Design and synthesis of unprecedented cyclic γ-AApeptides for antimicrobial development

Article abstract of DOI:10.1039/c2sc20428b

Antimicrobial drug resistance is one of the greatest threats facing mankind. Antimicrobial peptides (AMPs) can potentially circumvent drug resistance, probably through a bacterial membrane-disruption mechanism. However, they suffer from low in vivo stability, potential immunogenicity, and difficulty in optimization. The development of antimicrobial peptidomimetics is therefore an emerging research area as they avoid the potential disadvantages of AMPs. Cyclic peptidomimetics are of significant interest since constraints induced by cyclization are expected to further improve their antimicrobial activity. Nonetheless, the report of cyclic oligomeric peptidomimetics for antimicrobial development is rare. Herein, for the first time, we report the design and synthesis of cyclic γ-AApeptides via an on-resin cyclization. These cyclic γ-AApeptides are potent and broad-spectrum active against fungus and multi-drug resistant Gram-positive and Gram-negative bacterial pathogens. Our results demonstrate the potential of cyclic γ-AApeptides as a new class of antibiotics to circumvent drug resistance by mimicking the bactericidal mechanism of AMPs. Meanwhile, the facile synthesis of cyclic γ-AApeptides may further expand the applications of γ-AApeptides in biomedical sciences. The Royal Society of Chemistry 2012.

Full text of DOI:10.1039/c2sc20428b

View Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Chemical Science
Cite this: Chem. Sci., 2012, 3, 2570
www.rsc.org/chemicalscience
EDGE ARTICLE
Design and synthesis of unprecedented cyclic g-AApeptides for antimicrobial
development†
Haifan Wu,a Youhong Niu,a Shruti Padhee,a Rongsheng E. Wang,a Yaqiong Li,a Qiao Qiao,a Ge Bai,a
b
Chuanhai Cao and Jianfeng Cai
a
*
Received 5th April 2012, Accepted 17th May 2012
DOI: 10.1039/c2sc20428b
Antimicrobial drug resistance is one of the greatest threats facing mankind. Antimicrobial peptides
(AMPs) can potentially circumvent drug resistance, probably through a bacterial membrane-disruption
mechanism. However, they suffer from low in vivo stability, potential immunogenicity, and difficulty in
optimization. The development of antimicrobial peptidomimetics is therefore an emerging research
area as they avoid the potential disadvantages of AMPs. Cyclic peptidomimetics are of significant
interest since constraints induced by cyclization are expected to further improve their antimicrobial
activity. Nonetheless, the report of cyclic oligomeric peptidomimetics for antimicrobial development is
rare. Herein, for the first time, we report the design and synthesis of cyclic g-AApeptides via an on-resin
cyclization. These cyclic g-AApeptides are potent and broad-spectrum active against fungus and multi-
drug resistant Gram-positive and Gram-negative bacterial pathogens. Our results demonstrate the
potential of cyclic g-AApeptides as a new class of antibiotics to circumvent drug resistance by
mimicking the bactericidal mechanism of AMPs. Meanwhile, the facile synthesis of cyclic
g-AApeptides may further expand the applications of g-AApeptides in biomedical sciences.
thereby considered the driving force of their selectivity for
Introduction
bacteria against mammalian cells, in which anionic phospho-
lipids only exist in the inner leaflet of membrane, and the overall
charge of their membranes is zwitterionic.4,5 Due to this unique
antimicrobial mechanism which depends on the peptide’s global
chemical properties instead of their specific sequences, AMPs are
difficult for bacteria to develop resistance against.1 Further,
AMPs exhibit broad-spectrum activities against bacteria (both
Gram-positive and Gram-negative), fungi, and even viruses.1,2
Hence, they are regarded as ideal agents to supplement current
antimicrobial remedy.3
One of the major health concerns for eukaryotes is the infection
by pathogenic bacteria, which became acute in recent years due
to the bacteria’s rapid development of multi-drug resistance to
conventional antibiotics.1,2 While chemotherapeutic agents
primarily target the specific metabolic processes in bacteria, there
is also a class of ‘‘host-defense’’ peptides produced in the
eukaryotic innate immune response, which kill invading patho-
gens mainly through disruption of bacterial membranes.1,2 It is
generally accepted that these antimicrobial peptides (AMPs) can
be induced to segregate their cationic and lipophilic side-chain
functional groups upon binding to negatively charged bacterial
membranes, and the resulting globally amphipathic conforma-
tion leads to the disruption of bacterial membranes.3,4 Although
the details of membrane disruption remain elusive, it is widely
recognized that the hydrophobic part of AMPs drives the pep-
tide’s penetration through membrane via a hydrophobic inter-
action.4,5 The penetration process causes the depolarization of
bacterial membranes and often leads to their cell death.5 The
electrostatic interaction of AMPs with bacterial membrane is
Despite all the promising potentials, AMPs’ intrinsic peptidic
nature severely limits their practical application as therapeutics,
by making them immunoreactive, and susceptible to proteolytic
degradation.6 To circumvent the drawbacks, non-natural pepti-
domimetics were recently developed to mimic AMPs, in terms of
both antimicrobial activities and intrinsic mechanisms, while
their backbones were modified to be protease-resistant.7 Over the
last decade, non-natural antimicrobial oligomers have been
extensively investigated, such as b-peptides, peptoids, aryl-
amides, and other synthetic polymers.5,8 Whereas initial
approaches focused on peptide mimics that can adopt regular
helical conformations, it was later discovered to be unnecessary
for potent antimicrobial activity.9,10 Despite the growing interest
and intensive studies on linear peptidomimetics, it remains
a challenge to introduce a diverse set of functional groups to tune
their activity and selectivity, with the reported structure–activity
aDepartment of Chemistry, University of South Florida, 4202 E. Fowler
Ave, Tampa, FL 33620. E-mail: jianfengcai@usf.edu
bCollege of Pharmacy, University of South Florida, 4202 E. Fowler Ave.,
Tampa, FL 33620
† Electronic supplementary information (ESI) available: Experimental
procedure, characterization of g-AApeptide building blocks and cyclic
g-AApeptides. See DOI: 10.1039/c2sc20428b
2570 | Chem. Sci., 2012, 3, 2570–2575
This journal is ª The Royal Society of Chemistry 2012

View Online
relationship sometimes being inconsistent.11 In addition, some
active peptidomimetics do not exhibit good selectivity between
bacteria and mammalian cells,12 and there is still much room to
further increase their antimicrobial activity. Exploration of
antimicrobial agents with potent, broad-spectrum anti-bacteria
activity, while maintaining an excellent selectivity against
bacteria, is an urgent goal in chemical biology.
cyclic peptides or peptide–peptoid chimeras, whose structures, as
revealed by NMR, were unordered in water, but readily inducible
to form regular b-hairpins in the membrane-mimicking envi-
ronment. It is believed that the conformational bias induced by
the constrained template would stabilize the hairpin structure,
leading to the cluster of functional groups to form a hydrophobic
and
a hydrophilic face, upon interaction with bacteria
We recently reported a new class of peptidomimetics, termed
‘‘g-AApeptides’’,13 because they contain
membrane.21,27 Whereas the hemolytic activity of peptide oligo-
mers is determined by their lipophilicity, the antimicrobial effects
are always mediated by peptide charges and global amphiphi-
licity.27 It is therefore very interesting to investigate the antimi-
crobial activity of the cyclic peptidomimetics, as they are
expected to be more stable against proteolysis than cyclic
peptides, and more potent than linear peptidomimetics.28
a N-acylated-N-
aminoethyl amino acid unit (Fig. 1). In this unit, one side chain is
connected to the g-C in relation to carbonyl group, and the other
side chain is linked to the central N through acylation. As shown
in Fig. 1, g-AApeptides are able to project an identical number of
side chains as natural peptides of the same length; as such, they
have great potential to be developed for peptide mimicry.
Compared to conventional peptides, g-AApeptides are much
superior in both their limitless potential for diversification and
their inherent resistance to biodegradation. The synthesis of g-
AApeptides is straightforward, which facilitates library devel-
opment for drug screening purposes. Certain g-AApeptides have
been demonstrated to disrupt protein–protein interactions,13 and
to mimic a protein’s binding activity with nucleic acids.14 A small
library of linear g-AApeptides was designed, which led to the
identification of a potent g-AApeptide (g5) that displays signif-
icant activity against both bacteria and fungi, including the
multi-drug resistant clinically-relevant strains.15 In the effort of
our continuous exploration of antimicrobial g-AApeptides, we
herein report for the first time the design, synthesis, and evalu-
ation of cyclic g-AApeptides, some of which display antimicro-
bial activities superior to the previously reported linear ones.
Herein we designed cyclic antimicrobial g-AApeptides based
on the simple rationale that we previously developed to
successfully generate linear antimicrobial AApeptides
(Fig. S1†).15,29 In this rationale, potent antimicrobial activity can
be achieved by joining amphiphilic building blocks together to
form a globally amphipathic conformation upon the interaction
with bacterial membranes. The activity and selectivity can be fine
tuned by varying the ratio of cationic/hydrophobic groups. To
achieve a global distribution of cationic and hydrophobic groups
along the backbone, we prepared amphiphilic building blocks
with a cationic group and a hydrophobic group on either side
(Fig. S1A). By joining these building blocks together and
cyclizing the resulting oligomer (Fig. S1B†), a global amphiphi-
licity is expected to be achieved upon binding to bacterial
membranes.15 As such, the amphiphilic building block 2 was
prepared according to previously published procedure,15 in
which the amino acid is lysine, and the phenyl ended side chain is
appended to the amine (Fig. 2a). Given that an introduction of
a hydrophobic building block can tune the overall amphiphilicity
of g-AApeptides and improve their antimicrobial activity,15 we
also prepared building blocks 3 (Fig. 2) based on the reported
procedure. To facilitate the on-resin cyclization of g-AApeptide,
Results and discussion
Cyclic antimicrobial peptides were commonly observed in
nature, such as gramicidin S, tyrocidine, polymyxin B, and
protegrin I, which generally adopt semi-rigid backbone confor-
mations, with substituents positioned in well-defined space.16
Such a semi-rigid backbone as a result of cyclization favors the
binding event in entropy, while still possessing some flexibility to
optimize their conformations for binding.16,17 It was shown that
the lack of disulfide bonding diminished their hairpin confor-
mation, and reduced the membranolytic activity.18–20 Hence, the
cyclic peptides may have enhanced antimicrobial activity
compared to the linear ones. There has been significant effort in
the development and investigation of cyclic antimicrobial
peptides.16,21–26 For instance, Robinson et al.21,22 prepared several
Fig. 2 g-AApeptide building blocks used in the preparation of cyclic
g-AApeptides. a, The structures of the building blocks; b, the synthesis of
building blocks 1 and 3.
Fig. 1 General structures of a-peptides and g-AApeptides.
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci., 2012, 3, 2570–2575 | 2571

View Online
a special g-AApeptide building block 1 was designed. While the
synthesis was carried out similarly to the previously reported,15
the mono-allyl succinate was employed to modify the amine
(Fig. 2).
After building block 1 was conjugated to the beads, other
building blocks were sequentially assembled via standard Fmoc
chemistry.15 In the end, the allyl ester from the peptide was
deprotected by standard Pd(PPh3)4/PhSiH3 reduction, and the
resulting free carboxylate group reacted with the primary amino
group of the last assembled building block, which resulted in
protected cyclized g-AApeptides on the resin (Fig. 3a). The
desired cyclic g-AApeptides (Fig. 3b) were obtained upon
treatment with TFA and HPLC purification. The structure of the
previously prepared linear g5 is shown in Fig. 3c for comparison.
Cyclized g-AApeptides (HW-B-3, -4, -5) comprising of four,
five, and six amphiphilic building blocks were prepared as an
initial attempt, and tested for their antimicrobial activities
against a series of Gram-positive and Gram-negative bacteria
and fungi, many of which are multi-drug resistant and clinically-
relevant strains (Table 1). The oligomers’ hemolytic activities
towards human red blood cells were also measured, as an indi-
cation of their selectivity. For comparison, Pexiganan, a phase
III antimicrobial peptide drug candidate,3,30–32 as well as g5, the
most potent linear g-AApeptide,15 were both used as controls.
Similar to a linear g-AApeptide, which appeared to be more
potent with a longer sequence,15 the cyclic g-AApeptide with an
increasing ring size (from HW-B-3 to HW-B-5) tended to
augment the antimicrobial activities (Table 1). The most potent
cyclic g-AApeptide HW-B-5 has a similar activity to the well-
known Pexiganan, though it is still inferior to g5. It is notable
that the hemolytic activity of HW-B-5 is much less than Pex-
iganan and g5, implying the potential to improve its anti-bacteria
activities through the introduction of hydrophobic building
blocks.15
Given that linear g5 bears hydrophobic building blocks, and
thus exhibited much more potent antimicrobial activities than
other linear g-AApeptides completely made from amphiphilic
building blocks,15 we attempted this similar effort for our cyclic
g-AApeptides. HW-B-11, -12, and -13 were thereby prepared to
incorporate the same number of building blocks as HW-B-5, but
have one or two amphiphilic building blocks replaced by the
hydrophobic ones (Fig. 3b). As a result, HW-B-11, with the
change of only one building block, showed enhanced antimi-
crobial activities, especially against Gram-positive bacteria,
which are comparable to linear g5, and better than Pexiganan
(Table 1). In spite of its weaker activity towards Gram-negative
strain K. pneumoniae, HW-B-11 has a stronger inhibition of
fungus C. albicans than both Pexiganan and linear g5, with
a significant minimum inhibitory concentration (MIC) value of
5 mg mLÀ1. In addition, HW-B-11 still possesses a low hemolytic
activity, thereby making it a promising antimicrobial agent with
comparable or even better selectivity than Pexiganan and linear
g5. Following this success, HW-B-12 was developed to incor-
porate two hydrophobic building blocks, which are separated by
two hydrophilic building blocks, in a way similar to the design of
linear g5. Surprisingly, its antimicrobial activities were not
improved, or even weaker with respect to HW-B-11. In order to
assess whether the diminuendo of activity is due to the incor-
poration of more hydrophobic building blocks or is caused by
Fig. 3 a, General synthesis of cyclic g-AApeptides via on-resin cycli-
zation. b, The structures of cyclic g-AApeptides. c, The structure of linear
g-AApeptide g5.
their relative positions in the ring, HW-B-13 was synthesized by
placing two hydrophobic building blocks adjacent to each other
(Fig. 3b). As a result, HW-B-13 exhibited even better activities
than HW-B-11, Pexiganan, and linear g5 to arrest the growth of
both Gram-positive and Gram-negative bacteria pathogens, as
2572 | Chem. Sci., 2012, 3, 2570–2575
This journal is ª The Royal Society of Chemistry 2012

View Online
Table 1 The antimicrobial and hemolytic activities of oligomers. The MICs of Pexiganan and Linear g5 (obtained in the listed references) were shown
for comparison
MIC (mg mLÀ1
)
Organism
HW-B-3
HW-B-4
HW-B-5
HW-B-11 HW-B-12 HW-B-13a HW-B-14a Pexiganan3,30–32 Linear g515
Gram-positive
B. subtilis
S. epidermidis (MRSE) >50
25–50
10
10
20
>50
5
8
20
2
2
15
5
5
5
8
6
1
2
5
1
2
2
5
3
4
8
32
16
2
5
5
5
E. faecalis (VREF)
S. aureus (MRSA)
Gram-negative
K. pneumoniae
P. aeruginosa
>50
>50
25–50
>50
>50
>50
20
>50
18
20
10
5
>50
8
8
10
10
8
8-16
5
>50
Fungus
C. albicans
Hemolysis (H10/H50
>50
>50
>50
5
5
>500/>500 >500/>500 >500/>500 200/>500 150/450
2
40/100
4
45/300
124
181/495
8
75/300
)
a
Sequences showing the most broad-spectrum antimicrobial activity.
well as in Fungus (Table 1). To the best of our knowledge, HW-
B-13 is among the most potent antimicrobial peptidomimetics
with broad-spectrum activity. Especially towards two of the most
clinically relevant strains S. aureus (MRSA) and P. aeruginosa
(PA), HW-B-13 achieved a MIC value of 1 mg mLÀ1 and 8 mg
mLÀ1, respectively; which is at least 5-fold more potent than the
linear g5. Though HW-B-13 appears to be more hemolytic, its
overall selectivity in several important pathogens is still
improved relative to linear g5 and Pexiganan.
disruption events within bacteria membranes. On the contrary,
the amphiphilic topology of HW-B-12 may be scrambled by the
separated hydrophobic building blocks. Though linear g-
AApeptides with scrambled amphipathicity can still be induced
by a membrane to adopt a global amphiphilicity,15 the rigid
structure of cyclic g-AApeptides compromises their conforma-
tional flexibility.
The cytotoxicity of these cyclic g-AApeptides towards
mammalian cells (N2a/APP cell line) was also evaluated (Fig. S7,
ESI†). The results are quite consistent to hemolytic results, with
more hemolytic sequences being more toxic. However, the EC50s
(the effective concentration to cause half of the cells death) are all
above 12.5 mg mLÀ1, indicating cyclic g-AApeptides are very
selective towards bacteria over mammalian cells. For instance,
HW-B-13 is at least 12.5-fold more selective towards MRSA.
With further development and optimization, there is great
promise to identify more selective and potent antimicrobial g-
AApeptides in the future.
To further investigate the effect of hydrophobicity and to tune
the activity, HW-B-14 (Fig. 3b) with three hydrophobic building
blocks was developed, which, however, resulted in a slightly
decreased antimicrobial activity and hemolytic activity (Table 1)
in comparison to HW-B-13. Nevertheless, the activity and
selectivity of HW-B-14 are still generally comparable, or superior
to linear g5, against several strains including MRSA and PA.
Thus it is also a promising candidate for future antibiotic
development. Based on the results, it appears that the inclusion
of two neighbouring hydrophobic building blocks brings in the
optimal antimicrobial activity. The structure–activity studies of
HW-B-5, -11, -13, and -14 suggest that a higher percentage of
hydrophobic groups in g-AApeptides lead to a higher antimi-
crobial activity. It is well accepted that more lipophilicity would
lead to an increased hemolytic activity.3,9 While the result of our
structure–activity studies generally supports this rule, the slightly
decreased hemolytic activity demonstrated by HW-B-14, which
has one more hydrophobic building block than HW-B-13, is
quite unexpected. It suggests that besides the absolute hydro-
phobicity of functional groups, the overall conformations of
molecules may also affect their hemolytic activity. In this case,
multiple neighbouring hydrophobic building blocks in HW-B-14
promote the hydrophobic interactions among their side chains,
which limit their efficient contact with red blood cells. Finally,
the distinct activities between HW-B-12 and HW-B-13 suggest
the importance of position for hydrophobic building blocks. A
preliminary computer modelling of HW-B-13 reveals that the
cyclic g-AApeptide naturally adopts a globally amphipathic
conformation, with cationic side groups clustered at the bottom
left face of the ring, and the majority of hydrophobic groups at
the top face of the ring (Fig. S4†). Such a constrained structure
with predefined amphiphilicity may favor the binding and
In order to understand the antimicrobial mechanisms of cyclic
g-AApeptides, the most active ones, HW-B-11, -13, and -14,
were used to investigate their effects in cytoplasmic membrane
disruption through the depolarization of the S. aureus membrane
(Fig. 4).5 The membrane potential-sensitive dye DiSC3 was used,
the distribution of which between the medium and the cell inte-
rior reflects the membrane potential.5 The loss of membrane
potential as a result of membrane permeation/disruption will
lead to a dramatic increase in fluorescent intensity.5 Although the
oligomer concentration needed for depolarization is actually
higher than the oligomers’ MIC values, which is consistent to the
previous report,5 generally more active antimicrobial oligomers
with lower MIC values tended to reach a high percentage of
depolarization at a lower concentration. Such a trend was clearly
demonstrated by HW-B-11, -13 and -14, which supports the
membrane disruption mechanism of cyclic g-AApeptides
(Fig. 4).
The antimicrobial mechanism of cyclic g-AApeptide was
further assessed by fluorescence microscopy, in which B. subtilis
was treated with the most potent HW-B-13, and in the meantime
stained with DAPI and PI33 dyes (Fig. 5). DAPI stains all
bacterial cells irrespective of their viability, while PI selectively
stains injured or dead cells with damaged membranes.34,35
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci., 2012, 3, 2570–2575 | 2573

View Online
different functional groups, which could be easily achieved due
to the limitless potential of derivatization of g-AApeptides. In
addition, with a specifically designed building block for cycli-
zation, the corresponding synthesis is facile and straightfor-
ward, which may propel the application of cyclic g-AApeptides
to a broad field of biomedical research. The finding of enhanced
activity through the cyclization of g-AApeptide may also shed
light on the design and optimization of other non-natural
oligomers for future development of promising antimicrobial
agents. Introduction of diverse functional groups into cyclic g-
AApeptides for antimicrobial development, and other biolog-
ical applications of cyclic g-AApeptides are currently being
explored.
Acknowledgements
This work is supported by the USF Start-up fund.
Fig. 4 Depolarization of the cytoplasmic membrane of S. aureus by
cyclic g-AApeptides.
References
1 A. K. Marr, W. J. Gooderham and R. E. Hancock, Curr. Opin.
Pharmacol., 2006, 6, 468–472.
2 R. E. Hancock and H. G. Sahl, Nat. Biotechnol., 2006, 24, 1551–
1557.
3 N. P. Chongsiriwatana, J. A. Patch, A. M. Czyzewski, M. T. Dohm,
A. Ivankin, D. Gidalevitz, R. N. Zuckermann and A. E. Barron, Proc.
Natl. Acad. Sci. U. S. A., 2008, 105, 2794–2799.
4 B. Mensa, Y. H. Kim, S. Choi, R. Scott, G. A. Caputo and
W. F. DeGrado, Antimicrob. Agents Chemother., 2011, 55, 5043–
5053.
5 S. Choi, A. Isaacs, D. Clements, D. Liu, H. Kim, R. W. Scott,
J. D. Winkler and W. F. DeGrado, Proc. Natl. Acad. Sci. U. S. A.,
2009, 106, 6968–6973.
6 M. Zaiou, J. Mol. Med., 2007, 85, 317–329.
7 A. Violette, S. Fournel, K. Lamour, O. Chaloin, B. Frisch,
J. P. Briand, H. Monteil and G. Guichard, Chem. Biol., 2006, 13,
531–538.
8 G. N. Tew, R. W. Scott, M. L. Klein and W. F. Degrado, Acc. Chem.
Res., 2010, 43, 30–39.
9 B. P. Mowery, A. H. Lindner, B. Weisblum, S. S. Stahl and
S. H. Gellman, J. Am. Chem. Soc., 2009, 131, 9735–9745.
10 M. A. Schmitt, B. Weisblum and S. H. Gellman, J. Am. Chem. Soc.,
2004, 126, 6848–6849.
11 S. A. Fowler and H. E. Blackwell, Org. Biomol. Chem., 2009, 7, 1508–
1524.
Fig. 5 Fluorescence micrographs of B. subtilis treated with 5 mg mlÀ1
cyclic g-AApeptide HW-B-13 for 2 h. a1, Control, no treatment, DAPI
stained; a2, control, no treatment, PI stained; a3, control, no treatment,
the merged view. b1, HW-B-13 treatment, DAPI stained; b2, HW-B-13
treatment, PI stained; b3, HW-B-13 treatment, the merged view. Scale
bar: 2 mm.
12 E. A. Porter, X. Wang, H. S. Lee, B. Weisblum and S. H. Gellman,
Nature, 2000, 404, 565.
13 Y. Niu, Y. Hu, X. Li, J. Chen and J. Cai, New J. Chem., 2011, 35, 542–
545.
14 Y. Niu, A. J. Jones, H. Wu, G. Varani and J. Cai, Org. Biomol. Chem.,
2011, 9, 6604–6609.
15 Y. Niu, S. Padhee, H. Wu, G. Bai, L. Harrington, W. N. Burda,
L. N. Shaw, C. Cao and J. Cai, Chem. Commun., 2011, 47, 12197–
12199.
16 D. Obrecht, J. A. Robinson, F. Bernardini, C. Bisang,
S. J. DeMarco, K. Moehle and F. O. Gombert, Curr. Med. Chem.,
2009, 16, 42–65.
17 C. A. Olsen, H. L. Ziegler, H. M. Nielsen, N. Frimodt-Moller,
J. W. Jaroszewski and H. Franzyk, ChemBioChem, 2010, 11, 1356–
1360.
18 H. Tamamura, R. Ikoma, M. Niwa, S. Funakoshi, T. Murakami and
N. Fujii, Chem. Pharm. Bull., 1993, 41, 978–980.
19 H. Tamamura, T. Murakami, S. Horiuchi, K. Sugihara, A. Otaka,
W. Takada, T. Ibuka, M. Waki, N. Yamamoto and N. Fujii, Chem.
Pharm. Bull., 1995, 43, 853–858.
Whereas little PI staining (red fluorescence) was observed in the
control group, B. subtilis incubated with HW-B-13 for 2 h dis-
played a strong red staining by PI, indicating the significant
disruption of bacterial membrane by HW-B-13. The aggregation
of dead bacterial cells after the oligomer treatment is consistent
with literature reports,15,35 an indication of loss of membrane
potential.
Conclusion
In summary, we have shown the design, synthesis and evalua-
tion of a class of cyclic g-AApeptides for antimicrobial devel-
opment. Following a proper tuning of hydrophobicity, several
cyclic g-AApeptides turned out to be superior to previously
reported linear g-AApeptides and peptide drug candidates
under clinic testing, in terms of antimicrobial activity and
selectively. It is conceivable that the antimicrobial activity,
selectivity, and overall drugability of cyclic g-AApeptides can
be further improved in the future through the introduction of
20 S. Thennarasu and R. Nagaraj, Biochem. Biophys. Res. Commun.,
1999, 254, 281–283.
21 S. C. Shankaramma, Z. Athanassiou, O. Zerbe, K. Moehle,
C. Mouton, F. Bernardini, J. W. Vrijbloed, D. Obrecht and
J. A. Robinson, ChemBioChem, 2002, 3, 1126–1133.
2574 | Chem. Sci., 2012, 3, 2570–2575
This journal is ª The Royal Society of Chemistry 2012

View Online
22 S. C. Shankaramma, K. Moehle, S. James, J. W. Vrijbloed,
D. Obrecht and J. A. Robinson, Chem. Commun., 2003, 1842–1843.
23 Y. Q. Tang, J. Yuan, G. Osapay, K. Osapay, D. Tran, C. J. Miller,
A. J. Ouellette and M. E. Selsted, Science, 1999, 286, 498–502.
24 S. Fernandez-Lopez, H. S. Kim, E. C. Choi, M. Delgado,
J. R. Granja, A. Khasanov, K. Kraehenbuehl, G. Long,
D. A. Weinberger, K. M. Wilcoxen and M. R. Ghadiri, Nature,
2001, 412, 452–455.
25 V. Dartois, J. Sanchez-Quesada, E. Cabezas, E. Chi, C. Dubbelde,
C. Dunn, J. Granja, C. Gritzen, D. Weinberger, M. R. Ghadiri and
T. R. Parr, Jr., Antimicrob. Agents Chemother., 2005, 49, 3302–3310.
26 J. T. Fletcher, J. A. Finlay, M. E. Callow, J. A. Callow and
M. R. Ghadiri, Chem.–Eur. J., 2007, 13, 4008–4013.
29 S. Padhee, Y. Hu, Y. Niu, G. Bai, H. Wu, F. Costanza, L. West,
L. Harrington, L. N. Shaw, C. Cao and J. Cai, Chem. Commun.,
2011, 47, 9729–9731.
30 N. P. Chongsiriwatana, T. M. Miller, M. Wetzler, S. Vakulenko,
A. J. Karlsson, S. P. Palecek, S. Mobashery and A. E. Barron,
Antimicrob. Agents Chemother., 2011, 55, 417–420.
31 Y. Ge, D. L. MacDonald, K. J. Holroyd, C. Thornsberry, H. Wexler
and M. Zasloff, Antimicrob. Agents Chemother., 1999, 43, 782–788.
32 R. P. Hicks, J. B. Bhonsle, D. Venugopal, B. W. Koser and
A. J. Magill, J. Med. Chem., 2007, 50, 3026–3036.
33 C. A. Lipinski, F. Lombardo, B. W. Dominy and P. J. Feeney, Adv.
Drug Delivery Rev., 2001, 46, 3–26.
34 T. Matsunaga, M. Okochi and S. Nakasono, Anal. Chem., 1995, 67,
4487–4490.
35 C. Chen, F. Pan, S. Zhang, J. Hu, M. Cao, J. Wang, H. Xu, X. Zhao
and J. R. Lu, Biomacromolecules, 2010, 11, 402–411.
27 J. A. Robinson, Acc. Chem. Res., 2008, 41, 1278–1288.
28 M. L. Huang, S. B. Shin, M. A. Benson, V. J. Torres and
K. Kirshenbaum, ChemMedChem, 2012, 7, 114–122.
This journal is ª The Royal Society of Chemistry 2012
Chem. Sci., 2012, 3, 2570–2575 | 2575