Synergistic effect of clinically used antibiotics and peptide antibiotics against Gram-positive and Gram-negative bacteria

Ribosomally synthesized (natural) peptides demonstrate antimicrobial potency and may represent a novel therapeutic approach for the treatment of infections. The aim of the present study was to investigate the interaction between polycationic peptides and clinically used antimicrobial agents in the treatment of clinical isolates of Gram-positive and Gram-negative aerobic bacteria in vitro, using the microbroth dilution method. The combination studies demonstrated synergies between ranalexin and polymyxin E, doxycycline and clarithromycin. Similarly, magainin II was demonstrated to be synergistic with ceftriaxone, amoxicillin clavulanate, ceftazidime, meropenem, piperacillin and β-lactam antibiotics. Buforin II, cecropin P1 and indolicidin were not observed to be synergistic with the clinically used antibiotics, but demonstrated additive effects with them. Notably, no antagonistic effects were identified in all the combinations examined.


Introduction
The discovery of numerous peptide antibiotics has resulted in a novel area of research into antimicrobial agents. In particular, ribosomally synthesized (natural) peptides, due to their antimicrobial potency, may represent a novel therapeutic approach for the treatment of infections (1)(2)(3)(4). Antimicrobial peptides are produced in nature as a major component of the natural host defense molecules of a wide range of animals, plants and bacterial species (2,3,(5)(6)(7)(8)(9). Studies have demonstrated that this group of peptides possesses a broad spectrum of antibacterial activity, the site of action of which is the cytoplasmic membrane (2,3,(10)(11)(12)(13). Various mechanisms have been suggested to explain the mode of action of these compounds on the membranes of bacteria. These mechanisms include the binding of monomers to the membrane and insertion into the membrane to form an ion-channel pore that spans the membrane; and the carpet model in which peptide molecules saturate the surface of the membrane prior to extensively disrupting the permeability barrier (1,2). The lethal event that occurs at the cytoplasmic membrane is not fully understood; however, it has been indicated that peptides cause channel formation in the cytoplasmic membrane, resulting in cell death (2,3,11,13,14). In addition, it has been shown that peptides allow maximal entry of several hydrophobic substrates into the cell and exert synergistic effects with lipophilic and amphiphilic agents, including rifampin, macrolides, fusidic acid and novobiocin (13,15).
The present study investigated the interaction between five polycationic peptides and several clinically used antibiotics in the treatment of clinical isolates of Gram-positive and Gram-negative bacteria in vitro. Antibacterial susceptibility testing. The minimal inhibitory concentrations (MICs) of all compounds were determined using a microbroth dilution method with Mueller-Hinton broth (Becton Dickinson Italia, Milan, Italy) and an initial inoculum of 5x10 5 cfu/ml. The tests were conducted according to the procedures outlined by the National Committee for Clinical Laboratory Standards (16). Polystyrene 96-well plates (Becton Dickinson, Franklin Lakes, NJ, USA) were incubated for 18 h at 35˚C in air. The MIC was considered to be the lowest drug concentration at which observable growth was inhibited.

Synergistic effect of clinically used antibiotics and peptide antibiotics against Gram-positive and Gram-negative bacteria
Interaction studies. In the interaction studies, the four control strains and two representative strains for each species of Gram-negative and Gram-positive bacteria were selected.  (17).

Results
The in vitro activities of buforin II, cecropin P1, indolicidin, magainin II and ranalexin against different bacteria are presented in Table II. The peptides demonstrated various ranges of inhibitory values in the different species of bacteria. Overall, Gram-negative strains were more susceptible to buforin II and cecropin P1, but less susceptible to ranalexin. Furthermore, E. coli strains were highly susceptible to the peptides, whereas the isolates of P. aeruginosa were scarcely susceptible. Buforin II and cecropin P1 both demonstrated MIC 50 values of 0.50 µg/ml for E. coli and 8 µg/ml for P. aeruginosa. However, buforin II and ranalexin were the most active compounds against the staphylococcal isolates. They inhibited MS S. aureus strains at concentrations of 0.5-8 and 0.25-8 µg̸ml, respectively, and MR S. aureus strains at concentrations of 1-8 µg/ml. The combination studies indicated synergies between ranalexin and polymyxin E, doxycycline and clarithromycin, and the FIC indices aid in the quantification of the degree of synergy. The FIC indices demonstrated that the activity of ranalexin combined with polymyxin E, doxycycline or clarithromycin was four-to eight-fold greater compared with that of ranalexin alone (Tables III and IV). Furthermore, magainin II demonstrated synergistic effects with ceftriaxone, amoxicillin-clavulanate, ceftazidime, meropenem and piperacillin (Tables III and IV). The β-lactam antibiotics amoxicillin-clavulanate, ceftriaxone and meropenem decreased the MIC value of magainin II by a factor of eight (from 8 to 1 µg/ml) for the oxacillin-resistant control strain of S. aureus, ATCC 38591. Moreover, the FIC indices showed that ranalexin and magainin II acted synergistically with clinically used antibiotics against Gram-positive and Gramnegative organisms (Tables III and IV). Additive effects, but no synergy, were demonstrated by combinations of buforin II, cecropin P1 and indolicidin with the clinically used antibiotics. Notably, antagonism was not identified in the combinations studied (data not shown).

Discussion
Gram-negative and Gram-positive bacteria cause severe infectious diseases in mammals, and the emergence of their antimicrobial resistance is an increasing problem in human medicine. Throughout nature, polycationic peptides represent a conserved theme in antimicrobial defense. They may provide a new structural class of highly active antimicrobial agents, and potentially act as a resource for the development of novel anti-infective agents. Studies have indicated that polycationic peptides have variable antibacterial, antifungal and antiprotozoal activity in vitro (3,7,(18)(19)(20)(21)(22). Therefore, these compounds may be very valuable as adjuvants for antimicrobial chemotherapy. Furthermore, studies have demonstrated that to exert their antimicrobial and anti-endotoxic activity, the peptides must initially bind to lipopolysaccharides (LPSs) of the outer membrane of Gram-negative bacteria (2,3,13). LPSs trigger the acute phase response to infection and the molecules involved in the recognition process have been Table I. Primary structures of the five peptides.
Ranalexin, a 20-residue peptide, demonstrates structural similarity to the polymyxins, a class of membrane-active antibiotics (8). The polymyxins are a group of cyclic polycationic peptides originally derived from Bacillus polymyxa. Similar Table II. In vitro susceptibilities to polycationic peptides.    to the polymyxins, ranalexin is an amphipathic compound with a cationic heptapeptide ring at its carboxyl terminus (the molecule contains two cysteine residues in positions 14 and 20, linked by a disulfide bridge) and a hydrophobic region at its amino terminus. Polymyxins and polymyxin-like peptides act synergistically with lipophilic and amphiphilic agents, including rifampin, macrolides, fusidic acid and novobiocin (10,12). Furthermore, it has been demonstrated that polymyxin-like peptides allow maximal entry of hydrophobic substrates into the cell (15). The positive interaction between magainin II and β-lactam antibiotics may be due to an increased access of magainin II to the cytoplasmic membrane, following the breakdown of peptidoglycan by the β-lactam. However, other mechanisms may be involved in this interaction; the magainins may be membrane disruptive (6) and yield to the uncoupling of oxidative respiration (3,13). The peptide sequence of magainin II reveals that the molecule may exhibit large hydrophobic moments. When the peptide adopts an α-helical conformation in the solution, it is strongly amphiphilic, exhibiting a hydrophobic surface on one face and a hydrophilic surface on the other. It has been demonstrated that magainins are extremely surface active (6).
Studies have identified that buforin II may interact with biological membranes, in a similar manner to magainin; however, the mechanism of membrane interaction may be different. The length of its amphipathic region is ~24 residues, which is approximately two-thirds that of magainin II (35 residues). The amphipathic region of buforin II is unable to span the whole biological membrane (the hydrophobic region is >30 residues); therefore, it is not possible to directly apply the ion channel model that was suggested for magainin II to this peptide (26).
In conclusion, a number of polycationic peptides have been demonstrated to bind to the LPSs of Gram-negative bacteria and to self-promote their uptake into bacteria (3). However, some of these molecules have markedly lower affinities for LPS binding, but are still effective permeabilizers, potentially through a related but distinguishable method (3).
The results regarding the activity of clarithromycin against Gram-negative organisms are particularly noteworthy. Macrolides may inhibit protein synthesis by binding to the transpeptidation site of the larger ribosomal subunit. However, large hydrophobic antibiotic molecules such as macrolides are mostly ineffective against Gram-negative bacteria, as they are unable to diffuse across the outer membrane (9,27,28). Notably, the present study identified that clarithromycin and ranalexin had synergistic activity against Gram-negative bacteria, including P. aeruginosa. There is increasing support for the activity of macrolides as antipseudomonal agents. Studies have suggested that macrolides may inhibit P. aeruginosa cell growth in vitro; although clarithromycin and erythromycin (at 2 mg/l) demonstrated no effect on cell growth at 24 h, they were bactericidal against P. aeruginosa when the incubation was continued for 48 h (24).