Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
Mark D.P. Willcox, Institute for Eye Research, Level 4, North Wing, Rupert Myers Building, Gate 14, Barker Street, The University of New South Wales, Sydney, NSW 2052, Australia. E-mail: firstname.lastname@example.org
Aims: To develop an antimicrobial peptide with broad spectrum activity against bacteria implicated in biomaterial infection of low toxicity to mammalian cells and retaining its antimicrobial activity when covalently bound to a biomaterial surface.
Methods and Results: A synthetic peptide (melimine) was produced by combining portions of the antimicrobial cationic peptides mellitin and protamine. In contrast to the parent peptide melittin which lysed sheep red blood cells at >10 μg ml−1, melimine lysed sheep red blood cells only at concentrations >2500 μg ml−1, well above bactericidal concentrations. Additionally, melimine was found to be stable to heat sterilization. Evaluation by electron microscopy showed that exposure of both Pseudomonas aeruginosa and Staphylococcus aureus to melimine at the minimal inhibitory concentration (MIC) produced changes in the structure of the bacterial membranes. Further, repeated passage of these bacteria in sub-MIC concentrations of melimine did not result in an increase in the MIC. Melimine was tested for its ability to reduce bacterial adhesion to contact lenses when adsorbed or covalently attached. Approximately 80% reduction in viable bacteria was seen against both P. aeruginosa and S. aureus for 500 μg per lens adsorbed melimine. Covalently linked melimine (18 ± 4 μg per lens) showed >70% reduction of these bacteria to the lens.
Conclusions: We have designed and tested a synthetic peptide melimine incorporating active regions of protamine and mellitin which may represent a good candidate for development as an antimicrobial coating for biomaterials.
Significance and Impact of the Study: Infection associated with the use of biomaterials remains a major barrier to the long-term use of medical devices. The antimicrobial peptide melimine is an excellent candidate for development as an antimicrobial coating for such devices.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
It is predicted that every person in a technologically advanced society will host an implant at least once (Gristina 1987). These implants range from the simple use of contact lenses or catheters to life-saving devices such as the artificial heart. Although implants have led to a better quality of life and longer patient survival, device-related infection has emerged as a significant problem. Such infections are currently a major barrier to the long-term use of medical devices in treating various diseases and abnormalities. Reports of biomaterial infection rates for catheters reaching over 25% (Maki et al. 1988; Flowers et al. 1989) with cardiac abdominal and extremity vascular prostheses related infections resulting in the death of more than 30% of recipients (Masur and Johnson 1980; Bandyk et al. 1984). Similarly, it is estimated that 40–50% of patients with device-related endocarditis do not survive the first hospitalization (Douglas and Cobbs 1992). No effective therapies are currently available for the treatment of device-related infections and treatment usually requires removal of the device.
While many bacteria are particularly aggressive pathogens in their own right, once bacteria colonize a surface and differentiate into complex communities or biofilms, they become especially difficult to eradicate. Biofilms are a differentiated, high-density population of bacteria that are embedded in an extracellular polysaccharide matrix that protects the cells from stressful conditions such as desiccation and nutrient limitation (Costerton et al. 1981). Biofilms represent a particular challenge for antibiotic therapy. Cells within a biofilm can be up to 1000 times more resistant to antibiotics than planktonic forms (Costerton et al. 1999).
A variety of micro-organisms have been implicated in device-related infections including Staphylococcus aureus, Pseudomonas aeruginosa, coagulase-negative Staphylococci Enterococcus faecalis, Streptococcus viridans, Escherichia coli and Proteus mirabilis (Donlan 2001), with staphylococci and pseudomonas being among the most common. Current preventative measures to reduce the incidence of device-related infections focus on reducing contamination during implantation and care, prompt treatment of peripheral infections, antibiotic prophylaxis and biomaterial modification. Several antibiotic coatings including cefazolin (Kamal et al. 1991), minocycline–rifampin (Darouiche et al. 1999), teicoplanin (Bach et al. 1996), and vancomycin (Thornton et al. 1996) have been tested on biomaterials, particularly catheters. Alternative antimicrobial coatings such as silver (Maki et al. 1988; Jones et al. 2006), salicylic acid (Bryers et al. 2006), quaternary ammonium compounds (Ravikumar et al. 2006) and chlorhexidine (Leung et al. 2005) have also been trialled. However, in many cases development of microbial resistance (Neu 1992) makes these forms of therapy short-lived. Further, chlorhexidine and silver coatings show no clinical advantage over uncoated materials (Ciresi et al. 1996; Bong et al. 2003) and chlorhexidine may induce anaphylaxis, while silver and quaternary ammonium compounds can be cytotoxic (Lam et al. 2004). Other common problems with antimicrobial coating strategies are loss of activity after covalent attachment to devices and the inability to sterilize the coatings once attached.
Many cationic peptides or proteins are antimicrobial and probably mediate their activity by forming pores in the lipid bilayers of bacteria (Bechinger 1997). To increase the diversity of bacteria that can be inhibited by cationic peptides, several experiments have been performed using mixtures of peptides. Synthetic peptides containing lysine as the cationic moiety were shown to have greater activity when used in combination (Mor et al. 1994). The incorporation of active moieties from different cationic peptides in one single molecule has been examined. For example, a series of peptides made from combinations of cecropin A and melittin retain most of their antibacterial efficacy (Boman et al. 1989). Nos-Barbera et al. (1996) have used a hybrid of cecropin A and melittin in an experimental model of microbial keratitis and demonstrated that these peptides were able to reduce signs of infection and inflammation in this model. However, some portions of melittin used in this study (from GIGAVLKVLTTGLPALIS to VLKVL) would be expected to induce toxicity in mammalian cells.
The aim of the current study was to develop an antimicrobial peptide which has broad spectrum activity towards bacteria implicated in biomaterial infection, but is of low toxicity to mammalian cells. Further, this peptide must retain its antimicrobial activity when bound to a biomaterial surface. Additionally, the antimicrobial preferably should be stable under conditions that are used to heat-sterilize biomaterials and should have limited capacity to induce resistance in micro-organisms. We have designed a peptide incorporating active regions of protamine (from salmon sperm) and mellitin (from bee venom) which meets these criteria.
Materials and methods
Bacterial strains and growth conditions
Bacterial strains and their sources are listed in Table 1. The majority of the cationic proteins/peptides screening was performed with strains P. aeruginosa 6294 and S. aureus 31. Other listed strains were used to validate results or assess the effectiveness of the compounds over a range of bacteria.
Bacteria were grown overnight in Tryptone Soya Broth (TSB; Oxoid, Basingstoke, UK) and then washed three times in phosphate buffered saline (PBS; NaCl 8 g l−1, KCl 0·2 g l−1, Na2HPO4 1·15 g l−1, KH2PO4 0·2 g l−1). Bacteria were then resuspended in TSB to an OD660nm of 0·5 for S. aureus 31 and 0·1 for Pseudomonas and other species.
Synthesis of peptides
Peptides (Table 2) were synthesized by conventional solid-phase peptide synthesis protocols and were obtained from Auspep (West Melbourne, Vic., Australia). Peptides that were >80% pure were used in experiments. Protamine was obtained from Sigma (St Louis, MO, USA).
Table 2. Peptides
T L I S W I K N K R K Q R P R V S R R R R R R G G R R R R
G I G A I L K V L A T G L P T L I S W I K N K R K Q
P R R R R S S S R P V R R R R R P R V S R R R R R R G G R R R R
Attachment of protein to contact lenses
Peptides were prepared in distilled water. Etafilcon A lenses (CAS-61463-79-4; Vistakon, a division of Johnson & Johnson Vision Care Inc., Jacksonville, FL, USA) were removed from the manufacturer’s vials, washed three times in 1 ml PBS. The lenses were dried and then various amounts (final quantity: 125, 250 or 500 μg per lens) of peptides were pipetted onto both surfaces of the contact lens in a volume of 50 μl per lens and then left to dry overnight at room temperature in a laminar flow hood. Postabsorption, lenses were rehydrated in PBS for 10 min then washed four times in PBS with shaking.
For covalent attachment of melimine onto a contact lens, contact lenses (etafilcon A) were washed twice in 0·1 mol l−1 sodium acetate buffer pH 5·0, soaked in 2 ml 0·1 mol l−1 sodium acetate buffer pH 5·0 with 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) at a final concentration of 2 mg ml−1 and then allowed to react for 15 min at room temperature. Contact lenses were then washed three times in PBS pH 7·4 and then resuspended in 1 mg ml−1 melimine in PBS and incubated for 2 h at 37°C with mixing. Lenses were then washed four times in PBS.
To estimate the amount of attached peptide (either adsorbed or covalently attached), a method adapted from Cole and Ralston (1994) was used which has the advantage of revealing the topography of the peptide attachment to the lens surface. Briefly, contact lenses were stained for 24 h using filtered 0·025% Coomassie Blue in 10% acetic acid and 10% isopropanol at 37°C then destained in 10% acetic acid and 10% isopropanol at 37°C. Lenses were then extracted in 25% pyridine overnight. The extracted solutions were analysed spectrophotometrically at 600 nm using 25% pyridine as a blank. Melimine quantification was determined by correlating extracts against a standard curve constructed by pipetting known amounts of melimine on semi-dried acrylamide gels and extracting as above.
Bacterial colonization of contact lenses
Bacteria (1 ml) prepared as described above was allowed to adhere to contact lenses that had been treated with peptides (either adsorbed or covalently attached) or control contact lens not treated with peptides or an EDC process control respectively. Relative adherence of bacteria to test materials was determined using a modification of a previously described method (Williams et al. 2003). Briefly, lenses were homogenized using a tissue homogenizer (Ultra Tarax T8 dispersing tool, IKA, Rawang, Malaysia) until the contact lens disintegrated. Serial dilutions were made in PBS and aliquots (20 μl) plated out on nutrient agar. After incubation overnight at 37°C, viable bacteria were enumerated as colony forming units mm−2. Results are expressed as the average percent reduction in adherent bacteria (compared to the uncoated control lens) of triplicate measurements performed on a minimum of four separate occasions.
The data were log transformed prior to data analysis. The groups were compared using analysis of variance. Post hoc multiple comparisons were performed if the overall anova result was significant. For the overall test result, P ≤ 0·05 was considered as statistically significant. For the post hoc testing, P < 0·01 was considered as statistically significant after adjusting for multiple comparison correction using Tukey’s correction.
Heat stability of melimine
A heat stability study was conducted to evaluate whether exposure of melimine to autoclaving (121°C for 15 min) would alter the ability of the peptide to inhibit bacterial growth.
Melimine was dissolved in PBS to a concentration of 1000 μg ml−1 and autoclaved at 121°C for 15 min. Equal volumes of bacterial suspension (in PBS at optical densities described above) and melimine were then mixed and incubated at 35°C for periods of up to 72 h. Controls were unautoclaved melimine. Aliquots of the bacteria/melimine solutions were removed after 24, 48 and 72 h and viable bacteria enumerated using a standard Miles and Misra plate count assay.
Ability of bacteria to become resistant to melimine in vitro
Prior to melimine exposure, minimal inhibitory concentrations (MICs) were determined for P. aeruginosa 6294 and S. aureus 31. Strains were grown for 18 h in TSB at 35°C, then washed twice in PBS and resuspended in TSB to an OD660 of 0·1. This suspension was then diluted 1 : 100 in TSB. The cationic peptide was serially diluted from 15 mg ml−1 and dilutions were added to wells of a 96-well microtitre plate. Controls were bacteria grown in the absence of peptides. Bacteria were incubated for 18 h at 35°C and turbidity was measured at 660 nm. The lowest concentration of peptide resulting in a lack of turbidity (corresponding to no cell growth) was considered to be the MIC.
Bacteria were grown overnight in TSB and inoculated 1 : 100 into fresh TSB containing subinhibitory levels of melimine (1 mg ml−1 for P. aeruginosa and 30 μg ml−1 for S. aureus) and incubated overnight. Bacteria from broth containing subinhibitory levels were then re-incubated in fresh TSB with the same concentration of melimine. This was repeated for 30 consecutive days. MIC for the bacteria from the final passage was determined. Any increase in MIC would indicate a potential for bacteria to become resistant.
Lysis of sheep red blood cells by peptides
Sheep red blood cells (Sigma) were washed and resuspended in PBS. Peptides at various concentrations ranging from 5 to 5000 μg ml−1 were added to the red blood cells and incubated at 37°C for 4 h. One hundred percent (100%) lysis was obtained using distilled water and 0% lysis was obtained using PBS. Lysis was measured by determining the concentration of haemoglobin released into solution (as the increase in absorbance at 540 nm) following centrifugation to remove whole red blood cells.
Exponential phase bacterial suspensions (OD660 1·0) were incubated with melimine at the MIC or at 0·5 × MIC for 30 min at 37°C (Friedrich et al. 2000). Bacteria were fixed with 2·5% glutaraldehyde in 0·1 m PBS and embedded in an agar plug. Plugs were postfixed with 1% OsO4 in 0·2 mol l−1 PBS, stained with 1% uranyl acetate and then dehydrated through a graded series of aqueous ethanols, embedded in resin and thin sections cut. Sections were viewed under a Phillips transmission electron microscope (400×) with standard operating conditions and examined for structural changes.
Effect of adsorption of peptides to contact lenses on bacterial colonization
Protamine or melittin or a combination of melittin with protamine was adsorbed at noncytotoxic concentrations onto contact lenses and the ability of bacteria to colonize these lenses was assessed. Protamine (500 μg per lens) was effective at reducing the colonization of P. aeruginosa 6294 by approx. 65% but did not reduce colonization by S. aureus (Table 3). Melittin (15 μg per lens) was effective at reducing colonization of S. aureus 31 by approx. 95% but reduced P. aeruginosa colonization by less than 10% (Table 3). A combination of protamine (500 μg per lens) and melittin (15 μg per lens) resulted in significant reduction in only P. aeruginosa colonization at these concentrations (P < 0·01) but did not reduce levels of S. aureus.
Table 3. Effect of adsorption of peptides onto contact lenses on colonization of biomaterial by Pseudomonas aeruginosa strain 6294 and Staphylococcus aureus strain 31
P. aeruginosa percentage reduction
S. aureus percentage reduction
*P < 0·01 for P. aeruginosa adhesion compared to its appropriate control.
†P = 0·008 for S. aureus compared to its appropriate control.
‡P = 0·004 for a comparison of efficacy between mellitin and protamine + melittin.
§P < 0·001 when compared to its appropriate control.
¶P < 0·001 when efficacy is compared between P. aeruginosa and S. aureus.
Results are expressed as a percentage reduction in adherent bacterial numbers compared to an untreated contact lens.
Protamine (500 μg per lens)
65 ± 10*
Melittin (15 μg per lens)
17 ± 5
98 ± 16†‡
Protamine (500 μg per lens) + melittin (15 μg per lens)
64 ± 12*
0 ± 9†‡
Melimine (125 μg per lens)
60 ± 39§
5 ± 12¶
Melimine (250 μg per lens)
83 ± 11§
Melimine (500 μg per lens)
92 ± 5§
76 ± 20§
Effect of adsorbed melimine on bacterial colonization
The synthetic peptide melimine was adsorbed on to contact lenses at concentrations of 125, 250 and 500 μg per lens. No significant increase in reduction in the colonization of the contact lens by P. aeruginosa was observed (P < 0·001) between the three concentrations of melimine and a reduction of greater than 90% was found at 500 μg per lens (Table 3). In the case of S. aureus there was no significant reduction in colonization observed at 125 and 250 μg per lens, but a significant reduction of approx. 80% was found at 500 μg per lens (Table 3). At this concentration of melamine, there was no significant difference in its efficacy between the two strains of bacteria.
The effect of adsorbing melimine (500 μg per lens) on the ability to inhibit colonization of contact lenses by other strains of bacteria was examined (Fig. 1). At this concentration, melimine inhibited all of the bacterial strains tested and was particularly effective against strains of P. aeruginosa and S. aureus inhibiting colonization by at least 80% and was also able to reduce colonization of S. pneumoniae by more than 65% (Fig. 1).
Effect of covalently attached melimine on bacterial colonization
Determination of melimine covalently attached to the contact lenses using Coomassie Blue was found to be 18 ± 4 μg per lens and melamine was shown to be evenly distributed over both surfaces of the contact lens. When tested for efficacy against P. aeruginosa and S. aureus it was found to inhibit bacterial colonization by approx. 70% for both bacteria (Table 4).
Table 4. Effect of the peptide melimine covalently attached to the surface of a contact lens on bacterial colonization by Pseudomonas aeruginosa strain 6294 and Staphylococcus aureus strain 31
P. aeruginosa CFU mm−2
S. aureus CFU mm−2
Values are averages with standard deviations of triplicate measurements conducted on a minimum of four separate occasions.
11 240 ± 640
338 ± 300
3480 ± 140
45 ± 29
Heat stability of melimine
No differences were observed in the bacteriocidal efficacy of heat-treated melimine compared to control melimine solutions for either P. aeruginosa or S. aureus at any of the time points for up to 72 h (Fig. 2).
Ability of bacteria to become resistant to melimine in vitro
The MIC for P. aeruginosa was found to be 4 mg ml−1 while that for S. aureus was 125 mg ml−1. These values did not change over time suggesting limited potential for development of resistance to melimine.
Lysis of sheep red blood cells by peptides
Melittin caused extensive haemolysis of red blood cells and approx. 90% of cells were haemolysed at a concentration of 372 μg ml−1 (Fig. 3). In contrast, melimine did not cause any haemolysis at this concentration. The first appreciable haemolysis of red blood cells by melimine (10%) was observed at a concentration of 2500 μg ml−1, and exposure to 5000 μg ml−1 melimine resulted in approx. 50% haemolysis (Fig. 3). These concentrations are many times greater than the MICs for melimine. Protamine gave similar results to melimine (Fig. 3).
Exposure of bacteria to a concentration of melimine 0·5 times the MIC did not result in any structural changes of the bacterial cell wall (data not shown) compared to cells incubated with PBS (Fig. 4a,c). In contrast, bacteria exposed to melimine at the MIC for 30 min showed changes in morphology (Fig. 4b,d). Pseudomonas aeruginosa cells showed membrane blebbing at the cell surface, loss of membrane integrity, condensation of cytoplasmic contents and separation of the membrane from the cell wall (black arrow, Fig. 4b). Staphylococcus aureus (Fig. 4d) also exhibited changes in ultrastructure. Fibres can be seen extending from the cell surface (arrow, Fig. 4d), cytoplasmic contents were condensed and variability in the cell wall thickness was observed in many cells. Some cells also showed abnormal septal wall formation (Fig. 4d).
There is a clear need for a broad spectrum antimicrobial that prevents colonization of biomaterials, but does not damage mammalian cells, does not induce resistance and is stable throughout the sterilization process. Cationic peptides have the potential to meet these criteria and here we report development of one such synthetic peptide. It has been shown that a peptide incorporating the active regions of two peptides with differing spectra of activity can result in improved broad spectrum activity (Boman et al. 1989). However, many of these peptides/proteins are toxic to mammalian cells (Juvvadi et al. 1996; Subbalakshmi et al. 1999). This limits their usefulness as in vivo antibiotic-like substances. Recently, Subbalakshmi et al. (1999) have shown that the C-terminal 15 amino acid residues of melittin, a potent cationic peptide showing antibacterial and haemolytic activity, retains its antibacterial activity but has greatly reduced haemolytic activity compared to the full length melittin. Placing the C-terminal of melittin at the N-terminal of synthetic peptides was shown to produce antimicrobially active peptides with reduced mammalian cell cytotoxicity (Juvvadi et al. 1996). However, the range of bacteria that were inhibited by the C-terminal peptide was decreased and the amount of peptide needed to inhibit those bacteria was increased (Subbalakshmi et al. 1999). For use in vivo the toxic portion of the peptide fragment was removed (Subbalakshmi et al. 1999) and the remaining peptide contributed to melittin. A protamine fragment was synthesized to obtain sufficient sequence that could be linked to the melittin fragment whilst retaining the antimicrobial activity resulting in the synthetic peptide melimine (Table 3). As shown by electron microscopy, this peptide retained activity against both P. aeruginosa and S. aureus separately in solution resulting in a loss of membrane integrity and ultrastructure (Fig. 4). This synthetic peptide had the advantage over combined use of protamine and melittin in that activity against S. aureus was retained (Table 3). These findings were consistent with changes observed by others in bacteria exposed to effective antimicrobial cationic peptides (Shimoda et al. 1995; Friedrich et al. 2000; Mangoni et al. 2004). The peptide also showed a concentration-dependent ability to reduce bacterial colonization when adsorbed to a biomaterial with good broad spectrum efficacy at 500 μg per lens (Fig. 1).
Additionally, we have demonstrated that this peptide retains its activity when covalently linked to the polymer via an EDC coupling reaction. Coupling of the peptide to the surface by this method will result in attachment via reactive groups located at the termini or within the peptide, our findings of retained activity differ from those of Haynie et al. (1995) where activity was retained only when peptides were coupled via the carboxyl terminus. This may reflect differences between the methods of attachment or result from the different peptide sequences used. This is an important finding as covalent linkage would be required for development of an antimicrobial biomaterial. The covalent coupling, in fact, shows greater activity as 20 μg per lens gave an approx. 70% reduction in bacterial adhesion for both the Gram-positive and Gram-negative bacteria tested (Table 3). This apparent increase in efficacy may be related to the relative surface availability of the peptide compared to the adsorption process where there may be aggregation of the peptide and the possibility of uneven peptide distribution compared to the homogeneous distribution of the peptide after covalent coupling. Additionally, some peptide may have leached from the lens surface during the rehydration and washing steps following adsorption.
Melimine was also found not to induce resistance after repeated exposure of the bacteria to sub-inhibitory concentrations of the peptide. This development of resistance is a major problem for currently available antibiotic coatings such as cefazolin, minocycline–rifampin and vancomycin and other conventional antibiotics (Neu 1992). Recently, gentamicin-resistant staphylococci were recovered from gentamicin-loaded PMMA beads implanted in a patient after arthroplasty (Neut et al. 2003), highlighting the dangers of the development of resistance from exposure to biomaterial-associated antimicrobials. Such development of resistance is also associated with antibiotic-loaded bone cements (van de Belt et al. 2001) and represents a major barrier to the use of antimicrobial coatings on biomaterials.
A further requirement for a useful antimicrobial coating for a biomaterial is that it is nontoxic to mammalian cells. We have shown that melimine (even at high concentrations) does not cause the lysis of red blood cells, an accepted indicator of cytotoxicity. This is in contrast to melittin which results in almost complete lysis at concentrations at which melimine does not show any lysis (Fig. 3). Another advantage of the peptide melimine is its stability to heat (Fig. 2). Considered together, our findings suggest that the synthetic peptide melimine is a good candidate for further development as an antimicrobial coating for biomaterials having the properties of broad spectrum activity, biocompatibility, lack of induction of resistance and retention of activity when covalently linked to a biomaterial surface.
This research was supported under Australian Research Council’s Discovery Projects funding scheme (Project number DP0663368). The authors thank Dr Judith Flanagan for assistance with the preparation of this manuscript.