Development of Therapeutic Gramicidin S Analogues Bearing Plastic β,γ-Diamino Acids

Article abstract of DOI:10.1002/cmdc.202000097

Gramicidin S (GS), one of the most widely investigated antimicrobial peptides (AMPs), is known for its robust antimicrobial activity. However, it is restricted to topical application due to undesired hemolytic activity. With the aim of obtaining nontoxic GS analogues, we describe herein a molecular approach in which the native GS β-turn region is replaced by synthetic β,γ-diamino acids (β,γ-DiAAs). Four β,γ-DiAA diastereomers were employed to mimic the β-turn structure to afford GS analogues GS3–6, which exhibit diminished hemolytic activity. A comparative structural study demonstrates that the (βR,γS)-DiAA is the most-stable β-turn mimic. To further improve the therapeutic index (e. g., high antibacterial activity and low hemolytic activity) and to extend the molecular diversity, GS5 and GS6 were used as structural scaffolds to introduce additional hydrophobic or hydrophilic groups. We show that GS6K, GS6F and GS display comparable antibacterial activity, and GS6K and GS6F have significantly decreased toxicity. Moreover, antibacterial mechanism studies suggest that GS6K kills bacteria mainly through the disruption of the membrane.

Full text of DOI:10.1002/cmdc.202000097

Accepted Article
Title: Development of Therapeutic Gramicidin S Analogues Bearing
Plastic β,γ-Diamino Acids
Authors: Yang Wan, Qinkun Guan, Kaisen Chen, Qiang Chen, Jianguo
Hu, Keguang Cheng, Chengfei Hu, Jibao Zhu, Emeric Miclet,
and Valerie Alezra
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemMedChem 10.1002/cmdc.202000097
Link to VoR: https://doi.org/10.1002/cmdc.202000097
01/2020

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
Development of Therapeutic Gramicidin S Analogues Bearing
Plastic β,γ-Diamino Acids
Qinkun Guan,[a]＃ Kaisen Chen,[b]＃ Qiang Chen,[b] Jianguo Hu,[c] Keguang Cheng,[d] Chengfei Hu,[a]
Jibao Zhu,[a] Yi Jin,[a] Emeric Miclet,[e] Valérie Alezra,[f] Yang Wan[a][d][f]
*
[a]
Q. Guan, C. Hu, J. Zhu, Prof. Dr. Y. Jin, Dr. Y. Wan
National Pharmaceutical Engineering Center for Solid Preparation in Chinese Herbal Medicine
Jiangxi University of Traditional Chinese Medicine
1688 Meiling Avenue, WanLi, Nanchang 330004, P. R. China
E-mail: wy15506@gmail.com
[b]
[c]
[d]
[e]
[f]
Prof. Dr. K. Chen, Q. Chen
Department of clinical laboratory
The First Affiliated Hospital of Nanchang University
17 Yongwaizheng Street, Donghu, Nanchang 330006, P. R. China
Dr. J. Hu
College of Pharmacy
Jiangxi University of Traditional Chinese Medicine
1688 Meiling Avenue, WanLi, Nanchang 330004, P. R. China
Prof. Dr. K. Cheng, Dr. Y. Wan
State Key Laboratory for Chemistry and Molecular Engineering of Medicinal Resources
Guangxi Normal University
15 Yuchai Road, Guilin 541004, P. R China
Prof. Dr. E. Miclet
Laboratoire des Biomolécules
Sorbonne Université, PSL University, CNRS
4 Place Jussieu, Paris F-75005, France
Prof. Dr. V. Alezra, Dr. Y. Wan
Laboratoire
- -Saclay
＃ These authors contributed equally to this work.
Supporting information for this article is given via a link at the end of the document.
Abstract: Gramicidin S (GS), one of the most widely investigated
antimicrobial peptides (AMPs), is known for its robust antimicrobial
activity. However, it is restricted to topical applications due to
undesired hemolytic activity. With the aim to obtain non-toxic GS
analogues, we describe herein a molecular approach where the
G β- b β γ-diamino acids
β γ-D AA . β γ-DiAA diastereomers were employed to mimic
β-turn structure to afford GS analogues GS3-6 that exhibit
there is an urgent need for novel anti-infection agents that
possessing alternative mechanism of action to combat
antimicrobial resistance.[4] Delightfully, among the new emerging
categories of potential antibiotics, antimicrobial peptides (AMPs)
are considered to be one of the most promising alternatives to
classic antibiotics. Many AMPs are known to display an
extraordinarily broad range of antimicrobial activities covering
both Gram-positive and Gram-negative bacteria as well as
viruses, fungi and tumors.[5] More importantly, it is common
knowledge that AMPs kill microorganism through the disruption
of lipid bilayers of cell membrane, which makes it very difficult
for microorganisms to develop resistance.
diminished hemolytic activity.
A comparative structural study
m βR γS)-D AA m b β-turn
mimic. To further improve the therapeutic index (e.g. high
antibacterial activity and low hemolytic activity) and to extend the
molecular diversity, GS5 and GS6 were used as structural scaffolds
to introduce additional hydrophobic or hydrophilic groups. We show
that GS6K, GS6F and GS display comparable antibacterial activity
while GS6K and GS6F possess significantly decreased toxicity.
Moreover, antibacterial mechanism studies suggest that GS6K kills
bacteria mainly through the disruption of membrane.
Gramicidin S (GS) is an amphiphilic cyclodecapeptide which
adopts a rigid C2- mm β-hairpin structure with the primary
sequence
cyclo-(Pro-Val-Orn-Leu-DPhe)2.[6–9]
Its
strong
amphiphilicity results from spatially segregated hydrophobic
residue V L β-
strands. GS was isolated from Bacillus brevis in the Soviet
Union in 1942 and was primary applied to treat infected
wound.[10] There is a general agreement that GS kills bacteria by
the disruption of membrane bilayer, though its precise
mechanism of action is still debated. Proposed models include
the formation of discrete pores or the membrane destruction in a
detergent-like manner.[11–16] More recent studies showed that GS
can not only induce the disruption of membrane but can also
delocalize the peripheral membrane proteins.[13,15] Unfortunately,
GS is not selective toward the lipid bilayer which causes severe
toxicity towards human erythrocyte.[17] In order to obtain non-
toxic GS analogues, GS has been the subject of extensive
Introduction
Bacterial resistances are increasing while the development of
novel antibiotics is declining.[1] In Europe, approximately 25 000
people die every year due to antibiotic-resistant infections.[2]
Without urgent measures, we may ente ‘ - b
era, in which none of the known antibiotics will be active. In
addition to a tight control of the use of the existing antibiotics,[3]
1
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
− A .[18,19] It has become
clear that the biological activity of GS is determined by a panel
of physicochemical parameters: cationicity, hydrophobicity,
amphiphilicity and hydrophobic aromaticity.[18] In the past
decades, many strategies, including ring size variation,[20–22] β-
stabilize well-defined conformations.[40,41] f β γ-DiAAs
m b z β-turn region of GS and in
the same time to maintain its global amphiphilic character. The
f β γ-DiAA is
a source of
hydrosolubility or further functionalization, facilitating the
stepwise structural optimization. In the last work, we have
synthesized a pair of GS analogues GS1 and GS2 (Figure 1)
w β- b w β γ-DiAAs.[23] Both
turn modification,[23–30] β-strand modification[31,32] and light-
[33–35]
b “ m ”
have been recognized as
different routes to conceive new potent antimicrobial molecules
endowed with low toxicity against human cells. However, to date, analogues displayed totally diminished hemolytic activity while
there is no clinically applicable antibiotic related to GS and thus
a constant investment in this field is required.
maintained bactericidal activity against a panel of pathogenic
strains. Nevertheless, the conformational features of those
w tails and their antibacterial
activities were weak compared to GS. In this study, we
synthesized four additional analogues (GS3-6, Figure 1 w β-
m b b β γ-DiAA that differing in
stereochemistry. After conformational and biological evaluation,
GS5 and GS6 were selected as molecular scaffolds for further
modification to obtain analogues conferred with high therapeutic
index. The biological studies demonstrate that GS6K and GS6F
are highly toxic to bacterial strains, but are much less toxic to
human blood cells.
Among the molecular approaches, we are particularly
interested in the modification of DPhe- m f β-turn
region. Several studies have shown that even subtle changes in
β-turn region could have a dramatic impact on the biological
profile,[26,27,32,36] suggesting this region is a promising target for
m G . α-amino acids and
unnatural dipeptide surrogates have been employed to
substitute the native DPhe-Pro motif.[32,37,38] Interestingly,
m m k β-turn by unnatural dipeptide surrogates often
f β-hairpin structure and
subsequently leads to an optimal amphiphilicity.[26,39]
In our group, we have also contributed in the development of
GS-based antibiotics.[18,23] Our strategy involves the replacement
f G β-turn DPhe- b β γ-diamino acids
β γ-DiAAs) that can possess combination of stereochemistry
and hydrophilicity/hydrophobicity. We have previously shown
that when introduced into short p β γ-DiAAs are capable
to induce stable C9 or C12 intra-molecular H-bonds which finally
Results and Discussion
Scope of Lead Peptide
Molecular Design and Synthesis. Initially, four diastereomeric
GS analogues (GS3-6, Figure 1) were synthesized to evaluate
b b f m . β γ-DiAAs
Figure 1. GS and GS analogue 1-6 with β-turn region mono- or disubstituted with β γ-DiAAs
Scheme1. Synthesis of Fmoc-protected β γ-diAAs. Synthetic conditions: (a) LiOH, H2O/THF; (b) Pd/C, H2, MeOH; (c) FmocOSu, K2CO3, 1,4-
Dioxane/water.
2
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
Scheme 2. Synthesis of GS3-6. X β γ-DiAA, HBTU, HOBt, DiPEA, DMF) and deprotection (20% piperidine/DMF, v/v)
steps; (b) 1% TFA/DCM, v/v, 5x 10 min; (c) PyBOP, HOBt, DiPEA, DMF; (d) 90% TFA/water, v/v.
were prepared using the synthetic strategy previously described
b f m α-amino acids.[42] With respect to the
m β-turn region,[36,39] D- and L-
phenylalanine were chosen as starting materials (Scheme 1).
Cbz-protected intermediates 7, 8, 9 and 10 were prepared as
previously described.[23,43] Since the configuration of an
analogue of compound 9 has been previously determined by
crystallographic structure,[43] the other compounds could thereby
be differentiated by the comparison of their NMR spectrum with
their enantiomers. Noteworthy, the separation of the
diastereomers is rather difficult when directly applying the silica
gel chromatography. Alternatively, the minor diastereomers (8
and 10) can be easily precipitated in the presence of cold diethyl
ether, and it consequently allows the major diastereomers (7
and 9) being purified by chromatography. To allow the solid
phase synthesis, Cbz protecting group was further transferred to
Fmoc protecting group (Scheme 1).
various degrees. GS5 and GS6 preserve interesting activity
against several strains such as S. bovis and E. faecium while
GS3 is absolutely inactive against all strains.
Table 1. Antibacterial activity of GS3-6 and GS[a][b]
Peptide
Gram-negative
bacteria
Gram-positive bacteria
E.
P.
S.
B.
S.
E.
E.
coli
aeruginosa
aureus
subtilis
bovis
faecalis
faecium
GS
50
50
3.13
100
50
3.13
100
50
6.25
100
50
6.25
100
25
3.13
100
25
GS3
GS4
GS5
GS6
100
50
100
50
50
100
50
50
100
50
25
25
12.5
25
25
25
12.5
25
[a] measured as MIC μ /mL . [b] μ /mL w m x m
concentration.
Previous studies have demonstrated that GS and its
analogues can be readily synthesized by solution-phase
cyclization of the corresponding linear precursor peptides with
protected side chains.[44] To this end, the linear precursors were
assembled by following an either conventional or microwave-
assistant solid phase Fmoc/tBu strategy. All Fmoc- α-
amino acids are commercially available. Each projected peptide
was assembled stepwise on the solid support starting from
Fmoc-Pro-2-Chlorotrityl Chloride (CTC) Resin. During the
sequential coupling, HBTU/HOBt (conventional synthesis) or
HBTU (microwave-assistant synthesis) was used for activation,
DiPEA as base, DMF as solvent and 20% piperidine for Fmoc
deprotection. Resin cleavage was achieved in the condition of 1%
TFA/DCM (v/v) to keep the side-chain protecting groups intact.
The subsequent macrocyclization was achieved in diluted
solution (2 mg/mL) to lower the hetero-condensation. After
global deprotection in the presence of TFA/H2O (9:1, v/v), crude
peptides were pre-purified by the precipitation in the presence of
cold diethyl ether and then further purified by reverse-phase high
performance liquid chromatography (RP-HPLC, Scheme 2).
Antibacterial Activity. With cyclic peptides in hand, we
evaluated the antibacterial activity of GS3-6 and GS by
determination of the minimal inhibitory concentration (MIC) in a
standard screening assay in broth (Table 1). A representative
set of both Gram-positive and Gram-negative bacterial strains
were selected for antibacterial testing. In line with literature,[17]
GS exhibits potent activity against Gram-positive strains while
less efficacy against Gram-negative ones. In comparison to
parent molecule, GS3-6 all show in general less activity in
Figure 2. Hemolytic activity of GS3-6 and GS. Percentage hemolysis, mean ±
value, and n = 3.
Hemolytic Activity. The ability of GS3-6 to lyze human red
blood cell was then evaluated in comparison to GS (Figure 2).
Hemolysis was determined by the exposure of fresh prepared
human blood to a range of two-folded peptide concentrations
[17]
(15.6- μ /mL . A k w
GS induces severe
hemolysis in a very low concentration. However, GS3-6 all exert
negligible hemolysis in low concentration and the intensity of
hemolysis grows slowly as the peptide concentration increases.
NMR Spectroscopy. To learn how the structural variations
influence the biological profile, we made
a comparative
structural study between GS3-6 and GS by NMR spectroscopy
in aqueous solution. The complete 1H NMR resonance
assignment of each peptide was performed using COSY,
3
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
TOCSY and NOESY experiments. Large spectral dispersion is
observed in HN regions for all peptides except for GS3 (See
Supporting Information), suggesting stable conformations form
in GS4-6 and GS. As for GS3, the observation of overlapped NH
signals and the absence of long-range NOE are attributable to
the fact that no stable secondary structure is formed. These
findings are reminiscent of our previous results that instead of
intra-molecular interaction, inter-molecular H-bonds are formed
w βR γR)-DiAA 3 is introduced into short peptides.[45] As can
3
be found in Table 2, the J αH-NH coupling constants of the
Leu, Orn and Val (ranging between 8.6 and 9.1 Hz) of GS
β-sheet structure.[46] A set of comparable
values were calculated in GS6 while not in other analogues,
signifying an analogous secondary structure between GS6 and
GS. NOESY spectroscopy was also recorded as it is an
excellent option for deriving the conformational preferences of
cyclic peptides. Figure 3 shows that the observed long-range
NOEs (e.g. NH-Leu4 and CβH-Val7 β-turn region of
GS4-6 are comparable to that of native GS. Nevertheless, the
conformational preferences of the modified region of GS4-6 are
significantly distinct. The detected long-range NOEs between
NH-Orn3 and CβH or CγH f βS γS)-DiAA in GS5 indicate a
highly bended skeleton conformation, which is consistent with
our previous experimental data that multiple H-bonds are formed
βS γS)-DiAA when it is incorporated into short
peptides.[41] On the other hand, the long-range NOEs around
β γ-DiAA residue in GS4 and GS6 highlight a C9 H-bond
formation between NH-Val2 and CO-Leu9.[40] Overall, GS6 has a
m β-sheet character which is most likely similar to that of
GS.
Figure 3. Observed long-range NOEs (mixing time 500 ms) for GS4-6 and GS.
Solid double arrows indicate strong NOEs, dotted double arrows indicate weak
NOEs. NOESY spectra were recorded in H2O/D2O (9:1, v/v, containing 20 mM
PBS buffer) at pH 7.0 and 4 oC.
Table 2. 3J αH-NH coupling constants (Hz) measured for GS3-6 and GS in
1D 1H NMR spectra recorded in H2O/D2O (9:1, v/v, containing 20 mM PBS
buffer) at pH 7.0 and 4 oC.
Peptide
GS
GS3
GS4
GS5
GS6
Val2
9.1[a]
6.9
5.6
6.5
Orn3
9.0
6.8
7.7
6.5
7.2
Leu4
8.6
8.2
8.5
4.9
8.2
DPhe5
broad
8.4
Val7
9.1[a]
5.2
5.6
6.5
Orn8
9.0
Leu9
8.6
nd[b]
7.8
nd[b]
nd[b]
4.8
broad
nd[b]
4.5
8.3
8.0
broad
7.2
broad
[a] Measured in 25 oC, data not shown. [b] Not determined because of signal
overlapping.
Circular Dichroism (CD) Spectroscopy. To evaluate the
content of the secondary structure, CD spectra of GS3-6 and GS
were measured in methanol and water (Figure 4). In both
solvents, the spectral characteristics of GS3-6 are less
detectable in comparison to GS. However, GS6 displays a
similar feature to that of GS, albeit the excitation is weaker.
Similar to NMR spectroscopy analysis, these experiment data
also serves to demonstrate an analogous conformation between
GS6 G . f β- β-turn in the
secondary structure are confirmed by the two negative
shoulders around 207 and 225 nm.[47] Compared to spectra in
methanol, similar spectra characteristics were documented in
aqueous solution except an overall attenuation in excitation,
highlighting a more induced and stable secondary structure in
organic solvent.
Figure 4. CD spectra of GS3-6 and GS at 0.1 mM in (A) methanol and (B)
water.
RP-HPLC Analysis. The estimation of amphiphilic and
hydrophobic characteristics of AMPs is interesting as they are
highly associated with the biological activities. Since GS3-6
share the same hydrophilic and hydrophobic residues, the
difference of biological activities can be partially attributed to the
variable amphiphilic characteristics.[48] As an empirical method,
the amphiphilicity of GS3-6 were estimated using the RP-HPLC
retention times. As shown in Figure 5, GS4-6 display ignorable
4
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
difference with regard to retention time whereas GS3 has an
apparent shorter value. This result signifies that GS3 possesses
(e.g. β γ-DiAA motif may lead to potent
GS-related antibiotics endowed with low side effects. Given
GS6 m β-sheet structure and GS5
stands for a different type of structural conformation, they were
both employed as structural scaffolds for further optimizations
(Figure 6).
a
disordered amphiphilic structure, which is line with
aforementioned structural studies. When compared to GS, GS3-
6 all show much less hydrophobic/amphiphilic content as GS
has a greatly extended retention time. It is common knowledge
that hydrophobicity of AMPs is closely correlating with hemolytic
activity.[47,49] Therefore, the distinction of retention time serves to
explain why GS3-6 exhibit significantly reduced hemolytic
activity.
k β γ-DiAA, it
allows the introduction of additional group with relative ease.
Installation of additional residue in GS5 or GS6 can be normally
achi w : f ‘ m - m b
f z β γ-DiAA during Fmoc solid phase peptide
synthesis or on-resin introduction of functional groups. In our
f β- f z β γ-DiAA
requires a number of protection and deprotection steps which
may heavily hamper the efficiency, particularly considering the
poor solubility of minor diastereomer (compound 6) in most
organic solvents (except for DMF). For these reasons, the on-
resin introduction of functional groups is a more practical
strategy as it simplifies the synthetic procedure and shortens the
purification steps. To allow this, Boc protecting group in amino
acid 5 or 6 was switched to Alloc protecting group, delivering
β γ-DiAA 5a and 6a that are readily for Fmoc solid phase
peptide synthesis (Scheme 3). The subsequent peptide
synthesis followed f m . Af β γ-DiAA
coupling, Alloc group was removed in the presence of Pd(Ph3)4,
followed by the introduction of additional Boc-protected amino
acid. Total four analogues GS5K, GS5F, GS6K and GS6F were
synthesized within reasonable yields (Scheme 4).
Figure 5. Schematic representation of the retention times observed for GS
and GS3–6 on analytical RP-HPLC. Experiments were performed on a Welch
XB- m .6 x mm μm H3CN (30-
70%) into H2O (0.1% TFA, v/v) over 60 min at 1 mL/min flow rate, with UV
detection at 221 nm.
Development of Further Optimized GS Analogues
Molecular Optimization and Synthesis. Notwithstanding GS3-
6 show much decreased hemolytic effect, their bactericidal
m m f tion. To
improve the therapeutic index and to extend the molecular
diversity, further structural optimizations of GS analogues were
conducted. Previous studies have demonstrated that the
f b m β-turn region is
important to preserve the antibacterial activity[36,39] whereas
some others argued that the enhancement of cationicity in the
same region may lower the hemolytic activity without hampering
the antibacterial activity.[50] We thereby envisioned that tethering
additional hydrophobic/aromatic (e.g. phenylalanine) or cationic
Scheme 3. Synthesis of Alloc- β γ-diamino acids. (a) 50% TFA/DCM,
v/v; (b) AllocOsu, Na2CO3, 1,4-Dioxane/water.
Figure 6. Structure of the further optimized GS analogues.
5
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
Scheme 4. Synthesis of the further optimized G . X β γ-DiAA, HBTU, HOBt, DiPEA, DMF) and deprotection (20%
piperidine/DMF, v/v) steps; (b) Pd(Ph3)4, NMM, AcOH, CHCl3; (d) 1% TFA/DCM, v/v, 5x 10 min; (e) PyBOP, HOBt, DiPEA, DMF; (f) 90% TFA/water, v/v.
Antibacterial Activity. We first tested the antibacterial ability of
the newly synthesized GS analogues following the
aforementioned protocol. In general, the second batch of GS
analogues exhibit superior activity than GS4-6. Significantly,
although differing in molecular approach, both GS6K and GS6F
exerted comparable antibacterial activities against several
pathogenic strains (e.g. S. aureus and S. bovis) to that of native
GS (Table 3). Presumably, the improved antibacterial activity of
GS6F b b m β-turn
w f π-cation interaction during peptide
penetration.[36] On the other hand, the enhancement of
cationicity may serve to strengthen the electrostatic interaction
between peptide and cell membrane and thus increases the
antibacterial ability.[51]
Figure 7. Hemolytic activity the second batch of GS analogues. Percentage
hemolysis, mean ± value, and n = 3.
Table 3. Antibacterial activity of the second batch of GS analogues[a][b]
Cytotoxicity. Non-toxicity to mammalian cells is also important
for the clinical application of AMPs. In this regard, we further
evaluated the cytotoxicity of GS6K by the determination of cell
viability. LO2 cell was used to mimic human liver cell. Cell
Counting Kit-8 (CCK-8) assays were carried out in cell medium
supplemented with 1% fetal bovine serum. Figure 8 shows that
GS6K exerts negligible toxicity towards LO2 cells except for a
slightly decrease of cell viability at concentration higher than 10
μ . G severe cytotoxicity at low
< μ .
Peptide
Gram-negative
bacteria
Gram-positive bacteria
E.
P.
aeruginosa
50
S.
aureus
3.13
25
25
12.5
25
B.
subtilis
3.13
6.25
6.25
6.25
3.13
S.
E.
faecalis
6.25
50
25
12.5
12.5
E.
faecium
3.13
12.5
12.5
coli
50
50
50
50
50
bovis
6.25
12.5
12.5
3.13
6.25
GS
GS5K
GS5F
GS6K
GS6F
100
100
50
50
12.5
12.5
[a] m μ /mL . [b] μ /mL w m x m
concentration.
Hemolytic Activity. Next, the hemolytic activities of the second
batch of GS analogues were measured. Intriguingly, in spite of
the substantially enhanced antibacterial ability occurs, no
evident increase of hemolytic activity is gained for all analogues.
This is probably due to the maintained or even increased
number of cationicity in GS5F/GS6F and GS5K/GS6K,
respectively. In general, all GS analogues display hemolytic
activity to a point that is much lower than GS. Having realized
both antibacterial and hemolytic profile, GS6K was selected as
the peptide candidate for further biological studies.
Figure 8. Cytotoxicity of GS6K and GS. Percentage cell viability, mean ±
value, and n = 3
6
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
Antimicrobial Mechanism
Cytoplasmic Membrane Depolarization. To understand
whether the antibacterial mechanism of the optimum peptide
GS6K resembles to that of GS, DiSC3-5, a membrane potential
sensitive dye, was used to assess bacterial membrane
depolarization on S. aureus and B. subtilis. The distribution of
DiSC3-5 between medium and cell membrane is highly
depending on the membrane potential. The fluorescence of
cationic dye DiSC3-5 is self-quenched when accumulating to
cytoplasmic membrane. However, upon membrane disruption,
DiSC3-5 is released and resuspended in buffer, leading to
fluorescence enhancement. Figure 9 shows the time course of
fluorescence over 30 min. As can be found in Figure 9, the
addition of GS6K (2x MIC) to the medium resulted an
immediately increase of fluorescence on both bacterial strains,
though the change on B. subtilis is less significant. Maximum
fluorescence was achieved after addition of GS (4x MIC). These
observations suggest that GS6K is highly membrane active and
behaves similar to that of GS.
Figure 10. SEM photographs of treated (left panel) and untreated (right panel)
bacteria with GS6K (2x MIC, 1 h).
primary synthesized to obtain non-toxic GS-based antibiotics.
W w βR γS)-D AA m b β-
turn mimic among four diastereomers by
a comparative
structural study. In spite of possessing relatively weak
bactericidal activity, GS3-6 all exhibit ignorable hemolytic
activities. After introduction of extra hydrophobic or hydrophilic
β γ-DiAA motif through on-resin synthesis, we were
able to obtain non-toxic peptide GS6K and GS6F conferred with
potent antibacterial activity, highlighting the structural plasticity
of our building blocks. DiSC3-5 fluorescence and SEM
experiments suggest that the most potent analogue GS6K kills
bacteria mainly, if not only, through the disruption of cell
membrane. This observation is significant as antibiotics
possessing such model of action are unlikely to provoke
bacterial resistance. It is interesting to find that both GS6K and
GS6F exhibit improved biological profile in comparison to lead
peptide GS6 given the additionally tethered groups
encompassing converse physiochemical characters, which
highlights the molecular potential of peptide GS6. In general, we
have shown that it is possible to improve the therapeutic index of
GS-based antibiotics by the replacement of DPhe-Pro motif by a
β γ-DiAA. Considering there is virtually no
resistance observed for cyclopeptide antibiotics such as
gramicidin S and tyrocidines that have been topically applied for
over seventy years, it is of great interest to reinvestigate these
old but still potent AMPs to combat globally rising bacterial
Figure 9. Cytoplasmic membrane depolarization of (A) S. aureus and (B)
B.subtilis by the optimum peptide GS6K.
Scanning Electron Microscopy (SEM). To further elucidate the
mechanism of action, GS6K was used to perform SEM covering
both Gram-positive and Gram-negative pathogenic strains. SEM
of pathogen cell morphologies is shown in Figure 10. Compared
to the intact controls (right panel), microbial cells treated with
GS6K exhibit distinct changes in their morphology (left panel).
Unlike the smooth surfaces of the control cells, wrinkled cell
walls are observed on E. faecium and S.aureus cells. While for
the rest of bacterial strains, cell membrane was impaired to the
point that cell contents leaked out (arrows in Figure 10). These
results are likely attributable to the fact that GS6K kills bacteria
through the disruption of cell membrane.
Conclusion
In this work, four diastereomeric GS analogues GS3-6 w β-
m b b β γ-DiAAs were
7
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
water/dioxane (50 mg/mL, 1/1, v/v) followed by addition of AllocOSu (1.3
equiv) and Na2CO3 (1.0 equiv). The resulting mixture was allowed to stir
for 15 h at rt. After extraction with diethyl ether to remove the excess of
AllocOSu, it was acidified with 2N aqueous hydrochloric acid to pH 1 and
then extracted with ethyl acetate. The combined organic layers were
washed with water, dried with anhydrous MgSO4, filtered and
concentrated under reduced pressure to afford a crude product. The
crude product was purified by chromatography (SiO2, DCM/MeOH, 97/3,
v/v) to deliver a white solid. The yields of compounds 5a and 6a are 54%
and 69%, respectively.
resistance. We hope that our work will encourage and benefit
continuous investment in the field of AMP-based drug
development.
Experimental Section
General Remarks. Fmoc-Pro-(2-CTC)-resin (1% cross-linked, 100-200
mesh, 0.476 mmol/g) was purchased from GL Biochem (Shanghai) Ltd.
DMF was treated with CaH for overnight then distilled under reduce
pressure. THF was distilled over sodium/benzophenone under argon. All
other reagents and solvents were used without any further purification.
Flash chromatography was performed on silica gel (100-200 mesh).
Analytical thin-layer chromatography (TLC) was performed using
aluminum-backed silica gel plates coated with a 0.2 mm thickness of
GF254 silica gel, neutral. 1H, 13C, TOCSY, COSY, HSQC and NOESY
spectra were recorded on Bruker AV II 500 or Bruker AV II 600. Chemical
shifts are reported in δ units to 0.01 ppm precision with coupling
constants reported to 0.1 Hz precision using residual solvent (DMSO-d6:
1H: 2.50 ppm, 13C: 39.52 ppm; CDCl3: 1H: 7.26 ppm, 13C: 77.16) as an
internal reference. Multiplicities are reported as follows: s = singlet, d =
doublet, t = triplet, q = quadruplet, m = multiplet, bs = broad singlet. Mass
spectra were measured on a Microtof-Q Bruker Daltonics spectrometer
coupled with a LC equipment (U3000 Thermofisher). Analytical RP-HPLC
was performed on a Welch XB-C18 column (4.6 x mm μm
running linear gradients CH3CN (30-70%) into H2O (0.1% TFA, v/v) over
60 min at 1 mL/min flow rate, with UV detection at 221 nm. Semi-
preparative RP-HPLC purification was done on an YMC-Pack ODS-A
x mm μm s of CH3CN into H2O (0.1%
TFA, v/v) as indicated for each peptide, at a flow rate of 2 mL/min.
Compound 3 or 5. 1H NMR (600 MHz, DMSO): δ 11.74 (bs, 1H), 7.89 (d,
J = 7.5 Hz, 2H), 7.64 (d, J = 6.9 Hz, 2H), 7.43-7.40 (m, 2H), 7.35-7.31 (m,
3H), 7.24-7.19 (m, 4H), 7.16-7.14 (m, 1H), 6.78 (d, J = 5.8 Hz, 1H), 4.22-
4.11 (m, 3H), 4.04-3.98 (m, 1H), 3.94-3.87(m, 1H), 2.79-2.76 (m, 1H),
2.62-2.58 (m, 1H), 2.45-2.28 (m, 2H),1.42 (s, 9H). 13C NMR (150 MHz,
DMSO): δ 171.3, 156.8, 156.1, 144.9, 144.8, 141.7, 140.0, 130.0, 129.0,
128.6, 128.0, 126.9, 126.2, 121.1, 121.0, 78.8, 66.2, 56.6, 52.6, 47.7,
37.7, 37.3, 29.2. MS: (ESI+): m/z: calcd for C31H34N2NaO6+: 553.23
[M+Na]+; found: 553.25.
Compound 4 or 6. 1H NMR (600 MHz, DMSO): δ 12.03 (bs, 1H),7.87 (dd,
J1 = 3.6 Hz, J2= 7.8 Hz, 2H), 7.61 (d, J = 7.6 Hz, 2H), 7.42-7.38 (m, 2H),
7.35-7.28 (m, 3H), 7.25-7.18 (m, 4H), 7.13-7.10 (m, 1H), 6.94 (d, J = 9.2
Hz, 1H), 4.17-4.07 (m, 3H), 3.97-3.89 (m, 1H), 3.78-3.73 (m, 1H), 2.87
(dd, J1 = 2.8 Hz, J2= 14.3 Hz, 1H), 2.60 (dd, J1 = 11.4 Hz, J 2 = 14.3 Hz,
1H), 2.47 (dd, J1 = 3.8 Hz, J2 = 15.6 Hz, 1H), 2.35 (dd, J1 = 9.7 Hz, J2 =
15.6 Hz, 1H),1.40 (s, 9H). 13C NMR (150 MHz, DMSO): δ 173.8, 156.8,
156.2, 144.9, 144.7, 141.6, 140.3, 130.1, 128.9, 128.6, 128.0, 126.8,
126.2, 121.1, 121.0, 78.6, 66.2, 56.8, 52.6, 47.7, 37.7, 37.3, 29.2. MS:
(ESI+): m/z: calcd for C31H34N2NaO6+: 553.23 [M+Na]+; found: 553.08.
Compound 5a. 1H NMR (600 MHz, DMSO): δ 12.15 (bs, 1H), 7.89 (d, J =
7.4 Hz, 2H), 7.64-7.61 (m, 2H), 7.43-7.39 (m, 2H), 7.36-7.30 (m, 2H),
7.27-7.20 (m, 5H), 7.18-7.12 (m, 2H), 5.98-5.91 (m, 1H), 5.33 (d, J = 17.4
Hz, 1H), 5.20 (d, J = 10.5 Hz, 1H), 4.55-4.49 (m, 2H), 4.22-4.16 (m, 2H),
4.14-4.07 (m, 2H), 3.94-3.85 (m, 1H), 2.87-2.75 (m, 1H), 2.63-2.59 (m,
1H), 2.51-2.47 (m, 1H), 2.41-2.33 (m, 1H). 13C NMR (150 MHz, DMSO): δ
172.8, 156.3, 156.0, 144.3, 141.1, 139.3, 134.2, 129.5, 128.5, 128.1,
127.5, 126.5, 125.7, 120.6, 117.3, 65.9, 64.8, 55.7, 51.4, 47.1, 37.5, 36.9.
MS: (ESI+): m/z: calcd for C30H31N2O6+: 515.22 [M+H]+; found: 514.92.
β,γ-Diamino
Acids
Synthesis.
Cbz-(β,γ)-DiAA(β-Boc)-OCH3
(Compound 7, 8, 9 and 10). Compound 7, 8, 9 and 10 were synthesized
from Cbz-D-Phenylalanine or L-Phenylalanine, according the previously
described procedure.[23,43]
Fmoc-(β,γ)-DiAA(β-Boc)-OH (Compound 3, 4,[43] 5, and 6). A solution of
Cbz- β γ -D AA β-Boc)-OCH3 (1.0 equiv.) and LiOH.H2O (10 equiv., 0.5
N final conc.) in a mixture of THF/water (2/1, v/v) was stirred for 2 h at rt.
After volatile removal, the aqueous layer was acidified with 2N aqueous
hydrochloric acid to pH 1 and then extracted with ethyl acetate. The
combined organic layers were washed with water, dried with anhydrous
MgSO4, filtered and concentrated under reduced pressure to afford a
white solid that was used without any further purification.
Compound 6a. 1H NMR (600 MHz, DMSO): δ 12.15 (bs, 1H), 7.88-7.87
(m, 2H), 7.62 (d, J = 7.2 Hz, 2H), 7.43-7.38 (m, 4H), 7.34-7.29 (m, 2H),
7.27-7.20 (m, 4H), 7.14-7.12 (m, 1H), 5.96-5.90 (m, 1H), 5.32 (d, J = 17.2
Hz, 1H), 5.18 (d, J = 10.3 Hz, 1H), 4.51 (d, J = 4.68 Hz, 2H), 4.19-4.08
(m, 3H), 4.04-3.98 (m, 1H), 3.80-3.75 (m, 1H), 2.88-2.85 (m, 1H), 2.64-
2.60 (m, 1H), 2.53-2.49 (m, 1H), 2.40-2.36 (m, 1H). 13C NMR (150 MHz,
DMSO): δ 173.2, 156.3, 156.1, 144.3, 144.2, 141.1, 139.6, 134.3, 129.6,
128.4, 128.1, 127.5, 126.3, 125.8, 120.5, 117.2, 65.7, 64.7, 56.2, 52.5,
47.2, 37.1, 36.7. MS: (ESI+): m/z: calcd for C30H31N2O6+: 515.22 [M+H]+;
found: 514.92.
A solution of previously saponified compound (1.0 equiv.) in dry MeOH
(100 mg/mL) was stirred for 30 min in the presence of 10 wt. % Pd/C (10
mol %) at rt under H2 (1 bar). Filtration of the catalyst through celite pad
and concentration under reduced pressure gave
quantitative yield.
a white solid in
To a stirred solution of the hydrogenated compound (1.0 equiv.) in a
mixture of water/dioxane (50 mg/mL, 1/1, v/v) was added FmocOSu (1.3
equiv.) and K2CO3 (5 equiv.). The resulting mixture was allowed to stir for
15 h at rt. After extraction with diethyl ether to remove the excess of
FmocOSu, it was acidified with 2N aqueous hydrochloric acid to pH 1
and extracted with ethyl acetate. The combined organic layers were
washed with water, dried with anhydrous MgSO4, filtered and
concentrated under reduced pressure to afford a white solid. Crude
compound 3 and 5 were purified by chromatography (SiO2, DCM/MeOH,
97/3, v/v). Compound 4 and 6 was washed with diethyl ether to remove
the residue of FmocOsu. Compound 3, 4, 5 and 6 were all obtained in
reasonable yields: 39%, 84%, 35% and 85%, respectively.
Peptide Synthesis. Peptide length elongation was started from Fmoc-
Pro-(2-CTC)–Resin, either by Manual Peptide Synthesis or Automated
Peptide Synthesis.
Manual Peptide Synthesis. The manual peptide synthesis was performed
using the following protocol:
(1) Swelling in DMF for 20 min
(2) 2× cleavage of the Fmoc protection group with 20% piperidine in DMF
for 15 min
Fmoc-(β,γ)-DiAA(β-Alloc)-OH (Compound 5a and 6a). A solution of 5 or 6
in a mixture of TFA/DCM (100 mg/mL, 1/1, v/v) was stirred for 1 h at rt.
(3) washing sequentially with DMFx 2, MeOHx 2, DCMx 2, DMF
After volatile removal, the residue was dissolved in
a mixture of
8
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
(4) Coupling with Fmoc-AA-OH (4.0 equiv), HOBt (4.0 equiv), HBTU (4.0
equiv), and DIPEA (8.0 equiv) in DMF for 60 min
1109.42 [M+Na]+; for NMR data see Tables S1 in the Supporting
Information.
(5) washing sequentially with DMFx 2, MeOHx 2, DCMx 2, DMF
cyclo(Pro-DPhe-Leu-Orn-Val-(βS,γS)-DiAA-Leu-Orn-Val-) (GS5). GS5
was synthesized using Manual Peptide Synthesis on a 0.25 mM scale
(43 mg, 0.040 mmol, 16% overall yield, 95% purity). Analytic RP-HPLC,
linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA) for 60
min. Detections were done at 221 nm using a photodiode array detector.
tr = 20.82 min; MS: (ESI+): m/z: 544.25 [M+2H]2+, 1087.25 [M+H]+
1109.33 [M+Na]+; for NMR data see Tables S1 in the Supporting
Information.
− w f m w m
assembled.
Automated Peptide Synthesis. Peptides were synthesized by
microwave-assisted peptide synthesizer (Discovery Bio, CEM) using the
following cycles of deprotection, coupling and washing.
a
cyclo(Pro-DPhe-Leu-Orn-Val-(βR,γS)-DiAA-Leu-Orn-Val-) (GS6). GS6
was synthesized using Manual Peptide Synthesis on a 0.25 mM scale
(27 mg, 0.025 mmol, 10% overall yield, 99% purity). Analytic RP-HPLC,
linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA) for 60
min. Detections were done at 221 nm using a photodiode array detector.
tr = 20.73 min; MS: (ESI+): m/z: 544.33 [M+2H]2+, 1087.42 [M+H]+
1109.50 [M+Na]+; for NMR data see Tables S1 in the Supporting
Information.
m : = ° ΔT = 1 °C, Pmicrowave = 30 W, t = 210 s
with 20% piperidine in DMF, 4 mL/deprotection.
: = ° ΔT = 1 °C, Pmicrowave = 30 W, t = 600 s
with Fmoc-AA-OH (5 equiv.), HBTU (5 equiv.) and DiPEA (0.3 M in DMF,
12 equiv., 4 mL/coupling)
(3) Washing: after each deprotection or coupling, resin was washed with
DMFx 3 (3 mL/time).
cyclo(Pro-DPhe-Leu-Orn-Val-(βS,γS)-DiAA(β-Lys)-Leu-Orn-Val-) (GS5K).
GS5K was synthesized using Automated Peptide Synthesis on a 0.1 mM
scale (21 mg, 0.017 mmol, 17% overall yield, 93% purity). Analytic RP-
HPLC, linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA)
for 60 min. Detections were done at 221 nm using a photodiode array
detector. tr = 16.43 min; MS: (ESI+): m/z: 608.42 [M+2H]2+, 1215.42
[M+H]+; for NMR data see Tables S1 in the Supporting Information.
Resin Cleavage. Resin cleavage was performed with 1% TFA in DCM
(5x 10 min) to keep the side-chain protecting groups untouched. After
cleavage, the product solution was coevaporated with toluene (1/1, v/v)
to give the side-chain-protected linear peptides as a solid which was
used without any further purification.
cyclo(Pro-DPhe-Leu-Orn-Val-(βS,γS)-DiAA(β-Phe)-Leu-Orn-Val-) (GS5F).
GS5F was synthesized using Automated Peptide Synthesis on a 0.1 mM
scale (15 mg, 0.012 mmol, 12% overall yield, 91% purity). Analytic RP-
HPLC, linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA)
for 60 min. Detections were done at 221 nm using a photodiode array
detector. tr = 22.24 min; MS: (ESI+): m/z: 618.08 [M+2H]2+, 1234.58
[M+H]+; for NMR data see Tables S1 in the Supporting Information.
Peptide Cyclization. To a solution of linear peptide in DMF (2 mL/mg)
was added PyBOP (5.0 equiv.), HOBt (5.0 equiv.) and DiPEA (15.0
equiv.). After stirred for 15 h at rt, the solvent was removed under
reduced pressure.
Global Deprotection and Precipitation. The crude cyclized peptide was
dissolved in TFA/water (9/1, v/v, 10 mg/mL) and allowed to stir for 1 h at
rt. After evaporation of the TFA/water from the global deprotection, equal
volume (0.5 mL) aliquots of TFA solution were poured into two 15 mL
conical tubes. Cold diethyl ether (14 mL) was then dumped into each
conical tube. Precipitation was observed immediately and let for
incubation for another 15 min in ice bath. Precipitate was pelleted by
centrifugation (3000 rpm, 15 min) and transferred to round flask after
removal of the ether supernatant. Purification by semi-preparative RP-
HPLC and lyophilization gave the GS and analogues as a white powder.
cyclo(Pro-DPhe-Leu-Orn-Val-(βR,γS)-DiAA(β-Lys)-Leu-Orn-Val-) (GS6K).
GS6K was synthesized using Automated Peptide Synthesis on a 0.1 mM
scale (38 mg, 0.031 mmol, 31% overall yield, 97% purity). Analytic RP-
HPLC, linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA)
for 60 min. Detections were done at 221 nm using a photodiode array
detector. tr = 17.40 min; MS: (ESI+): m/z: 608.42 [M+2H]2+, 1215.67
[M+H]+; for NMR data see Tables S1 in the Supporting Information.
cyclo(Pro-DPhe-Leu-Orn-Val-)2 (GS). GS was synthesized using Manual
Peptide Synthesis on a 0.25 mM scale (57 mg, 0.050 mmol, 20% overall
yield, 98% purity). Analytic RP-HPLC, linear 30-70% CH3CN (0.1% TFA)
gradient into H2O (0.1% TFA) for 60 min. Detections were done at 221
nm using a photodiode array detector. tr = 47.88 min; MS: (ESI+): m/z:
571.33 [M+2H]2+, 1141.42 [M+H]+; for NMR data see Tables S1 in the
Supporting Information.
cyclo(Pro-DPhe-Leu-Orn-Val-(βR,γS)-DiAA(β-Phe)-Leu-Orn-Val-) (GS6F).
GS6F was synthesized using Automated Peptide Synthesis on a 0.1 mM
scale (15 mg, 0.012 mmol, 12% overall yield, 92% purity). Analytic RP-
HPLC, linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA)
for 60 min. Detections were done at 221 nm using a photodiode array
detector. tr = 27.01 min; MS: (ESI+): m/z: 618.00 [M+2H]2+, 1234.67
[M+H]+; for NMR data see Tables S1 in the Supporting Information.
cyclo(Pro-DPhe-Leu-Orn-Val-(βR,γR)-DiAA-Leu-Orn-Val-) (GS3). GS3
was synthesized using Manual Peptide Synthesis on a 0.25 mM scale
(30 mg, 0.028 mmol, 11% overall yield, 96% purity). Analytic RP-HPLC,
linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA) for 60
min. Detections were done at 221 nm using a photodiode array detector.
tr = 18.13 min; MS: (ESI+): m/z: 544.33 [M+2H]2+, 1087.33 [M+H]+; for
NMR data see Tables S1 in the Supporting Information.
NMR Spectroscopy. The NMR samples contained GS (2 mM) or GS3-6
(10 mM) were dissolved in 0.5 mL of H2O/D2O (9/1, v/v, containing 20
mM PBS buffer) at pH 7.0. The rest of GS analogues were prepared at
10-20 mM peptide concentration in 0.5 mL of DMSO-d6. NMR spectra
were acquired on a Bruker AV 500 or 600 MHz spectrometer at 4 oC or
o
25 C. 1H, COSY, TOCSY and NOESY spectroscopies were recorded to
fully assign the cyclic peptides complex data. Spectra were acquired by
standard pulse sequences using presaturation of the water signal if
necessary. The mixing times for TOCSY and NOESY are 80 ms and 500
ms, respectively.
cyclo(Pro-DPhe-Leu-Orn-Val-(βS,γR)-DiAA-Leu-Orn-Val-) (GS4). GS4
was synthesized using Manual Peptide Synthesis on a 0.25 mM scale
(45 mg, 0.041 mmol, 17% overall yield, 99% purity). Analytic RP-HPLC,
linear 30-70% CH3CN (0.1% TFA) gradient into H2O (0.1% TFA) for 60
min. Detections were done at 221 nm using a photodiode array detector.
tr = 21.76 min; MS: (ESI+): m/z: 544.25 [M+2H]2+, 1087.33 [M+H]+,
Circular Dichroism Spectroscopy. Samples for CD spectra were
prepared at 0.1 mM solutions in water or methanol. All spectra were
measured on a MOS-450 Spectrometer using a circular quartz cell with a
9
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
path length of 1 cm at 25 oC. Spectra were documented with a band
width of 1 nm, a duration time of 1 s and a scan speed of 100 nm/min.
Molar ellipticity [θ] was used to evaluate the Cotton effect. Each
measurement was repeated 3 times to calculate the mean value. The
spectra from solvents were subtracted as background in data analysis.
fluorescence were recorded with
a
CaryEclipse fluorescence
spectrophotometer (Aginent, USA). After the addition of peptides (2x
MIC), membrane depolarization was determined by monitoring the
change in fluorescence emission intensity of the DiSC3-5 dye at
excitation and emission wavelengths of 622 and 670 nm, respectively.
The addition of gramicidin S (4x MIC) was used to achieve maximum
fluorescence.
Antibacterial Assay. The following bacterial strains were used:
Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC
27853), Staphylococcus aureus (ATCC 29213), Bacillus subtilis (ATCC
6633), Streptococcus bovis (ATCC 33317), Enterococcus faecalis (ATCC
29212) and Enterococcus faecium (ATCC 35667). All culture media used
for bacteria incubation were freshly prepared. Bacteria were grown at
Tryptic Soy Broth (TSB) under constant shaking at 200 rpm for 6 h to
mid-logarithmic phase and harvested by centrifugation (10000x g for 3
min). The cells were washed three times with PBS buffer (10 mM, pH
7.4) and resuspended to A620 0.02 in TSB broth. GS and GS analogues
were dissolved in DMSO at 1 mg/mL as stock solution (2 mg/L solution
was used for P. aeruginosa). The stock solution was diluted in DMSO to
obtain a series of 2-fold dilutions. 96 Microtiter wells were filled with 10
μL μL f P. aeruginosa) GS or GS analogues solution in DMSO and
μL μL f P. aeruginosa) suspension of a mid-logarithmic phase
culture of bacteria. Cultures were carried out in triplicate. DMSO was
used as negative control. The MICs were assessed by measuring
turbidity at 620 nm after 48 h incubation at 37 oC. The 620 nm
absorbance was measured on a Multiskan FC microplate reader (Thermo
Scientific).
Scanning Electron Microscopy. Pathogens were grown at TSB broth
under constant shaking at 200 rpm for 6 h to mid-logarithmic phase and
harvested by centrifugation (10000x g for 3 min). The cells were washed
three times with PBS buffer (10 mM, pH 7.4) and resuspended to A620 0.4
in the same medium. Peptide treatment of the bacterial cells was carried
out at 37 oC for 1 h at their respective 2x MICs. Bacteria incubated with
PBS served as the control. After the incubation, the cells were harvested
by centrifugation (8000x g for 5min) and washed with PBS three times.
Bacterial cells were then fixed with 2.5% (w/v) glutaraldehyde at 4 oC for
6 h. Before centrifugation at 8000x g for 3min, the cells were mixed by
gently inverting the tube up and down for several minutes to prevent
clumping of the cells. After three washes with PBS, the bacterial pellets
were dehydrated for 15 min with a series of graded ethanol solutions (30,
50, 70, 80, 90 and 100%), the cells were mixed and centrifuged at 8000x
g for 3 min. Following dehydration, the dried bacterial cells were
transferred to dry alcohol. A volume of 10 uL of cell suspension was
transferred to the cover slide and allowed to oven-dried at 60 oC for 5 min.
The slides were coated and visualized under a scanning electron
microscope (FEI Field Electron Microscope, Quanta 250, USA).
Hemolysis Assay. Human blood solution, 10% (w/v) SDS solution and
PBS (10 mM, pH 7.4) solution were freshly prepared.GS and GS
analogues were dissolved in DMSO at 10 mg/mL as stock solution. The
stock solution was diluted in DMSO to obtain a series of 2-fold dilutions.
Fresh human blood solution was diluted to 5x105 erythrocytes per
m . w μL μL μL
diluted blood solution was added. DMSO was used as negative control.
SDS (0.1% w/v final concentration) was used as 100% hemolysis control.
Cultures were carried out in triplicate and incubated for 30 min at 37 oC.
Erythrocytes were separated by centrifugation (300 rpm, 10 min). A
m f μL f w f w 6
microtiter plate to measure the turbidity at 450 nm.
Acknowledgements
This work was financially supported by the National Natural
Science Foundation of China (No. 21967012), Health and
Family
Planning
Commission
of
Jiangxi
Province
(20195647,2018B012), start-up grant (No. 2018BSZR007) from
Jiaxngxi University of Traditionnal Chinese Medicine and State
Key Laboratory for Chemistry and Molecular Engineering of
Medicinal Resources (Guangxi Normal University, CMEMR2019-
B07). We thank Dr. Jingxian Yu (The University of Adelaide,
Australia) for attempted DFT calculations.
Cytotoxicity Assay. Peptides were dissolved in DMSO at 2.5 mM as
stock solution. The stock solution was diluted in cell medium (RPMI 1640,
μ /mL m m
with 1% fetal bovine serum to obtain a series of dilutions (0.1, 1, 5, 10 or
μ . L w w k w
medium supplemented with 1% fetal bovine serum in a humidified
incubator at 37 oC in 5% CO2. The cells (5 x103 cells per well) were
seeded in a 96-well plate in cell medium overnight. After culture medium
m μL m m ff ons of peptide
was added. Cultures were carried out in triplicate. After incubation for 24
m m w m μL f K- m m μL
m m μL K-8 solution) (Medchem Express, USA) was
added to each well and the samples were incubated for another 2 h. The
absorbance at 450 nm was measured by a SpectraMax i3x microplate
reader (MD, USA).
Keywords: m b • m • m mb
• m • β γ-diamino acids
[1]
[1]
L. Czaplewski, R. Bax, M. Clokie, M. Dawson, H. Fairhead, V. A.
Fischetti, S. Foster, B. F. Gilmore, R. E. W. Hancock, D. Harper, I. R.
Henderson, K. Hilpert, B. V. Jones, A. Kadioglu, D.Knowles, S.
Ólafsdóttir, D. Payne, S. Projan, S. Shaunak, J. Silverman, C. M.
Thomas, T. J. Trust, P. Warn, J. H. Rex, Lancet Infect. Dis. 2016, 16,
239–251.
[2]
[3]
European Centre for Disease Prevention and Control, Antimicrobial
resistance surveillance in Europe 2015, Annual report of the European
Antimicrobial Resistance Surveillance Network (EARS-Net), 2017
D. A. Goff, R. Kullar, E. J. C. Goldstein, M. Gilchrist, D. Nathwani, A. C.
Cheng, K. A. Cairns, K. Escandón-Vargas, M. V. Villegas, A. Brink, D,
van den Bergh, M. Mendelson, Lancet Infect. Dis. 2017, 17, e56–e63.
M. A. Cooper, D. Shlaes, Nature 2011, 472, 32–32.
Cytoplasmic Membrane Depolarization. S. aureus and B. subtilis were
grown at Luria Bertani (LB) broth under constant shaking at 200 rpm for 6
h to mid-logarithmic phase and harvested by centrifugation (10000x g for
3 min). The cells were washed three times with wash buffer (5 mM
HEPES, 20 mM glucose, pH 7.4) and resuspended to A620 0.005 (S.
aureus) or A620 0.03 (B. subtilis) in the same buffer. Then, 0.1 M KCl was
added to equilibrate the cytoplasmic and external K concentration. A
volume of 2 mL of cell suspension was placed in a 1 cm cuvette followed
by the addition of a DiSC3-5 stock solution (4 mM in HEPES buffer) to a
final concentration of 200 nM. The resulting solution was incubated until
[4]
[5]
M. Mahlapuu, J. Håkansson, L. Ringstad, C. Björn, Front. Cell. Infect.
Microbiol. 2016, 6, 194.
[6]
[7]
[8]
A. L. Llamas-Saiz, G. M. Grotenbreg, M. Overhand, M. J. van Raaij,
Acta Crystallogr. D Biol. Crystallogr. 2007, 63, 401–407.
S. E. Hull, R. Karlsson, P. Main, M. M. Woolfson, E. J. Dodson, Nature
1978, 275, 206–207.
A. Asano, M. Doi, X-Ray Struct. Anal. Online 2019, 35, 1–2.
a
stable reduction of fluorescence was achieved, indicating the
successful uptake of the dye into the bacterial membrane. Changes in
10
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
[9]
K. Yamada, M. Unno, K. Kobayashi, H. Oku, H. Yamamura, S. Araki, H.
[31] V. V. Kapoerchan, A. D. Knijnenburg, P. Keizer, E. Spalburg, A. J. de
Neeling, R. H. Mars-Groenendijk, D. Noort, J. M. Otero, A. L. Llamas-
Saiz, M. J. van Raaij, G. A. van der Marel, H. S. Overkleeft, M.
Overhand, Bioorg. Med. Chem. 2012, 20, 6059–6062.
Matsumoto, R. Katakai, M. Kawai, J. Am. Chem. Soc. 2002, 124,
12684–12688.
[10] G. G. Gause, M. G. Brazhnikova, Nature 1944, 154, 703.
[11] S. Afonin, U. H. N. Dürr, P. Wadhwani, J. Salgado, A. S. Ulrich, in
Bioact. Conform. II (Ed.: T. Peters), Springer Berlin Heidelberg, Berlin,
Heidelberg, 2008, pp. 139–154.
[32] M. Tamaki, T. Harada, K. Fujinuma, K. Takanashi, M. Shindo, M.
Kimura, Y. Uchida, Chem. Pharm. Bull. (Tokyo) 2012, 60, 1134–1138.
[33] O. Babii, S. Afonin, M. Berditsch, S. Reisser, P. K. Mykhailiuk, V. S.
Kubyshkin, T. Steinbrecher, A. S. Ulrich, I. V. Komarov, Angew. Chem.-
Int. Ed. 2014, 53, 3392–3395; Angew. Chem. 2014, 53, 3392–3395.
[34] Y. Q. Yeoh, J. Yu, S. W. Polyak, J. R. Horsley, A. D. Abell,
ChemBioChem 2018, 19, 2591–2597.
[12] Md. Ashrafuzzaman, O. S. Andersen, R. N. McElhaney, Biochim.
Biophys. Acta BBA - Biomembr. 2008, 1778, 2814–2822.
[13] M. Wenzel, M. Rautenbach, J. A. Vosloo, T. Siersma, C. H. M.
Aisenbrey, E. Zaitseva, W. E. Laubscher, W. van Rensburg, J. C.
Behrends, B. Bechinger, L. W. Hamoena, mBio 2018, 9, e00802-18.
[14] T. Katsu, T. Imamura, K. Komagoe, K. Masuda, T. Mizushima, Anal.
Sci. 2007, 23, 517–522.
[35] O. Babii, S. Afonin, A. Yu. Ishchenko, T. Schober, A. O. Negelia, G. M.
Tolstanova, L. V. Garmanchuk, L. I. Ostapchenko, I. V. Komarov, A. S.
Ulrich, J. Med. Chem. 2018, 61, 10793–10813.
[15] M. Wenzel, A. I. Chiriac, A. Otto, D. Zweytick, C. May, C. Schumacher,
R. Gust, H. B. Albada, M. Penkova, U. Krämer, R. Erdmannj, N.
Metzler-Nolteh, S. K. Strausk, E. Bremerl, D.Becherc, H. Brötz-
Oesterheltf, H. Sahlb, J. E. Bandowa, Proc. Natl. Acad. Sci. 2014, 111,
E1409–E1418.
[36] C. Solanas, B. G. de la Torre, M. Fernández-Reyes, C. M. Santiveri, M.
Á. Jiménez, L. Rivas, A. I. Jiménez, D. Andreu, C. Cativiela, J. Med.
Chem. 2009, 52, 664–674.
[37] C. Solanas, B. G. de la Torre, M. Fernández-Reyes, C. M. Santiveri, M.
Á. Jiménez, L. Rivas, A. I. Jiménez, D. Andreu, C. Cativiela, J. Med.
Chem. 2010, 53, 4119–4129.
[16] M. Berditsch, S. Afonin, J. Reuster, H. Lux, K. Schkolin, O. Babii, D. S.
Radchenko, I. Abdullah, N. William, V. Middel, U. Strähle, A. Nelson, K.
Valko, A. S. Ulrich, Sci. Rep. 2019, 9, 1–15.
[38] M. Tamaki, K. Fujinuma, T. Harada, K. Takanashi, M. Shindo, M.
Kimura, Y. Uchida, Bioorg. Med. Chem. Lett. 2012, 22, 106–109.
[39] G. M. Grotenbreg, A. E. M. Buizert, A. L. Llamas-Saiz, E. Spalburg, P.
A. V. van Hooft, A. J. de Neeling, D. Noort, M. J. van Raaij, G. A. van
der Marel, H. S. Overkleeft, M. Overhand, J. Am. Chem. Soc. 2006,
128, 7559–7565.
[17] L. H. Kondejewski, S. W. Farmer, D. S. Wishart, R. E. w. Hancock, R. S.
Hodges, Int. J. Pept. Protein Res. 1996, 47, 460–466.
[18] Q. Guan, S. Huang, Y. Jin, R. Campagne, V. Alezra, Y. Wan, J. Med.
Chem. 2019, 62, 7603–7617.
[19] S. Pal, U. Ghosh, R. S. Ampapathi, T. K. Chakraborty, in Pept. II (Ed.:
W. Lubell), Springer International Publishing, Cham, 2015, pp. 159–202.
[20] A. D. Knijnenburg, E. Spalburg, A. J. de Neeling, R. H. Mars ‐
Groenendijk, D. Noort, G. M. Grotenbreg, G. A. van der Marel, H. S.
Overkleeft, M. Overhand, Helv. Chim. Acta 2012, 95, 2544–2561.
[21] M. Jelokhani-Niaraki, L. H. Kondejewski, L. C. Wheaton, R. S. Hodges,
J. Med. Chem. 2009, 52, 2090–2097.
[40] S. Thétiot-Laurent, F. Bouillère, J.-P. Baltaze, F. Brisset, D. Feytens, C.
Kouklovsky, E. Miclet, V. Alezra, Org. Biomol. Chem. 2012, 10, 9660–
9663.
[41] A. Stanovych, R. Guillot, C. Kouklovsky, E. Miclet, V. Alezra, Amino
Acids 2014, 46, 2753–2757.
[42] C. T. Hoang, F. Bouillère, S. Johannesen, A. Zulauf, C. Panel, A.
Pouilhès, D. Gori, V. Alezra, C. Kouklovsky, J. Org. Chem. 2009, 74,
4177–4187.
[22] M. Tamaki, I. Sasaki, M. Kokuno, M. Shindo, M. Kimura, Y. Uchida, Org.
Biomol. Chem. 2010, 8, 1791.
[43] N. Auberger, A. Stanovych, S. Thétiot-Laurent, R. Guillot, C.
Kouklovsky, S. C. des Combes, C. Pacaud, I. Devillers, V. Alezra,
Amino Acids 2016, 48, 2237–2242.
[23] Y. Wan, A. Stanovych, D. Gori, S. Zirah, C. Kouklovsky, V. Alezra, Eur.
J. Med. Chem. 2018, 149, 122–128.
[24] A. S. Kotmale, E. Sangtani, R. G. Gonnade, D. Sarkar, S. Burade, P. R.
Rajamohanan, G. J. Sanjayan, New J. Chem. 2018, 42, 1197–1201.
[25] S. Pal, G. Singh, S. Singh, J. K. Tripathi, J. K. Ghosh, S. Sinha, R. S.
Ampapathi, T. K. Chakraborty, Org. Biomol. Chem. 2015, 13, 6789–
6802.
[44] M. Tamaki, S. Akabori, I. Muramatsu, J. Am. Chem. Soc. 1993, 115,
10492–10496.
[45] Y. Wan, J.-P. Baltaze, C. Kouklovsky, E. Miclet, V. Alezra, J. Pept. Sci.
2019, 25, e3143.
[46] D. S. Wishart, B. D. Sykes, F. M. Richards, Biochemistry 1992, 31,
1647–1651.
[26] B. Legrand, L. Mathieu, A. Lebrun, S. Andriamanarivo, V. Lisowski, N.
Masurier, S. Zirah, Y. K. Kang, J. Martinez, L. T. Maillard, Chem. – Eur.
J. 2014, 20, 6713–6720.
[47] L. H. Kondejewski, S. W. Farmer, D. S. Wishart, C. M. Kay, R. E. W.
Hancock, R. S. Hodges, J. Biol. Chem. 1996, 271, 25261–25268.
[48] 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.
[27] A. D. Knijnenburg, A. W. Tuin, E. Spalburg, A. J. de Neeling, R. H.
Mars-Groenendijk, D. Noort, J. M. Otero, A. L. Llamas-Saiz, M. J. van
Raaij, G. A. van der Marel, H. S. Overkleeft, M. Overhand, Chem. – Eur.
J. 2011, 17, 3995–4004.
[49] V. V. Kapoerchan, A. D. Knijnenburg, M. Niamat, E. Spalburg, A. J. de
Neeling, P. H. Nibbering, R. H. Mars-Groenendijk, D. Noort, J. M. Otero,
A. L. Llamas-Saiz, M. J. van Raaij, G. A. van der Marel, H. S.
Overkleeft, M. Overhand, Chem. – Eur. J. 2010, 16, 12174–12181.
[50] M. Kawai, H. Yamamura, R. Tanaka, H. Umemoto, C. Ohmizo, S.
Higuchi, T. Katsu, J. Pept. Res. 2004, 65, 98–104.
[28] C. Priem, A. Wuttke, M. Berditsch, A. S. Ulrich, A. Geyer, J. Org. Chem.
2017, 82, 12366–12376.
[29] G. Singh, S. Azmi, J. K. Ghosh, R. S. Ampapathi, S. Pal,
ChemistrySelect 2018, 3, 2272–2276.
[30] K. Yamada, S. Shinoda, H. Oku, K. Komagoe, T. Katsu, R. Katakai, J.
Med. Chem. 2006, 49, 7592–7595.
[51] M. R. Yeaman, N. Y. Yount, Pharmacol. Rev. 2003, 55, 27–55
11
This article is protected by copyright. All rights reserved.

10.1002/cmdc.202000097
ChemMedChem
FULL PAPER
Entry for the Table of Contents
E. coli
S. aureu
Text for Table of Contents
The native motif DPhe-Pro of antimicrobial peptide Gramicidin
S (GS) was
monosubstituted with a series of synthetic β γ)-diamino acids to afford non-toxic GS
analogues. The analogue GS6K included with βR γS)-diamino acid possesses sheet-
like conformation comparable to that of GS. Extensive biological evaluations
demonstrate that analogue GS6K displays both potent bactericidal activity and
diminished hemolytic effect. Moreover, antibacterial mechanism studies suggest that
GS6K kills bacteria mainly through the disruption of membrane.
12
This article is protected by copyright. All rights reserved.