Antimicrobial peptides properties beyond growth inhibition and bacterial killing

Survey methodology

To ensure an inclusive and unbiased analysis of literature and to accomplish the review’s objectives, a comprehensive analysis of published articles on the activity of antimicrobial peptides using the following online databases: Medline (PubMed), Science Direct (http://sciencedirect.com) database, Web of Science, Scopus, and Google Scholar system. Additionally, the following keywords were used: antimicrobial peptides, anti-virulence properties, quorum sensing, biofilms, targets together with Boolean operators such as “AND” and “OR”.

Anti-biofilm and anti-quorum sensing activity of AMPs

Biofilms are the preferred lifestyle of bacteria and are structured microbial aggregates, surrounded by a self-produced extracellular matrix, and attached to biotic or abiotic surfaces. Biofilms are involved in most chronic bacterial infections (Bjarnsholt, 2013). Moreover, they are crucial determinants of bacterial virulence. The biofilm matrix is formed by diverse components present in the extracellular polymeric substances: mainly proteins, polysaccharides, extracellular nucleic acids, and ions (Donlan, 2002). Biofilm formation is an ordered process, beginning with the initial contact and attachment to surfaces, mainly mediated by structures such as flagellum and fimbria, followed by micro-colony formation, maturation, and formation of the complex biofilm architecture, finally, detachment and dispersal of some cells from the biofilm occur (Sutherland, 2001).

Biofilms are pivotal for bacterial survival as they protect against adverse environmental conditions. They increase drug resistance by various mechanisms such as the decrease in the permeability of antibiotics, the promotion of dormancy and induction of bacterial persistence, the expression of the efflux pumps of antibiotics, and the synthesis of periplasmic glucans (aminoglycosides) that inactivate antibiotics (Hall & Mah, 2017). Biofilms also allow bacteria to evade the human defense mechanisms (Mirzaei et al., 2020) since several biofilm matrix proteins protect biofilms against human innate immune cells, opsonization, and phagocytosis (Lewis, 2008). Moreover, it has been demonstrated that some bacterial species, previously known as extracellular pathogens, can reside inside various host cells by adapting to intracellular life through the formation of microbial aggregates similar to bacterial biofilms, leading to their long-term survival inside the cells (Mirzaei et al., 2020).

Unlike antibiotics, AMPs are suitable for slowing growth and killing cells in the biofilm. Several examples of effective AMPs with this activity have been described that correlate with the ability of AMPs to resolve bacterial infections in vivo. For a recent full review of these activities and the translation potential of such peptides, see the work of Gislaine and coworkers (Silveira et al., 2021).

Since the aim of this work is to discuss the anti-virulence potential of AMPs, and one of the premises of anti-virulence therapies is not to affect directly bacterial growth and survival, most of the examples of AMPs with anti-biofilm activity discussed here will be peptides that inhibit biofilm formation at growth sub-inhibitory concentrations (Table 1).

AMP activity against biofilms is mediated by the degradation or destabilization of the extracellular matrix (Yasir, Willcox & Dutta, 2018). The PI peptide (derived from polyphemusin I) induces the degradation of the exopolysaccharides produced by Streptococcus mutans, causing the biofilm formation to be attenuated (Zhang et al., 2019). Also, an AMP complex produced by the insect Calliphora vicina promotes the degradation of the matrix of the biofilm produced by E. coli, Staphylococcus aureus, and Acinetobacter baumannii (Gordya et al., 2017). Hepcidin 20 from the human liver decreases the extracellular matrix and disrupts the architecture of S. epidermidis biofilms (Brancatisano et al., 2014). S4 (1-16) M4Ka (dermaseptin S4 derivative), which inhibits immature biofilms of Pseudomonas fluorescens (Quilès et al., 2016). Piscidin-3 is derived from fish, which degrades the extracellular DNA of P. aeruginosa biofilms (Libardo et al., 2017).

Biofilm inhibition by AMPs is also mediated by the downregulation of genes responsible for biofilm formation and transport of binding proteins; for example, in Staphylococcal biofilms, the β-defensin 3 from humans decreases the expression of the icaA, icaD, and icaR genes that codify enzymes responsible for the biosynthesis of the adhesin PIA, essential for biofilm formation (Rohde et al., 2010; Zhu et al., 2013). In addition, AMPs also inhibit genes that control the transport and binding proteins, such as ABC transporters that are involved in biofilm formation since they promote cell-to-surface and cell-to-cell interactions (Zhu et al., 2013; Wang et al., 2017).

Beyond inhibiting biofilm maturation, AMPs can also inhibit initial attachment and increase cell dispersal. One of the first discovered AMPs with the ability to eradicate biofilms was LL-37 (Overhage et al., 2008), derived from human cathelicidin, an amphipathic peptide widely distributed in body fluids (Burton & Steel, 2009). At low concentrations, LL-37 inhibits the adhesion of P. aeruginosa cells to surfaces, and at higher concentrations, it reduces the thickness of the biofilms (Hancock & Sahl, 2006). Moreover, LL-37 also eradicates P. aeruginosa biofilms in vivo (Chennupati et al., 2009). The anti-biofilm effects of LL-37 in P. aeruginosa at concentrations that do not affect viability and growth are related to the upregulation of the expression of type IV pili genes that lead to the promotion of twitching motility which is linked to biofilm dispersal and to the decrease in the expression of flagellar genes which leads to lower attachment to surfaces (Overhage et al., 2008). In addition, LL-37 treatment induces a strong down-regulation of the core genes of the Las and Rhl QS systems and the repression of genes that encode QS-dependent virulence factors such as LasB elastase and those responsible for the biosynthesis of rhamnolipids (Overhage et al., 2008). In addition, it inhibits the biofilm formation of other pathogens such as Francisella novicida and S. epidermidis (Amer, Bishop & van Hoek, 2010; Hell et al., 2010).

Other AMPs can prevent biofilm formation by inhibiting quorum sensing (Overhage et al., 2008). For example, Trp-containing peptides inhibit QS-regulated virulence and biofilm growth of multidrug-resistant P. aeruginosa. Significantly, peptides containing tryptophan at low concentrations reduced the production of virulence factors that regulate the gene expression of the Las and Rh1 systems. Biofilm formation was inhibited in a concentration-dependent manner, which was associated with inhibiting extracellular polysaccharide production by negatively regulating the transcription of pelA, algD, and pslA. These changes were correlated with alterations in the extracellular production of virulence and motility.

Also, two novel synthetic peptides (LIVRHK and LIVRRK) can inhibit biofilm formation of P. aeruginosa PA01, and QS-dependent phenotypes such as pyocyanin exoprotease, and rhamnolipid production were identified. In addition, a down-regulation of the expression of the core QS genes lasRI and rhlRI were observed, corroborating the inhibition of QS (Taha et al., 2019).

The discovery of the anti-biofilm and anti-QS properties of LL-37 led to the search for other natural and synthetic peptides with similar properties. De la Fuente and his colleagues in 2012 selected 50 small synthetic peptides and identified 16 with anti-biofilm activity against P. aeruginosa, with HH15 being one of the best. According to their sequence, 15 small peptides were designed, including peptide 1037 of only nine amino acids, reducing the biofilm formation of Burkholderia cenocepacia and Listeria monocytogenes (De La Fuente-Núñez et al., 2012).

The peptide 1037, like LL-37, stimulates twitching motility and decreases the expression of flagellar genes, leading to potent inhibition of swimming and swarming motilities (De La Fuente-Núñez et al., 2012). Comparison between the effect in gene expression of peptides (LL-37 vs. 1037) allowed the identification of ten common downregulated genes and four upregulated ones. The role of those genes in biofilm formation was confirmed using transposon mutants of each one, being nine of the ten mutants in the downregulated genes lower biofilm producers than the parental strain. Although the involved genes mainly were hypothetical proteins, the flagellar gene flgB, rhlB, involved in rhamnolipid biosynthesis and nirS encoding a nitrite reductase were identified. In addition, two of the four mutants in the upregulated genes (a hypothetical protein and actP, encoding an acetate permease have) higher biofilm producer than the parental strain (De La Fuente-Núñez et al., 2012).

In another study, it was shown that LL-37 exhibits anti-biofilm activity against S. epidermidis, where at low concentrations they prevent cell attachment, while at high concentrations, they prevent the maturation and establishment of biofilms. Also, LL-37 has a potent S. aureus biofilm eradication activity (Kang, Dietz & Li, 2019).

Other synthetic peptides such as D-Bac8c^2,5Leu, D-HB43, and D-ranalexin have effectively killed S. aureus biofilms. For example, the synthetic peptide D-Bac8c^2,5Leu, when applied as a catheter lock solution, has inhibitory activity on early and mature S. aureus biofilms in a rat venous catheter infection model (Zapotoczna et al., 2017).

Although some classic antibiotics at sub-MIC concentrations have shown anti-virulence and QS system regulation behaviors (Skindersoe et al., 2008; Zhang & Li, 2016), there is still little research on peptide antibiotics. However, bovicin HC5 (broad-spectrum lantibiotic) and nisin (polycyclic peptide antibiotic) have been reported to have anti-biofilm activity through QS interference from S. aureus (Pimentel-Filho et al., 2014). Similarly, subtilosin (cyclic lantibiotic) reduces violacein production in Chromobacterium violaceum (indicative of QS inhibition), as well as biofilm formation and autoinducer-2 (AI-2) production in Gardnerella vaginalis (Algburi et al. al., 2017) (Table 1).

Other anti-virulence targets of AMPs

Although biofilms and other QS-controlled phenotypes (exoproteases, phenazines, rhamnolipids, swarm motility) are essential for bacterial virulence, some important virulence factors are not positively regulated by QS. Eight secretion systems have been found in Gram-negative and Gram-positive bacteria. However, these systems are sometimes unregulated by QS or may even be downregulated, as in some vibrio species. Therefore, in cases where QS negatively regulates them, QS inhibition can promote virulence through secretion systems (Pena et al., 2019). Therefore, specific inhibitors of these systems in combination with QS inhibitors may be necessary to develop more robust antibacterial therapies (García-Contreras, 2016).

Accordingly, Zhang and coworkers elucidated the structural and functional details of the pseudopilus tip complex of the type II secretion system of P. aeruginosa, which is an essential component of the system that functions as a piston, allowing the export of multiple effectors (Zhang et al., 2018). Based on the structural details of the complex, two mimicking peptides [P1(EWESDNRLNEEQ) and P2 (TKLTRTWRQ)] that were able to compete with the binding of the XcpV and XcpW pseudolipins were designed, retaining the specific amino acids that allow the interaction between those pseudopilins and introducing other hydrophilic amino acids to enhance solubility. The utilization of those peptides precluded the formation of the core complex essential for the pseudopilus tip formation and strongly attenuated secretion through the type II secretion system (Zhang et al., 2018).

Interestingly, natural mammalian peptides, such as iron-binding lactoferrin, are potent inhibitors of T3SS in enteric bacteria (Salmonella, Shigella, and E. coli) by inducing translocon protein degradation. This activity is mediated by its binding to the lipopolysaccharide on the bacterial surface, destabilizing the protein-protein interactions essential for the system. Furthermore, lactoferrins have serine protease activity that can affect T3SS cleavage proteins (McShan & De Guzman, 2015).

Beyond the anti-T3SS of natural peptides, the strategy of using polypeptides that mimic some components of the systems and that compete with the binding of the natural bacterial components was effective to inhibit the system in Chlamydia, Salmonella, and Shigella, blocking their entrance to eukaryotic cells in cultures (McShan & De Guzman, 2015). Similarly, in enteropathogenic E. coli (EPEC), coiled-coil peptide mimetics, analogs of the EspA, EscF, and CesA proteins (CoilA, Coil B and CesA2) of its T3SS, inhibit the T3SS mediated hemolysis (Larzábal et al., 2019).

Another critical virulence determinant is the type VI secretion system, which delivers multiple effectors to prokaryotic and eukaryotic cells. Those effectors target cell walls, cell membranes, DNA, and to avoid self-poisoning, bacteria that produce them also produce neutralizing proteins that bid the effectors. Recently, in P. aeruginosa, the effector TplE, a lipolytic toxin effective against other bacteria and able to disrupt the endoplasmic reticulum in eukaryotic cells, had been characterized; this protein is neutralized by TplEi (Jiang et al., 2016).

Based on this interaction, Gao and coworkers generated a small peptide capable of competing with the TplEi-TplE interaction, for this TplE was hydrolyzed, generating a 26 amino acid fragment that strongly binds with TplEi, releasing the TplE toxin, and thus inducing the autointoxication of P. aeruginosa (Gao et al., 2017). This approach is attractive and represents a new concept for generating new inhibitors of secretory systems and other potential targets. A similar approach was recently used for the identification of small peptides that inhibit antitoxins that belong to the toxin-antitoxin systems (Lee et al., 2015; Sundar, Rajan & Piramanayagam, 2019), which are related to latency, persistence (Page & Peti, 2016) and bacterial virulence (Fernández-García et al., 2016). These systems are abundant in intracellular bacterial pathogens such as Mycobacterium tuberculosis (Sala, Bordes & Genevaux, 2014).

Additional anti-virulence activities of some AMPs, such as defensins are the capacity to bind and inhibit the activity of several bacterial toxins and related virulence factors (Table 1). Defensins are components of the innate immunity of mammals and are also found in invertebrates, plants, and fungi. Although these peptides had low sequence similarity, they share common structural features and display broad antibacterial and antiviral activity at high concentrations; in addition, they modulate inflammation and promote angiogenesis and wound healing. Moreover, they can neutralize several bacterial toxins, among them cytolysin, listeryolysin that promote pore formation, ribosyltransferase toxins, glycosylation promoting toxins, the MARTX toxins from Vibrio and Aeromonas, the Panton-Valentine leucocidin, staphylokinase from S. aureus, SIC which is the Streptococcal inhibitor of complement (Kudryashova, Seveau & Kudryashov, 2017). Upon binding to the toxins, defensins promote their unfolding, disrupting their secondary and tertiary structure, making them more susceptible to proteolysis and promoting their precipitation. Although the physicochemical properties that allow defensins to bind and neutralize a broad range of structurally diverse toxins are not completely understood, recent studies demonstrate that defensins act by recognizing regions of proteins showing structural plasticity and thermodynamic instability, features that are shared by a wide range of bacterial toxins (Kudryashova, Seveau & Kudryashov, 2017). In general, of the alpha-class such as HNP and HD5, they inhibit the lethal factor of Bacillus anthracis, the diphtheria toxin, the exotoxin A of P. aeruginosa, the cytotoxin B of Clostridioides difficile, among others (Kim et al. 2005, 2006; Giesemann, Guttenberg & Aktories, 2008). While those in the beta-class, such as hBD, inhibit the gonococcal toxin NarE from Neisseria gonorrhoeae and the lethal factor from B. anthracis (Rodas et al., 2016; Wei et al., 2009). In the case of those of the theta-class, the retrocyclins inhibit the lethal factor of B. anthracis and the vaginolysin of G. vaginalis (Wang et al., 2006; Hooven et al., 2012).

Beyond defensins, there are other notable examples of toxin-neutralizing peptides, such as the artificial peptide Pep19−2.5 and related ones, capable of inactivating lipopolysaccharides (LPS or endotoxin) and lipoproteins in vitro and in vivo. They also decrease inflammation mediated by the activation of signaling cascades (Heinbockel et al., 2018), and their efficacy has been reported in several mouse infection models, including endotoxemia and bacteremia (Heinbockel et al., 2013). Several other peptides with the ability to neutralize a wide variety of bacterial toxins have been described, for which we recommend consulting the following reviews (Jerala & Porro, 2005; Kudryashova, Seveau & Kudryashov, 2017; Schnell et al., 2019). The effect was not always determined at sub-MIC concentrations; however, the inhibition of toxins is a strategy contemplated within the anti-virulence targets.

Bacitracin is an antibiotic that inhibits cell wall synthesis, but recently it has also been reported to neutralize type A/B protein exotoxins by inhibiting pore formation, preventing translocation of the A subunit to the host cell cytosol. These toxins are made up of an enzymatic component (A subunit) and a binding/transport component (B subunit), such as the lethal factor of Bacillus anthracis, the toxin C2 of Clostridium botulinum, the CDT transferase of C. difficile, and epsilon toxin of Clostridium perfringens (Schnell et al., 2019). In addition to neutralizing bacterial toxins, some AMPs can inhibit exoproteases implicated in the generation of host damage during periodontal disease. For example, the salivary peptide histatin 5 inhibits the host metalloproteases and exoproteases from bacterial pathogens such as the gingipains produced by Porphyromonas gingivalis attenuating damage and inflammation (Gusman, Malonneet & Atassi, 2001). Moreover, histatin 5 also inhibits cysteine proteinases such as clostripain, which is produced by Clostridium histolyticum during gangrene, while other AMPs inhibit exoproteases such as subtilisin A, proteinase K, elastase, and chymotrypsin (Le, Fang & Sekaran, 2017).

Finally, a characteristic of some anti-virulence molecules is their adjuvant properties, which enable them to restore the activity of antibiotics on resistant strains (Díaz-Nuñez, García-Contreras & Castillo-Juárez, 2021). This strategy is very promising, and although it does not prevent the generation of resistance, it allows the reactivation of antimicrobials that are in danger of falling into disuse (González-Bello, 2017). In the case of AMPs at sub-MIC concentrations, some reports of adjuvant properties have been made in the literature, such as unarmycin A and C (Tanabe et al., 2007). These cyclopeptides isolated from marine bacteria are azole antifungal ejection pump inhibitors and restore fluconazole sensitivity of resistant strains and clinical isolates of Candida albicans (Tanabe et al., 2007). Also, plantaricin PLNC8 α β showed an adjuvant effect by potentiating the activity of conventional antibiotics (vancomycin, rifampicin, and gentamicin) against S. epidermidis, although the mechanism of action involved is unknown (Bengtsson et al., 2020).