Combating bacterial resistance by combination of antibiotics with antimicrobial peptides

Introduction

Humanity has benefited greatly through the discovery and advancement of antibiotics. However, these improvements in healthcare have slowed as novel mechanisms of resistance have developed and been able to spread throughout the world [1]. This has led to a situation whereby a minor infection with a resistant strain of bacteria can lead to death of the patient. The World Health Organisation (WHO) has discussed the likelihood of this increasing in the coming years [2], which has led to worldwide changes in regulations to stem the spread of these resistant genes from one bacterial species to another. The spread of these genes has become a well-documented event and has been reported for several traditional antibiotics [3]. Antimicrobial peptides (AMPs) are a novel class of alternatives to these traditional antibiotics and have been shown to possess similar or even more effective activity profiles [4]. AMPs are found throughout nature and are a part of the innate immune response in all organisms [5] as protection from invading pathogens [6], [7], e.g. AMPs are used by human immune cells to damage and destroy invading bacterial cells [8]. In humans, AMPs also are widely seen in epithelial surfaces such as skin, eyes, ears, mouth, gut, nervous and urinary systems [8]. Over the past two decades, hundreds of such peptides, many readily water-soluble, with potent antimicrobial activity have been isolated from single-celled organisms, invertebrates and vertebrates [9].

However, like the peptide hormone insulin, AMPs are subject to proteolytic and gastric degradation [10] which present a challenge for oral delivery. This has resulted in delivery through topical and sub-dermal routes to avoid the proteases in the gastrointestinal environment that enable most small molecule drugs to be distributed throughout the human body. As well as novel delivery methods, a number of approaches have emerged to improve stability, including constrained AMPs, cyclotides, hybrid AMPs, AMP conjugates, AMP mimetics and immobilised AMPs [11]. The pharmacokinetics of several AMPs were studied by Dosler and Mataraci who found that the minimum inhibitory concentration (MIC) and minimum biofilm-eradication concentration ratios of AMPs were less than those exhibited by traditional antibiotics [12]. Such exciting revelations give hope that AMPs can be developed as promising therapeutic agents against bacterial and fungal pathogens [11].

AMPs are active against a wide range of both Gram-positive and Gram-negative bacterial species and have an ability to penetrate the bacterial cell [13] as well as replicate the mechanisms of action utilized by traditional antibiotics [14]. These mechanisms overall lead to the death and lysis of the bacterial cell and include inhibiting cell wall and cell membrane production or preventing synthesis of folate, nucleic acids (DNA, RNA) and proteins [15]. The ability of an AMP to insert into the cell membrane and penetrate the defences of the bacteria is key to understanding how the peptide confers its antimicrobial activity. Many models for membrane activity have been developed over the years (Fig. 1) including the barrel stave, carpet model, and toroidal pore structures [16]. These forerunner models rely on the notion that these membrane-active peptides are attracted to the membrane surface by electrostatic forces and, once they reach a threshold concentration, are then able to insert into the membrane to form peptide-lined pore structures [17]. More recent modifications to the trans-membrane pore model have also been reported and advancements in solid-state NMR [18] have led to a number of studies using model bacterial membranes, which are now being extended to live bacterial cells [19].

Fig. 1:

Different modes of action that have been described for membrane-active AMPs. Adapted from L.T. Nguyen, E.F. Haney, H.J. Vogel, Trends in Biotechnology 2011, 9: 464 [16].

These membrane structural changes have been studied in detail using not only naturally occurring peptides but also the so-called “designer” peptides which have used rational design by selecting portions of active amino acid sequences of AMPs to develop peptides with the desired traits [20]. AMPs have been developed through natural selection of traits through evolution which has resulted in lengths of 10–50 amino acids, overall cationic charge and an amphipathic nature [21]. These peptides may counteract the resistance that bacteria have employed to circumvent the action of traditional antibiotics [22]. This has provided some hope that antibiotic resistance will not lead to the apocalyptic scenario that has been predicted and that AMPs could be used to fight not only resistant strains but also be employed to enhance the effectiveness of traditional antibiotics. Studies of synergism that involve the co-administration of traditional antibiotics and AMPs [23] as well as antibiotic-AMP conjugates, where an antibiotic is covalently bound to the peptide [24], have shown in several instances that these resistant strains can be targeted and killed.

The widespread appropriate and inappropriate use of antibiotics in humans and farm animals has resulted in the emergence of antimicrobial resistance (AMR). The rates and severity of infections caused by AMR bacteria are increasing and are becoming harder and more complicated to treat and manage [25]. In a 2013 report by the Centres for Disease Control, 2 million patients per year develop a bacterial infection and 23 000 people die from these infections [26]. Of concern is that the majority of these deaths were attributable to three multidrug-resistant (MDR) Gram-negative bacterial infections caused by Enterobacteriaceae spp., Acinetobacter baumannii and Pseudomonas aeruginosa. The ESCAPE pathogens (Enterococcus faecium, Staphylococcus aureus, Clostridium difficile, Acinetobacter spp., P. aeruginosa and Enterobacteriaceae spp.) are considered to be the most important AMR threats by Gram-negative bacteria, the last three, being prioritised by the WHO as the threats [27], [28].

Bacteraemia, urinary tract infections, intraabdominal infections and ventilator-associated pneumonia are the four major hospital acquired infections and are commonly caused by Gram-negative bacteria with Escherichia coli, Klebsiella pneumoniae and P. aeruginosa accounting for 70% of these infections [29]. These infections are becoming harder to treat as the prevalence of MDR is increasing, for example in Australia the prevalence of MDR in E. coli was 12.6%, Klebsiella spp. 10.6% and Enterobacter spp. 9.7% in 2011, an average rise of 2–3% from 2009 [30]. In the USA, MDR prevalence is generally higher: E. coli and Klebsiella spp. 13%, Enterobacter spp. 3.7%, P. aeruginosa 15.4% and A. baumannii 60% in 2010 [29]. As well as having a toll on human health, MDR Gram-negative bacterial infections are associated with significantly higher economic burden with treatment costs and lost productivity costing an estimated $55 billion in the USA in 2008 [26].

Resistance mechanisms (Fig. 2) in bacteria include the ability to utilise an efflux mechanism to remove the antibiotic [31], development of an immunity bypass system which confers an “immunity” to the antibiotic action [32], and an ability to block the antibiotic mode of action [33]. A common mechanism of resistance is modification of target sites, which allows the bacteria to literally metabolise the drug by modifying its binding site [34] or produce enzymes to degrade the antibiotic [35]. Conjugation of antibiotics to AMPs could allow synergistic effects to be utilised and bypass resistance mechanisms that have evolved in bacteria.

Fig. 2:

Antibiotic targets and mechanisms of resistance adapted from G.D. Wright, BMC Biology 2010, 8:123 [36].

Overview of antibiotics

Since the discovery of penicillin by Alexander Fleming in 1928, humanity has been able to be a step ahead of bacteria, with a myriad of life threatening diseases being readily controlled with antibiotics [37]. Penicillin was serendipitously discovered by Fleming, when he noticed a spore of an Ascomycetous mould showed an inhibition of growth when it had landed on a petri dish of S. aureus [37]. Later Howard Florey and others discovered that an antibacterial substance secreted by the mould was responsible for the antibacterial effect [38], which was later named penicillin. Production issues hampered the preparation of the drug but it found use in the later years of World War II after being tested on soldiers in North Africa in 1943 [39]. This also enabled the optimisation of production and purification which lead to mass production of penicillin in several countries and was eventually made available for civilian use [40]. Penicillin was the first of the β-lactam family of drugs which has grown to include an assortment of antibiotic compounds containing the beta lactam-ring core structure required for the antibacterial effects [41]. Modifications of the structures of these drugs have led to increased effectiveness [42], reduction in adverse side effects [43] and optimised safety profiles [44], as with most pharmaceutical discoveries. The utilisation of the pyrrolidine ring led to the sub-family carbapenems [45] while the modification of the side chains of the cephalosporin C structure has led to new generations of the original cephalosporin C isolated from a fungi of the genus Acremonium [46]. The natural evolution of drug discovery has led to a wide range of antibiotic classes including the aminoglycosides, the glycopeptides, the tetracyclines and the macrolides [46]. Each has similarities and variations in the mode of action that they utilise to produce their differing spectrums of activity [47].

Antibiotic mechanisms of action

Aminoglycosides such as streptomycin and kanamycin inhibit the synthesis of proteins used for cell wall production in the target bacteria [48], which leads to the death of Gram-positive and Gram-negative bacterial cells. Glycopeptides such as vancomycin are not effective against Gram negative bacteria as they target the peptidoglycan synthesis of the cell wall of the bacteria by binding to amino acids within the cell wall itself [49]. The tetracycline and macrolide classes of drugs inhibit the synthesis of proteins [50] but each utilise different biochemical pathways and thus are effective against divergent species of bacteria [51]. Other mechanisms (Fig. 2) that antibiotics utilise include inhibiting the synthesis of DNA and RNA [52], such as used by levofloxacin and folate [53] as seen in the use of trimethoprim, and depolarisation of the cell membranes of Gram-positive bacteria [54], which is exhibited by the antibiotic daptomycin, a lipopeptide. As antibiotic discovery and development has evolved, bacteria have adapted to the antibiotic environment pressures by developing mechanisms of resistance to antibiotic action [55].

These environmental pressures have led to rapid emergence of antibiotic resistance especially in the case of the quinolone antibiotics such as levofloxacin where, within 2 years, resistance developed after commercial release [56]. It has also led to the use of chloramphenicol being reduced, not only due to resistance [57] but also the increased risk to health with higher doses to kill emerging resistant strains [58]. Resistance has arisen in several ways due to factors such as the widespread use and abuse of antibiotics in healthcare [59] and particularly in the overuse of antibiotics in the agricultural industry [60]. The ability of bacteria to share genetic material has resulted in rapid spread of resistance mechanism genes and the emergence of multi-and extensively drug resistant (MDR and XDR, respectively) bacteria species [61].

AMP mechanisms of action

The mechanisms that antibiotics use to produce their action vary across the different classes and families of drugs. The same can be said for AMPs with a wide range of activities seen among the different peptide structures and particular amino acid sequences [62] but often with similar MIC values to that of conventional antibiotics (Table 1). There are several similarities that exist between AMP and antibiotic modes of action, as some antibiotics are in fact also peptides and, therefore, follow similar biochemical pathways as other AMPs [63]. The glycopeptide class of antibiotics contains a tricyclic peptide with a glycoside residue [64]. These drugs bind to the amino acids acyl-D-alanyl-D-alanine to inhibit the synthesis of peptidoglycan involved in the cell wall structure of the bacteria [65]. Most AMPs have been found to be membrane-active [66] and an understanding of their ability to enter a bacterial cell is key to understanding their antimicrobial effect. Model membranes have been used to increase this understanding by determining how a peptide inserts into the bacterial membrane [67] and what advantages or disadvantages each has on the effectiveness of the antimicrobial properties [68].

Several existing models for membrane activity, such as the carpet, barrel-stave and toroidal pore structures, have been used for many years [69]. This has led to a general understanding of how an AMP is able to exert its effects on different bacterial species. As previously noted, the peptides were thought to reach a threshold concentration and form the membrane model structures within the membrane bilayer to allow them to penetrate the cell [17]. The carpet model was first proposed by Shai [70] who concluded that the peptide cecropin P1 was found to only be active when it formed a “carpet” on the bilayer surface (Fig. 1). This structure was found to have similarities with the toroidal pore model where both effects were seen and exhibit transient holes in the membrane [71]. In 1995, Matsuzaki and co-workers [72] showed that the toroidal pore structure, however, caused the phospholipid headgroups to bend back on itself like a “torus”, making it distinct from the carpet model. In the toroidal pore, both lipid headgroups and peptides line the lumen whereas in the barrel-stave model only peptide line the pore, i.e. the peptides interact laterally which enable a channel in the bilayer to be made [73]. The peptides insert into the membrane such that the hydrophobic surfaces interact with the lipid core of the membrane and the hydrophilic surfaces point inwards producing an aqueous pore (Fig. 1) [74].

The peptides that exhibit these pore structures have been found to adjust the type of model based on the composition of the membrane. An example is seen in the channel-forming peptide, pardaxin, which exhibited a carpet or barrel-stave structure depending on the membrane composition [75]. An updated model describes toroidal pore and barrel-stave models as “transmembrane models” which form spontaneously above a critical peptide to lipid ratio [76]. These updated models consider the helical structures of the peptides and that AMPs may permeabilise the membrane using a non-pore forming mechanism, such as a carpet-like mechanism, or produce transmembrane pores depending upon the peptide alignment on the membrane surface [77]. Generally, shorter peptides below 20 amino acids seem to utilise a carpet-like mechanism while longer peptides prefer to interact by forming transmembrane pores [24].

Magainin 2 and maculatin 1.1 are antimicrobial peptides (Table 2) found in an African clawed frog and Australian tree frog, respectively. These α-helical membrane-active peptides form transmembrane pores in multiple ways. Magainin, for example, re-orientates itself in the bilayer to line a proteinaceous channel as described in the barrel-stave model or in the form of a peptide-lipid complex as in the toroidal model or carpet model [78]. The ability of these peptides to exhibit multiple membrane structures and modes of action is not just seen in amphibian AMPs but in a wide range of peptides including the designed peptide, A3-APO, which is a discontinuous dimer of the monomer, Chex1-Arg20 [79]. The peptide has been shown to exhibit a dual mode of action with both membrane penetration and intracellular target inhibition [79]. Even the short 13 amino acid peptide, indolocidin, isolated from bovine neutrophils has been shown to cause an increase in the transmembrane current of planar bilayers and induce the formation of discrete channels [80]. Ubiquicidin (UBI) is an AMP found in human macrophage cells and part of the first line of defence which are released during an acute infection [81]. Synthetic fragments of this peptide have been made and show remarkable effectiveness against methicillin resistant S. aureus (MRSA) [81]. They have been shown to exhibit an even greater activity when coupled to the antibiotic (Table 2) chloramphenicol [82].

Table 2:

Antimicrobial peptide (AMP) sequences and antibiotic structures.

Peptide sequences			Antibiotic structures
Magainin 2 GIGKFLHSAGKFGKAFVGEIMKS			Levofloxacin	Kanamycin
Maculatin 1.1 GLFGVLAKVAAHVVPAIAEHF
Chex1-Arg20 Chex-RPDKPRPYLPRPRPPRPVR
A3-APO			Cephalosporin	Vancomycin
Chex-RPDKPRPYLPRPRPPRPVR \ Chex-RPDKPRPYLPRPRPPRPVR /	Dab
Pexiganan (MSI-78) GIGKFLKKAKKFGKAFVKILKK
Indolocidin ILPWKWPWWPWRR
Ubiquicidin MQIFVKTLTGKTITLEVEPSDTIENVKAKIQDKEGIPPD QQRLIFAGKQLEDGRTLSDYNIQKESTLHLVLRLRGG			Chloramphenicol
1-amino-cyclohexane-carboxylic acid		2,4-diaminobutyric acid

Antibiotic coupling to AMPs

The synergistic effects of AMPs and antibiotics have been reported to be effective against a range of resistant strains of bacteria [83]. The activity profiles of AMPs conjugated with antibiotics follows this trend with an increase in the antimicrobial action [84]. Methods of conjugating peptides with antibiotics depend on the structures of both but include utilizing amide bond formation attached to linking group [85], a disulfide tether [86] and a glutaraldehyde linker [87]. As solid phase synthesis of peptides can produce an amidated terminus on the peptide, using this to link the peptide to a drug is logical [88]. The covalent conjugation of the cationic AMP, Chex1-Arg20, and several other AMPs, including the magainin analogue MSI-78 with 7-cephalosporanic acid and/or 7-aminodesacetoxycephalosporanic acid, was undertaken by Li et al. [89]. This was achieved by attaching the 7-amino group of cephalosporin to the N-terminus of the peptide. However, only the MSI-conjugated analogues were found to produce synergistic activity against A. baumannii and multidrug resistant (MDR) A. baumannii 156 [89]. The antibiotic levofloxacin was conjugated to the peptide indolicin by Ghaffar et al. [90] using an amide bond or labile ester linkage, which showed little change in activity. Brezden et al. coupled kanamycin with a broad-spectrum AMP, P14LRR, with and without a disulfide linkage to produce a compound which releases kanamycin and P14SH as it metabolizes [91].

The conjugation of magainin with vancomycin via a copper(I)-catalyzed alkyne-azide cycloaddition (CuAAC) “click reaction”, however, showed the same or greater activity than vancomycin alone [92]. This increase in activity was also noted in the UBI variant coupled to the antibiotic chloramphenicol (CAP) via a glutaraldehyde linker, which not only showed enhanced activity against E. coli and S. aureus but led to a reduction in the toxicity against human cells compared with CAP alone [87]. This indicates that the synergistic effect that was seen in the combination of these drugs is also translated when they are conjugated to each other [93]. The modification of these compounds, however, has had issues arise when the active site for antimicrobial effects is blocked leading to a reduction in activity [94]. The conjugation of a drug or antibiotic onto a cell penetrating peptide has been investigated previously with the intention to use as drug carriers with a self-immolative linker that will liberate its cargo in situ [95] with enhanced specificity and less side effects.

Combination therapy and antibiotic adjuvants

A more commonly used treatment for AMR infections is combination therapy where multiple antimicrobial compounds are co-administered during treatment with the aim of producing a synergistic effect between the antimicrobials and enhancing activity [96]. Combination therapy is being used widely in the treatment of many health conditions [97] and has recently been regarded as a promising and cost-effective solution for bacterial infections to overcome the inadequacies of antibiotic monotherapy [98]. Through the combination of antimicrobial drugs that interact with different and multiple bacterial targets, combination therapy could have profound implications in reducing the likelihood of resistance acquisition [98], [99], re-sensitizing MDR bacteria to antibiotics that were otherwise ineffective [98], and mitigating toxic side-effects to the body as similar levels of antimicrobial efficacy could be attained with lower drug concentrations [99].

While synergistic antibiotic-antibiotic combinations are the most commonly reported [98], most antibiotic cocktails tend to be only effective against selective bacterial species and such combinations have also been reported to be associated with greater resistance due to increased selective pressure [97]. Therefore, recent studies have focused on combining antibiotics with a compound that potentiates or restores antibiotic activity towards MDR bacteria. These compounds, known as antibiotic adjuvants, are divided in to two classes: Class I adjuvants work with antibiotics on bacterial targets; and Class II adjuvants enhance the host’s antimicrobial systems resulting in the potentiation of antibiotic activity [100]. Class I adjuvants are divided into Class I.A compounds that directly inhibit antibiotic resistance mechanisms, e.g. inactivating enzymes or efflux pump systems, and Class I.B adjuvants that enhance antibiotic activity by circumventing intrinsic resistance mechanisms, e.g. enhancing membrane permeability or inducing metabolic pathways. AMPs have drawn great interest as a new class of therapeutics as their mode of action is distinct from currently used antibiotics and does not readily induce resistance [101]. The differences in the mode of action of AMPs to that of antibiotics has led to increasing interest in using AMPs as antibiotic adjuvants as there is potential for bactericidal synergism [102] (Fig. 3).

Fig. 3:

AMPs as antibiotic adjuvants.

The idea to use a cell-penetrating AMP to allow antibiotic drugs to traverse eukaryotic membranes has been mentioned above and was used by Brezden et al. [91] who conjugated kanamycin to the AMP, P14LRR, which allowed the drug to enter mammalian cells where the tether was then reduced within the cells to release the antibiotic. Synergistic antibacterial activity was achieved and the mechanism of the adjuvant appears to be an effective tool in combatting intracellular pathogenic bacteria [91].

Perspectives and future work

There is a clear need for the development of novel classes of antibacterials and treatment modalities for the use of current antibiotics and one area on which research should focus is AMP compounds and materials. As well as developing monotherapies based on AMPs, polypeptides such as structurally nano-engineered antimicrobial poly-peptides (SNAPPs) and oligomers [79], [103] exhibit activity against a range of bacteria, including the ESCAPE and MDR pathogens, and seem to proceed via a multimodal mechanism of bacterial cell death by targeting and destabilization of the inner membrane [103]. Other novel strategies include lipidation where an AMP has a fatty acid attached to the N-terminus [104], photodynamic therapy which utilises a photosensitiser that produces a reactive oxygen species that damages bacterial proteins when exposed to light [105], and organometallic complexes where a metal centre in the drug confers new functionalities and reactivities [106]. A highly novel approach as discussed here is to combine these new mono-therapeutics with traditional antibiotics in a combined therapeutic approach. These new strategies are a promising array of potential treatments which continually need to be developed to ensure humanity can adapt to new threats that arise. A recent editorial brought into focus the challenges the world faces currently and into the future in regard to death due to cancer and AMR bacteria [107]. The need for an interconnected research strategy to target cancer treatment and to address the global emergence of AMR was highlighted. Multidisciplinary approaches to these issues are leading to discovery and production of novel therapeutic materials [107]. Increasing discussion has been stimulated and should lead to further research to effectively take on this ever-growing world health issue of bacterial resistance.