Synthesis of cationic antimicrobial β2′2-amino acid derivatives with potential for oral administration

Article abstract of DOI:10.1021/jm101327d

We have prepared a series of highly potent achiral cationic β ^{2, 2}-amino acid derivatives that fulfill the Lipinski's rule of five and that contain the basic structural requirements of short cationic antimicrobial peptides. Highest antimicrobial potency was observed for one of the smallest β^{2, 2}-amino acid derivatives (M_w 423.6exhibiting a MIC of 3.8 μM against methicillin-resistant Staphylococcus aureus (MRSA), methicillin-resistant Staphylococcus epidermidis (MRSE), and Staphylococcus aureus, and 7.7 μM against Escherichia coli. The β ^{2, 2}-amino acid derivatives were shown to have similar absorption properties as several commercially available drugs, and the results implied a resembling membrane disrupting mechanism of action as reported for much larger cationic antimicrobial peptides. By their high potency, nontoxicity, absorption properties, and ease of synthesis, the β^{2, 2}-amino acid derivatives demonstrate a way to modify a vastly investigated class of cationic antimicrobial peptides into small druglike molecules with high commercial potential.

Full text of DOI:10.1021/jm101327d

858 J. Med. Chem. 2011, 54, 858–868
DOI: 10.1021/jm101327d
Synthesis of Cationic Antimicrobial β2,2-Amino Acid Derivatives with Potential for
Oral Administration
Terkel Hansen, Dominik Ausbacher, Gøril E. Flaten, Martina Havelkova, and Morten B. Strøm*
Department of Pharmacy, Faculty of Health Sciences, University of Tromsø, N-9037 Tromsø, Norway
Received October 14, 2010
We have prepared a series of highly potent achiral cationic β2,2-amino acid derivatives that fulfill the
Lipinski’s rule of five and that contain the basic structural requirements of short cationic antimicrobial
peptides. Highest antimicrobial potency was observed for one of the smallest β2,2-amino acid derivatives
(Mw 423.6) exhibiting a MIC of 3.8 μM against methicillin-resistant Staphylococcus aureus (MRSA),
methicillin-resistant Staphylococcus epidermidis (MRSE), and Staphylococcus aureus, and 7.7 μM
against Escherichia coli. The β2,2-amino acid derivatives were shown to have similar absorption prop-
erties as several commercially available drugs, and the results implied a resembling membrane disrupt-
ing mechanism of action as reported for much larger cationic antimicrobial peptides. By their high
potency, nontoxicity, absorption properties, and ease of synthesis, the β2,2-amino acid derivatives dem-
onstrate a way to modify a vastly investigated class of cationic antimicrobial peptides into small drug-
like molecules with high commercial potential.
Introduction
India and Pakistan.10 The plasmid associated NDM-1 gene
may easily be transferred to other bacteria through horizontal
gene transfer and confers resistance to a number of antimi-
crobial drugs, such as fluoroquinolones, aminoglycosides,
and all β-lactams, including the carbapenems which belong
to the drugs of last resort. Thereis therefore an urgent need for
new anti-infective drugs with new targets, distinctive modes of
action, and low probability of inducing resistance in order to
face the present and forthcoming challenges of multiresistant
bacteria.1,4,11,12
An extensively investigated class of anti-infective candi-
dates is cationic antimicrobial peptides (AMPs), also known
as host defense peptides (see Giuliani for an excellent
review).13 So far, the AntimicrobialPeptideDatabaseincludes
more than 1200 AMPs originating from the innate immunity
of mammals, amphibians, and insects.14,15 AMPs show a
unique mode of action by targeting and rapidly destroying
the bacterial cell membrane and are thereby less prone to
induce bacterial resistance compared to conventional antimi-
crobial drugs.13,16,17 Encouragingly, a number of small to
medium-large pharmaceutical companies have been estab-
lished as a result of the extensive research on these promising
molecules, and several drug candidates are in clinical trials.18
However, an imperative obstacle to success when develop-
ing clinical useful AMP-based drugs is to solve the inappropri-
ate pharmacokinetic properties of AMPs involving low
metabolic stability, troublesomeabsorptionkinetics, and risks
of unwanted immunological side effects unrelated to their
antimicrobial activity.19 As a result, the majority of AMP-
based drug candidates that are currently in clinical trials are
directed toward topical treatment of bacterial infections, to be
used as antiseptics for implants, or for sterilization of catheter
sites.13,18,20 Noteworthy, Brouwer et al. have demonstrated
target-identification by AMPs in vivo by showing that certain
AMPs are able to localize at the site of infection in mice.21
Rapid increase in bacterial resistance has become a major
public concern by escalating alongside a lack of development
of new anti-infective drugs.1 During the last few decades, only
four new antimicrobial drugs, linezolid, daptomycin, tigecy-
cline, and most recently telavancin have been approved for
treatment of serious pathogenic infections.2 At present, only a
handful of the major pharmaceutical companies have R&D
programs on anti-infective agents, and the low interest has
been defended based on simple economics.3 A modern drug-
development program for an antimicrobial agent is estimated
to cost more than 1 billion US$, which should be balanced
within a 10-15 year time frame. In addition, development of
new anti-infective agents is associated with high risks of induc-
ing bacterial resistance within a few years, as shown by history,
ortobe restricted foruseasadrug of last resort.4,5 Meanwhile,
the Gram-positive bacterium methicillin-resistant Staphylo-
coccus aureus (MRSA)a has become a leading hospital- and
community-acquired pathogen causing life-threatening cuta-
neous infections, and MRSA is currently outnumbering deaths
in US hospitals caused by HIV/AIDS and tuberculosis
together.6-9 Multiresistant Gram-negative enterobacteria-
ceae Escherichia coli and Klebsiella pneumoniae containing
the New Dehli metallo-β-lactamase 1 (NDM-1) gene were
recently isolated from patients who had been hospitalized in
*To whom inquiries should be directed. Associate Professor Morten
B. Strøm, Department of Pharmacy, Faculty of Health Sciences,
University of Tromsø, N-9037 Tromsø, Norway. Phone: þ47 77 64 40
85. Fax: þ47 77 64 61 51. E-mail: morten.strom@uit.no.
a Abbreviations: AMP, cationic antimicrobial peptide; DIPEA, di-
isopropyethylamine; MBC, minimal bactericidal concentration; MRSA,
methicillin-resistant Staphylococcus aureus; MRSE, methicillin-resis-
tant Staphylococcus epidermidis; NDM-1, New Dehli metallo-β-lacta-
mase 1; Ra/Ni, Raney nickel catalyst; RBC, red blood cells; TEA,
triethylamine; TFFH, fluoro-N,N,N0,N0-tetramethylformamidinium
hexafluorophosphate; TI, therapeutic index; TIS, triisopropylsilane.
r
pubs.acs.org/jmc
Published on Web 01/10/2011
2011 American Chemical Society

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3 859
There is therefore a great potential for systemic treatment of
bacterial infections by AMPs once the pharmacokinetic issues
have been solved.
final coupling step, and the β2,2-amino acids were also pre-
activated for 2 h with TFFH instead of the standard 10 min.27
The coupling reactions were followed by mass spectrometric
analysis and terminated by addition of aqueous Na2CO3.
However, due to sterical hindrance the coupling time had to
be extended up to 7 days for attachment of the C-terminal
cationic groups. Despite these measures and addition of
excessive TFFH, small amounts of uncoupled β2,2-amino
acids were nevertheless detected during purification by RP-
HPLC.
We have recently reported the antimicrobial activity of a
series of highly potent β-peptidomimetics consisting of a
lipophilic β2,2-amino acid coupled to a C-terminal L-arginine
amide residue that fulfill the minimal pharmacophore require-
mentsfor anti-Staphylococcalactivityof shortcationic AMPs,
that is, having an overall net charge of þ2 and containing two
sufficiently bulky and lipophilic residues.22 The β-peptidomi-
metics display exceptional proteolytic stability and low he-
molytic activity, and by the ease of synthesis allowing amend-
ment of their structural properties, the β-peptidomimetics
demonstrate a way to solve many of the obstacles associated
with developing small AMP-based drugs.
In the present study, we have further addressed the “drug-
likeness” of small antimicrobial β-peptidomimetics by pre-
paring a series of small β2,2-amino acid derivatives that, in
addition to fulfilling the pharmacophore model of short
AMPs, also fulfill the Lipinski’s rule of five which is used as
guidelines for design of peroral active drugs.23 A series of 16
achiral cationic β2,2-amino acid derivatives were synthesized
and screened for antimicrobial activity against the Gram-
positive bacteria MRSA, methicillin-resistant Staphylococcus
epidermidis (MRSE), Staphylococcus aureus, and the Gram-
negative bacterium Escherichia coli. The seven most potent
β2,2-amino acid derivatives against E. coli were furthermore
screened for possible lead-compound identification against
the Gram-negative bacteriumPseudomonas aeruginosa, which
is associated with cystic fibrosis. The peroral drug potential of
the β2,2-amino acid derivatives was furthermore addressed
both through theoretical calculations and by using a recently
developed phospholipid vesicle based barrier model for study-
ing passive diffusion across biological membranes, such as the
intestinal epithelia.24
Antimicrobial Activity. With a few exceptions, the anti-
microbial potency against all the test bacteria increased with
decreasing size of the C-terminal cationic groups. As shown
in Table 1, there was an increasing number of β2,2-amino acid
derivatives displaying MIC values below 10 μM against the
Gram-positive bacteria and E. coli when comparing series 4
and 5 with the smaller compounds of series 6 and 7. Im-
portant exceptions were observed against MRSA, which in
general was surprisingly susceptible to nearly all of the β2,2
-
-
amino acid derivatives. Thus, all four series 4-7 of β2,2
amino acid derivatives included compounds with MIC val-
ues below 10 μM against MRSA. Noteworthy, the two most
potent β2,2-amino acid derivatives 4c and 7d against MRSA
belonged to two structurally quite different classes of deriv-
atives.
As a general trend, the antimicrobial potency against the
Gram-positive bacteria and E. coli increased also with respect
to side-chain structure in the order a, b, c, and d. The β2,2
-
amino acid derivatives 4a, 5a, 6a, and 7a displayed in general
poor antimicrobial activity and were also among the least
lipophilic derivatives, as observed by measuring the overall
lipophilicity of the derivatives on an analytical RP-HPLC
C18-column (Figure 2). Highest antimicrobial potency against
all test bacteria was displayed by the β2,2-amino acid deriv-
atives 4d, 5d, 6d, and 7d, which belonged to the most
lipophilic derivatives. Both our previous report on β-pepti-
domimetics and reports by Haug et al. have shown a strong
correlation between high overall lipophilicity and high anti-
microbial potency.22,29 The results of the present study was
however unable to detect a similar relationship, as observed
when comparing overall lipophilicity as displayed in Figure 2
with antimicrobial potency shown in Table 1. Clearly, over-
all lipophilicity was not the sole determinant for antimicro-
bial potency.
When considering antimicrobial potency against the Gram-
positive test bacteria MRSE and S. aureus, 6d, 7c, and 7d
were the most potent β2,2-amino acid derivatives against
MRSE, and the two latter compounds 7c and 7d were also
the most potent derivatives against S. aureus. We observed a
somewhat lower potency against E. coli compared to the
Gram-positive bacteria, although the trends in antimicrobial
potency were similar. The two most potent β2,2-amino acid
derivatives against E. coli were 6d and 7d, which displayed
similar potency as against the Gram-positive bacteria.
On the basis of the results from screening of the β2,2-amino
acid derivatives against the Gram-negative bacterium E. coli,
the seven most potent β2,2-amino acid derivatives with MIC
values below 25 μM against E. coli were also screened for
antimicrobial activity against the Gram-negative bacterium
P. aeruginosa for possible lead-compound identification (Table 1).
The antimicrobial potency against P. aeruginosa was in
general lower than against E. coli, but three β2,2-amino acid
derivatives 6d, 7c, and 7d were identified as promising lead
molecules against P. aeruginosa. The remaining β2,2-amino
The present study demonstrates how a vastly investigated
class of peptides can be turned into small drug-like molecules
by confining the pharmacophore of small AMPs into a single
achiral β2,2-amino acid derivative. The derivatives displayed
very low hemolytic activity and comparable or even improved
antimicrobial potency compared to much larger AMPs. The
β2,2-amino acid derivatives demonstrate furthermore the po-
tential for oral administration of future AMP-based drugs by
having improved pharmacokinetic properties, as revealed by
the results from the permeability studies and by fulfilling the
Lipinski’s rule of five.23
Results
Synthesis. The β2,2-amino acid derivatives (Figure 1) were
synthesized according to the strategies shown in Scheme
1.22,25-27 For an excellent review covering a wide range of
methods for preparing β-amino acids, we highly recommend
the review by Abele and Seebach.28
The strategies chosen in the current study enabled synthe-
sis of the β2,2-amino acid derivatives by similar reaction
conditions and allowing the use of parallel synthesis equip-
ment ensuring high-throughput synthesis. The Boc-pro-
tected β2,2-amino acids 3a-d were synthesized by the same
overall method that we have recently reported, and in the
following step coupled to four different C-terminal cationic
groups (Scheme 1).22 In order to improve the yields, 1.5 equiv
of the coupling reagent fluoro-N,N,N0,N0-tetramethylform-
amidinium hexafluorophosphate (TFFH) was used in the

860 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3
Hansen et al.
Figure 1. Structure of the antimicrobial β2,2-amino acid derivatives that were prepared for development of more drug-like AMP-based
molecules. All the β2,2-amino acid derivatives were isolated as their ditrifluoroacetate salts.
Scheme 1. Strategies for Synthesis of All the Antimicrobial β2,2-Amino Acid Derivativesa
a (a) NaOMe (1 equiv), R-Br (1 equiv), performed twice, 78 °C s: MeOH. (b) Ra/Ni, H2(g), 45 °C, 5 days, s: MeOH containing 2% acetic acid. (c) TEA,
pH 8, Boc2O (1.2 equiv), r.t, 18 h, s: H2O/dioxane (1:5). (d) LiOH (6 equiv), 18 h, 100 °C, s: H2O/dioxane (1:3). (e) DIPEA (3 equiv), TFFH (1.5 equiv), r.
t., 2 h, then addition of the desired amine (2 equiv), up to 7 days, s: DMF. (f) TFA/TIS/H2O (95:2.5:2.5), r.t, 2 h, s: DCM. All β2,2-amino acid derivatives
were isolated as their di-trifluoroacetate salts.
acid derivatives tested against P. aeruginosa showed no or
poor antimicrobial potency.
derivatives displayed hemolytic activity above 100 μM, which
was well above their MIC values. A therapeutic index (TI)
was calculated by dividing the hemolytic activity (EC50) of
the β2,2-amino acid derivatives by the MIC value against
Selectivity. The hemolytic activity of the β2,2-amino acid
derivatives was in general very low, and except for 7d all

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3 861
Table 1. Minimal Inhibitory Concentration (MIC in μM) against MRSA, MRSE, S. aureus, E. coli, and P. aeruginosa, and Hemolytic Activity (EC50 in
μM) against Human RBC for a Series of Small β2,2-Amino Acid Derivatives Prepared for Development of More Drug-like AMP-Based Moleculesa
c
MICb
EC50
therapeutic index
entry
MRSAd
MRSEe
S. aureusf
E. colig
P. aeruginosah
RBCi
MRSA
MRSE
S. aureus
E. coli
P. aeruginosa
j
4a
4b
4c
4d
72
6.5
3.4
13
72
65
14
13
289
45
289
129
48
-
1116
409
172
120
26
17
29
26
25
20
26
8.6
8.5
26
20
13
13
255
337
1.3
5a
5b
5c
5d
315
14
315
70
315
49
315
70
-
525
303
348
38
41
23
7.5
20
48
11
20
23
7.5
14
23
7.4
15
15
7.2
15
15
22
15
-
289
1.2
6a
6b
6c
6d
56
7.1
7.5
7.4
56
21
56
14
160
50
507
468
117
274
9.1
66
16
37
9.1
22
16
74
9.1
33
16
37
3.2
9.4
5.1
37
7.5
3.7
7.5
7.4
23
7.4
300
74
0.4
3.7
7a
7b
7c
7d
59
7.4
7.8
3.8
25
15
71
15
336
52
1377
425
457
18
23
57
55
28
19
28
4.1
8
3.9
3.8
3.9
3.8
24
7.7
55
23
59
4.7
117
4.7
117
4.7
19
2.3
8.3
0.8
a A therapeutic index was calculated by dividing the EC50 RBC values by the MIC values against each bacterial strain. Highest concentrations tested
were b200 μg/mL and c1000 μg/mL. All β2,2-amino acid derivatives were isolated as their ditrifluoroacetate salts, and the molar concentrations were
calculated as such. d Methicillin-resistant Staphylococcus aureus (ATCC 33591). e Methicillin-resistant Staphylococcus epidermidis (ATCC 27626).
i
f Staphylococcus aureus (ATCC 25923). g Escherichia coli (ATCC 25922). h Pseudomonas aeruginosa (ATCC 27853). Human red blood cells. j The
notation “-“ denotes no detectable activity (MIC or EC50) within the concentration range tested. Please note that the aminoglycoside antibiotic
gentamicin was used as positive control and had a MIC value of 8.4 μM against S. aureus (ATCC 25923).
E. coli, the highly potent β2,2-amino acid derivative 6d
displayed the highest TI, but 4d, 5d, and 7c were also
promising compounds for future studies based on their
high TIs. Among the seven β2,2-amino acid derivatives tested
against P. aeruginosa, 7c displayed the highest TI followed by
6d, whereas the TI for the remaining four active compounds
was poor.
€
Drug-likeness and Oral Absorption. By using the Schroding-
€
er’s QikProp application that is included in the Schroding-
er’s Maestro software v9.1, an evaluation of the drug-like-
ness of the β2,2-amino acid derivatives with respect to the
Lipinski’s rule of five was assessed together with an estimation
of percentage oral absorption in humans (Table 2).
Figure 2. Retention time (Rt) on an analytical RP-HPLC C18
-
The calculations concerning hydrogen bond donor (HBD)
and hydrogen bond acceptor (HBA) groups were based on
the number of hydrogen bonds that could be formed with
water molecules in an aqueous solution, as determined by the
software. This explains the noninteger values in Table 2 for
the HBA groups of β2,2-amino acid derivatives confined in
the 6a-d and 7a-d series.
column for all β2,2-amino acid derivatives prepared as an estimation
of overall lipophilicity. The Rt was determined using a linearly
increasing gradient of acetonitrile in water containing 0.1% tri-
fluoroacetic acid.
each bacterial strain (Table 1). The results showed that the
β2,2-amino acid derivatives 4b and 4c displayed the upper-
most selectivity for MRSA compared to human red blood
cells (RBC). Although we observed a general trend of in-
creased antimicrobial potency of the β2,2-amino acid deriva-
tives by decreasing size of the C-terminal cationic group, in
the case of MRSA, a large C-terminal cationic group was
advantageous and resulted in both high antimicrobial po-
tency and very low RBC toxicity. Further inspections of the
TI with respect to MRSA revealed that 6b, 7b, and 7c also
displayed profound potency and selectivity for MRSA com-
pared to human RBC and are thereby promising β2,2-amino
acid derivatives for future studies.
The results from the calculations revealed that all the β2,2
-
amino acid derivatives fulfilled the Lipinski’s rule of five
since none of the compounds violated more than one of the
four rules.23 A single rule was violated by 5 of the 16 β2,2
-
amino acid derivatives prepared, in which three of these
violations were due to a molecular mass above 500 (4b, 4c,
and 4d), while the other two were due to a calculated log P
value (of the un-ionized compound) above the limit of 5 (5d
and 6d).
Calculation of percentage oral absorption in humans by
the software concluded that all the β2,2-amino acid deriva-
tives were likely to be absorbed rather well by having a
calculated oral absorption in the range of 62-89% (Table 2).
Highest percentage oral absorption was calculated for the
β2,2-amino acid derivatives 5a, 5b, 5c, 6b, and 6c, which all
were in the range of 86-89% oral absorption.
A fairly different TI-profile was observed for the β2,2
-
amino acid derivatives with respect to the other two Gram-
positive strains MRSE and S. aureus, in which the highly
potent compounds 6d and 7c displayed the highest selectivity
for these strains. In case of the Gram-negative bacterium

862 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3
Hansen et al.
Table 2. Evaluation of Drug-Likeness of the Prepared β2,2-Amino Acid
Derivatives with Respect to the Lipinski’s Rule of Five and Calculation of
Potential Oral Absorption in Humans
between the experimental permeability studies and the
calculations.30
Transmission Electron Microscopy. As a preliminary study
of the mechanism of action of the β2,2-amino acid deriva-
tives, S. aureus was incubated with two different concentra-
tions of 7c and investigated by transmission electron micro-
scopy (TEM) (Figure 4). In order to get representative
pictures, a much higher number of bacteria had to be used
compared to what was used in the MIC determinations, and
this also explains the high concentration of compound 7c
that was necessary in the TEM experiments.
The TEM experiments revealed formation of mesosome
structures arising from the cell wall or the septum of the
treated bacteria, both at the lowest concentration of 7c treated
for 3 h and at the highest concentration of 7c when treated for
2 h. The mesosome structures indicated membrane damage
and were absent in pictures of the untreated S. aureus control
cells. In addition to mesosome structures, lysed bacteria and
debris were observed in the pictures of S. aureus treated for 3
h with the highest concentrations of 7c.
Lipinski’s rule of five (highest permitted value)
calc oral
comp Mwa (500) HBDb (5) HBAc (10) CLogPd (5) abse (%)
4a
4b
4c
4d
464.7
544.6
508.7
520.8
1
1
1
1
7
7
7
7
3.8
4.2
4.3
4.8
72
62
63
66
5a
5b
5c
5d
407.6
487.5
451.6
463.7
1
1
1
1
5
5
5
5
3.9
4.4
4.3
5.1
86
88
87
79
6a
6b
6c
6d
395.6
475.5
439.6
451.7
2
2
2
2
4.5
4.5
4.5
4.5
3.8
4.6
4.7
5.3
84
89
89
80
7a
7b
7c
7d
367.5
447.4
411.5
423.6
4
4
4
4
3.5
3.5
3.5
3.5
2.9
3.4
3.4
3.7
73
71
72
73
Discussion
a Molecular weights were calculated for nonionized β2,2-amino acid
derivatives. b Calculated number of hydrogen bond donor (HBD)
groups. c Calculated number of hydrogen bond acceptor (HBA) groups.
The values were averages over a number of configurations, hence the
noninteger values. d Calculated octanol-water partition coefficient
CLogP of the neutral species of the compounds e Calculated percent
oral absorption in humans. All predictions were calculated using the
Synthesis. We have recently reported the synthesis of a
series of highly potent small β-peptidomimetics designed to
fulfill the minimal pharmacophore model of short cationic
AMPs with anti-Staphylococcal activity.22 During this work,
we discovered that lipophilic 2,2-disubstituted propionic
acid derivatives, or β2,2-amino acids, that are linked to an
L-arginine amide residue are excellent building blocks for
preparing highly potent AMP-based molecules. The β-pep-
tidomimetics had the size of a dipeptide but the side-chain
functionalities of a tripeptide and showed superior proteo-
lytic stability compared to many naturally occurring AMPs.
In the present study, we have investigated whether even
simpler scaffolds based on the same lipophilic β2,2-amino
acids, and which fulfilled the same pharmacophore model,
could provide even more drug-like AMP-based compounds
while retaining the high antimicrobial potency of the pre-
viously reported β-peptidomimetics (Figure 1). A prerequi-
€
€
Schrodinger QikProp application included in the Schrodinger’s Maestro
software v9.1.
Permeability. Encouraged by the theoretical calculations,
the permeability of the β2,2-amino acid derivatives was fur-
ther investigated using a recently established phospholipid
vesicle based barrier model.24 The model has been developed
for estimation of drug permeability by passive diffusion across
biological barriers such as the intestinal epithelia. Four β2,2
-
amino acid derivatives (4c, 5c, 6c, and 7c), which showed
similar potency against MRSA and against E. coli were
investigated (Figure 3). All test compounds contained the
same lipophilic 2-naphthalene methylene side-chains but
different C-terminal cationic groups. On the basis of the
models classification of absorption, all four compounds
showed permeability equivalent to being moderately absorbed
in humans.24
However, the rank order of the compounds based on
permeability through the barrier was a bit different from
the rank order based on the calculations of human absorp-
tion, as observed when comparing Figure 3 and Table 2. For
the experimental permeability values, compound 5c showed
highest permeability followed by compounds 6c, 4c, and 7c.
Especially compound 7c showed a much lower experimental
permeability than expected from the calculated absorption
values, but we also experienced difficulties with dissolving 7c
in the buffer solution used in the experiment. Because of the
low solubility of 7c the experiment had therefore to be
performed with pH 5.5 in the donor compartment, and not
pH 7.4 as for the other compounds. The lower pH in the
donor compartment increased the solubility of compound
7c, but might also have resulted in a somewhat higher ioniza-
tion of 7c disfavoring permeation across the phospholipid
barrier. Together, the low solubility of 7c and lowered pH
in the donor compartment may explain the discrepancy
site when designing the scaffolds was therefore that the β2,2
-
amino acid derivatives should fulfill the Lipinski’s rule
of five, which is used as a guideline to identify drug candi-
dates with potential for high bioavailability following oral
administration.23
The scaffolds prepared in the present study were based on
four different lipophilic β2,2-amino acid residues selected
among the most potent β-peptidomimetics from our pre-
vious study, and coupled to four different C-terminal cation-
ic groups: 1-[2-(dimethylamino)ethyl]piperazine (4a-d), N-
methylpiperazine (5a-d), N,N-dimethylendiamine (6a-d),
and 1,2-diaminoethane (7a-d).
It should be noted that the β2,2-amino acid building blocks
were themselves achiral, and when selecting cationic groups
this benefit was kept by choosing four achiral cationic groups
from a commercial supplier. Compared to synthesis of larger
AMPs in which control of epimerization in each coupling
step is crucial, achirality is clearly an advantage and excludes
the necessity of separating steroisomers that may be the
source of unfavorable pharmacological effects. Further-
more, preparation of the compounds was accomplished from
cheap starting materials and avoiding expensive resins and
equipment used for solid phase peptide synthesis of AMPs.
The reported antimicrobial β2,2-amino acid derivatives should

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3 863
Figure 3. Permeability with standard deviation of the four β2,2-amino acid derivatives 4c, 5c, 6c, and 7c across a phospholipid vesicle based
barrier as a model for estimation of passive diffusion across, for example, intestinal epithelia.24 *Because of low solubility of 7c at pH 7.4, the
experiment had to be run with pH 5.5 in the donor compartment. The pH in the acceptor compartments was always 7.4.
Figure 4. Transmission electron microscopy (TEM) study of the effects of the β2,2-amino acid derivative 7c on S. aureus. Top left: untreated S.
aureus control cells. Top right: S. aureus incubated with 117 μM of 7c (30 ꢀ MIC) for 3 h. Bottom left: S. aureus incubated with 234 μM of 7c (60
ꢀ MIC) for 2 h. Bottom right: S. aureus incubated with 234 μM of 7c (60 ꢀ MIC) for 3 h. Mesosome structures (arrows) as an indication of
membrane damage were observed at all stages of incubation with 7c, whereas debris from lysed bacteria was only observed in pictures with the
highest concentration of 7c.
therefore be highly suitable for large-scale industrial manu-
facture.
Antimicrobial Activity. The results of the study were a
series of highly potent β2,2-amino acid derivatives against both
Gram-positive and Gram-negative bacteria. The antimicrobial
potency of the most potent compounds 4c, 6d, 7c, and 7d
revealed MIC values below 4 μM against MRSA, MRSE,
and S. aureus, and MIC values of less than 8 μM against
E. coli for 6d and 7d (Table 1). With respect to E. coli, the high
potency of the smallest β2,2-amino acid derivatives may have

864 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3
Hansen et al.
been a result of easier penetration through the porins of the
outer LPS barrier of E. coli, and thereby easier accessibility
to the cytoplasmic membrane. However, the Gram-negative
bacterium P. aeruginosa was somewhat less susceptible with
a MIC value of 23 μM for the most potent derivative 7d.
Whether the lower potency against P. aeruginosa was an
effect of differences in membrane composition or defensive
mechanisms was not investigated. However, P. aeruginosa is
reported to have much lower outer membrane permeability
than E. coli, and to possess a rich arsenal of efflux pumps that
can transport anti-infective agents such as tetracyclines and
fluoroquinolones out of the periplasmic space, and together
these effects could explain the reduced susceptibility to our
small compounds.31,32
the four C-terminal cationic groups, but also the structure,
primary vs tertiary amine, pKa, rotational freedom and flex-
ibility of the cationic groups, and the ability to form addi-
tional noncharged interactions, such as reinforced hydrogen-
bond interactions with the bacterial surface.
The β2,2-amino acid derivatives 4a-d contained a C-termi-
nal 1-[2-(dimethylamino)ethyl]piperazine group, which was
probably capable of forming the most complex charge-
hydrogen bond interactions with the bacterial surface, and
thereby perhaps being the one closest to resembling the arginine
residue of the recently reported β-peptidomimetics.22 As
described above, this motif was highly successful against
MRSA, and 4b and 4c were both highly potent and displayed
the overall highest TI with respect to MRSA. The β2,2-amino
acid derivatives 5a-d contained a C-terminal N-methylpi-
perazine group which resulted in a more constrained C-
terminus. The reduced rotational freedom and lack of other
HBD groups could explain the somewhat overall low anti-
microbial potency of the 5a-d series. The C-terminal N,N-
dimethylendiamino group of β2,2-amino acid derivatives
6a-d ensured enhanced flexibility and an extra HBD group
that could explain some of the improved potency of 6a-d
compared to the β2,2-amino acid derivatives in series 4 and 5.
The smallest series of β2,2-amino acid derivatives 7a-d con-
tained a 1,2-diaminoethane group and were the only com-
pounds containing a C-terminal primary amino group. The
profound high antimicrobial potency displayed especially by
compounds 7b, 7c, and 7d may be a combination of small
size, flexible C-terminal primary amino group, and two extra
HBD groups.
Selectivity. Hemolytic activity was used as a measurement
of toxicity of the prepared β2,2-amino acid derivatives. The
erythrocyte surface consists of glycoproteins bearing sialic
acids that provide a negative surface charge, which contri-
butes to efficient blood flow and repulsive forces between
blood cells and endothelial cells.33-35 The negative charge on
the erythrocytes are also anchoring sites for positively charged
compounds such as AMPs to initiate potential hemolytic
action. However, the only β2,2-amino acid derivatives that
could be classified as hemolytic was 7d. All the other β2,2
-
amino acid derivatives displayed EC50 values above 100 μM,
which was well above their MIC values (Table 1).
We were a bit surprised by the high hemolytic activity
displayed by 7d compared to 4d, 5d, and 6d, which all
contained the same tert-butyl benzylic side-chains, but dif-
ferent C-terminal cationic groups (Figure 1). The high
hemolytic activity of 7d emphasizes an observation we made
from our series of β-peptidomimetics showing that small
changes in structure can have a surprisingly high impact on
hemolytic activity, whereas antimicrobial potency is much
less affected.22
The little hemolytic activity that was detected was however
used to calculate a preliminary therapeutic index (TI) by
dividing the EC50 values by the MIC values against each
bacterial strain. On the basis of antimicrobial potency and
TI, the most promising candidate against MRSA infections
was thereby 4b, with 4c as a good second choice (Table 1).
When regarding the other two Gram-positive bacterial strains
MRSE and S. aureus, compound 7c emerged as the most
promising candidate based on its high potency and TI. In
case of the Gram-negative bacteria, compound 6d displayed
the highest TI with respect to E. coli, while for P. aeruginosa
compound 7c displayed the highest TI.
C-Terminal Cationic Groups. When considering target
interactions, the cationic groups in AMPs certainly have an
important role by attracting AMPs to the negatively charged
cell surfaces of bacteria, and thereby giving rise to the
specificity for bacterial cells over mammalian cells.36 In most
synthetic derivatives of naturally occurring AMPs, lysine,
ornithine, and arginine residues are used to provide the
cationic charge since these residues are relatively cheap and
are easily incorporated into AMPs.13 Noteworthy, Svenson
et al. have recently investigated unnatural cationic amino
acid side-chains in short AMPs and have shown that the
choice of cationic group plays an important role for anti-
microbial potency and proteolytic stability.37,38
Although a complex system, investigating different cation-
ic groups can be a powerful tool for optimizing AMP-based
compounds, and may improve potency, stability, safety,
strain specificity, and last but not least, the pharmacokinetic
properties of AMPs as discussed more in detail below.
Lipophilic Side-Chains. The overall lipophilicity of the
β2,2-amino acid derivatives was estimated by investigating
the retention time on a RP-HPLC C18 column (Figure 2).
The results showed that compounds 4d, 5d, 6d, and 7d with
two para-tert-butyl benzyl side-chains were most lipophilic,
whereas β2,2-amino acid derivatives 4a, 5a, 6a, and 7a with
two benzyl propyl side-chains were the least lipophilic deriv-
atives, which also correlated well with antimicrobial po-
tency. Thus, 4d, 5d, 6d, and 7d displayed high antimicrobial
potency, whereas the potency of 4a, 5a, 6a, and 7a was poor.
However, except for these derivatives, there was little corre-
lation between overall lipophilicity and antimicrobial po-
tency, as observed by the generally higher potency of β2,2
-
amino acid derivatives 4c, 5c, 6c, and 7c containing two
2-naphthyl methylene side-chains compared to the more
lipophilic derivatives 4b, 5b, 6b, and 7b containing two para-
trifluoromethyl benzyl side-chains (Table 1 and Figure 2).
The results were thereby somewhat contradictory with re-
spect to earlier reported studies on small AMP-based com-
pounds and our own β-peptidomimetics, in which much
stronger correlations between antimicrobial potency and
overall lipophilicity have been observed.22,29
However, there are other effects to consider, such as the
structure and conformations of the lipophilic side-chains of
the β2,2-amino acids. The close proximity of the lipophilic
side-chains causes steric repulsions between the ortho-pro-
tons of the aromatic side-chains b, c, and d.39 This may
induce quite rigid conformations with close to perpendicular
orientation of these side-chains that affect antimicrobial
In the current study, we investigated four different C-term-
inal cationic groups, and as a general rule, the antimicrobial
potency increased with reduction in the size of the C-terminal
cationic groups. However, not only the size differed between

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3 865
potency. The low potency of 4a, 5a, 6a, and 7a can in part be
explained by the reduced ability of their side-chains to form
such rigid conformations together with an overall too low
lipophilicity. Noteworthy, Tew et al. observed a major
improvement in antimicrobial potency by including groups
that enabled formation of intramolecular hydrogen bonds,
and which thereby rigidified their foldamers.40
commercial drugs.24 The permeability barrier consisted of
phospholipid vesicles made from Lipoid E 80, which is an egg
lipid extract with 82% (w/w) phosphatidylcholine in addi-
tion to other zwitterionic phospholipids, triglycerides, and
cholesterol.24 Four β2,2-amino acid derivatives (4c, 5c, 6c,
and 7c) with comparable potency against MRSA and against
E. coli were selected since they represented each of the
C-terminal cationic groups, but all contained the same
lipophilic 2-naphthyl methylene side-chains.
Furthermore, the para-trifluoromethyl benzyl side-chains
of 4b, 5b, 6b, and 7b had a somewhat profound effect on
MRSA, resulting in a 2-10 times higher potency against
MRSA compared to the two other Gram-positive bacteria
MRSE and S. aureus (Table 1). We made the same observa-
tion against MRSA for our β-peptidomimetics containing
this specific side-chain.22 In drug development, substitution
of aromatic or benzylic hydrogen atoms with fluorine is
frequently used to optimize the pharmacokinetic properties
of drug candidates.41-43 The contributions of fluorine may
involve improved membrane permeability due to increased
lipophilicity and altered electronic distribution due to the
high electronegativity of the fluorine atom. The high anti-
microbial potency displayed by 4b, 5b, 6b, and 7b against
MRSA may have been a result of a combination of these
effects. Additionally, by having an effective van der Waals
radius similar to an isopropyl group, the trifluoromethyl
group increased the bulk of the β2,2-amino acid side-chains
which may have affected the membrane interaction of the
compounds.44,45 Noteworthy, Tew et al. have reported a
series of small amphipathic foldamers, in which one of their
most potent compounds currently undergoing clinical trials
contains two trifluoromethyl phenyl groups.40 Thus, triflu-
oromethyl substitution may be an efficient way of improving
potency, safety, and pharmacokinetic properties of AMP-
based drug candidates.
The results showed that all the tested β2,2-amino acid
derivatives were within the models range of moderate absorp-
tion, in which highest drug-permeability was displayed by the
β2,2-amino acid derivative 5c followed by 6c and 4c
(Figure 3). The smallest derivative 7c was just above the
borderline between poor and moderate absorption. How-
ever, due to low solubility the permeability of 7c was mea-
sured with pH 5.5 in the donor compartment compared to
pH 7.4 for the others, which may have decreased the perme-
ability through the barrier due to increased ionization of 7c
(as described in detail in the Results section).
The measured absorption of the β2,2-amino acid deriva-
tives was nevertheless in the same range as for several
commercial available drugs that have been investigated by
the same model, such as the antihypertensive drug enalapril,
the β-blocker atenolol, and the antiulcer drug ranitidine,
which all are administered by the oral route.24
Noteworthy, we observed no lysis of the phospholipid vesicles
during the experiments, as we have occasionally observed for
more hemolytic peptides, which supported the hemolytic
results showing that the β2,2-amino acid derivatives were
safe to eukaryotic cell membranes (results not shown).37
Transmission Electron Microscopy. We have so far not
identified any intracellular targets for the β2,2-amino acid
derivatives, but at the present stage we cannot exclude pos-
sible inhibition of metabolic enzymes in the bacteria, or a
mechanism of action involving a combination of membrane
disruption and inhibition of crucial bacterial enzymes. Cer-
tainly, the cationic groups of the β2,2-amino acid derivatives
would be susceptible to interact with negatively charged
intracellular molecules, such as DNA and RNA once dif-
fused across the membrane.
As a brief study of the mechanism of action, the effects of
treating S. aureus bacteria with the β2,2-amino acid derivative
7c was studied by transmission electron microscopy (TEM)
(Figure 4). Compound 7c was selected due to its high potency
and selectivity for S. aureus. As described in the Results
section, we had to increase the number of bacteria and hence
also the concentration of 7c in order to get representative
TEM pictures.
Lipinski’s Rule of Five and Permeability Across Biological
Barriers. The drug-likeness and potential for oral adminis-
tration of the β2,2-amino acid derivatives was estimated by
the Lipinski’s rule of five as well as calculation of oral
€
absorption using the Schrodinger QikProp application
(Table 2). The rules state that an oral active drug should
not violate more than one of the following four criteria:
(1) the octanol-water partition coefficient log P should be
less than 5, (2) the molecular mass (Mw) should not exceed
500 Da, (3) a maximum of 5 hydrogen bond donor (HBD)
groups is tolerated, and (4) there should be no more than
10 hydrogen bond acceptor (HBA) groups. However excep-
tions to the Lipinski’s rule of five are known and involve
drugs that are transported across membranes by carrier
proteins, such as the highly polar antibiotic erythromycin.46
The issue of active transport of the β2,2-amino acid deriva-
tives was not investigated in the current study.
By using TEM to assess the impact of peptide treatment,
we observed a membrane effect on S. aureus treated with
7c by the appearance of mesosome structures (Figure 4).
Mesosome structures have earlier been regarded as structural
artifacts thought to be caused by chemical fixation of the
bacteria prior to resin embedding and investigation by TEM.47
However, several groups have pointed out that mesosome
structures must be regarded as cytoplasmic membrane al-
terations, and have demonstrated similar mesosome struc-
tures after treatment of S. aureus with different types of
antimicrobial peptides such as the 26 amino acid R-helical
peptide CP29, the cecropine derivative CP11CN, and the
linear bactenecin derivative Bac2A-NH2.48,49 None of the
control cells displayed mesosome structures although they
were prepared in the same way as the treated bacteria. We
The calculations showed that all the β2,2-amino acid
derivatives fulfilled the Lipinski’s rule of five even if five
compounds violated one single rule (Table 2).23 Three of
these violations were due to molecular weights above 500 (4b,
4c, and 4d), whereas two were owing to a calculated log P
above 5 (5d and 6d). Furthermore, the calculated oral
absorption was good and ranged from 62 to 89%, with only
three β2,2-amino acid derivatives having a calculated oral
absorption below 70% (Table 2).
On the basis of these encouraging theoretical results, we
measured the permeability of the β2,2-amino acid derivatives
by passive diffusion through a phospholipid vesicle based
barrier model that has proven to give strong coherence be-
tween experimental and measured oral bioavailability of

866 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3
Hansen et al.
therefore suspect that the mesosome structures were precur-
sors of membrane disintegration. TEM pictures of S. aureus
treated with the highest concentrations of 7c revealed lysed
cells and formation of debris. We therefore anticipate that
the TEM pictures of mesosome structures, lysed bacteria,
and formation of debris reflected subsequent stages in the
mechanism of action of 7c, and that membrane disruption
was a plausible mode of action. As determined by reincubat-
ing treated bacteria on agar plates, the β2,2-amino acid deriv-
atives were bactericidal and not bacteriostatic (results not
shown).
RP-HPLC was carried out on a Waters system equipped with a
˚
Sunfire C18 100 A, 5 μm, 19 ꢀ 250 mm column, and eluted with
acetonitrile and water, both containing 0.1% TFA. Analytical
HPLC was carried out on a Waters 2695 HPLC equipped with a
˚
Kromasil C18, 300 A, 5 μm, 4.6 ꢀ 250 mm column with PDA
detector spanning from wavelength 210 to 310 nm. Compounds
undergoing biological evaluation were at least 95% pure as
determined by analytical HPLC at 214 and 254 nm. All com-
pounds were prepared by using parallel reaction carousels from
Radleys.
Antimicrobial Activity. The antimicrobial screening was con-
ducted at TosLab A/S (Tromsø, Norway). Each compound was
tested in duplicate at 200, 100, 50, 35, 15, 10, 5, 2.5, 1, 0.5 μg/mL.
All tested compounds were ditrifluoroacetic acid salts.
Hemolytic Activity. The hemolytic activity was assessed as
described earlier.22 In brief, the plasma fraction of heparinized
human blood was first removed by centrifugation and three
additional washing steps with 37 °C prewarmed phosphate
buffered saline (PBS). Subsequently, the human RBCs were
diluted to 10% hematocrit and the β2,2-amino acid derivatives
were dissolved in PBS providing concentrations ranging from 1
to 1000 μg/mL. The diluted RBC were added to the compound
solutions to a final erythrocyte concentration of 1% v/v. PBS
and Triton X-100, in a final concentration of 0.1% v/v, were
included as negative and positive control. After 1 h agitated
incubation at 37 °C the samples were centrifuged at 4000 rpm for
5 min. Release of hemoglobin was determined by measuring the
absorbance of the supernatant at 405 nm. Hemolytic activity
was calculated as the ratio of treated sample with β2,2-amino
acid derivative and Triton X-100 treated sample according to
the formula:
Conclusion
Naturally occurring cationic AMPs have a general size of
12-50 amino acids and are composed of approximately 50%
lipophilic amino acids.50 The current β2,2-amino acid deriva-
tives forced this motif to its limits, in which the structural
requirements of 50% lipophilic amino acids and a net positive
charge was certified by two lipophilic side-chains within a
single β2,2-amino acid unit, and with two cationic groups of
diverse structures on each sides (Figure 1).
Both β2,2-amino acid derivatives with a broad spectrum and
a narrow spectrum of activity were developed, and all com-
pounds but one were nontoxic toward human RBC (Table 1).
Most notable was the high potency and therapeutic index
displayed by the β2,2-amino acid derivatives against MRSA,
especially for 4b and 4c. The β2,2-amino acid derivatives
displayed high specificity, in which the most potent and selec-
tive compound against MRSE and S. aureus was 7c, whereas
6d displayed the highest potency and selectivity againstE. coli.
The TEM studies on S. aureus treated with 7c indicated a
similar membrane disrupting mode of action as reported for
much larger AMPs, although interaction with intracellular
targets, or a combination of both, cannot be ruled out at the
present stage (Figure 4).
Abs½β2,2-amino acidꢁ - Abs½negative controlꢁ
½%ꢁhemolysis ¼
Abs½positive controlꢁ - Abs½negative controlꢁ
ꢀ 100
Permeability Experiments. The phospholipid vesicle-based
barriers were prepared according to previous reported pro-
cedures.24,51 The lipid used was Lipoid E 80, which is an egg
lipid extract with approximately 82% (w/w) phosphatidylcho-
line, 9% phosphatidylethanolamine, 3% lysophosphatidylcho-
line, 2% sphingomyelin, 3% triglycerides, and 1% cholesterol
(according to the manufacturer Lipoid GmbH, Ludwigshafen,
Germany). In brief, liposome dispersions extruded through
filters with a pore size of 800 nm and 800 þ 400 nm were
deposited on a filter support by use of centrifugation. The
liposomes were added in consecutive steps, first the smaller
liposomes and then the larger ones. Freeze-thaw cycling was
then used to promote fusion of the liposomes and hence produce
a tight permeation barrier.
Permeability experiments were performed at room tempera-
ture by moving inserts, containing the different β2,2-amino acid
derivatives dissolved in phosphate buffer, at certain time inter-
vals to wells containing an equal quantity of fresh phosphate
buffer pH 7.4.24 For 4c, 5c, and 6c phosphate buffer pH 7.4 was
also used on the donor side while for 7c phosphate buffer pH 5.5
was used to be able to dissolve 7c. UV absorbance (Spectra-
max190; Molecular devices, Molecular Device Corporation,
Sunnyvale, CA, USA) at a wavelength of 230 nm was used to
quantify the amount of peptides in the different acceptor
compartments giving values for the time points 1, 2, 3, 3.5, 4,
4.5, and 5 h. The electrical resistance across the lipid barriers was
measured (Millicell-ERS, Millipore, Billerica, MA, USA) im-
mediately after completion of the permeation studies to control
the integrity of the barrier.
All β2,2-aminoacid derivatives fulfilled the Lipinski’s ruleof
five, which is used as guidelines to ensure high peroral bio-
availability of drug candidates.23 Studies using a phospholipid
vesicle based barrier model supported these results and re-
vealed that the absorption properties of the β2,2-amino acid
derivatives were comparable to several commercially avail-
able drugs.
The β2,2-amino acid derivatives were achiral and prepared
from cheap starting materials by methods that can easily be
scaled up for industrial manufacture. The high potency, non-
toxicity, absorption properties, and ease of synthesis empha-
sizethe commercial potential of this new class ofantimicrobial
compounds.
Experimental Section
All chemicals used in the synthesis and biological evaluation of
the β2,2-amino acid derivatives were purchased from Sigma-
Aldrich and used without further purification. Lipoid E-80 was
obtained from Lipoid GmbH, Ludwigshafen, Germany. Filter
inserts (Transwell, d = 6.5 mm) and plates were purchased from
Corning Inc. (Lowell, MA, USA), and the mixed cellulose ester
filters (0.65 μm pore size) were purchased from Millipore
(Billerica, MA, USA). 1H and 13C NMR spectra were recorded
on a 600 MHz Varian spectrometer, and chemical shifts are
expressed in ppm relative to methanol (1H 3.31 ppm, 13C 39.0
ppm). The values are given in δ scale. Mass spectra were obtained
on a Micromass Quattro LC (Micromass, Manchester, UK).
High-resolution mass spectra were obtained on a Waters Micro-
mass LCT Premier (Micromass, Manchester, UK). Preparative
Transmission Electron Microscopy (TEM). Approximately
4 ꢀ 108 bacteria were incubated with the β2,2-amino acid deriva-
tive 7c for a maximum time range of 3 h. Suspensions were
prefixated with Karnovsky’s fixative and stored in pure fixative

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3 867
at 4 °C overnight. After postfixating with ferrocyanide-reduced
osmium tetroxide and dehydration in a graded series of ethanol,
samples were infiltrated with a 1:1 mixture of acetonitril/Epon
resin (AGAR 100, DDSA, MNA, and DMP-30) overnight. Pure
resin was applied the following day and after polymerization
ultrathin sections were prepared. Microscopy was carried out
on a JEOL JEM-1010 TEM and representative pictures were
taken.
Acknowledgment. We thank Lytix Biopharma A/S for
financial support and Espen Hansen (Marbio, Tromsø Norway)
for assistance with HRMS.
Supporting Information Available: Spectroscopic data for all
the compounds 4a-d, 5a-d, 6a-d, and 7a-d, as well as general
procedures for the synthesis of 1a-d, 2a-d, and 3a-d. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Synthesis of β2,2-Amino Acid Derivatives. Synthesis of the
Boc-protected β2,2-amino acids was conducted in accordance
with our previous publication; please see Hansen et al. for details
and yields, and the Supporting Information for general pro-
cedures.22
References
(1) Bad Bugs, No Drugs; Infectious Diseases Society of America: Arlington,
VA, July, 2004; p 37.
(2) Corey, G. R.; Stryjewski, M. E.; Weyenberg, W.; Yasothan, U.;
Kirkpatrick, P. Telavancin. Nat. Rev. Drug Discovery 2009, 8, 929–
930.
(3) Boucher, H. W.; Talbot, G. H.; Bradley, J. S.; Edwards, J. E.;
Gilbert, D.; Rice, L. B.; Scheld, M.; Spellberg, B.; Bartlett, J. Bad
Bugs, No Drugs: No ESKAPE! An Update from the Infectious
Diseases Society of America. Clin. Infect. Dis. 2009, 48, 1–12.
(4) Peet, N. P. Drug resistance: a growing problem. Drug Discovery
Today 2010, 15, 583–586.
General Procedure for Synthesis of 4a-d, 5a-d, 6a-d, and
7a-d (GP1). The synthesis was based on the textbook of Chan
and White.27 In brief, the Boc-protected β2,2-amino acids (3a-d)
(typically 0.2 mmol) were dissolved in DMF (0.02 M) and
DIPEA (3 equiv) was added along with TFFH (1 equiv). The
β2,2-amino acids were activated for 2 h before the desired amine
was added (2 equiv). Each reaction was followed by MS, and
allowed to react for up to 7 days before it was diluted with ethyl
acetate and washed with brine. The organic phase was dried over
MgSO4, filtered, and evaporated to dryness. The Boc-protected
β2,2-amino acid derivatives were deprotected by dissolving them
in DCM, adding an equivalent volume of TFA/TIS/water
(95:2.5:2.5) and stirred at r.t. for 2 h before evaporation to
dryness. The crude products were purified by preparative RP-
HPLC and lyophilized. The purity of the compounds was
checked by analytical HPLC with a PDA detector spanning
from 210 to 310 nm. All compounds possessed purity above
95%, as determined by analytical HPLC-PDA at 214 and
254 nm.
Synthesis of 3-Amino-1-(4-(2-(dimethylamino)ethyl)piperazin-
1-yl)-2,2-bis(4-(trifluoromethyl)benzyl)propan-1-one (4b). The
synthesis was conducted according to GP1. 1H NMR (CD3OD):
δ 2.94-3.04 (9H, m); 3.12 (2H, d, J = 14.9 Hz); 3.19 (2H, br t);
3.50 (2H, br t); 3.55 (2H, d, J = 14.9 Hz); 7.44 (4H, d, J =
8.1 Hz); 7.67 (4H, d, J = 8.1 Hz). 13C NMR (CD3OD): δ 39.6;
43.9; 45.1; 51.9; 52.3; 53.5; 118.2; 124.7; 126.7; 130.7; 131.0;
132.0; 140.9; 172.5. HRMS-ESIþ: [M þ H]þ calcd: 545.2713
found: 545.2718, C27H35F6N4O.
Synthesis of 3-Amino-1-(4-(2-(dimethylamino)ethyl)piperazin-
1-yl)-2,2-bis(naphthalen-2-ylmethyl)propan-1-one (4c). The synthe-
sis was conducted according to GP1. 1H NMR (CD3OD): δ 2.98
(6H, s); 3.07 (2H, s); 3.25 (2H, br d); 3.44 (2H, t, J = 6.9 Hz); 3.61
(2H, t, J = 6.9 Hz); 3.67 (2H, d, J = 14.8 Hz); 7.36 (2H, dd J =
8.4 Hz, J = 1.1 Hz); 7.46-7.52 (4H, m); 7.72 (2H, s); 7.83-7.88
(6H, m). 13C NMR (CD3OD): δ 40.3; 43.9; 45.8; 52.0; 52.4; 52.9;
53.4; 127.3; 127.6; 128.6; 128.8; 129.1; 129.5; 130.1; 133.9; 134.0;
134.9; 173.6. HRMS-ESIþ: [M þ H]þ calcd: 509.3278 found:
509.3283, C33H41N4O.
Synthesis of 3-Amino-2,2-bis(4-tert-butylbenzyl)-N-(2-(dimethyl-
amino)ethyl)propanamide (6d). 1H NMR (CD3OD): δ 1.28 (18H,
s); 2.89-2.92 (8H); 2.97 (2H, s); 3.17 (2H, d, J = 14.2 Hz); 3.30
(2H, t, J = 6.1 Hz); 3.60 (2H, t, J = 6.0 Hz); 7.14 (4H, d J =
8.1 Hz); 7.36 (4H, d, J = 8.1 Hz). 13C NMR (CD3OD): δ 31.7
(q, J = 21 Hz); 35.2; 36.2 (t, J = 20 Hz); 40.5; 43.2; 43.8; 50.6;
58.4 (t, J = 20 Hz); 126.6; 131.2; 133.3; 151.4; 177.8. Multiplets
probably occurring due to insufficient carbon decoupling.
HRMS-ESIþ: [M þ H]þ calcd: 452.3639 found: 452.3650,
C29H46N3O.
(5) Collier, R. Drug development cost estimates hard to swallow.
CMAJ 2009, 180, 279–280.
(6) Boucher, H. W.; Corey, G. R. Epidemiology of methicillin-resis-
tant Staphylococcus aureus. Clin. Infect. Dis. 2008, 46 (Suppl 5),
S344–S349.
(7) King, M. D.; Humphrey, B. J.; Wang, Y. F.; Kourbatova, E. V.;
Ray, S. M.; Blumberg, H. M. Emergence of community-acquired
methicillin-resistant Staphylococcus aureus USA 300 clone as the
predominant cause of skin and soft-tissue infections. Ann. Intern.
Med. 2006, 144, 309–317.
(8) Moran, G. J.; Krishnadasan, A.; Gorwitz, R. J.; Fosheim, G. E.;
McDougal, L. K.; Carey, R. B.; Talan, D. A. Methicillin-resistant
S. aureus infections among patients in the emergency department.
N. Engl. J. Med. 2006, 355, 666–674.
(9) Taubes, G. The bacteria fight back. Science 2008, 321, 356–361.
(10) Karthikeyan, K. K.; Mark, A. T.; Timothy, R. W.; Jay, B.;
Fafhana, B.; Ravikumar, B.; Uma, C.; Michel, D.; Christian,
G. G.; Seema, I.; Padma, K.; Anil, V. K.; Sunil, M.; Shazad, M.;
Tabassum, N.; David, L. P.; Andrew, P.; Claire, P.; Rachel, P.;
Bhargavi, R.; Ujjwayini, R.; Jayanta, B. S.; Madhu, S.; Elizabeth,
S.; Mandayam, A. T.; Jane, T.; Supriya, U.; Marina, W.; William,
W.; David, M. L.; Neil, W. Emergence of a new antibiotic
resistance mechanism in India, Pakistan, and the UK: a molecular,
biological, and epidemiological study. Lancet Infect. Dis. 2010, 10,
597–602.
(11) Opar, A. Bad bugs need more drugs. Nat. Rev. Drug Discovery
2007, 6, 943–944.
(12) Todd, B. Beyond MRSA: VISA and VRSA: what will ward off
these pathogens in health care facilities? Am. J. Nurs. 2006, 106
28–30.
(13) Giuliani, A.; Pirri, G.; Nicoletto, S. F. Antimicrobial peptides: an
overview of a promising class of therapeutics. Cent. Eur. J. Biol.
2007, 2, 1–33.
(14) Wang, G.; Li, X.; Wang, Z. APD2: the updated antimicrobial
peptide database and its application in peptide design. Nucleic
Acids Res. 2009, 37, D933–937.
(15) Wang, G. The Antimicrobial Peptide Database. http://aps.unmc.edu/
AP/main.php (2010.09.28).
(16) Hancock, R. E. Peptide antibiotics. Lancet 1997, 349, 418–422.
(17) Yeaman, M. R.; Yount, N. Y. Mechanisms of antimicrobial
peptide action and resistance. Pharmacol Rev 2003, 55, 27–55.
(18) Hancock, R. E. R.; Sahl, H.-G. Antimicrobial and host-defense
peptides as new anti-infective therapeutic strategies. Nat. Biotech-
nol. 2006, 24, 1551–1557.
(19) Welling, M. M.; Hiemstra, P. S.; van den Barselaar, M. T.;
Paulusma-Annema, A.; Nibbering, P. H.; Pauwels, E. K.; Calame,
W. Antibacterial activity of human neutrophil defensins in experi-
mental infections in mice is accompanied by increased leukocyte
accumulation. J. Clin. Invest. 1998, 102, 1583–1590.
(20) Koczulla, A. R.; Bals, R. Antimicrobial peptides: current status
and therapeutic potential. Drugs 2003, 63, 389–406.
Synthesis of 3-Amino-N-(2-aminoethyl)-2,2-bis(4-tert-butyl-
benzyl)propanamide (7d). The synthesis was conducted accord
ing to GP1. 1H NMR (CD3OD): δ 1.31 (18H, s); 2.91 (2H, d,
J = 14.1 Hz); 2.96 (2H, s); 3.10 (2H, t, J = 6.2 HZ); 3.16 (2H,
d, J = 14.1); 3.51 (2H, t, J = 6.2 Hz); 7.15 (4H, d, J = 8.2 Hz);
7.38 (4H, d, J = 8.3 Hz). 13C NMR (CD3OD): δ 31.7; 35.3;
38.6; 40.6; 43.4; 50.5; 126.6; 131.2; 133.5; 151.5; 177.3.
HRMS-ESIþ: [M þ H]þ calcd: 424.3326 found: 424.3324,
(21) Brouwer, C. P.; Sarda-Mantel, L.; Meulemans, A.; Le Guludec, D.;
Welling, M. M. The use of technetium-99m radiolabeled human
antimicrobial peptides for infection specific imaging. Mini Rev.
Med. Chem. 2008, 8, 1039–1052.
(22) Hansen, T.; Alst, T.; Havelkova, M.; Strøm, M. B. Antimicrobial
C
27H42N3O.
activity of small β-peptidomimetics based on the pharmacophore

868 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 3
Hansen et al.
model of short cationic antimicrobial peptides. J. Med. Chem.
2010, 53, 595–606.
(23) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Experimental and computational approaches to estimate solubility
and permeability in drug discovery and development settings. Adv.
Drug Delivery Rev. 2001, 46, 3–26.
(24) Flaten, G. E.; Dhanikula, A. B.; Luthman, K.; Brandl, M. Drug
permeability across a phospholipid vesicle based barrier: a novel
approach for studying passive diffusion. Eur. J. Pharm. Sci. 2006,
27, 80–90.
of short cationic antimicrobial peptides by incorporation of argi-
nine analogues. Mol. Pharm. 2009, 6, 996–1005.
(38) Svenson, J.; Vergote, V.; Karstad, R.; Burvenich, C.; Svendsen,
J. S.; De Spiegeleer, B. Metabolic fate of lactoferricin-based anti-
microbial peptides: effect of truncation and incorporation of amino
acid analogs on the in vitro metabolic stability. J. Pharmacol. Exp.
Ther. 2010, 332, 1032–1039.
(39) Andersen, N. H.; Kier, B.; Shu, I.; Williams, V. Tryptophan
interactions that stabilize folding motifs: a guide to placement,
dynamics applications, and expectations for fold stabilization. In
European Peptide Symposium (EPS-31), Kobenhagen, 2010; p 465.
(40) Tew, G. N.; Scott, R. W.; Klein, M. L.; Degrado, W. F. De novo
design of antimicrobial polymers, foldamers, and small molecules:
from discovery to practical applications. Acc. Chem. Res. 2010, 43,
30–39.
(25) Cronin, J. R.; Yuen, G. U.; Pizzarello, S. Gas chromatographic-
mass spectral analysis of the five-carbon β-, γ-, and δ-amino
alkanoic acids. Anal. Biochem. 1982, 124, 139–149.
€
(26) Seebach, D.; Abele, S.; Sifferlen, T.; Hanggi, M.; Gruner, S.; Seiler,
P. Preparation and structure of β-peptides consisting of geminally
disubstituted β2,2- and β3,3-amino acids: a turn motif for β-pep-
tides. Helv. Chim. Acta 1998, 81, 2218–2243.
(41) Hagmann, W. K. The many roles for fluorine in medicinal chem-
istry. J. Med. Chem. 2008, 51, 4359–4369.
(27) Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis: A
Practical Approach; Oxford University Press: Oxford, 2000; p 346.
(28) Abele, S.; Seebach, D. Preparation of achiral and of enantiopure
geminally disubstituted β-amino acids for β-peptide synthesis. Eur.
J. Org. Chem. 2000, 2000, 1–15.
(29) Haug, B. E.; Stensen, W.; Svendsen, J. S. Application of the Suzuki-
Miyaura cross-coupling to increase antimicrobial potency gener-
ates promising novel antibacterials. Bioorg. Med. Chem. Lett. 2007,
17, 2361–2364.
(30) Flaten, G. E.; Luthman, K.; Vasskog, T.; Brandl, M. Drug perme-
ability across a phospholipid vesicle-based barrier 4. The effect of
tensides, co-solvents and pH changes on barrier integrity and on
drug permeability. Eur. J. Pharm. Sci. 2008, 34, 173–180.
(31) Hancock, R. E. W. Resistance mechanisms in Pseudomonas
aeruginosa and other nonfermentative Gram-negative bacteria.
Clin. Infect. Dis. 1998, 27, S93–S99.
(32) McMurry, L.; Petrucci, R. E., Jr.; Levy, S. B. Active efflux of
tetracycline encoded by four genetically different tetracycline
resistance determinants in Escherichia coli. Proc. Natl. Acad. Sci.
U. S. A. 1980, 77, 3974–3977.
(42) Bohm, H. J.; Banner, D.; Bendels, S.; Kansy, M.; Kuhn, B.; Muller,
K.; Obst-Sander, U.; Stahl, M. Fluorine in medicinal chemistry.
ChemBioChem 2004, 5, 637–643.
(43) Gimenez, D.; Andreu, C.; del Olmo, M.; Varea, T.; Diaz, D.;
Asensio, G. The introduction of fluorine atoms or trifluoromethyl
groups in short cationic peptides enhances their antimicrobial
activity. Bioorg. Med. Chem. 2006, 14, 6971–6978.
(44) Binkert, C.; Frigerio, M.; Jones, A.; Meyer, S.; Pesenti, C.; Prade,
L.; Viani, F.; Zanda, M. Replacement of isobutyl by trifluoro-
methyl in pepstatin A selectively affects inhibition of aspartic
proteinases. ChemBioChem 2006, 7, 181–186.
(45) Smart, B. E. Fluorine substituent effects (on bioactivity). J. Fluor-
ine Chem. 2001, 109, 3–11.
(46) Takano, M.; Hasegawa, R.; Fukuda, T.; Yumoto, R.; Nagai, J.;
Murakami, T. Interaction with P-glycoprotein and transport of
erythromycin, midazolam and ketoconazole in Caco-2 cells. Eur.
J. Pharmacol. 1998, 358, 289–294.
(47) Ebersold, H. R.; Cordier, J. L.; Luthy, P. Bacterial mesosomes:
method dependent artifacts. Arch. Microbiol. 1981, 130, 19–22.
(48) Friedrich, C. L.; Moyles, D.; Beveridge, T. J.; Hancock, R. E.
Antibacterial action of structurally diverse cationic peptides on
Gram-positive bacteria. Antimicrob. Agents Chemother. 2000, 44,
2086–2092.
(33) Cook, G. M.; Heard, D. H.; Seaman, G. V. Sialic acids and the
electrokinetic charge of the human erythrocyte. Nature 1961, 191,
44–47.
(34) Eylar, E. H.; Madoff, M. A.; Brody, O. V.; Oncley, J. L. The
contribution of sialic acid to the surface charge of the erythrocyte.
J. Biol. Chem. 1962, 237, 1992–2000.
(35) Godin, C.; Caprani, A. Interactions of erythrocytes with an
artificial wall: influence of the electrical surface charge. Eur.
Biophys. J. 1996, 25, 25–30.
(49) Shimoda, M.; Ohki, K.; Shimamoto, Y.; Kohashi, O. Morphology
of defensin-treated Staphylococcus aureus. Infect. Immun. 1995,
63, 2886–2891.
(50) Shai, Y.; Oren, Z. From “carpet” mechanism to de-novo designed
diastereomeric cell-selective antimicrobial peptides. Peptides 2001,
22, 1629–1641.
(36) Zasloff, M. M. Antimicrobial peptides of multicellular organisms.
(51) Flaten, G. E.; Bunjes, H.; Luthman, K.; Brandl, M. Drug perme-
ability across a phospholipid vesicle-based barrier 2. Characteriza-
tion of barrier structure, storage stability and stability towards pH
changes. Eur. J. Pharm. Sci. 2006, 28, 336–343.
Nature 2002, 415, 389–395.
(37) Svenson, J.; Karstad, R.; Flaten, G. E.; Brandsdal, B. O.; Brandl,
M.; Svendsen, J. S. Altered activity and physicochemical properties