ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Apr. 2005, p. 1340–1345 Vol. 49, No. 4
0066-4804/05/$08.00
0 doi:10.1128/AAC.49.4.1340–1345.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Bactericidal and Antiendotoxic Properties of Short Cationic Peptides
Derived from a Snake Venom Lys49 Phospholipase A2
Carlos Santamarı́a,1,2 Silda Larios,3 Steve Quirós,1 Javier Pizarro-Cerda,4 Jean-Pierre Gorvel,5
Bruno Lomonte,1* and Edgardo Moreno2
Instituto Clodomiro Picado, Facultad de Microbiologı́a, Universidad de Costa Rica, San José,1 and Programa de Investigación
en Enfermedades Tropicales, Escuela de Medicina Veterinaria, Universidad Nacional, Heredia,2 Costa Rica; Departamento
de Microbiologı́a, Escuela de Medicina, Universidad Nacional Autónoma de Managua, León, Nicaragua3; and
Institut Pasteur, Paris,4 and Centre d’Immunologie de Marseille-Luminy,5 Marseille, France
Received 30 June 2004/Returned for modification 8 September 2004/Accepted 23 November 2004
The activities of short synthetic, nonhemolytic peptides derived from the C-terminal region of myotoxin II,
a catalytically inactive phospholipase A2 homologue present in the venom of the snake Bothrops asper, have
been shown to reproduce the bactericidal activity of the parent protein. They combine cationic and hydropho-
bic-aromatic amino acids, thus functionally resembling the antimicrobial peptides of innate defenses. This
study evaluated the antimicrobial and antiendotoxic properties of a 13-mer derivative peptide of the C-terminal
sequence from positions 115 to 129 of myotoxin II, named pEM-2. This peptide (KKWRWWLKALAKK) showed
bactericidal activity against both gram-positive and gram-negative bacteria. In comparison to previously
described peptide variants derived from myotoxin II, the toxicity of pEM-2 toward eukaryotic cells in culture
was significantly reduced, being similar to that of lactoferricin B but lower than that of polymyxin B. The all-D
enantiomer of pEM-2 [pEM-2 (D)] retained the same bactericidal potency of its L-enantiomeric counterpart,
but it showed an enhanced ability to counteract the lethal activity of an intraperitoneal lipopolysaccharide
challenge in mice, which correlated with a significant reduction of the serum tumor necrosis factor alpha levels
triggered by this endotoxin. Lethality induced by intraperitoneal infection of mice with Escherichia coli or Salmo-
nella enterica serovar Typhimurium was reduced by the administration of pEM-2 (D). These results demon-
strate that phospholipase A2-derived peptides may have the potential to counteract microbial infections and
encourage further evaluations of their actions in vivo.
Microbicidal cationic peptides have emerged as promising PLA2 protein was demonstrated for the first time and was
therapeutic alternatives to cope with the increasing rates of mapped to a specific membrane-damaging protein site (21, 29).
antibiotic resistance encountered worldwide (15). In higher ani- Although p115-129, which is derived from a snake venom
mals, a number of cationic peptides are localized at sites highly PLA2, does not have significant sequence homology with any
exposed to microorganisms, such as the skin; intestine; lungs; known cationic peptides of the innate defenses, it shows func-
and body secretions, such as tears, sweat, and saliva (16, 27). tional similarities with them. Previous studies showed that this
They are now considered key components of the innate im- peptide interacts with lipopolysaccharide (LPS) and lipid A
mune defenses, with physiologically relevant activities against from different gram-negative bacteria or with lipoteichoic acid
both gram-negative and gram-positive bacteria, protozoa, and from Staphylococcus aureus and relies on a membrane-perme-
fungi (39). More than 500 cationic peptides have been isolated abilizing mechanism to exert its bactericidal effects (29). LPS is
from mammals, amphibians, arthropods, plants, bacteria, and a complex molecule composed of a fatty acid (lipid A), an O-
even viruses (15). polysaccharide chain, and a core sugar, inserted into the outer
On the other hand, several types of mammalian secreted membrane of gram-negative bacteria. It plays a major role in
phospholipases A2 (PLA2s; EC 3.1.1.4) have been reported to the pathophysiology of septic shock and promotes a massive
exert potent bactericidal actions that are dependent on their activation of macrophages, endothelial cells, and fibroblasts,
enzymatic activities (5, 18, 19, 20, 30, 37). Notably, a group of which release diverse proinflammatory mediators, such as tu-
PLA2 homologues present in snake venoms, known as Lys49 mor necrosis factor alpha (TNF-), interleukin-1 (IL-1),
PLA2s (13, 24), have also been shown to be bactericidal, even IL-6, and nitric oxide (1, 14). Severely affected patients present
though they lack enzymatic activity (29, 34). In the case of with fever, shock, disseminated intravascular coagulation, mul-
myotoxin II, a Lys49 PLA2 isolated from the venom of the tiple-organ failure, and death (1, 11).
snake Bothrops asper, this bactericidal effect was reproduced by Variants of the original p115-129 sequence have recently
a synthetic 13-mer peptide corresponding to the sequence from been tested in vitro with the aim of optimizing its bactericidal
positions 115 to 129 located at its C-terminal loop (p115-129). activity while minimizing its toxicity toward eukaryotic cells.
Thus, an enzymatically independent bactericidal effect of a Such studies identified a peptide variant, named pEM-2, as an
interesting candidate for further evaluation of its antimicrobial
potential (32). Therefore, in the present investigation, the bac-
* Corresponding author. Mailing address: Instituto Clodomiro Pi-
cado, Facultad de Microbiologı́a, Universidad de Costa Rica, San José tericidal activities and antiendotoxic properties of synthetic
2060, Costa Rica. Phone: (506) 229-0344. Fax: (506) 292-0485. E-mail: pEM-2 in both its L- and D-enantiomeric forms were charac-
blomonte@cariari.ucr.ac.cr. terized in vitro and in vivo.
1340
VOL. 49, 2005 BACTERICIDAL PEPTIDES FROM A Lys49 PHOSPHOLIPASE A2 1341
(This work was performed in partial fulfillment of the re-
quirements for the M.Sc. degree of C. Santamarı́a at the Uni-
versity of Costa Rica.)
MATERIALS AND METHODS
Peptide synthesis. Peptides were derived from the sequence from positions
115 to 129 (KKYRYYLKPLCKK) of B. asper myotoxin II (29), with some
modifications. They were synthesized by N-(9 fluorenyl)-methoxycarbonyl chem-
istry, as described previously (8). Peptide p115-W3 (KKWRWWLKPLCKK)
corresponds to a peptide with a triple tyrosine-to-tryptophan substitution (23),
whereas in pEM-2 (KKWRWWLKALAKK) the proline and cysteine residues of
p115-W3 were each replaced by an alanine residue. The latter peptide was
synthesized in its L- and D-enantiomeric forms. All peptides were purified by
reverse-phase high-performance liquid chromatography and were obtained with
95% final purity. Their observed molecular masses corresponded to the ex-
pected theoretical values. Peptides were stored dry at 20°C and were dissolved
in sterile pyrogen-free saline solution (0.15 M NaCl) immediately before use.
Bacteria, LPS, and antimicrobials. Bacteria were maintained at 70°C in
Trypticase soy broth (TSB) containing 10% (vol/vol) glycerol. Frozen stocks were
thawed and cultured for 4 h (enterobacteria) or overnight (Brucella spp.) in TSB. FIG. 1. Bactericidal activities of synthetic myotoxin II peptide
Purified Escherichia coli O111:H4 LPS was kindly provided by A. Weintraub p115-129 (F) and modified peptide p115-W3 (■) against S. aureus
(Karolinska Institute, Stockholm, Sweden). Lyophilized LPS was diluted in en- ATCC 23923 (A), S. enterica serovar Typhimurium ATCC 14028 (B),
dotoxin-free saline and was dispersed by sonication for 30 s, immediately before B. abortus 2308 (C), and B. abortus 45/20 (D). Each point represents
use. Polymyxin B sulfate was obtained from Sigma (St. Louis, Mo.), whereas the mean  SD of triplicate viable counts, as described in Materials
lactoferricin B was kindly provided by W. Bellamy (Morinaga Dairy Co., Hi- and Methods.
gashihara, Japan).
Bactericidal assay. Bactericidal activities were assayed as described previously
(29). Briefly, log-phase bacteria were obtained from TSB cultures, and their tides (0,12.5, 25, 50, 100, and 200 g/ml) in 500 l of DMEM. After 10 min, LPS
concentration was adjusted to 4  106 CFU/ml in 0.01 M phosphate buffer (pH (100 ng/ml) was added to the wells, and the mixture was incubated for 3 h at
7.4) containing 1% peptone (buffer A) by reading the absorbance at 540 nm 37°C. The supernatants were collected and stored at 70°C until they were
(enterobacteria) or 420 nm (Brucella spp.). One hundred microliters of these assayed. TNF- levels in serum or the cell culture supernatants were determined
suspensions, which contained 4  105 CFU, was incubated with 10 l of various by an enzyme-linked immunosorbent assay (R&D Systems, Milwaukee, Wis.).
concentrations of peptides or buffer A alone for 20 min. Then, the remaining Samples were diluted 1:6 and assayed in duplicate wells in at least two indepen-
viable bacteria were counted by plating appropriate dilutions on Trypticase soy dent experiments. All experiments were performed with a minimum of 90%
or blood agar after 24 h of growth at 37°C. The minimal microbicidal concen- viability, as determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazo-
tration (MMC) was defined as the lowest peptide concentration that resulted in lium bromide (MTT) reduction assay (28).
the complete absence of bacterial growth after the bacteria were plated on agar. Endotoxic shock model. In order to corroborate the neutralization of LPS by
Cytolytic activity. The cytolytic effects of the peptides were determined with pEM-2 observed in vitro by the LAL assay, experiments were also carried out
the murine skeletal muscle myoblast cell line C2C12 (ATCC CRL-1772), as with CD-1 mice. All in vivo assays were approved by the Committee for Animal
described previously (22). C2C12 is the most sensitive target for Lys49 PLA Care and Use, University of Costa Rica. Endotoxic shock was induced by intra-2-
derived peptides and their parent proteins. In brief, cells were grown in 96-well peritoneal (i.p.) injection of E. coli O111:H4 LPS. Mice were presensitized with
plates in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with an i.p. injection of 15 mg of D-galactosamine 1 h before the LPS challenge (10).
penicillin-streptomycin solution (100 U of penicillin per ml, 100 g of strepto- Three groups of mice (body weight, 16 to 18 g) received an i.p. injection con-
mycin per ml) and 15% fetal calf serum (FCS) in an atmosphere with 7% CO taining (i) 10 g of LPS alone, (ii) 50 g of pEM-2 alone, or (iii) a mixture of LPS2
at 37°C. Immediately before the experiment, the growth medium was removed (10 g) and pEM-2 (50 g) preincubated for 20 min at 37°C. One hour after
and then various amounts of the peptides (12.5, 25, 50, and 100 g) were added injection, a blood sample was collected by retroorbital bleeding and serum was
to the assay medium (DMEM with 1% FCS). After 3 h at 37°C, the supernatants obtained in order to quantify TNF- levels, as described above. Lethality was
were collected and the activity of lactate dehydrogenase released from the recorded after 24, 48, and 72 h.
damaged cells was determined (Sigma kit 500). Controls for 0 and 100% toxicity Bacterial peritonitis model. Three groups of CD-1 mice (body weight, 21 to
8 6
values consisted of assay medium and 0.1% Triton X-100 in assay medium, 26 g) received an i.p. injection of (i) 5  10 E. coli or 2  10 Salmonella enterica
respectively. serovar Typhimurium CFU (obtained from log-phase TSB cultures) in 0.25 ml of
LAL assay. In vitro neutralization of LPS activity was assessed by the Limulus saline, (ii) the bacteria followed immediately by 100 g of the all-D enantiomer
amoebocyte lysate (LAL) chromogenic assay (BioWhittaker, Walkersville, Md.). of pEM-2 [pEM-2 (D)] in 0.25 ml of saline, or (iii) 100 g of pEM-2 (D) alone.
In brief, E. coli LPS was solubilized at a concentration of 5 endotoxin units/ml by Lethality was recorded daily over an observation period of 5 days.
sonication in endotoxin-free water. Peptides were prepared at concentrations of Statistical analyses. All endotoxin and TNF- experiments were performed in
12.5, 25, 50, 100, and 200 g/ml. Fifty microliters of endotoxin was mixed with triplicate. At least two independent assays were performed for all data collection.
50 l of the different peptide solutions in a 96-well plate, and the plate was in- Values are expressed as means  standard deviations (SDs) and were statistically
cubated at 37°C for 20 min. Then, 100 l of LAL reagent was added to each well, compared by Fisher’s exact test, an unpaired Student’s t test, or analysis of
and the plate was incubated for 10 min at 37°C. Finally, 100 l of substrate (the variance, as appropriate, by using InStat (version 2.04) GraphPad software.
colorless peptide acetyl-Ile-Glu-Ala-Arg-para-nitroaniline) was added, and the
plate was incubated at 37°C for 6 min. The reactions were stopped with 100 l
of 25% (vol/vol) acetic acid. The final absorbances at 405 nm were determined RESULTS
on a microplate reader (Labsystems). Assays were performed in triplicate wells, Microbicidal activity of p115-129 variants. As assessed by
and the results were expressed as the percent inhibition compared with the
control endotoxin activity values. the broth microdilution method, a triple Trp3Tyr substitution
LPS stimulation of RAW 264.7 cells. The murine macrophage cell line RAW in p115-129 enhanced its microbicidal action against both
264.7 (ATCC TIB-71) was grown in DMEM, as described above. Twenty-four gram-negative (S. enterica serovar Typhimurium) and gram-
hours before stimulation, the cells were washed with DMEM containing 15% positive (S. aureus) bacteria or some resistant bacteria, such as
FCS and detached by gentle scraping. Then, 5  105 cells were seeded into
24-well plates and were allowed to adhere overnight. Immediately before stim- Brucella abortus (Fig. 1). The greatest enhancement in the
ulation, the growth medium was removed, and the cells were washed three times activity of p115-W3 was observed against S. enterica serovar
in serum-free DMEM and then incubated with various concentrations of pep- Typhimurium, suggesting a stronger interaction with compo-
1342 SANTAMARÍA ET AL. ANTIMICROB. AGENTS CHEMOTHER.
TABLE 1. Microbicidal spectrum of pEM-2 (L)
against various bacteria
Bacterium MMC(g/ml)a
Pseudomonas aeruginosa ATCC 27853 ......................................... 1
Vibrio cholerae Ogawa IMS 124..................................................... 1
Shigella sonnei ATCC 25931 .......................................................... 1
Escherichia coli ATCC 25922......................................................... 5
Salmonella enterica serovar Typhimurium D984 ......................... 5
Enterococcus faecalis ATCC 29212................................................ 5
Staphylococcus aureus ATCC 23923.............................................. 5
Klebsiella pneumoniae ATCC 13883.............................................. 20
Brucella abortus 45.20...................................................................... 250
a FIG. 3. Cytotoxic activities of p115-W3 (F) and pEM-2 (Œ) onAssays were performed in triplicate. MMCs refer to the minimum peptide
concentrations that completely killed bacteria in liquid medium, as evidenced by C2C12 myoblasts. Polymyxin B (■) and lactoferricin B (}) were in-
the lack of subsequent growth after plating of appropriate dilutions in agar, as cluded for comparison. Each point represents the mean  SD of
described in Materials and Methods. triplicate cultures.
nents of the gram-negative bacterial membrane. Two addi- Comparative cytolytic activities of p115-W3 and pEM-2. It
tional changes were introduced in the sequence of p115-W3. was previously reported that even though p115-W3 is nonhe-
Its cysteine and proline were replaced by two alanine residues, molytic, it displays a marked toxic effect toward eukaryotic
respectively, in order to avoid oxidative dimerization (Cys) and cells in comparison to that of the original p115-129 of myotoxin
to facilitate synthesis (Pro). This peptide, named pEM-2, dis- II (23). As shown in Fig. 3, 150 g of p115-W3 per ml caused
played bactericidal activity against S. aureus and S. enterica 100% lysis of murine muscle cells, an effect more potent than
serovar Typhimurium (Table 1) similar to that of p115-W3. In that of polymyxin B. In contrast, pEM-2 exerted significantly
addition, pEM-2 showed strong action against other gram- weaker activity against these cells, reaching only half-maximal
negative bacteria, such as Pseudomonas aeruginosa, Shigella toxicity at the highest dose tested (600 g/ml). The minimal
sonnei, and Vibrio cholerae, with MMCs of 1 g/ml, and a lower concentration of pEM-2 required to induce complete lysis of
level of activity against Klebsiella pneumoniae, with an MMC of the cultures in vitro was estimated to be 	2.7 mg/ml (data not
20 g/ml (Table 1). In comparison, pEM-2 (D) was evaluated shown). The toxic activity of pEM-2 in this assay was similar to
for its bactericidal activity. As shown in Fig. 2, pEM-2 (D) that of lactoferricin B (Fig. 3), a bactericidal peptide derived
displayed microbicidal activity similar to that of its L-enantio- from mammalian lactoferrin.
meric counterpart against S. enterica serovar Typhimurium and In vitro endotoxin-neutralizing activities of pEM-2 enanti-
S. aureus. Both enantiomers of pEM-2 showed higher levels of omers. In order to evaluate the endotoxin-neutralizing ability
activity than lactoferricin B (Fig. 2). of pEM-2, both of its enantiomers were tested in the LAL
assay. pEM-2 (L) showed significant LPS-neutralizing activity,
similar to that of polymyxin B, whereas pEM-2 (D) also inhib-
ited endotoxin, although to a lower degree (Fig. 4). Control
peptides alone did not induce a response in the LAL assay.
Effects of pEM-2 enantiomers on TNF-
 secretion in vitro.
In order to assess the inhibition of LPS-induced TNF- release
by pEM-2, RAW 264.7 macrophages were challenged with 100
ng of LPS per ml, either alone or in combination with different
concentrations of peptides. Control experiments indicated that
at the concentrations used the peptides alone were not toxic
FIG. 2. Bactericidal activity of the L enantiomer (F) and the D
enantiomer (■) of synthetic peptide pEM-2 against S. enterica serovar FIG. 4. Inhibition of the endotoxin activity of LPS in vitro (by the
Typhimurium ATCC 14028 (A) and S. aureus ATCC 23923 (B). Poly- chromogenic LAL assay) by the L enanatiomer (F) and the D enan-
myxin B (Œ) and lactoferricin B (E) were included for comparison. tiomer (■) of pEM-2. Polymyxin B (Œ) was included as a reference.
Each point represents the mean  SD of triplicate viable counts. Each point represents the mean  SD of triplicate assays.
VOL. 49, 2005 BACTERICIDAL PEPTIDES FROM A Lys49 PHOSPHOLIPASE A2 1343
TABLE 3. Effect of pEM-2 (D) on the lethality induced by
peritoneal infection of gram-negative bacteriaa into mice
Lethality
Mouse treatment group (no. of deaths/
total no. of mice)b
Saline control ........................................................................... 0/4
pEM-2 (D) 
 saline ................................................................ 0/8
E. coli 
 saline......................................................................... 12/12
E. coli 
 pEM-2 (D)................................................................ 9/12c
FIG. 5. Inhibition of endotoxin-induced TNF- release in RAW S. enterica serovar Typhimurium 
 saline............................ 5/5
c
264.7 cells by L and D enantiomers of pEM-2. E. coli O111:H4 LPS S. enterica serovar Typhimurium 
 pEM-2 (D)................... 4/5
(100 ng/ml) was incubated at 37°C for 30 min alone or with the pep- a E. coli ATCC 25922 and S. enterica serovar Typhimurium ATCC 14028.
tides (100 g/ml) and was then added to the cell cultures, as described b All deaths occurred within the first 24 h.
in Materials and Methods. 
, statistically significant difference (P 
 c Nonsignificant (P  0.05) reduction, by Fisher’s exact test.
0.05) between LPS alone and pEM-2 (D).
for these cells, since there was no TNF- release, and the mice that were immediately treated with 100 g of pEM-2 (D)
viability (as judged by MTT-reducing activity) was nearly 100% by the i.p. route, 25% survival was recorded after 72 h of ob-
in all wells. Only a weak inhibition of LPS-induced cytokine servation. However, this reduction in lethality was not statisti-
release was observed with pEM-2 (D) in these in vitro experi- cally significant by Fisher’s extact test (P  0.05). Similar re-
ments. At the highest dose tested (100 g/ml), pEM-2 (D) re- sults were obtained in experiments in which peritonitis was
duced the level of TNF- secretion triggered by LPS by 25%, induced with S. enterica serovar Typhimurium, in which treat-
whereas virtually no inhibition was recorded for pEM-2 (L) ment of the mice with pEM-2 (D) rescued one of five animals
(Fig. 5). from lethality (P  0.05). No signs of acute toxicity were
Effects of pEM-2 enantiomers on LPS-induced lethality and observed in the control mice that received the peptide alone
serum TNF-
 levels. Injection of 10 g of E. coli LPS by the (Table 3).
i.p. route caused 100% lethality in galactosamine-sensitized
CD-1 mice within the first 24 h. As shown in Table 2, no re- DISCUSSION
duction in lethality was observed when the mice were chal- The bactericidal activity of myotoxin II against gram-positive
lenged with the same dose of LPS if it was previously incubated and gram-negative bacteria has been reproduced by a short
with pEM-2 (L). In these animals, the serum TNF- levels 1 h segment (positions 115 to 129) near its C-terminal loop (29)
after challenge were not statistically different from those in the which has characteristics that functionally resemble those of
control group receiving LPS alone. On the other hand, pEM-2 natural antimicrobial cationic peptides. The original sequence
(D) prevented the endotoxin-induced lethality by nearly 40% in from positions 115 to 129 was initially modified to enhance its
three independent experiments. Moreover, serum TNF- lev- hydrophobicity and ability to interact with membranes by re-
els in pEM-2 (D)-treated mice decreased by more than 40% placing its three tyrosine residues by tryptophan. This change
compared to those in the nontreated control group receiving (p115-W3) increased the bactericidal potency against E. coli by
LPS alone (Table 2). 1 order of magnitude, but toxicity toward eukaryotic cells was
Effect of pEM-2 (D) on E. coli or S. enterica serovar Typhi-
8 also drastically potentiated (23). Such an enhancement of themurium-induced lethal peritonitis. Injection of 5  10 E. coli bactericidal effect of p115-W3 was confirmed in the present
cells into the peritoneal cavities of CD-1 mice resulted in 100% study with a broader range of bacteria, including both gram-
lethality, and this always occurred within 24 h (Table 3). In positive and gram-negative species, as well as Brucella spp.,
which are known to be highly resistant to cationic peptides (9,
26). Nevertheless, due to the high degree of toxicity of p115-
TABLE 2. Effects of pEM-2 (L) or pEM-2 (D) on serum TNF- W3, a series of modifications were examined for their effects
levels and lethality induced by LPS challenge in mice on the bactericidal and cytolytic potencies of the peptides (32).
Mouse treatment Lethality Serum From such analyses, pEM-2, a modified synthetic peptide,
group (no. of deaths/ TNF- levela was selected for further characterization. The present re-total no. of mice) (pg/ml)
sults showed that pEM-2 displays lower cytolytic activity
pEM-2 (L) 0/6 
40 than p115-W3 toward myoblasts, with a toxicity comparable to
LPS control 10/10 10,412  5,286 that of lactoferricin B, a well-characterized cationic peptide
LPS 
 pEM-2 (L) 10/10 9,515  3,643
with a high degree of bactericidal action and a low-level cyto-
pEM-2 (D) 0/6 
40 toxic effect (3, 40). Interestingly, pEM-2 showed lower toxicity
LPS control 10/10 12,276  5,438 toward myoblasts than polymyxin B, a peptide restricted to
LPS 
 pEM-2 (D) 6/10b 7,241  4,189c topical therapeutic use (17).
a All deaths occurred within the first 24 h of observation. Despite its lower toxicity, pEM-2 conserved the microbicidal
b Statistically significant difference (P 
 0.01) from the results for the control characteristics of p115-W3, suggesting the importance of its
group receiving LPS alone, by Fisher’s exact test.
c Statistically significant difference (P 
 0.01) from the results for the control tryptophan residues in affecting prokaryotic membranes. In
group receiving LPS alone, by Student’s t test. agreement with this, it was previously observed that if the
1344 SANTAMARÍA ET AL. ANTIMICROB. AGENTS CHEMOTHER.
N-terminal hydrophobic segment of p115-W3, which contains tantly with a significant reduction in the levels of TNF- in
the tryptophan cluster, is replaced by the same cationic se- serum), no significant differences in lethality or in TNF- con-
quence of its C-terminal half (KKALAKLKALAKK), the pep- centrations were observed in pEM-2 (L)-treated mice. These
tide loses all bactericidal activity (32). results suggest that pEM-2 (L) might be rapidly eliminated by
The microbicidal spectrum of pEM-2 appears to be broad proteolysis or renal clearance, similar to other small cationic
and includes several gram-positive and gram-negative bacteria. L-peptides (2, 6, 35, 38), whereas the half-life of pEM-2 (D)
Thus, the actions of pEM-2 are similar to those of other cat- might be more prolonged in vivo.
ionic peptides such as lactoferricin (25) and P-113, a derivative In the present work, pEM-2 was preincubated with LPS
of histatin 5, which have lengths comparable to that of pEM-2 before i.p. injection in mice in order to assess its antiendotoxic
(31). ability independently of pharmacokinetic considerations. Also,
Although many investigations on the in vitro actions of cat- the potential of pEM-2 (D) to counteract a lethal bacterial
ionic peptides have been performed, there are still disadvan- peritonitis process was evaluated immediately after adminis-
tages when they are evaluated in vivo with animal models. tration of the bacteria. The modest protection from lethality
Their absorption is extremely poor, thus limiting oral admin- observed in both models, with preincubation and by indepen-
istration, and high local concentrations of these compounds dent administration, encourages the performance of further
are difficult to achieve. Two additional limitations include their studies with delayed administration of peptides to evaluate
relatively high levels of toxicity as well as their rapid degrada- their actual performance in established cases of endotoxemia
tion by natural proteases (16). The former disadvantage has or sepsis. It was recently shown that the results derived from
been approached by attempting to make modifications to their septic shock models based on LPS injection are very similar to
amino acid sequences in order to improve their selectivities those derived from other experimental models, such as cecal
toward prokaryotes (7). On the other hand, the introduction of ligation and puncture, which better reflect the conditions ob-
D-amino acids in their sequences circumvents the actions of served in patients with sepsis (12). This suggests that pEM-2
proteases, while it preserves their bactericidal properties (36). (D) could also be assessed in such models, for example. Further
As observed in this study, pEM-2 (D) retained the bactericidal preclinical investigations are needed to evaluate the potential
potency of pEM-2 (L) against S. aureus and S. enterica serovar of these peptides for the treatment of bacterial infections.
Typhimurium, suggesting that use of the D enantiomer would
be an interesting alternative for exploration of the possible ACKNOWLEDGMENTS
antiendotoxic properties of pEM-2 in vivo. We gratefully acknowledge the financial support for these studies by
Septic shock is the most common and, unfortunately, the NeTropica (grant 01-R-2003), the Lindbergh Foundation, the Ameri-
most dangerous complication of bacterial infection. LPS, a can Society for Microbiology-MIRCEN, the Florida, Ice & Farm of
complex macromolecule from the outer membranes of gram- Costa Rica, the CR-USA Foundation, and the University of Costa
Rica.
negative bacteria, is central to the pathophysiology of sepsis, Thanks are also due to P. Fourquet (CIML, Marseille, France) for
and lipid A is particularly central to the pathophysiology of peptide synthesis, A. Tarkowski and M. Bokarewa (University of Göte-
sepsis (1, 4, 11, 14, 33). The LAL assay was used to assess the borg, Göteborg, Sweden) for transcription factor expression analy-
endotoxin-neutralizing potentials of pEM-2 enantiomers. In ses, Y. Angulo for collaboration in cytotoxicity experiments, and E.
this test, lipid A triggers a coagulation event based on an Chaves-Olarte for critical reading of the manuscript.An INSERM patent (patent 04165A10) has been submitted.
enzymatic cascade. Therefore, peptides interacting with lipid A
of LPS can inhibit the LAL reaction. The results showed that REFERENCES
both pEM-2 (L) and pEM-2 (D) interact with LPS in vitro and 1. Adrie, C., and M. R. Pinsky. 2000. The inflammatory balance in human
inhibit its triggering of the LAL reaction. The inhibition ob- sepsis. Intensive Care Med. 26:364–375.2. Battafarano, R., P. Dahlberg, C. Ratz, J. Johnston, B. Gray, J. Haseman, K.
served with the L enantiomer was very similar to that caused by Mayo, and D. Dunn. 1995. Peptide derivatives of three distinct lipopolysac-
polymyxin B, suggesting an interesting antiendotoxin potential charide binding proteins inhibit lipopolysaccharide-induced tumor necrosis
of the pEM-2 sequence. Nevertheless, pEM-2 ( ) was unable factor-alpha secretion in vitro. Surgery 118:318–324.L 3. Bellamy, W., M. Takase, W. Wakabayashi, K. Kawase, and M. Tomita. 1992.
to prevent the LPS-triggered TNF- release from murine mac- Antibacterial spectrum of lactoferricin B, a potent bactericidal peptide de-
rophages in vitro, whereas pEM-2 (D) presented a significant, rived from the N-terminal region of bovine lactoferrin J. Appl. Bacteriol. 73:
472–479.
although partial, inhibition (	25%) in comparison to that of 4. Beutler, B., and A. Poltorak. 2001. Sepsis and evolution of the innate im-
control cultures stimulated with LPS alone. These results agree mune response. Crit. Care Med. 29:S2–S7.
with the observation that pEM-2 ( )- or pEM-2 ( )-treated 5. Buckland, A., and D. Wilton. 2000. The antibacterial properties of secretedL D phospholipases A2. Biochem. Biophys. Acta 1488:71–82.
macrophage cultures showed similar levels of the transcription 6. Dankesreiter, S., A. Hoess, J. Schneider, H. Wagner, and T. Miethke. 2000.
factors NF-B and AP-1 compared to those in the control Synthetic endotoxin-binding peptide block endotoxin-triggered TNF- pro-
cultures stimulated with LPS alone (unpublished data). Some duction by macrophages in vitro and in vivo and prevent endotoxin-mediatedtoxic shock. J. Immunol. 164:4804–4811.
studies have described similar findings, i.e., a weak inhibition 7. Fernández-López, S., H. Kim, E. Choi, M. Delgado, J. Granja, A. Khasanov,
of TNF- release in vitro, despite the stronger inhibition ob- K. Kraehenbuehl, G. Long, D. Weinberger, K. Wilcoxen, and M. R. Ghadiri.
2001. Antibacterial agents based on the cyclic D,L--peptide architecture.
served in vivo (35). Nature 412:452–455.
The differences in the inhibition of TNF- release caused by 8. Fields, G. B., and R. L. Noble. 1990. Solid phase peptide synthesis utilizing
the enantiomers might be explained by the possible degrada- 9-fluorenylmethoxycarbonyl amino acids. Int. J. Pept. Protein Res. 35:161–214.
tion of pEM-2 (L) in the macrophage cultures. Experiments 9. Freer, E., E. Moreno, I. Moriyón, J. Pizarro-Cerdá, A. Weintraub, and J. P.
with mice confirmed the better performance of pEM-2 (D) in Gorvel. 1996. Brucella-Salmonella lipopolysaccharide chimeras are less per-
comparison to that of pEM-2 ( ). Whereas the former de- meable to hydrophobic probes and more sensitive to cationic peptides andL EDTA than are their native Brucella sp. counterparts. J. Bacteriol. 178:
creased the rate of LPS-induced mortality by 40% (concomi- 5867–5876.
VOL. 49, 2005 BACTERICIDAL PEPTIDES FROM A Lys49 PHOSPHOLIPASE A2 1345
10. Galanos, C., M. A. Freudenberg, and W. Reutter. 1979. Galactosamine- 26. Martı́nez de Tejada, G., J. Pizarro-Cerdá, E. Moreno, and I. Moriyón. 1995.
induced sensitization to the lethal effects of endotoxin. Proc. Natl. Acad. Sci. The outer membranes of Brucella spp. are resistant to bactericidal cationic
USA 76:5939–5941. peptides. Infect. Immun. 63:3054–3061.
11. Gando, S., S. Nanzaki, and S. Sasaki. 1998. Activation of the extrinsic 27. Miyakawa, Y., P. Ratnakar, R. Gurara, M. Costello, O. Mathieu-Costello, R.
coagulation pathway in patients with severe sepsis and septic shock. Crit. Lehrer, and A. Catanzaro. 1996. In vitro activity of the antimicrobial pep-
Care Med. 26:2005–2009. tides human and rabbit defensins and porcine leukocyte protegrin against
12. Giacometti, A., O. Cirioni, R. Ghiselli, F. Mocchegiani, M. S. Del Prete, C. Mycobacterium tuberculosis. Infect. Immun. 64:926–932.
Viticchi, W. Kamysz, E. Lempicka, V. Sabba, and G. Scalise. 2002. Potential 28. Mosmann, T. 1983. Rapid colorimetric assay for cellular growth and survival:
therapeutic role of cationic peptides in three experimental models of septic application to proliferation and cytotoxicity assays. J. Immunol. Methods 65:
shock Antimicrob. Agents Chemother. 46:2132–2136. 55.
13. Gutiérrez, J. M., and B. Lomonte. 1997. Phospholipase A2 myotoxins from 29. Páramo, L., B. Lomonte, J. Pizarro-Cerdá, J. A. Bengoechea, J. P. Gorvel,
Bothrops snake venoms, p. 321–352 In R. M. Kini (ed.), Venom phospho- and E. Moreno. 1998. Bactericidal activity of Lys-49 and Asp-49 myotoxic
lipase A2 enzymes: structure, function, and mechanism. John Wiley & Sons, phospholipases A2 from Bothrops asper snake venom: synthetic Lys-49 myo-
Chichester, England. toxin II-(115–129) peptide identifies its bactericidal region. Eur. J. Biochem.
14. Hack, C. E., and S. Zeerleder. 2001. The endothelium in sepsis: source of and 253:452–461.
a target for inflammation. Crit. Care Med. 29:S21–S27. 30. Qu, X. D., and R. I. Lehrer. 1998. Secretory phospholipase A is the principal
15. Hancock, R. E., and D. Chapple. 1999. Peptide antibiotics. Antimicrob. 2bactericide for staphylococci and other gram-positive bacteria in human
Agents Chemother. 43:1317–1323. tears. Infect. Immun. 66:2791–2797.
16. Hancock, R. E., and M. Scott. 2001. The role of antimicrobial peptides in 31. Rothtein, D., P. Spacciapoli, L. Tran, T. Xu, D. Roberts, M. Dalla Serra, D.
animal defenses. Proc. Natl. Acad. Sci. USA 97:8856–8861. Buxton, F. Oppenheim, and P. Friden. 2001. Anticandida activity is retained
17. Hellman, J., and S. H. Warren. 1999. Antiendotoxin strategies. Infect. Dis. in P-113, a 12-amino-acid fragment of histatin 5. Antimicrob. Agents Che-
Clin. N. Am. 13:371–386. mother. 45:1367–1373.
18. Horwitz, S. S., L. Tan, X. Qu, Y. Cho, P. B. Eisenhawer, and R. I. Lehrer. 32. Santamarı́a, C., S. Larios, Y. Angulo, J. Pizarro, J. P. Gorvel, E. Moreno,
1995. Bactericidal properties of murine intestinal PLA2. J. Clin. Investig. 95: and B. Lomonte. Antimicrobial activity of myotoxic phospholipases A2 from603–610. crotalid snake venoms and synthetic peptide variants derived from their
19. Kini, R. M., and H. Evans. 1989. A common cytolytic region in myotoxins, C-terminal region. Toxicon, in press.
hemolysins, cardiotoxins and antibacterial peptides. Int. J. Protein Res. 34:
277–286. 33. Sasse, K. C., E. Nauenberg, and A. Long. 1995. Long-term survival after
20. Koduri, R. S., J. O. Gronroos, V. J. Laine, C. Le Calvez, G. Lambeau, T. J. intensive care units admission with sepsis. Crit. Care Med. 23:1040–1047.
Nevalainen, and M. H. Gelb. 2002. Bactericidal properties of human and 34. Soares, A. M., R. Guerra-Sá, C. Borja-Oliveira, V. M. Rodrigues, L. Ro-
murine groups I, II, V, X, and XII secreted phospholipases A . J. Biol. drigues-Simioni, V. Rodrigues, M. R. M. Fontes, B. Lomonte, J. M. Gutiér-2
Chem. 277:5849–5857. rez, and J. R. Giglio. 2000. Structural and functional characterization of
21. Lomonte, B., E. Moreno, A. Tarkowski, L. Å Hanson, and M. Maccarana. BnSP-7, a lysine-49 myotoxic phospholipase A2 homologue from Bothrops
1994. Neutralizing interaction between heparins and myotoxin II, a Lys49 neuwiedi pauloensis venom. Arch. Biochem. Biophys. 378:201–209.
phospholipase A2 from Bothrops asper snake venom. Identification of a 35. Uknis, M., K. Wasiluk, R. Acton, H. Klaerner, P. Dahlberg, E. Ilyina, J.
heparin-binding and cytolytic toxin region by the use of synthetic peptides Haseman, B. Gray, K. Mayo, and D. Dunn. 1997. Design of a potent novel
and molecular modeling. J. Biol. Chem. 269:29867–29873. endotoxin antagonist. Surgery 122:380–385.
22. Lomonte, B., Y. Angulo, S. Rufini, W. Cho, J. R. Giglio, M. Ohno, J. J. 36. Wade, D., A. Boman, B. Wahlin, C. M. Drain, D. Andreu, H. Boman, and B.
Daniele, P. Geoghegan, and J. M. Gutiérrez. 1999. Comparative studies of Merrifield. 1990. All D amino acid-containing channel-forming antibiotic
the cytolytic activity of myotoxic phospholipases A on mouse endothelial peptides. Proc. Natl. Acad. Sci. USA 87:4761–4765.2
(tEnd) and skeletal muscle (C2C12) cells in vitro. Toxicon 37:145–158. 37. Weinrauch, Y., P. Elsbach, L. M. Madsen, A. Foreman, and J. Weiss. 1996.
23. Lomonte, B., J. Pizarro-Cerdá, Y. Angulo, J. P. Gorvel, and E. Moreno. 1999. The potent anti-Staphylococcus aureus activity of a sterile rabbit inflamma-
Tyr3Trp-substituted peptide 115–129 of a Lys49 phospholipases A ex- tory fluid is due to a 14 kD phospholipase A2. J. Clin. Investig. 97:250–257.2
presses enhanced membrane-damaging activities and reproduces its in vivo 38. Weiss, C., K. Wasiluk, T. Kellogg, and D. Dunn. 2000. Bactericidal and
myotoxic effect. Biochim. Biophys. Acta 1461:19–26. endotoxin neutralizing activity of a peptide derived from Limulus antilipo-
24. Lomonte, B., Y. Angulo, and L. Calderón. 2003. An overview of lysine-49 polysaccharide factor. Surgery 128:339–344.
phospholipase A2 myotoxins from crotalid snake venoms and their structural 39. Zasloff, M. 2002. Antimicrobial peptides of multicellular organism. Nature
determinants of myotoxic action. Toxicon 42:885–901. 415:385–395.
25. Lupetti, A., K. Paulusma-Annema, M. Welling, S. Senesi, J. van Diessel, and 40. Zhang, G., D. Mann, and C. Tsai. 1999. Neutralization of endotoxin in vitro
P. Nibbering. 2000. Candidacidal activity of human lactoferrin peptides and in vivo by a human lactoferrin-derivated peptide. Infect. Immun. 67:
derived from the N terminus. Antimicrob. Agents Chemother. 44:3257–3263. 1353–1358.