|Home | About | Journals | Submit | Contact Us | Français|
Biofilms are sessile microbial communities that cause serious chronic infections with high morbidity and mortality. In order to develop more effective approaches for biofilm control, a series of linear cationic antimicrobial peptides (AMPs) with various arginine (Arg or R) and tryptophan (Trp or W) repeats [(RW)n-NH2, where n = 2, 3, or 4] were rigorously compared to correlate their structures with antimicrobial activities affecting the planktonic growth and biofilm formation of Escherichia coli. The chain length of AMPs appears to be important for inhibition of bacterial planktonic growth, since the hexameric and octameric peptides significantly inhibited E. coli growth, while tetrameric peptide did not cause noticeable inhibition. In addition, all AMPs except the tetrameric peptide significantly reduced E. coli biofilm surface coverage and the viability of biofilm cells, when added at inoculation. In addition to inhibition of biofilm formation, significant killing of biofilm cells was observed after a 3-hour treatment of preformed biofilms with hexameric peptide. Interestingly, treatment with the octameric peptide caused significant biofilm dispersion without apparent killing of biofilm cells that remained on the surface; e.g., the surface coverage was reduced by 91.5 ± 3.5% by 200 μM octameric peptide. The detached biofilm cells, however, were effectively killed by this peptide. Overall, these results suggest that hexameric and octameric peptides are potent inhibitors of both bacterial planktonic growth and biofilm formation, while the octameric peptide can also disperse existing biofilms and kill the detached cells. These results are helpful for designing novel biofilm inhibitors and developing more effective therapeutic methods.
Antimicrobial peptides (AMPs) are promising alternatives to traditional antibiotics (5). Native AMPs are part of the host defense in organisms ranging from bacteria to insects, plants, and animals (14). They are capable of eliminating a broad spectrum of microorganisms, including viruses, bacteria, and fungi (4, 14). Compared with widespread antibiotic resistance (38), resistance to AMPs is rare, possibly because AMPs directly target cell membranes that are essential to microbes (14, 29). In addition, no cross-resistance has been observed in clinic due to the diversity of peptide sequences (42). Thus, native and synthetic AMPs offer potential alternatives to antibiotics for treating drug-resistant infections (3, 26, 27).
In mammalian innate immune systems, some AMPs are produced constitutively, while others are inducible within hours after detection of invading microorganisms (4, 13). Although the detailed mechanism of AMPs' activities remains elusive (5), AMPs are known to disrupt cell membranes of microbes, interfere with metabolism, and/or target cytoplasmic components (41). Most known AMPs are cationic and amphiphilic (29). It is hypothesized that the initial interaction occurs via an electrostatic attraction between the AMP molecule and microbial membrane. Cationic AMPs can cover bacterial membranes, disrupt the membrane potential, create pores across the membrane, and consequently cause the leak of cell contents and cell death (27, 41). AMPs are relatively selective in targeting microbes rather than mammalian cells, most likely because of the fundamental differences between microbial and host membranes (41), e.g., a higher abundance of negatively charged phospholipids and an absence of cholesterol in microbial membranes.
Known AMPs vary dramatically in sequence, size (from 12 to 50 amino acids), and structure (α-helices or β-sheets) (23). However, most AMPs have two types of side chains with relatively conservative sequences: positively charged basic residues, containing arginine (R), lysine (K), and/or histidine (H), that presumably mediate the interaction with the negatively charged microbial membrane, and bulky hydrophobic residues, rich in tryptophan (W), proline (P), and/or phenylalanine (F), that facilitate permeabilization and membrane disruption (26).
Although AMPs are promising agents for antimicrobial therapies (15), only a few have made it to clinical trials and applications, with varied success (15, 42). There are several issues that need further development. First, the MICs of AMPs are relatively high compared to those of conventional antibiotics. Recent studies suggest that the peptide/lipid (P/L) ratio needs to be higher than a threshold to allow the AMPs to be oriented perpendicular to the membrane so that pores can be created to kill bacteria (22, 30). Thus, an optimization of peptide structure and size may improve their antimicrobial activities. In addition to the high MICs, the wide application of AMPs is also hindered by their high manufacturing costs and the cytotoxicity of some AMPs.
Given the limit of currently available AMPs, it is important to develop more effective AMPs with reduced manufacturing cost and enhanced activity (17, 26, 28, 39). Strøm et al. (39) chemically synthesized a series of short cationic AMPs containing repeating R and W residues in order to identify the minimal pharmacophore with high antimicrobial activities. The data suggest that tetrapeptides or capped tripeptides are effective and there is no correlation between the order of amino acids and antimicrobial activity. Liu et al. (26) analyzed the effects of chain length on the activities of AMPs with repeating pharmacophore sequences (RW)n-NH2 (n = 1, 2, 3, 4, or 5). The tests of antimicrobial activities and the hemolysis of red blood cells suggest that (RW)3-NH2 has the optimal chain length. Although longer chains are more potent antimicrobials, they can stimulate hemolysis.
Most of the AMP studies to date are focused on planktonic bacteria. However, the majority of pathogenic bacteria tend to adhere to surfaces and form sessile microbial communities with highly hydrated structures of secreted polysaccharide matrix, collectively known as biofilms (9). Biofilms can tolerate up to 1,000 times more antibiotics and disinfectants than their planktonic counterparts (2, 7, 8). For example, Folkesson et al. (12) reported that biofilm formation of E. coli K-12 increases its tolerance to polymyxin E, a polypeptide antibiotic that kills Gram-negative bacteria by disrupting membranes (34, 40). Since biofilms are involved in 80% of human bacterial infections (1), it is necessary to study biofilm inhibition and dispersion by AMPs.
In this study, a series of linear peptides (RW)n-NH2 (where n = 2, 3, or 4) were studied for the effects of their activities on planktonic cells and biofilms of E. coli to understand the structural effects on the antimicrobial activities of AMPs. We chose E. coli RP437 in this study because it is one of the model strains for biofilm research and allows us to compare the data with those of our previous studies (6, 16, 19, 20).
One of the model strains for biofilm formation, Escherichia coli RP437 [thr-1(Am) leuB6 his-4 metF159(Am) eda-50 rpsL1356 thi-1 ara-14 mtl-1 xyl-5 tonA31 tsx-78 lacY1 F−] (31), was kindly provided by John S. Parkinson (University of Utah) and used in this study. E. coli RP437 was grown in Luria-Bertani (LB) medium containing 10 g/liter tryptone, 5 g/liter yeast extract, and 10 g/liter NaCl (33) at 37°C, either with shaking at 200 rpm for overnight cultures and growth tests or without shaking for biofilm formation. For the motility assays, semisolid LB plates were prepared by adding 4 g/liter agar (for swarming assay) or 3 g/liter agar (for swimming assay) in LB medium before autoclaving.
Three short cationic antimicrobial peptides, including tetrameric peptide (RW)2-NH2, hexameric peptide (RW)3-NH2, and octameric peptide (RW)4-NH2 (Fig. (Fig.1),1), were prepared as described previously (26, 28). The molecular weight of each peptide was confirmed using matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) (mass spectrometry) as described previously (26). Briefly, the (RW)n-NH2 sequences were assembled on Rink amide resin from Nova Biochem (San Diego, CA) with a Rainin Instrument PS3 solid-phase synthesizer (Woburn, MA) using Fmoc (9-fluorenylmethoxycarbonyl) chemistry. Fmoc-Trp(butoxycarbonyl [Boc])/Arg(2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl [Pbf]), the coupling reagent HBTU [2-(1H-benzotriazol-1-yl) 1,1,3,3-tetramethyluroniumhexafluoro phosphate], and HOBT (N-hydroxybenzotriazol) were also purchased from Nova Biochem. Cleavage of the peptides from the resin was performed with 95% trifluoroacetic acid (TFA) in the presence of the scavenger, 2.5% triisopropylsilane (TIS), and 2.5% H2O. After precipitation with cold ether, samples were purified on a reverse-phase high-performance liquid chromatography C18 preparative column (2.2 by 25 cm, 300 Å; Grace Vydac Co., Hesperia, CA) with water and acetonitrile as eluents. Fractions containing product were pooled and lyophilized. The molecular weight of each peptide was confirmed by a Bruker matrix-assisted laser desorption ionization-time-of-flight mass spectrometer (Billerica, MA). All peptides were dissolved in Tris-buffer (12.1 g/liter Tris base and 8.8 g/liter sodium chloride, pH 7.2) to a final concentration of 5 mg/ml as the stock solution and stored at 4°C.
To investigate the antimicrobial activities of the three peptides, the planktonic growth of E. coli RP437 in the absence and presence of each peptide at different concentrations was studied. An overnight culture of E. coli RP437 grown in LB medium was used to inoculate a 96-well plate (containing 200 μl sterile LB medium in each well) to an optical density at 600 nm (OD600) of 0.05. The peptides were added to a final concentration of 0, 25, 50, 100, or 200 μM. Six replicates were tested for each condition. The plate was incubated at 37°C, with shaking at 200 rpm. OD600 readings were measured every 30 to 60 min to monitor bacterial growth using a microplate reader (EL808; BioTek Instruments, Inc., Winooski, VT). To understand if the peptides are bactericidal or bacteriostatic, an E. coli RP437 overnight culture was washed with 0.85% NaCl buffer and diluted to an OD600 of 1.0 in 0.85% NaCl buffer. Hexameric and octameric peptides were then added as 25 μM concentrations. The cultures were incubated at 37°C, with shaking at 200 rpm for 3 h. The cells before and after treatment were spread on LB plates and incubated for 24 h to count the number of CFU.
To quantify biofilm formation of E. coli in the absence and presence of peptides, a microplate-based assay was performed by following the protocols reported previously (25, 32), with slight modifications. Briefly, an overnight culture of E. coli RP437 grown in LB medium was used to inoculate fresh LB medium to an OD600 of 0.05. Each peptide was added to a final concentration of 0, 25, 50, 100, or 200 μM. Four replicates were tested for each condition. The plates were incubated at 37°C without shaking for 24 h. Planktonic cells were then carefully removed by pipetting, and the plates with biofilms were washed three times with deionized (DI) H2O and dried by gently patting them on a piece of paper towel. To quantify biofilms, each well was stained with 300 μl of 0.1% crystal violet and incubated for 20 min at room temperature. The plates were then washed three times with DI H2O again to remove extra dye. An OD540 was measured to quantify the biofilms on the bottom of each well (bottom biofilm). Then, 300 μl of 95% ethanol was added to dissolve all the absorbed dye, and an OD540 was measured again to quantify the total biomass of liquid-solid biofilm and air-liquid biofilm (total biofilm) in each well.
In addition to the above-described experiment of biofilm formation in the presence and absence of peptides, another three experiments were conducted to understand the effects of peptides on biofilm cells specifically. In the first experiment, the biofilms were inoculated as described above and grown for 24 h. The planktonic cells were then removed by pipetting, and the biofilms were washed with 0.85% NaCl. The biofilms were then treated with different concentrations of peptides in 0.85% NaCl buffer for 3 h and analyzed after staining with 0.1% crystal violet as described above. To corroborate the results, cell adhesion and biofilm development were also studied using LB medium supplemented with 0.5% glucose in 96-well plates with a high inoculation cell density (OD600 of 1.0). At 3 h after inoculation, the planktonic cells were removed by washing the plates with 0.85% NaCl buffer gently. The peptides were then added at different concentrations, and biofilms were analyzed by staining with 0.1% crystal violet as described above after 3 and 24 h of incubation.
To visualize the viability of biofilm cells and surface coverage in the absence and presence of peptides, 24-hour biofilms of E. coli RP437 on 316L stainless steel surfaces were stained with the Live/Dead BacLight bacterial viability kit (Invitrogen Corporation, Carlsbad, CA) and analyzed using an AXIO Imager M1 microscope (Carl Zeiss, Inc., Germany). E. coli biofilms were formed on 316L stainless steel coupons (1/4 in. by 1/2 in. by 1/6 in.) by following a procedure described previously (11), with slight modifications. Briefly, an overnight culture of E. coli RP437 was used to inoculate fresh LB medium in a petri dish with coupons to an OD600 of 0.05. The peptides were added to a final concentration of 0, 25, 50, 100, or 200 μM. The cultures were then incubated at 37°C without shaking for 24 h. Four coupons were tested for each condition.
To analyze biofilms using fluorescence microscopy, 24-hour biofilms were gently washed three times with 0.85% NaCl buffer to remove planktonic cells and stained in 1 ml of 0.85% NaCl buffer containing 0.6 μl of 3.34 mM SYTO 9 (to stain live cells) and 2.4 μl of 20 mM propidium iodide (to stain dead cells) in the dark for 15 min. Five spots on each coupon were randomly picked and imaged using an Axio Imager M1 microscope. The surface coverage of biofilms was then calculated using the COMSTAT software (18).
Besides the biofilm formation in the absence and presence of peptides, the killing and/or dispersion of preformed biofilms by peptides were also investigated using live/dead staining. Biofilms were allowed to form on 316L stainless steel coupons using the procedure described above, except that no peptide was added at inoculation. After 24 h of incubation, the coupons were washed gently with 0.85% NaCl buffer three times to remove planktonic cells. They were then transferred to 0.85% NaCl buffer supplemented with or without peptides (0, 25, 50, 100, or 200 μM) and incubated at 37°C without shaking for 3 h. The coupons were then transferred immediately to fresh 0.85% NaCl buffer and stained with the Live/Dead staining kit for biofilm imaging as described above.
To corroborate the results of the live/dead staining assay, the viability of biofilm cells with and without treatment was also examined by counting the number of CFU of viable biofilm cells. The biofilms of E. coli were formed on 316L stainless steel coupons and treated with peptides as described above. The coupons were then sonicated for 2 min and vortexed for 1 min in 0.85% NaCl buffer in order to release biofilm cells from the surfaces. The released bacterial cells were diluted in 0.85% NaCl buffer, spread on LB agar plates, and incubated at 37°C for 24 h to count the number of CFU.
The swarming assay was performed as described previously (19), with slight modifications. Briefly, fresh LB plates with 4 g/liter agar and 0 or 25 μM tetrameric, hexameric, or octameric peptide were inoculated with E. coli RP437 using toothpicks. The plates were then incubated at 37°C, and the colony diameters were measured every 1 or 2 h.
The swimming assay was conducted by following the same procedure of the swarming test, except that the agar was added as 0.3%. This condition was found to support swimming (36, 37), with the cells moving in the agar medium rather than 0.4% agar, which allows the cells only to swarm on the surface of agar medium.
All statistical analyses were performed by using SAS 9.1.3, Windows version (SAS, Cary, NC).
Microplate assays were used to monitor the effects of the antimicrobial activities of the peptides on the planktonic growth of E. coli RP437. The peptides were tested by adding them at inoculation to final concentrations of 0, 25, 50, 100, or 200 μM. Six replicates were tested for each condition. As shown in Fig. Fig.2,2, tetrameric peptide did not inhibit the growth of E. coli RP437, since the specific growth rates were 0.48 ± 0.02, 0.43 ± 0.03, 0.44 ± 0.03, 0.45 ± 0.04, and 0.44 ± 0.03 h−1 with 0, 25, 50, 100, and 200 μM tetrameric peptide, respectively. Hexameric and octameric peptides at all tested concentrations showed slightly extended lag phases and reduced specific growth rates compared to the peptide-free control. For example, the hexameric peptide at 200 μM extended the lag phase to 2 h (no apparent lag phase for the peptide-free control) and reduced the specific growth rate by 36% (from 0.48 ± 0.02 h−1 to 0.31 ± 0.03 h−1) compared to the peptide-free control. The antimicrobial activities of peptides follow the order of tetrameric peptide < hexameric peptide ≈ octameric peptide (Fig. (Fig.2).2). These data suggest that a critical chain length may be required to obtain significant antimicrobial activities.
To understand if the antimicrobial peptides are bactericidal or bacteriostatic, E. coli RP437 cells at an OD600 of 1.0 were treated with and without 25 μM hexameric or octameric peptide in 0.85% NaCl buffer for 3 h. Treatments with hexameric and octameric peptides at 25 μM were found to reduce the number of viable cells by 90.4 ± 1.3% and 90.1 ± 1.5%, respectively, suggesting that both hexameric and octameric peptides are bactericidal.
A microplate-based assay using crystal violet staining was conducted to quantify biofilm formation in the absence and presence of antimicrobial peptides. Each of the three peptides was added at the time of inoculation in 96-well plates to concentrations of 0, 25, 50, 100, or 200 μM. Four replicates were tested for each condition. The plates were incubated at 37°C for 24 h and then stained with 0.1% crystal violet as described in Materials and Methods. Both the bottom biofilms and total biofilms were quantified. For convenience of comparison, the biofilm data were normalized by the highest value of OD540, which is the biomass in the presence of 25 μM tetrameric peptide for bottom biofilms and 50 μM tetrameric peptide for total biofilms. Similar results of biofilm inhibition by peptides were observed for bottom biofilms (Fig. (Fig.3)3) and total biofilms (see Fig. S1 in the supplemental material). Tetrameric peptide did not show any significant inhibition of biofilm formation (hierarchy model, two-way analysis of variance [ANOVA], P = 0.97), which is consistent with the result for planktonic growth. The hexameric and octameric peptides showed similar dose-dependent inhibition (hierarchy model, two-way ANOVA adjusted by Bonferroni method, P < 0.01 for all tested conditions). For example, hexameric peptide inhibited bottom biofilm formation by 61.3 ± 2.4% at 100 μM and 95.0 ± 1.1% at 200 μM, while octameric peptide inhibited biofilm formation by 39.5 ± 3.8% at 100 μM and 84.4 ± 6.8% at 200 μM (Fig. (Fig.3).3). To corroborate the above results, the peptides were also tested with adhesion in LB medium supplemented with 0.5% glucose at a higher cell density (OD600 of 1.0). Treatments with hexameric and octameric peptides for 3 and 24 h were found to significantly reduce the biomass of attached cells (see Fig. S2 in the supplemental material). For example, treatment with 100 μM octameric peptides for 3 and 24 h reduced the biomass by 89.4 ± 9.9% and 93.8 ± 3.9%, respectively.
Mature biofilms are known to be tolerant to antimicrobial treatments (10). To understand if the synthetic peptides are also effective against preformed biofilms, the peptides were tested by adding them to 24-hour biofilms. Both hexameric and octameric peptides were able to reduce biofilm mass significantly, suggesting that these two peptides are also effective against existing biofilms (Fig. (Fig.4).4). For example, 100 μM octameric peptide reduced the biomass of bottom biofilms by 82.8 ± 10.8% (hierarchy model, two-way ANOVA adjusted by Bonferroni method, P < 0.0001). Similar results were obtained for total biofilms treated under the same conditions (see Fig. S3 in the supplemental material).
To gain insight into the antimicrobial effects of peptides on biofilms, a live/dead staining assay was used to visualize biofilm cells on 316L stainless steel coupons with fluorescence microscopy. Representative fluorescence images of biofilms after 24-hour incubation in the absence and presence of peptides are shown in Fig. Fig.5,5, and the surface coverage of live biofilm cells calculated from the images are shown in Fig. Fig.6.6. Compared to untreated samples, biofilms formed in the presence of tetrameric peptide showed slightly lower surface coverage (e.g., 23.0 ± 8.1% reduction at 25 μM [hierarchy model, two-way ANOVA adjusted by Bonferroni method, P = 0.0001]), but there was no apparent dose-dependent inhibition with the concentrations tested (correlation coefficient = −0.29, P = 0.13). In contrast, the hexameric and octameric peptides showed strong and dose-dependent killing of biofilm cells (hierarchy model, two-way ANOVA adjusted by Bonferroni method, P < 0.0001 for all the conditions tested). For example, the hexameric peptide at 25, 50, 100, and 200 μM reduced live biofilm cells by 59.9 ± 11.5%, 63.3 ± 8.7%, 77.1 ± 6.2%, and 94.7 ± 3.0%, respectively, compared to the peptide-free control. Addition of octameric peptide also caused significant reduction of live biofilm cells, e.g., 42.4 ± 11.3% and 85.1 ± 4.7% reductions at 50 μM and 100 μM, respectively, compared to the peptide-free control.
To understand the effects of peptides on preformed biofilms and corroborate the results based on 96-well plates, the 24-hour E. coli biofilms formed on 316L stainless steel coupons were treated with antimicrobial peptides for 3 h, stained with the Live/Dead staining kit, and analyzed using fluorescence microscopy. Representative fluorescence images of biofilms are shown in Fig. Fig.7,7, and the surface coverage of live biofilm cells is shown in Fig. Fig.8.8. Consistent with the 96-well plate results, the tetrameric peptide did not show any significant killing or dispersion of biofilms (hierarchy model, two-way ANOVA, P = 0.43), while the hexameric peptide showed significant (hierarchy model, two-way ANOVA adjusted by Bonferroni method, P < 0.0001 for all the conditions tested) and dose-dependent (correlation coefficient of −0.79, P = 0.0001) killing of biofilm cells. For instance, the live biofilm cells were reduced by 25.3 ± 11.2%, 28.8 ± 16.4%, 58.4 ± 7.6%, and 63.6 ± 8.0% after 3-hour treatments with hexameric peptide at 25, 50, 100, and 200 μM, respectively, compared to the peptide-free control (Fig. (Fig.77 and and8).8). Interestingly, octameric peptide induced significant dispersion of biofilms without the apparent killing of the biofilm cells that remained on the surface (hierarchy model, two-way ANOVA adjusted by Bonferroni method, P < 0.0001 for all the conditions tested). For example, 47.4 ± 6.2%, 71.5 ± 9.2%, 87.9 ± 8.0%, and 91.5 ± 3.5% of the biofilm was removed after treatment for 3 h with 25, 50, 100, and 200 μM octameric peptide, respectively (Fig. (Fig.77 and and88).
To confirm that the films were indeed detached by octameric peptide rather than any artifacts due to interaction between the dyes and the peptide, a CFU assay was performed to quantify the viable biofilm cells before and after treatment. Treatments with tetrameric peptide at 25 μM and 200 μM were also included as negative controls. The results of viable biofilm cells (%) normalized by the number of CFU indicate that no significant reduction of viable biofilm cells was observed for the treatments with tetrameric peptide at 25 μM and 200 μM (hierarchy model, two-way ANOVA, P = 0.48), while the treatments with octameric peptide at 25 μM (P = 0.01) and 200 μM (P < 0.0001) significantly (both are based on a hierarchy model, two-way ANOVA adjusted by the Bonferroni method) reduced the number of attached viable cells (by 26.8 ± 7.3% and 97.1 ± 1.3%, respectively). Thus, these results corroborate the live/dead staining data and confirm that 87% biofilm cells were dispersed and killed by 200 μM octameric peptide (see Fig. S4 in the supplemental material). Although the octameric peptide did not kill cells that remained attached, the dispersed biofilm cells were effectively killed by this peptide. For example, only 4.9 ± 2.7% of the cells dispersed by 200 μM octameric peptide were viable, while 98.2 ± 1.6% of the planktonic cells in the peptide-free control were viable based on live/dead staining. It remains unknown whether the octameric peptide kills the biofilm cells and then disperses them, or disperses them before killing the detached cells.
To better understand the mechanisms of biofilm inhibition by peptides with repeating Trp/Arg, we also evaluated the effects of peptides on the swarming motility of E. coli, which plays an important role in its biofilm formation (32). A motility assay was conducted to evaluate the swarming motility in the absence and presence of 25 μM of each antimicrobial peptide. The diameters of swarming colonies during the experimental period are shown in Fig. Fig.9.9. Hexameric peptide showed the strongest inhibition of the swarming among the peptides tested in this study; e.g., the swarming rate was 2.56 ± 0.02 mm/hour without peptide and 0.29 ± 0.01 mm/hour with hexameric peptide. The octameric peptide also significantly reduced the swarming rate to 1.35 ± 0.02. In comparison, tetrameric peptide slightly increased the swarming rate to 2.90 ± 0.03 (one-way ANOVA, Tukey test, P < 0.01 for all tested conditions).
The peptides were found to inhibit, in addition to swarming motility, the swimming motility of E. coli. This assay was conducted using 0.3% agar plates, which support swimming (cells entering the agar) as opposed to swarming (cells remaining on the surface of agar medium) seen with 0.4% agar. The diameters of swimming colonies were then measured. The tetrameric peptide did not exhibit significant effect (one-way ANOVA, Tukey test, P = 0.9), while hexameric and octameric peptides significantly inhibited the swimming motility (one-way ANOVA, Tukey test, P < 0.0001 for both peptides). The average swimming rates were 8.8 ± 0.2, 8.5 ± 0.4, 1.1 ± 0.2, and 2.3 ± 0.4 mm/hour without peptide and with 25 μM tetrameric, hexameric, and octameric peptide, respectively. Thus, the swimming motilities were in the order of control (no peptide) ≈ 25 μM tetramer > 25 μM octamer > 25 μM hexamer, which is consistent with the swarming data.
In this study, three short peptides of identical compositions but different chain lengths were rigorously compared for the effects of their activities on the planktonic growth and biofilm formation of E. coli RP437. The antimicrobial activities of the short cationic peptides against the planktonic growth of E. coli were found to follow the order of tetrameric peptide < hexameric peptide ≈ octameric peptide, which is consistent with previous results of 50% inhibitory concentrations (IC50s) for E. coli D31 (26, 28). In addition to the inhibitory effects on planktonic growth, the effects of these peptides on biofilms were studied for the first time. Hexameric and octameric peptides both inhibited biofilm formation when added at inoculation. These two peptides also exhibited inhibitory effects on preformed biofilms. However, while the hexameric peptide caused significant death of biofilm cells on the 316L stainless steel surface, the octameric peptide caused up to 91.5 ± 3.5% (at 200 μM) dispersion of preformed E. coli biofilms, while 95.1 ± 2.7% of the detached biofilm cells were dead.
The biofilm data suggest that these or related peptides may have promising applications in treating chronic infections. Mature biofilms are protected by the exopolysaccharide (EPS) matrix, which presents a barrier for drug penetration (24). As shown in the biofilm images of live/dead staining in Fig. Fig.7,7, the killing of preformed biofilm cells by hexameric peptide and dispersion of preformed biofilms by octameric peptide suggest that the peptides are able to penetrate the EPS matrix. In addition, a threshold concentration seems to exist for some peptides to kill biofilms significantly. For example, 71.1 ± 7.7% and 57.6 ± 11.3% of biofilm cells survived in the samples with 25 and 50 μM octameric peptide added at inoculation, respectively, compared to the control, while live biofilm cells were dramatically reduced to 14.9 ± 4.7% and 12.4 ± 6.3% with addition of 100 and 200 μM octameric peptide, respectively (Fig. (Fig.6).6). This finding suggests that a threshold concentration (between 50 and 100 μm) may exist. These results are consistent with the two-state models proposed by Huang (22) and Shai (35). According to these models, peptides first attach to the cell membrane via electrostatic interactions between negatively charged cell membranes and positively charged peptides. When the concentration of peptides reaches a threshold, peptides create holes and disrupt the membrane, leading to cell death. In our study, patches (stained red) were seen on the surfaces with peptide treatments (Fig. (Fig.5,5, panels C4 and D3 and 4), which may represent debris from lysed cells. This observation is consistent with the hypothesis of Liu et al. (26, 28) that the effectiveness of these short cationic peptides is due to a membranolytic mechanism, by which AMPs create transmembrane channels that cause depolarization of the membrane and cell death.
The inhibition of planktonic growth appears to be correlated with the chain length of peptides since the hexametric and octameric peptides significantly inhibited E. coli growth, while tetrameric peptides did not exhibit significant inhibition. However, hexametric and octameric peptides exhibited similar activities in inhibiting E. coli planktonic growth, suggesting that there might be an optimal chain length for the desired activities. In fact, difficulties may be encountered in penetrating cell membranes and biofilms if the chain of an AMP is too long. For this reason, more compact branched structures may offer advantages for peptides with higher molecular weights. Consistent with this hypothesis, we found recently that a small dendrimeric peptide, (RW)4D, has potent activities against both planktonic and biofilm cells of E. coli (21).
It is interesting to notice that a 3-hour treatment of preformed biofilms with octameric peptide can significantly induce biofilm dispersion. This result was further confirmed by the CFU data. The number of viable biofilm cells on the surface was significantly reduced by 97.1 ± 1.3% after a 3-hour treatment with octameric peptide at 200 μM compared with the peptide-free control. The mechanism of this dispersal is unknown. However, the motility data suggest that it is not due to induced swarming or swimming. Although the small number of cells that remained on the surface appeared to be alive, the majority (95.1 ± 2.7%) of detached cells were dead. Further study is necessary to understand if the octameric peptide killed the biofilm cells first or after detachment. The differences between hexameric and octameric peptides in their activities of killing biofilm cells, inhibiting swarming motility, and dispersing preformed biofilms also deserve further study. Synergistic effects between these peptides and other AMPs or antibiotics may also exist, and the study of these effects is part of our ongoing effort in developing more effective therapies.
In conclusion, we systematically compared the activities of three antimicrobial peptides against planktonic growth and biofilm formation, as well as the structure and viability of preformed biofilms of E. coli. The hexameric and octameric peptides were found to be potent biofilm inhibitors, and octameric peptide was also found to disperse existing biofilms and effectively kill the detached cells. Further study of the mechanism of these activities as well as the synergic effects with other antimicrobial agents will help design more effective antimicrobials and therapeutic approaches.
Shuyu Hou was supported by a graduate fellowship from the Syracuse Biomaterials Institute.
We thank Arne Heydorn at the Technical University of Denmark for providing the COMSTAT software, as well as John S. Parkinson at the University of Utah for providing the strain of E. coli RP437. We are grateful to Jiejing Qiu for helping with statistical analysis.
Published ahead of print on 22 January 2010.
†Supplemental material for this article may be found at http://aem.asm.org/.