Download amino acids. combinatorial libraries made up of

Identification of antimicrobial peptides by using
combinatorial libraries made up of unnatural
amino acids.
S E Blondelle, E Takahashi, P A Weber and R A Houghten
Antimicrob. Agents Chemother. 1994, 38(10):2280. DOI:
10.1128/AAC.38.10.2280.
These include:
CONTENT ALERTS
Receive: RSS Feeds, eTOCs, free email alerts (when new articles
cite this article), more»
Information about commercial reprint orders: http://journals.asm.org/site/misc/reprints.xhtml
To subscribe to to another ASM Journal go to: http://journals.asm.org/site/subscriptions/
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
Updated information and services can be found at:
http://aac.asm.org/content/38/10/2280
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Oct. 1994, p. 2280-2286
Vol. 38, No. 10
0066-4804/94/$04.00+0
Copyright C 1994, American Society for Microbiology
Identification of Antimicrobial Peptides by Using Combinatorial
Libraries Made Up of Unnatural Amino Acids
SYLVIE E. BLONDELLE,l* EMA TAKAHASHI,' PATRICIA A. WEBER,2
AND RICHARD A. HOUGHTEN" 2
Torrey Pines Institute for Molecular Studies' and Houghten Pharnaceutical Inc.,2
San Diego, California 92121
The use of water-soluble synthetic peptide combinatorial libraries permits the systematic examination of
tens to hundreds of millions of peptides in existing microdilution assays. In the present study, we prepared and
determined the antistaphylococcal activities of two new synthetic peptide combinatorial libraries (one
N-acetylated, the other not) composed of tetrapeptides having one position defined and the remaining three
positions made up of mixtures of L-, D-, and unnatural amino acids (a total of 58 different amino acids). These
libraries, when used in conjunction with an iterative selection process, allowed for the development of a series
of individually defined tetrapeptides with high levels of activity against Staphylococcus aureus. The activities of
the final individual peptides against two additional strains of gram-positive bacteria (methicillin-resistant S.
aureus and Streptococcus sanguis), a gram-negative bacterium (Escherichia coli), as well as the yeast Candida
albicans were also determined. Cell viability assays showed that the identified peptides are bacteriostatic
against both S.
aureus
and E. coli.
MATERIALS AND METHODS
Staphylococcus aureus continues to be a problematic pathobecause of the increasing occurrence of strains which are
resistant to commonly used antibiotics. Such resistant strains
lead to a constant need for new antistaphylococcal agents. In
recent years, peptides, either isolated from natural sources or
synthetically designed, have emerged as potent antimicrobial
agents (for a review, see reference 2). However, the process
involved in the development of such new antimicrobial peptides, i.e., their isolation and characterization and the synthesis
of large numbers of peptide analogs required for activity
optimization, is time-consuming and limited by the large
number of individual peptides which can be generated and
screened. Highly active antimicrobial peptides have recently
been identified through the use of soluble (i.e., non-supportbound), synthetic peptide combinatorial libraries (SPCLs [11,
12, 14, 15]) made up of L-amino acids. This approach permits
the systematic screening of hundreds of millions to billions of
peptides for the identification of optimal, biologically active
peptides (1, 7, 8, 11, 15, 18). Since these newly developed
antimicrobial peptides are composed entirely of L-amino acids
and are therefore susceptible to proteolytic breakdown and
rarely have activity when administered orally, their therapeutic
applicability is limited to topical or intravenous use. Antimicrobial peptides with D- and/or unnatural amino acids in their
sequences would be expected to enhance the stabilities of these
peptides toward proteolysis and, in turn, increase their duration of activity and applicability. In the present study, we
extended our search for new antimicrobial agents by determining the activities of two new SPCLs made up of L-, D- and
unnatural amino acids against S. aureus. The spectra of activity
of the final individual peptides, as well as their toxicities, are
described.
gen
Synthesis of SPCLs and individual peptides. The peptide
mixtures making up each of the two SPCLs, as well as those
used in the iterative process, were prepared by the process of
divide, couple, and recombine (9, 15, 17) in conjunction with
simultaneous multiple-peptide synthesis (10). All individual
peptides were prepared by simultaneous multiple-peptide synthesis. Peptide mixtures and individual peptides were cleaved
from the resulting resins by using low to high concentrations of
hydrogen fluoride (13, 20). Twenty-four peptides were individually cleaved simultaneously with HF by using a multiple-vessel
cleavage apparatus (Chiron Mimotopes Peptide Systems, San
Diego, Calif.) (13). Peptides were exhaustively extracted with
TABLE 1. Unnatural amino acids included in the two
tetramer SPCLs
Abbreviation
used in sequences
Bala ......
aABA ......
gABA ......
aAIB ......
eAca ......
7aHa ......
bAsp ......
gGlu ......
Cys[ACM] ......
elys........
oaFmoc-elys ......
MetO2 ......
Nle ......
Nve ......
Orn ......
dOrn ........
NO2F ........
Corresponding author. Mailing address: Torrey Pines Institute for
Molecular Studies, 3550 General Atomics Court, San Diego, CA
92121. Phone: (619) 455-3803. Fax: (619) 455-3804.
*
Hyp ......
Thiopro ......
2280
Protected amino acid used
during synthesis
Boc-,3-alanine
Boc-L-oa-amino butyric acid
Boc-L-'y-amino butyric acid
Boc-L-oa-amino isobutyric acid
Boc-L-e-amino caproic acid
Boc-7-amino heptanoic acid
Boc-L-aspartic acid (cs-Bzl)
Boc-L-glutamic acid (a-Bzl)
Boc-cysteine (acetamidomethyl)
N-e-Boc-N-cx-CBZ-L-lysine
N-E-Boc-N-a-Fmoc-L-lysine
Boc-L-methionine sulfone
Boc-L-norleucine
Boc-L-norvaline
N-a-Boc-N-8-CBZ-L-ornithine
N-8-Boc-N-oa-CBZ-L-ornithine
Boc-p-nitro-L-phenylalanine
Boc-hydroxyproline (Bzl)
Boc-L-thioproline
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
Received 19 April 1994/Returned for modification 6 June 1994/Accepted 22 July 1994
IDENTIFICATION OF NEW ANTIMICROBIAL PEPTIDES
VOL. 38, 1994
2281
I
80
60
% inhibition
A
20
0
AC DEFGHIKLMNPQRSTVWY
a cd e f h i k I m np q r s t vwy
12345678910111213141516171819
I
80
60
% inhibition
40
20
I
0
I
DEFGHIKLMNPQRSTVWY a cd e f h i k I m np q r s t vwy 12345678910111213141516171819
FIG. 1. Antistaphylococcal activities of the N-acetylated (A) and non-N-acetylated (B) SPCLs. The D-amino acids are represented by lowercase
letters; the unnatural amino acids are numbered as listed in Table 1. Each individual bar is segregated by the first amino acid in the peptide
mixtures (U) and represents the percent inhibition of S. aureus growth at a peptide mixture concentration of 2.5 mg/ml.
AC
water and lyophilized. The identities of the individual peptides
determined by laser desorption time-of-flight mass spectroscopy (MALDI-TOF mass spectrometer; Kratos Analytical,
Ramsey, N.J.). The peptide mixtures and individual peptides
were soluble in H20 at the concentrations tested.
Antimicrobial and antifungal assays. The strains S. aureus
ATCC 29213, methicillin-resistant S. aureus (MRSA) ATCC
33591, Streptococcus sanguis ATCC 10566, Escherichia coli
ATCC 25922, and Candida albicans ATCC 10231 were used in
were
the bioassays (American Type Culture Collection [ATCC],
Rockville, Md.). To initiate the exponential phase of bacterial
growth prior to the assay, a sample of bacteria grown overnight
at 37°C in Mueller-Hinton broth (MH [Becton Dickinson
Microbiology Systems] for E. coli and S. aureus) or brain heart
infusion broth (Becton Dickinson Microbiology Systems; for S.
sanguis) was reinoculated and incubated at 37°C. MRSA was
grown at 35°C in cation-adjusted MH broth (BBL, Cockeysville, Md.) in a similar manner. A final concentration of 1 x
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
40
2282
BLONDELLE ET AL.
RESULTS
Preparation of the two SPCLs. The two SPCLs used in the
present study, one N-acetylated, the other not, were composed
of tetrapeptides in which the first position was defined (U) and
three positions (Z) were an equimolar mixture of 56 different
L-, D-, and unnatural amino acids (19 L-, 18 D-, and 19 unnatural
amino acids; L- and D-cysteine were omitted from the mixture
positions but were included in the defined positions). The two
SPCLs can be represented by the general formula Ac-UZZZNH2 and UZZZ-NH2. The 19 unnatural amino acids used in
the study are listed in Table 1. These were chosen according to
their diversities and commercial availabilities. Each tetrapep-
TABLE 2. Antistaphylococcal activities of peptide mixtures derived
from iterative selection process
Position defined
Sequence
IC50 (Pxg/ml)
Second
(aFmoc-elys)WZZ-NH2
(aFmoc-Elys)cZZ-NH2
(otFmoc--lys)CZZ-NH2
(aFmoc-elys)wZZ-NH2
(aFmoc-Elys)yZZ-NH2
(aFmoc-elys)FZZ-NH2
(aFmoc-clys)(NO2F)ZZ-NH2
18
23
24
30
33
38
42
Third
(otFmoc-elys)WWZ-NH2
(aFmoc-elys)WFZ-NH2
(aFmoc-elys)WfZ-NH2
(otFmoc-Elys)W(Nle)Z-NH2
(aFmoc-slys)WLZ-NH2
(aFmoc-elys)WIZ-NH2
(aFmoc-rlys)WRZ-NH2
(otFmoc-elys)WrZ-NH2
(aFmoc-rlys)WpZ-NH2
(aFmoc-elys)W(Orn)Z-NH2
(oFmoc-elys)W(dOm)Z-NH2
(aFmoc-elys)WKZ-NH2
(aFmoc-Elys)WwZ-NH2
(aFmoc-Elys)WiZ-NH2
(aFmoc-Elys)W(aAIB)Z-NH2
(otFmoc-Elys)WYZ-NH2
(aFmoc-Elys)WsZ-NH2
(aFmoc-elys)W(Nve)Z-NH2
3
4
4
4
5
5
6
6
6
6
6
7
7
7
7
8
8
8
(tFmoc-Elys)cwZ-NH2
(otFmoc-elys)cfZ-NH2
7
(aFmoc-elys)c(NO2F)Z-NH2
(aFmoc-Elys)c(Nve)Z-NH2
(aFmoc-elys)c(Nle)Z-NH2
Fourth
(otFmoc-clys)WfR-NH2
(aFmoc-clys)Wfl-NH2
(aFmoc-elys)Wfw-NH2
(aFmoc-elys)Wfr-NH2
(aFmoc-elys)Wf(KCBZ)-NH2
(aFmoc-elys)Wf(Thiopro)-NH2
(aFmoc-Elys)WfL-NH2
(atFmoc-elys)WfP-NH2
(aFmoc-elys)Wfi-NH2
(aFmoc-elys)Wf(Orn)-NH2
(aFmoc-elys)Wf(dOrn)-NH2
(cLFmoc-elys)Wf(aAIB)-NH2
(aFmoc-elys)WKW-NH2
(otFmoc-elys)WKC-NH2
(aFmoc-elys)WYr-NH2
(aFmoc-elys)WY(aABA)-NH2
(aFmoc-rlys)cir-NH2
(aFmoc-elys)ciR-NH2
(otFmoc-elys)ciK-NH2
(etFmoc-elys)ci(aAIB)-NH2
(aFmoc-Elys)ci(Orn)-NH2
(aFmoc-elys)ci(dOm)-NH2
7
8
9
9
2
2
2
2
3
3
4
4
4
4
4
4
4
4
4
4
3
4
4
4
4
4
tide SPCL was therefore composed of 58 peptide mixtures,
each containing 175,616 (563) tetrapeptides, for a total of
10,185,728 (58 x 175,616) tetrapeptides per SPCL.
Activities of the two SPCLs against S. aureus. Each of the 58
peptide mixtures making up the two SPCLs was examined for
its ability to inhibit S. aureus growth in a microdilution assay.
The percent inhibition at a concentration of 2.5 mg of each
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
105 to 5 x 105 CFU/ml was used in all assays. Prior to the
antifungal assay, two colonies of newly grown C. albicans
culture were inoculated in 5 ml of phosphate-buffered saline
(PBS; 35 mM phosphate buffer, 0.15 M NaCl [pH 7.0]),
vortexed, and diluted 10-fold in yeast medium (YM) broth
(Difco Laboratories, Detroit, Mich), for an approximate final
concentration of 1 X 105 to 5 x 10 CFU/ml. The assays were
carried out in 96-well flat-bottom plates (Costar, Pleasanton,
Calif.). Bacterial or yeast suspension in 2x concentrated broth
was added to the peptide mixtures at concentrations that
varied from 2,500 to 1 ,ug/ml in H20 in serial twofold dilutions.
The plates were then incubated for 21 h at 37°C for the
bacteria or 48 h at 30°C for C. albicans. The relative percent
growth of the bacteria or yeast found for each peptide, as
determined by measuring the optical density at 620 nm with a
Titertek Multiskan Plus apparatus (Flow Laboratories,
McLean, Va.), was consistent in three separate determinations. The concentration necessary to inhibit 50% bacterial or
yeast growth (IC50) was then calculated by using sigmoidal
curve-fitting software (Graphpad; ISI, San Diego, Calif.). The
MICs were defined as the lowest concentrations of peptides at
which no change in the optical density at 620 nm occurred
between time zero and 21 or 48 h.
To evaluate whether the peptides were bacteriostatic or
bacteriocidal, killing assays were carried out as follows. Aliquots of cells (1 X 105 to 5 x 105 CFU/ml) of 200 ,ul in
suspension in MH or 2x concentrated YM broth were allowed
to bind to an Eppendorf tube for 15 min at room temperature.
An equal volume of peptide solution in H20 was then added,
and the tubes were incubated at 37°C for 1 to 24 h. Experiments with control tubes lacking peptide were carried out in an
identical manner. Following incubation, 100 ,lI of each solution was diluted in PBS (10-1 to 10-6 dilution) and plated onto
MH or YM agar plates.
The MICs in the presence of human serum were determined
by adding 50 ,ul of heat-inactivated (obtained following a
60-min treatment at 56°C), filtered human serum (Biowhittaker, Walkersville, Md.) or human serum albumin (Sigma, St.
Louis, Mo.) to each well and serially diluting the peptides in
the serum or albumin solution. The assays were then carried
out as described above.
Hemolytic assays. The peptides' hemolytic activities were
determined by using human erythrocytes (RBCs). The assays
were carried out in 96-well flat-bottom plates against a 0.25%
RBC suspension as described elsewhere (3). In brief, peptide
mixtures were added to the RBC solution at concentrations
that varied from 500 to 4 ,ug/ml in serial twofold dilutions.
Following a 1-h incubation at 37°C, the plates were centrifuged
at 2,800 rpm for 5 min. The supernatant was separated from
the pellet, and its optical density at 414 nm was measured. The
hemolytic dose to lyse 50% of RBCs (HD50) was calculated by
using the sigmoidal curve-fitting software Graphpad.
ANTIMICROB. AGENTS CHEMOTHER.
IDENTIFICATION OF NEW ANTIMICROBLAL PEPTIDES
VOL. 38, 1994
2283
TABLE 3. Spectra of activity of antistaphylococcal tetrapeptides
MIC range (,ug/ml) for
Tetrapeptide
MRSA
S. sanguis
E. coli
C. albicans
(oiFmoc-£lys)WfR-NH2
(aFmoc-£lys)Wfl-NH2
(aFmoc-elys)Wfw-NH2
(axFmoc-£lys)Wfr-NH2
(caFmoc-£lys)Wf(KCBZ)-NH2
(aFmoc-£lys)Wf(Thiopro)-NH2
(aFmoc-elys)WfL-NH2
(aFmoc- lys)WfP-NH2
(aFmoc-elys)Wfi-NH2
(aFmoc- 1ys)Wf(Orn)-NH2
(aFmoc-elys)Wf(dOrn)-NH2
4-8
3-4
3-4
(aFmoc--lys)Wf(aAIB)-NH2
48
8-16
4-8
5-8
8-16
8-16
8-16
8-16
8-16
>250
4-8
<2
4-8
8-16
3-5
31-62
125-250
4-8
4-8
8-16
8-16
4-8
3-4
3-4
3-4
5-8
5-8
5-8
5-8
8-16
5-8
5-8
8-16
31-62
62-125
62-125
16-31
31-62
31-62
31-62
31-62
>125
31-62
31-62
31-62
125-250
250-500
250-500
250-500
250-500
250-500
300-500
250-500
>500
250-500
125-250
125-250
(aFmoc-elys)WKW-NH2
(aFmoc-elys)WKC-NH2
5-8
5-8
>250
62-125
4-8
8-16
16-31
31-62
250-500
>500
(aFmoc-elys)WYr-NH2
5-8
5-8
62-125
8-16
5-8
8-16
31-62
62-125
250-500
250-500
4-8
8-16
8-16
8-16
16-32
>250
>250
>250
125-250
8-16
8-16
3-4
5-8
5-8
6-8
4-8
4-8
31-62
31-62
62-125
62-125
31-62
31-62
125-250
250-500
250-500
250-500
125-250
250-500
(aFmoc-elys)WY(aABA)-NH2
(aFmoc- Iys)cir-NH2
(aFmoc-elys)ciR-NH2
(aFmoc-elys)ciK-NH2
(aFmoc-elys)ci(aAIB)-NH2
(aFmoc-elys)ci(Orn)-NH2
8-16
(aFmoc-elys)ci(dOrn)-NH2
peptide mixture per ml is shown in Fig. 1. The IC50s were then
determined for those peptide mixtures which exhibited more
than 50% inhibition at 2.5 mg/ml. Overall, the non-N-acetylated peptide mixtures were found to be more active than
the corresponding N-acetylated ones. The peptide mixture
(aFmoc-elys)ZZZ-NH2 showed the greatest activity (IC50 =
44 ,ug/ml; MIC = 78 to 156 ,ug/ml). The next most active
peptide mixture, wZZZ-NH2, had an IC50 of 212 ,ug/ml and
an MIC ranging from 312 to 625 ,ug/ml [fivefold lower
than that of (ctFmoc-slys)ZZZ-NH2]. A single peptide mix-
ture, (aFmoc-elys)ZZZ-NH2, was chosen to be followed
through the iterative selection process.
Identification of individual antistaphylococcal peptides. The
identification of individual peptides from an SPCL mixture
involves ranking, selecting, and reducing the number of peptide sequences within the mixtures while sequentially defining
a single position at a time. By reducing the number of
individual peptides within a mixture, the actual concentration
of each individual peptide per milligram of each of the new
peptide mixtures increases 56-fold with each iteration. Thus,
o012
1010
D 108-
Elog
0106 ~
104
0
0.5
1
2
Incubation time (h)
by (otFmoc-elys)WfR-NH2. The killing of 105 CFU
24
of S. aureus per ml was evaluated at a peptide
FIG. 2. Killing curve of S. aureus
concentration of 15.6 jig/ml. The CFU per milliliter of the suspension with (filled squares) and without (open squares) peptide is plotted as a
function of the incubation time.
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
S. aureus
2284
BLONDELLE ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
TABLE 4. MICs of tetrapeptides for MRSA in presence of seruma
MIC (,ug/ml) in presence of serum
concn (%) of:
Tetrapeptide
(otFmoc-elys)Wfl-NH2
(otFmoc-elys)Wfi-NH2
(tFmoc-Elys)ci(Orn)-NH2
0
6.25
12.5
25
50
7.8
7.8
7.8
125
>250
250
125
>250
250
250
>250
250
>250
>250
250
a Heat-inactivated normal human serum pool.
1010
-
k
108
0
106
104
0
0.5
1
2
24
Incubation time (h)
Killing curve of E. coli by (aFmoc-elys)WfR-NH2. The killing of 105 CFU of E. coli per ml was evaluated at a peptide concentration
FIG. 3.
of 125 ug/ml. The CFU per milliliter of the suspension with (filled squares) and without (open squares) peptide is plotted as a function of the
incubation time.
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
we synthesized a new series of 58 peptide mixtures in which the
second position of (oaFmoc-slys)ZZZ-NH2 was defined. This
new series can be represented by the general formula (otFmocElys)UZZ-NH2. Each peptide mixture is then made up of 3,136
(56) individual tetrapeptides. The three most active peptide
mixtures were found to be (aFmoc-elys)WZZ-NH2, (aFmocElys)cZZ-NH2, and (otFmoc-elys)CZZ-NH2 (Table 2). The
iterative selection process was then repeated in order to define
the third positions of (otFmoc-e1ys)WZZ-NH2 and (otFmoc-Clys)
cZZ-NH2. The peptide mixtures of (xFmoc-elys)WUZ-NH2
and (otFmoc-elys)cUZ-NH2 then contained 56 individual tetrapeptides. Of the 58 peptide mixtures making up (aFmoc-elys)
WUZ-NH2 and of the 58 peptide mixtures making up (oxFmocslys)cUZ-NH2, 18 and 5 peptide mixtures, respectively, had
IC50s less than 10 ,ug/ml (Table 2) and MICs ranging from 5 to
18 ,ug/ml. The large number of active sequences indicates the
redundancy of the third position (i.e., low relative specificity of
each amino acid at this position) on the activities of the two
series of peptides being pursued. The follow-up of the iterative
process from these 23 active sequences would require the
synthesis of 1,334 individual tetrapeptides (23 x 58). In order
to decrease the number of peptides required, three main
criteria were considered: (i) level of antimicrobial activity, (ii)
toxicity as indicated by hemolytic activity, (iii) and differences
in the chemical functionalities of the residues making up the
sequences. Thus, in our first selection, we synthesized four
series of 58 individual tetrapeptides in which the last positions
of the active peptide mixtures (oxFmoc-e1ys)WKZ-NH2, (aoFmocelys)WYZ-NH2, (aFmoc-e1ys)WfZ-NH2, and (aFmoc-elys)
ciZ-NH2 were defined. The most active tetrapeptides identified from these series had IC50s ranging from 2 to 4 ,ug/ml
(Table 2) and MICs ranging from 3 to 16 ,ug/ml (Table 3). It is
noteworthy that none of the peptide mixtures and individual
peptides with antistaphylococcal activity had aspartic, P-aspartic, glutamic, or y-glutamic acid in their sequences.
Viable counts of S. aureus cells in the presence of (aFmocelys)WfR-NH2 were then carried out to determine if this
representative peptide was bacteriostatic or bactericidal. As
shown in Fig. 2, (ctFmoc-elys)WfR-NH2 had bacteriostatic
activity against S. aureus.
The tetrapeptides identified above were also tested against
an MRSA strain. Twelve of the identified peptides had similar
activities against the two strains of S. aureus, while decreases or
losses in activity were observed for the remaining nine peptides.
Activity in the presence of serum. The activities of the
tetrapeptides were also examined in the presence of various
concentrations in human serum (Table 4). This furthers the
measure of the therapeutic potentials of these peptides. The
activities of all of the peptides tested decreased in the presence
of 6.5% serum, with (aFmoc-elys)Wfl-NH2 having the greatest
activity. Human serum albumin was then used to investigate if
this decrease was due to enzyme inactivation or serum protein
binding. As shown in Table 5, higher levels of activity were
found solely in the presence of human serum albumin.
Spectra of activity of the antistaphylococcal peptides. The
antimicrobial activities of the peptides found to have antistaphylococcal activities were also assayed against a second
gram-positive bacterium (S. sanguis), a gram-negative bacterium (E. coli), and a fungal species (C. albicans). The MICs of
each of the tetrapeptides for these three microorganisms were
compared with the MICs for S. aureus. Overall, these peptides
were found to be more active against gram-positive bacteria.
Thus, as seen in Table 3, all of the peptides tested showed
VOL.
TABLE 5. MICs of (oaFmoc-elys)ci(Orn)-NH2 for S. aureus in
presence of serum or albumin
TABLE 6. Hemolytic activities of antistaphylococcal peptides
ID50
MIC (,ug/ml) in presence of:
Concn (%)
0
6.25
12.5
25
50
a
2285
IDENTIFICATION OF NEW ANTIMICROBLAL PEPTIDES
38, 1994
(4/ml)
Serum
Albumin
15.6
250
250
250
>250
15.6
31.2
31.2
62.5
62.5
Heat-inactivated normal human serum pool.
DISCUSSION
The widespread use of antibiotics has caused numerous
antibiotic-resistant strains to develop. For instance, MRSA is
acquiring resistance to antibiotics such as the fluoroquinolones
(for a review, see reference 5). With the emergence of vancomycin-resistant Enterococcus strains, there is concern that this
resistance will be passed to MRSA. Such new resistant strains
underscore the need for new antibiotics that are effective
against them. The rational design of new antimicrobial agents
requires years of effort to screen a large number of new
compounds. The introduction of soluble combinatorial peptide
and organic chemical libraries now enables the identification of
antimicrobial agents in a matter of weeks. A series of successful uses of such SPCLs has been reported for the identification
of biologically active peptide sequences in various assay systems such as enzyme-linked immunosorbent assays (15, 18),
radioreceptor assays (7), enzyme inhibitor assays (8), and
microdilution assays (11, 12, 14, 15). In particular, investigators
have identified a number of hexapeptides composed of Lamino acids which exhibit antimicrobial activities against S.
aureus (11, 12, 15) and E. coli (14). These new hexapeptides
have MICs that vary from 3 to 20 jig/ml. These activities are
greater or equivalent to those found for naturally occurring
peptides, which are generally greater in length (i.e., 10 residues
(aFmoc-elys)Wfl-NH2
(aFmoc-elys)Wfw-NH2
(aFmoc-elys)Wfr-NH2
(aFmoc-clys)Wf(KCBZ)-NH2
(aFmoc-elys)Wf(Thiopro)-NH2
(oaFmoc-elys)WfL-NH2
(aFmoc-EIys)WfP-NH2
(aFmoc-EIys)Wfi-NH2
(aFmoc-elys)Wf(Orn)-NH2
(aFmoc-elys)Wf(dOrn)-NH2
(aFmoc-elys)Wf(aAIB)-NH2
(aFmoc-elys)WKW-NH2
(aFmoc-elys)WKC-NH2
(aFmoc-clys)WYr-NH2
(aFmoc-elys)WY(aABA)-NH2
(aFmoc-elys)cir-NH2
(aFmoc-elys)ciR-NH2
(aFmoc-elys)ciK-NH2
(aFmoc-elys)ci(aAIB)-NH2
(aFmoc-elys)ci(Orn)-NH2
(aFmoc-clys)ci(dOrn)-NH2
16
>gml
17-46
18-26
100
100
100
100
10-22
100
0
49-76
69-98
13-20
3-4
0
0-1
3-4
0
6-7
0
0-1
0
0
5-7
0
32-37
40-43
80-82
56-77
10-12
8-9
31-66
114-143
100
27-51
7-8
48-66
50-77
89-94
100
4-5
79-173
70-71
28-53
100
77-114
77-119
54-79
122-130
197-208
36-63
42-54
52-55
45-52
294-320
40-58
>500
64-98
42-60
383-410
62-70
1-5
2-3
1-2
11-12
0
0-1
for the gramicidins, 23 for the magainins, 35 to 37 for the
cecropins, and 30 to 43 for the defensins; for a review, see
reference 2).
To expand the chemical diversity of the original SPCLs and
to decrease the risk of enzymatic degradation in in vivo
systems, D- and unnatural amino acids have been incorporated,
thus permitting the screening of a larger variety of peptides.
The generation of such libraries can be easily performed by
using the divide, couple, and recombine synthetic approach (9,
15, 17), which ensures the equimolarity of each peptide within
a mixture. The tetrapeptide format was chosen so that the
number of individual peptides in each peptide mixture was in
the same range as that in the initial hexapeptide library used
for the identification of new antimicrobial peptides (11, 12, 14,
15).
The individual tetrapeptides found in the current study to
exhibit antimicrobial activities against S. aureus were similar to
those described earlier, which were made up of L-amino acids
only. The greater antistaphylococcal activities observed for the
non-N-acetylated peptide mixtures compared with the activities of the equivalent N-acetylated peptide mixtures suggests
that hydrogen bonding or electrostatic interactions through the
protonated N terminus are involved in the antimicrobial
activities of these series of tetrapeptides. The lack of activity of
those tetrapeptides with either a glutamic acid or an aspartic
acid supports this hypothesis. The presence of negative charges
in such peptides may prevent the initial electrostatic interactions between the peptides and the negatively charged phospholipids that make up the bacterial cells. The higher level of
activity found for the (oFmoc-elys)ZZZ-NH2 mixture compared with that of the Ac-KZZZ-NH2 mixture, which differ
only by their protecting N-terminal groups, may be due to the
hydrophobicity of the Fmoc group relative to that of an acetyl
group. Longer Fmoc-protected peptides have been reported to
have potent antimicrobial activities (6).
The tetrapeptides studied here appear to exhibit their
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
similar activities against the two gram-positive bacteria. Lower
levels of activity, although similar within the series, were found
for all of the tetrapeptides against E. coli and C. albicans, with
the exception of (aFmoc-e1ys)Wfi-NH2 and (otFmoc-slys)
WKC-NH2. (otFmoc-e1ys)Wfi-NH2 was highly selective for
gram-positive bacteria, while (ctFmoc-e1ys)WKC-NH2 was active against bacteria but was inactive against C. albicans.
Kinetic studies of the activity of (aFmoc-e1ys)WfR-NH2
against E. coli (Fig. 3) were carried out in a manner similar to
that described above for S. aureus. Those studies indicated that
(aoFmoc-s1ys)WfR-NH2 exhibits bacteriostatic activity against
E. coli.
Hemolytic activity of the antistaphylococcal peptides. The
ability of the antistaphylococcal peptides to lyse eukaryotic
cells was determined by using human RBCs. The HD50 of each
peptide was calculated (Table 6). The HD5Os were generally
found to be greater than the MICs for the gram-positive
bacteria. At 16 j,g/ml, the concentration corresponding to the
MICs for gram-positive bacteria, the majority of the peptides
showed less than 5% hemolysis (Table 6). At a higher concentration (125 ,ug/ml), only (aFmoc-elys)Wfi-NH2 had no hemolytic activity, with (aFmoc-e1ys)WfL-NH2 and (aFmoc-elys)
Wf(aAIB)-NH2 being 10 to 20% hemolytic (Table 5). Interestingly, (otFmoc-s1ys)Wfi-NH2 was found to be the only
peptide of the series highly selective for gram-positive bacteria.
(oFmoc-elys)WfR-NH2
% Hemolysis at:
12 jml
2286
BLONDELLE ET AL.
8.
9.
10.
11.
12.
13.
ACKNOWLEDGMENTS
We thank Mary Buck and Adam Lucka for technical assistance and
Eileen Silva for editing the manuscript.
These studies were supported by Houghten Pharmaceuticals Inc.
1.
2.
3.
4.
5.
6.
7.
REFERENCES
Appel, J. R., C. Pinilla, and R. A. Houghten. 1992. Identification of
related peptides recognized by a monoclonal antibody using a
synthetic peptide combinatorial library. Immunomethods 1:1723.
Blondelle, S. E., and R A. Houghten. 1992. Progress in antimicrobial peptides. Annu. Rep. Med. Chem. 27:159-168.
Blondelle, S. E., L. R Simpkins, E. Perez-Paya, and R A.
Houghten. 1993. Influence of tryptophan residues on melittin's
hemolytic activity. Biochim. Biophys. Acta 1202:331-336.
Boman, H. G., I. Faye, G. H. Gudmundsson, J. Lee, and D.
Lidholm. 1991. Cell-free immunity in cecropia: a model system for
antibacterial proteins. Eur. J. Biochem. 201:23-31.
Brighty, K. E., W. Kohlbrenner, and P. R McGuirk 1993. Recent
developments in antibacterial resistance mechanisms. Annu. Rep.
Med. Chem. 28:141-150.
DeGrado, W. F. 1983. Solution phase synthesis of cecropin A 1-22
and potent analogues thereof using segments prepared on an
oxime solid phase support, p. 195-198. In J. H. Hruby and D. H.
Rich (ed.), Peptides: structure and function. Proceedings of the
Eighth American Peptide Symposium. Pierce Chemical Co.,
Rockford, Ill.
Dooley, C. T., N. N. Chung, P. W. Schiller, and R. A. Houghten.
1993. Acetalins: new opioid receptor antagonists determined
14.
15.
16.
17.
18.
19.
20.
through the use of synthetic peptide combinatorial libraries. Proc.
Natl. Acad. Sci. USA 90:10811-10815.
Eichler, J., and R A. Houghten. 1993. Identification of substrateanalog trypsin inhibitors through the screening of synthetic peptide combinatorial libraries. Biochemistry 32:11035-11041.
Furka, A., F. Sebestyen, M. Asgedom, and G. Dibo. 1991. General
method for rapid synthesis of multicomponent peptide mixtures.
Int. J. Pept. Protein Res. 37:487-493.
Houghten, R A. 1985. General method for the rapid solid-phase
synthesis of large numbers of peptides: specificity of antigenantibody interaction at the level of individual amino acids. Proc.
Natl. Acad. Sci. USA 82:5131-5135.
Houghten, R. A., J. R Appel, S. E. Blondelle, J. H. Cuervo, C. T.
Dooley, and C. Pinilla. 1992. The use of synthetic peptide combinatorial libraries for the identification of bioactive peptides.
BioTechniques 13:412-421.
Houghten, R A., S. E. Blondelle, and J. H. Cuervo. 1992. Development of new antimicrobial agents using a synthetic peptide
combinatorial library involving more than 34 million hexamers, p.
237-239. In R. Epton (ed.), Innovation and perspectives in solid
phase synthesis-peptides, polypeptides and oligonucleotides.
Solid Phase Conference Coordination, Oxford.
Houghten, R. A., M. K. Bray, S. T. De Graw, and C. J. Kirby. 1986.
Simplified procedure for carrying out simultaneous multiple hydrogen fluoride cleavages of protected peptide resins. Int. J. Pept.
Protein Res. 27:673-678.
Houghten, R A., K. T. Dhin, D. E. Burcin, and S. E. Blondelle.
1993. The systematic development of peptides having potent
antimicrobial activity against E. coli through the use of synthetic
peptide combinatorial libraries, p. 249-256. In R. H. Angeletti
(ed.), Techniques in protein chemistry IV. Academic Press, Inc.,
Orlando.
Houghten, R. A., C. Pinilla, S. E. Blondelle, J. R Appel, C. T.
Dooley, and J. H. Cuervo. 1991. Generation and use of synthetic
peptide combinatorial libraries for basic research and drug discovery. Nature (London) 354:84-86.
Jackson, M., H. H. Mantsch, and J. H. Spencer. 1992. Conformation of magainin-2 and related peptides in aqueous solution and
membrane environments probed by Fourier transform infrared
spectroscopy. Biochemistry 31:7289-7293.
Lam, K. S., S. E. Salmon, E. M. Hersh, V. J. Hruby, W. M.
Kazmierski, and R J. Knapp. 1991. A new type of synthetic
peptide library for identifying ligand-binding activity. Nature
(London) 354:82-84.
Pinilla, C., J. R. Appel, P. Blanc, and R A. Houghten. 1992. Rapid
identification of high affinity peptide ligands using positional
scanning synthetic peptide combinatorial libraries. BioTechniques
13:901-905.
Rana, F. R, E. A. Macias, C. M. Sultany, M. C. Modzrakowski,
and J. Blazyk. 1991. Interactions between magainin 2 and Salmonella typhimurium outer membranes: effect of lipopolysaccharide
structure. Biochemistry 30:5858-5866.
Tam, J. P., W. F. Heath, and R. B. Merrifield. 1983. SN2
deprotection of synthetic peptides with a low concentration of HF
in dimethyl sulfide: evidence and application in peptide synthesis.
J. Am. Chem. Soc. 105:6442-6455.
Downloaded from http://aac.asm.org/ on April 18, 2014 by PENN STATE UNIV
activities by inhibiting cell growth and development rather than
lysing the cells. This effect may be due to the shortness of these
peptides, which would prevent them from spanning the lipid
bilayer of a membrane and/or forming a channel through the
membranes, as has been proposed for longer antimicrobial
peptides such as magainins (16) and cecropins (4). Because of
their cationic character, these short peptides may bind to the
lipopolysaccharides of the bacterial cell membranes in a manner similar to that of the magainins (19), thereby increasing
cell permeability. This is supported by their higher levels of
activity against gram-positive bacteria, which have a less complex cell wall and lack an outer membrane, compared with
their activities against gram-negative bacteria. The results of
the experiments with serum suggest that the activities of the
tetrapeptides decrease because of degradation by heat-stable
proteases rather than binding to serum proteins. Further
studies are necessary to determine the cause of the decrease in
activity that was observed. The studies described above indicate that new potent antimicrobial compounds can be readily
and rapidly identified through the examination of extendeddiversity peptide libraries.
ANTIMICROB. AGENTS CHEMOTHER.