 12H); ES-MS
m/z 1341.78 (M+H 1341.68 calcd C64H90F5N14O12).
Retention time LCMS 9.80 min.
ES-MS
m/z
1361.03
(M+H
1360.62
C66H83F5N13O13). Retention time LCMS 19.4 min.
calcd
Compound 4: Isolated as a fine white powder (8.4 mg
2.7% Yield) after HPLC purification of the appropri-
1
ate fractions. H NMR (400 MHz, d6-DMSO, 25 ꢀC);

6676
M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6. Bax, R. P. Clin. Infect. Dis. 1997, 24, S151–S153.
5.11. Bacterial inhibition assays
7. Coates, A.; Hu, Y. M.; Bax, R.; Page, C. Nat. Rev. Drug
Dis. 2002, 1, 895–910.
The in vitro activities of Tyrocidine-A, 1–8, and
daptomycin were measured using a microdilution method
(CLSI) according to standard procedures.65 The com-
pounds were dissolved in cation adjusted Mueller-Hinton
broth (CAMHB) or a small amount of dimethylsulfoxide
(dmso), followed by broth. Serial twofold dilutions were
prepared in CAMHB in micro titer trays. Daptomycin
and vancomycin were also tested as controls.
8. Cohen, M. L. Science 1992, 257, 1050–1055.
9. Spratt, B. G. Science 1994, 264, 388–393.
10. Sefton, A. M. Drugs 2002, 62, 557–566.
11. Walsh, C. Nature 2000, 406, 775–781.
12. Silver, L.; Bostian, K. Eur. J. Clin. Microbiol. Infect. Dis.
1990, 9, 455–461.
13. Walsh, C. Nat. Rev. Microbiol. 2003, 1, 65–70.
14. Rosamund, J.; Allsop, A. Science 2000, 287, 1973–
1976.
15. Tettelin, F. Drug Dis. Today 2001, 887–892.
16. Gonzales, R. D.; Schreckenberger, P. C.; Graham, M. B.;
Kelkar, S.; DenBesten, K.; Quinn, J. P. Lancet 2001, 357,
1179.
17. Pai, M. P.; Rodvold, K. A.; Schreckenberger, P. C.;
Gonzales, R. D.; Petrolatti, J. M.; Quinn, J. P. Clin. Infect.
Dis. 2002, 35, 1269–1272.
18. Bax, R. P.; Anderson, R.; Crew, J.; Fletcher, P.; Johnson,
T.; Kaplan, E.; Knaus, B.; Kristinsson, K.; Malek, M.;
Strandberg, L. Nat. Med. 1998, 4, 545–546.
19. Souli, M.; Kontopidou, F. V.; Koratzanis, E.; Antonia-
dou, A.; Giannitsioti, E.; Evangelopoulou, P.; Kannavaki,
S.; Giamarellou, H. Antimicrob. Agents Chemother. 2006,
50, 3166–3169.
20. Rose, W. E.; Rybak, M. J. Pharmacotherapy 2006, 26,
1099–1110.
21. Fraise, A. P. J. Infect. 2006, 53, 293–300.
22. Doan, T.-L.; Fung, H. B.; Mehta, D.; Riska, P. F. Clin.
Therapeutics 2006, 28, 1079–1106.
23. Hancock, R. E. W.; Chapple, D. S. Antimicrob. Agents.
Chemother. 1999, 43, 1317–1323.
24. Zasloff, M. N. Engl. J. Med. 2002, 347, 1199–1200.
25. Zasloff, M. Nature 2002, 415, 389–395.
26. Hotchkiss, R. D.; Dubos, R. J. J. Biol. Chem. 1940, 136,
803–804.
5.11.1. Bacterial strains. The test organisms were Staph-
ylococcus aureus ATCC 33592 (oxacillin-resistant),
Staphylococcus aureus ATCC 29213 (oxacillin-suscepti-
ble), Staphylococcus epidermidis (clinical isolate MRL
19010), Enterococcus faecalis ATCC 29212, Enterococ-
cus faecalis (clinical isolate MRL 17001, vancomycin-
resistant), Bacillus subtilis (clinical isolate MRL
18734), Bacillus cereus (clinical isolate MRL 18731),
and Escherichia coli ATCC 25922.
5.11.2. Method. Overnight cultures of the strains were
suspended in saline to equal the 0.5 McFarland turbidity
standard. They were further diluted 1:100 in CAMHB
and added to the wells containing the diluted com-
pounds. Drug-free wells were quantitatively subcultured
to determine the actual colony counts in the test. The
trays were incubated overnight at 36 ꢀC. After growth
had occurred, they were examined using an inverted mir-
ror apparatus. The MIC was the concentration that
completely inhibited growth of the organisms.
Acknowledgments
27. Hotchkiss, R. D.; Dubos, R. J. J. Biol. Chem. 1941, 141,
155–162.
This work was funded by the National Institutes of
Health through the NIH Roadmap for Medical Re-
search, Grant (1R01GM075857-01). Information on
this RFA (RM-05-013) can be found at (http://grants.
nih.gov/grants/guide/notice-files/NOT-RM-05-013. html,
links to nine initiatives are found here http://nihroad-
map.nih.gov). We also thank the Baxter Foundation
for fellowship to M.A.M., and A.P. and K.W. for help
with instrumentation and fruitful conversation.
28. Jelokhani-Niaraki, M.; Prenner, E. J.; Kay, C. M.;
McElhaney, R. N.; Hodges, R. S. J. Pept. Res. 2002, 60,
23–36.
29. Jelokhani-Niaraki, M.; Prenner, E. J.; Kondejewski, L.
H.; Kay, C. M.; McElhaney, R. N.; Hodges, R. S.
Biophys. J. 2000, 78, 286A.
30. Jelokhani-Niaraki, M.; Prenner, E. J.; Kondejewski, L.
H.; Kay, C. M.; McElhaney, R. N.; Hodges, R. S. J. Pept.
Res. 2001, 58, 293–306.
31. Jones, C. R.; Sikakana, C. T.; Hehir, S.; Kuo, M. C.;
Gibbons, W. A. Biophys. J. 1978, 24, 815–832.
32. Jones, C. R.; Sikakana, C. T.; Kuo, M. C.; Gibbons, W.
A. J. Am. Chem. Soc. 1978, 100, 5960–5961.
33. Kato, T.; Izumiya, N. Biochimica. Et Biophysica. Acta
1977, 493, 235–238.
References and notes
1. Chang, S.; Sievert, D. M.; Hageman, J. C.; Boulton, M.
L.; Tenover, F. C.; Downes, F. P.; Shah, S.; Rudrik, J. T.;
Pupp, G. R.; Brown, W. J.; Cardo, D.; Fridkin, S. K. N.
Engl. J. Med. 2003, 348, 1342–1347.
34. Lewis, R.; Prenner, E. J.; Kiricsi, M.; Jelokhani-Niaraki,
M.; Hodges, R. S.; McElhaney, R. N. Biophys. J. 2005, 88,
575A.
35. Prenner, E. J.; Jelokhani-Niaraki, M.; Kondejewski, L.
H.; Hodges, R. S.; McElhaney, R. N. Biophys. J. 2000, 78,
321A.
36. Salgado, J.; Grage, S. L.; Kondejewski, L. H.; Hodges, R.
S.; McElhaney, R. N.; Ulrich, A. S. J. Biomol. NMR 2001,
21, 191–208.
37. Shai, Y. Biochimica. Et Biophysica. Acta-Biomembr. 1999,
1462, 55–70.
38. Shai, Y. Biopolymers 2002, 66, 236–248.
39. Prenner, E. J.; Lewis, R.; Jelokhani-Niaraki, M.; Hodges,
R. S.; McElhaney, R. N. Biochimica. Et Biophysica. Acta-
Biomembr. 2001, 1510, 83–92.
2. Jones, R. N.; Farrell, D. J.; Morrissey, I. Antimicrob.
Agents. Chemother. 2003, 47, 2696–2698.
3. Sievert, D. M.; Boulton, M. L.; Stoltman, G.; Johnson,
D.; Stobierski, M. G.; Downes, F. P.; Somsel, P. A.;
Rudrik, J. T.; Brown, W.; Hafeez, W.; Lundstrom, T.;
Flanagan, E.; Johnson, R.; Mitchell, J. JAMA 2002, 288,
824–825.
4. Tsiodras, S.; Gold, H. S.; Sakoulas, G.; Eliopoulos, G. M.;
Wennersten, C.; Venkataraman, L.; Moellering, R. C.;
Ferraro, M. J. Lancet 2001, 358, 207–208.
5. Cui, L. Z.; Tominaga, E.; Neoh, H. M.; Hiramatsu, K.
Antimicrob. Agents. Chemother. 2006, 50, 1079–1082.

M. A. Marques et al. / Bioorg. Med. Chem. 15 (2007) 6667–6677
6677
40. Fuchs, P. C.; Barry, A. L.; Brown, S. D. Diagn. Microbiol.
Infect. Dis. 2000, 38, 51–58.
52. Qin, C. G.; Zhong, X. F.; Bu, X. Z.; Ng, N. L. J.; Guo, Z.
H. J. Med. Chem. 2003, 46, 4830–4833.
41. Ball, L. J.; Goult, C. M.; Donarski, J. A.; Micklefield, J.;
Ramesh, V. Org. Biomol. Chem. 2004, 2, 1872–1878.
42. Ando, S.; Nishikawa, H.; Takiguchi, H.; Lee, S.; Sugihara,
G. Biochimica. Et Biophysica. Acta Biomembr. 1993, 1147,
42–49.
53. Ruttenbe, Ma.; Mach, B. Biochemistry 1966, 5, 2864.
54. Katayama, T.; Nakao, K.; Akamatsu, M.; Ueno, T.;
Fujita, T. J. Pharm. Sci. 1994, 83, 1357–1362.
55. Mukhtar, T. A.; Koteva, K. P.; Wright, G. D. Chem. Biol.
2005, 12, 229–235.
43. Jelokhani-Niaraki, M.; Kondejewski, L. H.; Farmer, S.
W.; Hancock, R. E. W.; Kay, C. M.; Hodges, R. S.
Biochem. J. 2000, 349, 747–755.
44. Kondejewski, L. H.; Farmer, S. W.; Wishart, D. S.;
Hancock, R. E. W.; Hodges, R. S. Int. J. Pept. Protein.
Res. 1996, 47, 460–466.
56. Kuo, M. C.; Drakenberg, T.; Gibbons, W. A. J. Am.
Chem. Soc. 1980, 102, 520–524.
57. Kuo, M. C.; Gibbons, W. A. Biochemistry 1979, 18, 5855–
5867.
58. Kuo, M. C.; Gibbons, W. A. J. Biol. Chem. 1979, 254,
6278–6287.
45. Kondejewski, L. H.; Farmer, S. W.; Wishart, D. S.; Kay,
C. M.; Hancock, R. E. W.; Hodges, R. S. J. Biol. Chem.
1996, 271, 25261–25268.
59. Kuo, M. C.; Gibbons, W. A. Biophys. J. 1980, 32, 807–836.
60. Trauger, J. W.; Kohli, R. M.; Walsh, C. T. Biochemistry
2001, 40, 7092–7098.
46. Kondejewski, L. H.; Jelokhani-Niaraki, M.; Farmer, S.
W.; Lix, B.; Kay, C. M.; Sykes, B. D.; Hancock, R. E. W.;
Hodges, R. S. J. Biol. Chem. 1999, 274, 13181–13192.
47. Kondejewski, L. H.; Lee, D. L.; Jelokhani-Niaraki, M.;
Farmer, S. W.; Hancock, R. E. W.; Hodges, R. S. J. Biol.
Chem. 2002, 277, 67–74.
61. Kohli, R. M.; Trauger, J. W.; Schwarzer, D.; Marahiel, M.
A.; Walsh, C. T. Biochemistry 2001, 40, 7099–7108.
62. Trauger, J. W.; Kohli, R. M.; Mootz, H. D.; Marahiel, M.
A.; Walsh, C. T. Nature 2000, 407, 215–218.
63. Bax, R.; Mullan, N.; Verhoef, J. Int. J. Antimicrob. Agents
2000, 16, 51–59.
48. McInnes, C.; Kondejewski, L. H.; Hodges, R. S.; Sykes, B.
D. J. Biol. Chem. 2000, 275, 14287–14294.
49. Bu, X. Z.; Wu, X. M.; Xie, G. Y.; Guo, Z. H. Org. Lett.
2002, 4, 2893–2895.
50. Kohli, R. M.; Walsh, C. T.; Burkart, M. D. Nature 2002,
418, 658–661.
51. Qin, C. G.; Bu, X. Z.; Wu, X. M.; Guo, Z. H. J. Comb.
Chem. 2003, 5, 353–355.
64. Kay, C.; Lorthioir, O. E.; Parr, N. J.; Congreve, M.;
McKeown, S. C.; Scicinski, J. J.; Ley, S. V. Biotech.
Bioeng. 2000, 71, 110–118.
65. CLSI, Performance Standards for Antimicrobial Suscep-
tibility Testing; Sixteenth Informational Supplement. In
CLSI Document M100-S16, 7th ed.; Clinical and Labora-
tory Standards Institute: Wayne, Pennsylvania, 2006; Vol.
26, pp M100–S16.