ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, June 1999, p. 1429–1434 0066-4804/99/$04.0010 Copyright © 1999, American Society for Microbiology. All Rights Reserved.
Vol. 43, No. 6
Characterization of Novel Antimicrobial Peptoids BOB GOODSON,1 ANTON EHRHARDT,2 SIMON NG,1 JOHN NUSS,1 KIRK JOHNSON,1 MARTY GIEDLIN,1 RALPH YAMAMOTO,1 WALTER H. MOOS,1† ANKE KREBBER,1 MARTHA LADNER,1 MARY BETH GIACONA,1 CHARLES VITT,1 AND JILL WINTER1* Chiron Corporation, Emeryville, California 94608-2916,1 and Center for Research in Anti-Infectives and Biotechnology, Department of Medical Microbiology, Creighton University School of Medicine, Omaha, Nebraska 681782 Received 5 August 1998/Returned for modification 1 October 1998/Accepted 7 April 1999
Peptoids differ from peptides in that peptoids are composed of N-substituted rather than alpha-carbonsubstituted glycine units. In this paper we report the in vitro and in vivo antibacterial activities of several antibacterial peptoids discovered by screening combinatorial chemistry libraries for bacterial growth inhibition. In vitro, the peptoid CHIR29498 and some of its analogues were active in the range of 3 to 12 mg/ml against a panel of gram-positive and gram-negative bacteria which included isolates which were resistant to known antibiotics. Peptoid antimicrobial activity against Staphylococcus aureus was rapid, bactericidal, and independent of protein synthesis. b-Galactosidase and propidium iodide leakage assays indicated that the membrane is the most likely target of activity. Positional isomers of an active peptoid were also active, consistent with a mode of action, such as membrane disruption, that does not require a specific fit between the molecule and its target. In vivo, CHIR29498 protected S. aureus-infected mice in a simple infection model. This paper describes a series of N-substituted glycine trimers (herein generically referred to as peptoids) with antibacterial activity. Peptoids (7) have been used for a variety of applications. Recently, lipid-modified peptoid molecules, referred to as liptoids, have been used to deliver DNA to mammalian cells (9) and peptoid-peptide conjugates have been demonstrated to bind to SH3 domains with a higher affinity than peptides alone (19). Combinatorial chemistry peptoid libraries are part of the collection of small molecule libraries that the Chiron Corporation synthesizes and screens for pharmaceutical activity. In another report we describe the peptoid library that led us to individual antibacterial peptoid molecules (18). In this paper we characterize the biological properties of some of these new antibacterial molecules. The structure of peptoids distinguishes them from the seemingly similar peptides and other molecules known to have antibacterial properties. Peptoids have been shown to be resistant to proteolysis (14), and it is likely that the pharmacokinetics and pharmacodynamics of these molecules will be different than those of previously described molecules. It is possible to robotically make a great variety of peptoids, and this diversity offers the potential for designing desirable properties into compounds from this class. The aim of this study was to characterize the biological properties of a few select antimicrobial peptoids in order to evaluate them as future drug development candidates.
Media for in vitro assays. Mueller-Hinton broth (MHB) and Mueller-Hinton agar (MHA) (Oxoid, Unipath Ltd., Basingstoke, Hampshire, England), MuellerHinton cation-adjusted broth (MHCA) (Becton Dickinson, Cockeysville, Md.), and brain heart infusion (BHI) broth (Difco, Detroit, Mich.) were used for in vitro assays. Peptoid synthesis. The choice of the substituents (amine building blocks) for the CHIR29498 and analogue molecules was described previously (18), and the chemistry for peptoid synthesis is described in detail in Figliozzi et al. (7). The structures for the peptoids described within this study are shown in Fig. 1. In vitro susceptibility testing. Custom microbroth dilution trays (MIC2000, Dynatech Laboratories, Inc., Chantilly, Va.) were prepared with MHB. Chiron compounds (concentrated stocks of 20 mM in 100% dimethyl sulfoxide [DMSO]) were diluted into MHB and tested in a doubling dilution series (0.6 to 40 mM) (16). The final DMSO concentration was below 1% in these assays, and appropriate DMSO controls showed no effect on growth (data not shown). The comparison agents vancomycin and gentamicin were tested in the range of 1 to 128 mg/ml. Bacterial strains were incubated overnight at 37°C in MHB and inoculated into the custom trays (final inoculum, 104 CFU). Trays were incubated at 37°C for 18 to 24 h before results were read. Time-kill studies in the presence and absence of chloramphenicol. S. aureus SA4 was grown overnight in MHB to a culture density of approximately 109 CFU/ml. Cells were then diluted 1:100 in sterile physiologic saline solution, and a 1-ml aliquot of this suspension was then added to 99 ml of prewarmed MHB or MHB containing 8 mg of chloramphenicol/ml to achieve initial culture densities of approximately 5 3 105 CFU/ml. Cultures were incubated at 37°C with agitation, and peptoid antimicrobials were added 5 min after inoculation. Sampling aliquots were removed from all cultures immediately after peptoid addition (t 5 0 h) and then again at 30 min and 1, 2, 4, 6, and 24 h after peptoid addition. Aliquots containing chloramphenicol were treated for 5 min at 37°C with chloramphenicol acetyltransferase in the presence of acetyl coenzyme A to inactivate the drug and prevent carryover. All other aliquots were treated in parallel by addition of an equal amount of sterile MHB. Treated samples were then serially diluted in sterile saline and plated in duplicate by adding 1-ml aliquots to molten MHA. Solidified plates were then incubated at 37°C for a minimum of 18 h before viable counts were determined. Data from duplicate plate counts were averaged, and the resulting values were plotted on a log scale against time. The lower limit of detection in this assay was approximately 1 CFU/ml. Leakage assays. E. coli ML-35 (lacI lacZ1 lacY) was utilized for an inner membrane leakage assay (21). Logarithmic-phase bacteria (6 3 106 CFU) were incubated in 0.6 ml of a 10 mM sodium phosphate buffer (pH 7.5) containing 100 mM NaCl and 1.5 mM o-nitrophenyl-b-D-galactopyranoside (ONPG) at 37°C. At time zero, defined amounts of test compounds were added, and the conversion of substrate to product was measured by spectrophotometric readings of absorbance at 405 nm over time. Sonicated ML-35 cells were used as a positive control for total lysis and to demonstrate that the test compounds did not stimulate or inhibit the b-galactosidase activity. Gramicidin S (GS; Sigma) was included in the assay as a positive control for drug-induced leakage (11). Flow cytometry. The flow cytometric determination of bacterial fluorescence was carried out according to Mason et al. (12, 13). Propidium iodide (PI; Molecular Probes, Inc.) was diluted from a 1-mg/ml stock solution to 100 mg/ml in MHCA medium. S. aureus ATCC 25923 was grown in MHCA broth and diluted
MATERIALS AND METHODS Bacterial strains. The strains obtained from the American Type Culture Collection (ATCC, Rockville, Md.) included Staphylococcus aureus strains (ATCC 25923, 29213, 13709) Streptococcus pneumoniae (ATCC 49619), Enterococcus faecalis (ATCC 29212), Enterococcus faecium (ATCC 35667), Escherichia coli (ATCC 25922, ATCC 43827 ML-35), and Pseudomonas aeruginosa (ATCC 27853). All other strains listed in the tables, figures, or text are from the strain collection of the Center for Research in Anti-Infectives and Biotechnology, Creighton University.
* Corresponding author. Mailing address: Chiron Corporation, 4560 Horton St., Emeryville, CA 94608-2916. Phone: (510) 923-3642. Fax: (510) 923-4115. E-mail:
[email protected]. † Present address: Mito Kor, San Diego, CA 92121. 1429
1430
GOODSON ET AL.
ANTIMICROB. AGENTS CHEMOTHER. Female CD-1 mice, weighing 25 g, were obtained from Charles River Laboratories and were quarantined for 1 week. Selection of doses for novel small molecule efficacy testing was based on the 50% lethal dose (LD50) from an initial acute toxicity study of three to five naive animals per group injected intraperitoneally (i.p.) with increasing concentrations of test compound. The apparent LD50 from this small trial was 60 mg/kg of body weight. The test compound was formulated in a 5% 2-hydroxy-propyl-b-cyclodextrin (HPbCD) solution (5% HPbCD, 20 mM sodium citrate, 85.6 mM NaCl, pH 6.0) (HPbCD solution). The high dose selected for evaluation of in vivo antibacterial efficacy was twice the apparent LD50. For the infection studies, S. aureus was prepared as described above and diluted 25-fold into fresh BHI, and 0.5 ml (108 CFU) was injected i.p. into mice. Mice were subsequently injected i.p. with 0.2 ml of either saline, HBbCD solution (drug-free control), antibacterial candidate in HBbCD solution, or ampicillin (0.5 mg/kg). Animals were housed 5 per cage, and each treatment group consisted of 5 to 10 mice in each experiment. The timing for administration of test compound varied from t 5 0 (immediately after infection) to approximately 2 h following infection and is indicated in Results. Animals were monitored at 4, 14, 20, 24, 36, 48, and 120 h following infection. S. aureusinfected control animals (injected with saline or HPbCD solution in place of antibacterial candidates) died within 24 h of bacterial infection. Infected animals treated with HPbCD solution only did not differ from infected saline control animals in terms of morbidity or mortality. Survival was defined as animals living 5 days and was obvious by 36 to 48 h as animals rarely succumbed between 48 and 120 h. Surviving animals that were observed for several weeks did not show any signs of abnormality as a result of the treatment.
RESULTS
FIG. 1. Structure of DHAA and the peptoids. CHIR26240 is a dimer peptoid, and the other peptoids shown are trimers. CHIR32133 is one of the series of six positional isomers shown in Table 1. The bold A, B, and C near the substituents of CHIR32133 are used to define substituent positions on the peptoid scaffold. By using this molecule to define the positions, the substituents in the positional isomers are thus described as e.g., BCA. The structures for the six positional isomers of CHIR32133 described in Table 1 are 32133 (shown above), ABC; 33090, ACB; 33091, BAC; 33092, BCA; 33670, CAB; and 33671, CBA. in MHCA to a final concentration of 108 CFU/ml. One milliliter of culture was incubated with the appropriate drug for 30 min or 2 h. The cultures were then centrifuged for 10 min at 10,000 3 g, the supernatant was aspirated, and the bacterial pellet was then resuspended in 1 ml of fresh MHCA broth. Two hundred microliters of the bacterial suspension was incubated in the dark for 5 min at room temperature with 0.8 ml of MHCA plus the PI working solutions. The stained S. aureus cells were analyzed on a flow cytometer (Coulter XLMCL) with System II software (Coulter Corporation). Polysciences Inc. 0.5-, 1.0-, and 2.0-mm-diameter beads were used to calibrate forward-scatter and sidescatter parameters. Medium or 80% ethanol-treated S. aureus was used to set negative and positive gates. Twenty thousand events were collected for each sample, and the results are expressed in percentages of positive events. Hemolysis assay. Blood from female BALB/c mice or male humans was drawn with heparin as the anticoagulant. The blood was processed and assayed (20) immediately. The erythrocytes (RBC) were spun at 1,000 3 g, washed three times with 10 volumes of phosphate-buffered saline (PBS), and then resuspended in PBS. Washed RBC were treated with diluted test compound or control fluid. Triton X-100 (0.1%) was used for positive-control hemolysis values, and DMSO was used as the negative control since all tested compounds were dissolved in DMSO. The final DMSO concentration for each test sample was 1%. The samples were incubated in Costar 96-well round-bottomed microtiter plates at 37°C for 60 min. Plates were centrifuged at 1,000 3 g for 5 min, and 100 ml of supernatant was transferred to deep-well 96-well plates and diluted with 900 ml of PBS. A 200-ml aliquot of the diluted supernatant was transferred to a Corning 96-well flat-bottomed plate and read in a spectrophotometer at 560 nm to evaluate heme release. Human blood was tested due to our interest in peptoids as future therapeutic molecules, and mouse blood was tested to determine whether hemolysis would be a problem in the animal model. Murine infection assay. Master stocks of S. aureus (Smith, ATCC 13709) were prepared by growth to log phase in BHI broth and then frozen in 0.5-ml aliquots at 280°C in 10% glycerol. On the morning of the experiment, 0.25 ml of thawed bacterial stock was inoculated into 20 ml of BHI and grown at 37°C for approximately 7 h until a density of ;5 3 109 CFU/ml was achieved as measured by absorption at 600 nm in relation to a growth curve plotted under identical conditions. Plate counts were made from stock dilutions to confirm the exact CFU per milliliter used in the experiment, and values ranged from 0.75 3 108 to 2.5 3 108 CFU/mouse.
Activity profiles. MIC data for dehydroabietylamine (DHAA) and the peptoid compounds are shown in Table 1. MICs for these compounds are presented as micromolar values to allow for direct comparison of the candidate drugs based on the number of molecules delivered rather than the weight of drug delivered. Peptoid trimers consist of a peptoid backbone (polyglycine) with three substituted nitrogens (Fig. 1). Likewise, dimer peptoids are defined as a peptoid backbone with two substituted nitrogens. DHAA makes up almost a third of the weight of the trimer peptoids described in this paper (Fig. 1). Totarol, a natural product that is similar to DHAA (totarol is an acid whereas DHAA is an amine) has antistaphylococcal activity (8), and thus we tested DHAA alone for activity. As shown in Table 1, DHAA inhibited the majority of the tested gram-positive strains at 40 mM (11.4 mg/ml) concentrations, while none of the gram-negative strains were inhibited at this level. Recent independent testing indicated that the gram-negative test strains used in this study were not inhibited by DHAA at a concentration of 100 mM. The differences in activity between DHAA and CHIR29498 against Enterococcus sp. strains as well as the gram-negative strains tested indicated that DHAA was more efficacious as a peptoid than as a free amine. In the interest of finding more potent molecules that were smaller than the peptoid trimer, several dimer compounds incorporating DHAA were synthesized and evaluated. CHIR26240 is shown as an representative example of the dimers (Fig. 1). CHIR26240 demonstrated generally enhanced activity (twofold) compared to that of DHAA against grampositive organisms but remained inactive against the gramnegative strains at the concentrations tested (Table 1). The trimer peptoids were thus further characterized. Against the gram-positive strains tested, the activities of CHIR29498, CHIR29496, and CHIR32133 (Fig. 1) were similar to each other and generally within one twofold dilution of the dimer (Table 1). Against gram-negative strains on the other hand, these trimers were more than twofold more active than either the dimer or DHAA itself (although Table 1 shows values up to 40 mM, DHAA and the dimer were tested at concentrations up to 100 mM, at which level they remained inactive against the gram-negative strains in the profile). To assess whether potency could be influenced by the rearrangement of substituents within a peptoid molecule, all posi-
VOL. 43, 1999
NOVEL ANTIMICROBIAL PEPTOIDS
1431
TABLE 1. Activities of DHAA and related peptoid compounds against gram-positive and gram-negative bacteriaa Strain (phenotype)b
S. aureus ATCC 29213 S. aureus ATCC 25923 S. aureus 188 (Mr Qr) S. aureus 244 (Mr) S. epidermidis CNS25 S. epidermidis CNS112 (Mr Qr) S. haemolyticus CNS66 (Mr Qr) Streptococcus pneumoniae ATCC 49619 Streptococcus pneumoniae BT2 (Pr) Enterococcus faecalis ATCC 29212 Enterococcus faecalis 141 (VanB Qr) Enterococcus faecium ATCC 35667 Enterococcus faecium 129 (VanA) E. coli ATCC 25922 E. coli Misc 239 (TEM-10) Klebsiella pneumoniae ThaiK21 (SHV-5 Qr) Enterobacter cloacae 122 (Qr) Serratia marcescens 69 Pseudomonas aeruginosa ATCC 27853 Pseudomonas aeruginosa 197 (Qr)
MIC (mM) of:
MIC (mg/ml) of:
DHAA
26240
29498
29496
32133
33090
33091
33092
33670
33671
Gentamicin
Vancomycin
20 20 20 20 20 20 40 20 40 40 .40 .40 40 .40 .40 .40 .40 .40 .40 .40
10 10 20 10 20 10 20 20 20 20 20 20 20 .40 .40 .40 .40 .40 .40 .40
20 20 10 10 10 10 10 10 10 10 10 10 5 20 20 20 20 40 20 40
20 20 10 10 10 10 10 10 10 10 5 5 20 10 20 40 .40 20 20 20
10 10 10 10 10 10 20 20 20 20 20 10 10 40 20 40 20 40 40 40
10 20 20 10 10 10 20 20 20 20 20 20 20 40 20 40 40 40 40 40
10 10 10 10 10 10 10 20 20 10 10 10 10 .40 .40 .40 .40 .40 .40 .40
10 10 10 10 10 10 20 20 20 20 20 20 10 .40 .40 .40 .40 .40 40 .40
5 5 5 5 5 5 5 10 10 5 5 5 5 10 10 20 10 20 20 20
10 20 10 10 10 10 20 20 20 20 20 20 20 40 20 40 40 40 40 .40
#1 4 128 8 8 32 64 2 8 32 .128 4 8 #1 #1 16 #1 32 2 .128
#1 #1 #1 #1 2 4 2 #1 #1 2 128 #1 .128 .128 .128 .128 .128 .128 .128 .128
a MICs are listed as micromolar values for DHAA and peptoid compounds to allow comparison of the same number of molecules per assay. For conversions, 1 mM 5 mol wt/1,000. Molecular weights of the compounds are as follows: DHAA, 285; CHIR26240, 525; CHIR32133 and isomers, 623; and CHIR29498, 629. MICs for the comparison agents vancomycin and gentamicin are expressed in micrograms per milliliter, as is standard for these agents. b Mr, methicillin resistant; Qr, quinolone resistant; Pr, penicillin resistant; VanA or VanB, vancomycin resistance phenotypes.
tional isomers of the trimer CHIR32133 (Fig. 1) were also synthesized. These rearrangements can be represented by the letter patterns ABC, ACB, BAC, BCA, CAB, and CBA, where A, B, and C represent the substituents that extend from the nitrogens of the peptoid backbone and the patterns correspond to the substituent orders found in CHIR32133, CHIR33090, CHIR33091, CHIR33092, CHIR33670, and CHIR33071, respectively. Differences in activity were observed among these isomers (Table 1). CHIR33670 was the most potent peptoid tested in this study, with MICs generally two- to fourfold below those of the other positional isomers against all tested bacterial strains (3.115 to 12.46 mg/ml). Conversely, CHIR33091 and CHIR33092 showed reduced activities against gram-negative strains. Despite these differences, the generally similar grampositive activities of the six positional isomers may be consistent with a nonspecific target mode of action for these compounds and led us to investigate the membrane as the potential target for peptoidal antimicrobial activity. Killing kinetics. Although CHIR29498 and CHIR29496 differ at the C terminus (Fig. 1), both peptoids demonstrated very similar growth inhibition profiles against the strains tested (Table 1), and both demonstrated rapid killing in assays in which treated S. aureus were plated minutes after treatment (data not shown). CHIR29496 was further analyzed by a comparison of the killing kinetics in the presence and absence of chloramphenicol (Fig. 2). Peptoid killing activity was not eliminated by protein synthesis inhibition. Incubation of SA4 cells with 8 mg of chloramphenicol/ml was used to produce bacteriostasis through reversible blockage of protein synthesis. Since viable counts from the initial (t 5 0 h) aliquots from peptoid-containing cultures were substantially below those of the growth control culture, all graph traces were plotted using the t 5 0 h growth control count as the initial point. This can be justified in that all cultures were inoculated with equal amounts of a common culture within a few seconds of each other. The lower counts in the peptoid-containing cultures are therefore the result of 5 min of peptoid activity on the culture (the time
required for chloramphenicol acetyltransferase inactivation of chloramphenicol prior to plating in drug-free agar). Loss of cell viability was extremely rapid (.100-fold within 5 min), and the curves probably do not accurately represent the slope of the initial killing due to the sampling protocol which specified a second sample at t 5 0.5 h, by which time viability had already reached counting levels that were unreliable. The peptoid was used at a concentration of 10 mM (6.29 mg/ml), which was twice the MIC for this organism (SA4). Thus the primary bactericidal mechanism of action of the peptoid on S. aureus appears to be protein synthesis independent. As shown in Fig. 2, killing is rapid and is essentially complete within the first 30 min after exposure. Leakage assays. To see if exposure to peptoid antimicrobials caused leakage of cellular contents, an inner membrane leak-
FIG. 2. Effect of protein synthesis inhibition on antibacterial activity. The data are numbers of viable cells (CFU/milliliter) in aliquots removed at various times after drug exposure compared to those for a drug-free control culture. This experiment utilizes the S. aureus strain SA4 obtained from the Center for Research in Anti-Infectives and Biotechnology, Creighton University, strain collection. It is a methicillin-susceptible S. aureus clinical isolate obtained in Florida in 1971.
1432
GOODSON ET AL.
ANTIMICROB. AGENTS CHEMOTHER.
FIG. 3. CHIR29498 and leakage of b-galactosidase from E. coli ML-35. (A) Profile of b-galactosidase activity in the supernatant of CHIR29498- or GS-treated cells compared to that for DMSO-treated cells. (B) b-Galactosidase activities from the supernatants of CHIR26240-, DHAA-, and DMSO-treated cells.
age assay was performed (21). E. coli ML-35 (ATCC43827) constitutively produces cytoplasmic b-galactosidase and is both lactose and permease deficient. Extracellular hydrolysis of the substrate ONPG is therefore only possible if b-galactosidase is released from the cells through membrane leakage or cellular lysis. Figure 3 shows steadily increasing levels of extracellular enzyme activity (measured by absorbance at 405 nm) when this strain was exposed to the trimer peptoid CHIR29498 at or above MIC levels. Given the rapidly bactericidal activities we observed with these compounds (Fig. 2), the slow, steadily increasing levels of extracellular activity suggest leakage from intact cells rather than release through complete lysis. Further, though levels of released enzyme were higher in the presence of higher concentrations of drug, the rates of release remained similar. The kinetics shown in Fig. 3 suggest that while both drug levels produced similar final rates of enzyme efflux, more time was required to achieve this level of permeabilization at the lower drug concentration. The dimer peptoid CHIR26240 produced an initial small release of b-galactosidase which leveled off within 10 min of drug addition. DHAA produced no increase in extracellular enzyme activity at the levels tested for both the dimer and trimer peptoids. The steadily increasing levels of released enzyme seem to be correlated with bactericidal activity as neither CHIR26240 nor DHAA was active against E. coli at the levels tested but CHIR29498 was. GS, which is known to cause leakage (11), was included for reference (Fig. 3). Since a gram-positive leakage indicator strain was not readily available, we used flow cytometry (5) to evaluate the membrane permeability of CHIR29498-treated S. aureus (ATCC 25923) to PI. PI is normally excluded from intact staphylococcal cells and can only enter if the cytoplasmic membrane has been damaged (23). As shown in Table 2, PI rapidly entered staphylococcal cells exposed to the trimer peptoid at one or two times the MIC level. DHAA-treated cells also took up PI, but at much reduced levels consistent with the lower activity of DHAA. Although death by any mechanism may
eventually cause membrane damage and subsequent influx of PI, we did not observe such influx (in the stated time frame) when cells were treated with ampicillin, a drug which is bactericidal to the test strain but does not target the plasma membrane (data not shown). Interestingly, 2 h of exposure to the active peptoid compounds yielded higher percentages of stained cells than did 30 min of exposure, a result not consistent with the idea that peptoids achieve a steady state of membrane permeabilization as suggested by the leakage kinetics observed with E. coli (Fig. 3). It is, of course, possible that not all cells in a population are permeabilized in a uniform manner or that the kinetics of uptake are substantially different in S. aureus cells than they are for E. coli. The extent and rapidity of killing observed in the S. aureus time-kill experiments, however, do not support the presence of a large subpopulation of cells which are only slowly damaged by these compounds (Fig. 2). Hemolysis assay. We utilized a simple assay in which washed mouse or human RBC were exposed to CHIR29498 or 0.1% Triton X-100 and released heme was measured (20). It was apparent (Fig. 4) that CHIR29498 caused heme leakage from TABLE 2. Effects of DHAA and the related peptoid CHIR29498 on S. aureus membrane permeability to PI Compound
Concn
b
Control DMSO DHAA DHAA CHIR29498 CHIR29498
1 2 1 2
1% 3 MIC 3 MIC 3 MIC 3 MIC
% Bacteria stained ata: 30 min
2h
0.0 0.2 1.3 5.7 54.4 74.2
0.0 0.1 4.1 45.7 78.1 94.1
a Values are percentages of total cells counted which were stained by PI uptake. Values were determined by flow cytometric methods (13, 14). The S. aureus strain was ATCC 25923. b Antimicrobial-free medium.
VOL. 43, 1999
NOVEL ANTIMICROBIAL PEPTOIDS
1433
TABLE 3. Survival of CD-1 mice infected i.p. with 108 CFU of S. aureus followed by i.p. injection of antimicrobial agents Compound
Survival (%)a
Dose (mg/kg)
Controlb DHAA
3 10 30
0 0 0 0
CHIR26240
3 10 30
0 10 20
CHIR29498
3 10 30
0 100 100
Ampicillin
0.5
100
a
Survival is expressed as the percentage of mice surviving 5 days after infection. b Antimicrobial-free HPbCD solution.
DISCUSSION
FIG. 4. Hemolysis measurements of washed human RBC treated with CHIR29498, CHIR26240, DHAA, ampicillin or DMSO. Values are represented as the percentages of total lysis compared to lysis by 0.1% Triton X-100.
washed RBC at some value greater than its corresponding MIC for staphylococci (Table 1). The hemolytic activity of CHIR29498 was fairly modest compared to that of other cellmembrane-active molecules such as magainin 2 (1) or the positive control GS. Murine infection assay. In the absence of detailed knowledge of the pharmacokinetics of CHIR29498, we sought in vivo verification of the efficacy of this compound in a model of S. aureus infection and treatment. In this model, CD-1 mice were infected via i.p. injection with S. aureus (Smith, ATCC 13709) and then either immediately (Table 3), or after a defined delay (Table 4), dosed with a single i.p. injection of the drug to be tested. In preliminary experiments, threefold inoculum increments between 3 3 106 and 3 3 108 CFU/mouse were tested, and the bacterial dose causing 100% lethality after i.p. infection was determined to be approximately 108 CFU/mouse. Test compounds were then evaluated based on their ability to protect mice from this otherwise lethal inoculum. The infection load was heavy to allow for unambiguous interpretation of the number of animals rescued from the infection by the treatment. Table 3 shows the results of treating infected animals at time zero with various levels of CHIR29498 (peptoid trimer), CHIR26240 (peptoid dimer), or DHAA compared to drugfree control or 0.5-mg/kg ampicillin. The results shown are from two separate experiments with 5 to 10 animals per treatment. Other than ampicillin, only CHIR29498 was effective in this single-dose model, protecting 100% of the animals at a dose of 10 mg/kg. Notably, CHIR29498 was also able to provide protection in a delayed treatment i.p.-i.p. model as shown in Table 4. In delayed treatment experiments, the test compounds were delivered as single-dose treatments 30 or 110 min after infection. Again, of the test compounds, only CHIR29498 was efficacious, protecting 50% of the animals when dosed (30 mg/kg) 110 min after i.p. infection (Table 4).
The peptoids described in this paper are easily synthesized molecules which display rapid bactericidal activities against a panel of gram-positive and gram-negative bacteria, including both drug-sensitive and multiply-resistant isolates. The peptoid compounds tested in these studies were generally slightly more active against gram-positive than gram-negative isolates. The bactericidal activities of the peptoids were found to be protein synthesis independent, and rearrangement within a peptoid molecule did not seem to greatly affect potency as positional isomers of a tested molecule retained most of the parent molecule’s activity. Antibacterial peptoids have an activity profile that is similar to that of the extensively described antibacterial peptides (cecropins, defensins, attacins, diptericins, magainins, protegrins [4, 10, 15, 17, 24]). The smallest natural antibacterial peptide to date is indolicidin (6), which is 13 amino acids in length. Screening of synthetic peptide libraries for antibacterial activity identified activity in 6-mer synthetic peptides which were most potent when unnatural amino acids were included in the synthesis (2, 3). By contrast, we have identified antibacterial peptoids, such as CHIR26240, which are smaller than these synthetic antibacterial peptides. Studies of antibacterial peptides have shown that the D and the L forms were equally potent, suggesting a target that does not require a lock-and-key interaction (22). Likewise, positional isomers of an antibacterial peptoid tested in this study retained most of the activity of TABLE 4. Survival of CD-1 mice infected i.p. with 108 CFU of S. aureus and treated by i.p. injection of antimicrobial agents at 0, 30, or 110 min post-infection
Compound
Dose (mg/kg)
b
Control CHIR26240 CHIR29498 Ampicillin
30 30 0.5
% Survival for mice treated with antimicrobial agent at the indicated times (min)a 0
30
110
0 20 10 100
0 20 80 80
0 0 50 50
a Survival is expressed as the percentage of mice surviving 5 days after infection. b Antimicrobial-free HPbCD solution.
1434
GOODSON ET AL.
the parent molecule. It is commonly accepted that antibacterial peptides cause membrane leakage, perhaps by interfering with the charge of the phosphate on phospholipids of the membranes. In some cases the amphipathic nature of the peptoids may lead directly to pore formation in membranes. In reality, the mechanism of the membrane disruption for membranedisrupting compounds has yet to be fully elucidated. However, a profile for such membrane-active compounds can be compiled from the peptide and peptoid literature: rapid bactericidal activity independent of protein synthesis, rapidly induced leakage (as measured by leakage assays), and active molecules with equally active isomers. CHIR29498 was not found to be aggressively hemolytic in vitro at concentrations that were effectively antimicrobial. It did, however, cause hemolysis at higher concentrations. We do not know if hemolysis is responsible for the high-dose (60 mg/kg) animal toxicity observed in this study (described in Materials and Methods). In vivo antimicrobial efficacy of the peptoid compounds was demonstrated in a simple S. aureus mouse infection model. This model is used when little is known about the pharmacokinetic properties of a lead molecule. To make the i.p.-i.p. model more stringent, we delayed the time between infection and drug treatment. CHIR29498 protected infected animals even when treatment was delayed by up to 110 min; however, the drop in efficacy upon time delay (Table 4) compared to that of immediate dosing (Table 3) indicates that CHIR29498 may require optimization for improved absorption or stability within the body. Although this is a very simple model, it was capable of discriminating among CHIR29498, CHIR26240, and DHAA, all of which were active in vitro. Finally, although we did not know that DHAA was antibacterial when it was first selected for the combinatorial libraries from which these peptoids were derived, adding properties to an already active moiety may be yet another useful application of peptoid chemistry. We have identified active peptoids that do not contain independently active substituents and thus would like to emphasize that it is possible to create antimicrobial peptoids from building blocks which do not have independent activity. ACKNOWLEDGMENTS We greatly appreciate the consultations with Fred Cohen (UCSF) and Robin Cooper. REFERENCES 1. Aboudy, Y., E. Mendelson, I. Shalit, R. Bessale, and M. Fridkin. 1994. Activity of two synthetic amphiphilic peptides and magainin-2 against herpes simplex virus types 1 and 2. Int. J. Peptide Prot. Res. 43:573–582. 2. Blondell, S. E., and R. A. Houghten. 1996. Novel antimicrobial compounds identified using synthetic combinatorial library technology. Trends Biochem. Sci. 14:60–65.
ANTIMICROB. AGENTS CHEMOTHER. 3. Blondell, S. E., E. Takahashi, P. A. Weber, and R. A. Houghton. 1994. Identification of antimicrobial peptides by using combinatorial libraries made up of unnatural amino acids. Antimicrob. Agents Chemother. 38: 2280–2286. 4. Boman, H. G. 1995. Peptide antibiotics and their role in innate immunity. Annu. Rev. Immunol. 13:61–92. 5. Durodie, J., K. Coleman, and L. Simpson. 1995. The application of flow cytometry to the estimation of bacterial antibiotic susceptibility. J. Antimicrob. Chemother. 36:443–444. 6. Falla, T. J., and R. E. W. Hancock. 1997. Improved activity of a synthetic indolicidin analog. Antimicrob. Agents Chemother. 41:771–775. 7. Figliozzi, G. M., R. Goldsmith, S. C. Ng, S. C. Banville, and R. Zuckermann. 1996. Synthesis of N-(substituted) glycine peptoid libraries. Methods Enzymol. 267:437–447. 8. Haraguchi, H., S. Oike, H. Muroi, and L. Kubo. 1996. Mode of antibacterial action of totarol, a diterpene from Podocarpus nagi. Planta Med. 62:122–125. 9. Huang, C.-Y., T. Uno, J. E. Murphy, S. Lee, J. D. Hamer, J. A. Escobedo, F. E. Cohen, R. Ramachandran, V. Dwarki, and R. Zuckermann. 1998. Liptoids—novel cationic lipids for cellular delivery of plasmid DNA in vitro. Curr. Biol. 5:345–354. 10. Kimbrell, D. A. 1991. Insect antibacterial proteins: not just for insects and against bacteria. BioEssays 13:657–663. 11. Kondejewski, L. H., S. W. Farmer, D. S. Wishart, R. E. W. Hancock, and R. S. Hodges. 1996. Gramicidin S is active against both gram-positive and gram-negative bacteria. Int. J. Pept. Protein Res. 47:460–466. 12. Mason, D. J., R. Almon, J. M. Stark, and D. Lloyd. 1994. Rapid estimation of bacterial antibiotic susceptibility. J. Microsc. 176:8–16. 13. Mason, D. J., E. G. M. Power, H. Talsania, I. Phillips, and V. A. Gant. 1995. Antibacterial action of ciprofloxacin. Antimicrob. Agents Chemother. 39: 2752–2758. 14. Miller, S. M., R. J. Simon, S. Ng, R. N. Zuckermann, J. M. Kerr, and W. H. Moos. 1994. Proteolytic studies of homologous peptide and N-substituted glycine peptoid oligomers. Biol. Med. Chem. Lett. 4:2657–2661. 15. Mor, A., K. Hani, and P. Nicolas. 1994. The vertebrate peptide antibiotics dermaseptins have overlapping structural features but target specific microorganisms. J. Biol. Chem. 269:31635–31641. 16. National Committee for Clinical Laboratory Standards. 1997. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 4th ed. Approved standard M7-A4 17. National Committee for Clinical Laboratory Standards. Wayne, Pa. 17. Nicolas, P., and A. Mor. 1995. Peptides as weapons against microorganisms in the chemical defense system of vertebrates. Annu. Rev. Microbiol. 49:227– 304. 18. Ng, S., B. Goodson, A. Ehrhardt, and J. Winter. Bioorganic and medicinal chemistry. Combinatorial discovery process yields antimicrobial peptoids. Submitted for publication. 19. Nguyen, J. T., C. W. Turck, F. E. Cohen, R. N. Zuckermann, and W. A. Lim. 1998. Exploiting the basis of proline recognition by SH3 and WW domains of N-substituted inhibitors. Science 282:2088–2092. 20. Salauze, D., and D. Decouvelaere. 1994. In vitro assessment of haemolytic potential of candidate drugs. Fr. Comp. Haematol. Int. 4:34–36. 21. Skerlavaj, B., D. Romeo, and R. Gennaro. 1990. Rapid membrane permeabilization and inhibition of vital functions of gram-negative bacteria by bactenecins. Infect. Immun. 58:3724–3730. 22. Wade, D., A. Boman, B. Wahlin, C. M. Drain, D. Andreu, H. G. Boman, and R. B. Merrifield. 1990. All D-amino acid containing channel forming antibiotic peptides. Proc. Natl. Acad. Sci. USA 87:4761–4765. 23. Waggoner, A. S. 1979. Dye indicators of membrane potential. Annu. Rev. Biophys. Bioeng. 8:47–49. 24. Zanetti, M., P. Storici, A. Tossi, M. Scocchi, and R. Gennaro. 1994. Molecular cloning and chemical synthesis of a novel antibacterial peptide derived from pig myeloid cells. J. Biol. Chem. 269:7855–7858.
