Antibacterial activity of human defensins against Staphylococcus aureus and Escherichia coli

Introduction

Rapid and widespread increase in the resistance of microorganisms to antimicrobial drugs is known to present a serious problem and challenge to modern medicine (Roca et al., 2015; Li & Webster, 2018). The threat of increasing antibiotic resistance and methods to combat it are under active discussion at the level of the World Health Organization and the United Nations; in 2016, the “Global action plan to combat antimicrobial resistance” has been published. According to this document, the key objectives to solve this problem are the optimization of the use of antimicrobial drugs, as well as the development of new drugs (World Health Organization, 2015). Over the past 10 years, only several new antibacterial drugs have been introduced to the pharmaceutical market (Bassetti et al., 2013; Andrei, Droc & Stefan, 2019). An increase in antimicrobial resistance naturally leads to a decrease in the effectiveness of therapy and, as a result, an increase in the duration of treatment, an increase in mortality and financial expenses on treatment (Fair & Tor, 2014; Rolain et al., 2016). For example, 19,000 people die annually in the United States from infections caused by methicillin-resistant strains of Staphylococcus aureus (MRSA) (Fischbach & Walsh, 2009), while the annual financial expenses on treatment of this infection comprise $3 billion. According to the latest report from the Centers for Disease Control and Prevention (USA), the financial burden associated with increasing microbial resistance comprises about $55 Billion and 8 Million additional bed days (US CDC, 2019). It is estimated that by 2050 more than 10 million people will die annually from infections caused by resistant strains and by that time the global economy will lose about US $100 trillion due to this problem (O’Neill, 2016).

The formation of resistance takes place due to various causes and mechanisms. This is known to be a natural evolutionary process of adaptation of microorganisms to frequent contact with substances possessing antimicrobial properties (Martinez et al., 2009). The wide spread of antibiotic resistance is due to two factors—mutations and horizontal gene transfer (Martinez & Baquero, 2000).

The human body is in continuous contact with a large number of pathogenic and non-pathogenic microorganisms. In the process of evolution, defense mechanisms have formed that allow first to identify the pathogen and then, if necessary, to exercise adequate control of its further penetration and spread. These tasks are accomplished through the innate immune system which is capable (unlike the adaptive immunity system) of immediately recognizing and destroying infectious agents of various nature (Iwasaki & Medzhitov, 2015). The most important component of innate immunity is antimicrobial peptides (AMPs) with a length of 5 to ~100 amino acid residues. These peptides have a broad spectrum of antimicrobial activity against various infectious agents: bacteria, viruses, fungi and protozoa. Among the six kingdoms (bacteria, archaea, protists, fungi, plants, and animals), more than 3,000 AMPs have been identified by now (Wang, Li & Wang, 2016). Among AMPs, of great interest are human defensins: human neutrophil peptide-1 (HNP-1), human beta-defensin-1 (hBD-1), and human beta-defensin-3 (hBD-3), since they have a wide spectrum of antimicrobial activity (Pachón-Ibáñez et al., 2017). Since the outer surface of all bacteria has a negative charge (due to the presence of lipopolysaccharides and/or teichoic acids), positively charged and hydrophobic AMPs (in particular, defensins) nonspecifically “accumulate” on the surface of both gram-positive and gram-negative microorganisms. The antibacterial activity of defensins is believed to be related to membrane permeabilization of microorganisms (Kagan et al., 1990; Wimley & Hristova, 2020). However, some AMPs have been found to use alternative mechanisms of antimicrobial action (Matsuzaki et al., 1991; Mor & Nicolas, 1994; Oren & Shai, 1998; Chan, Prenner & Vogel, 2006). It has also been shown that HNP-1 can inhibit the synthesis of the bacterial cell wall by binding to precursor lipid II (De Leeuw et al., 2010).

Unfortunately, the introduction of native AMPs into clinical practice as a monotherapy for bacterial infections has a number of limitations: high synthesis cost, hemolytic activity, cytotoxicity for macroorganism, immunogenicity, and pharmacokinetic specifics (Moravej et al., 2018; Lei et al., 2019). To solve these problems, two approaches have been proposed: (i) modifying native AMPs (or designing new peptides with antimicrobial activity) (Lei et al., 2019), and (ii) using native AMPs at low doses in combination with conventional antibiotics (Zharkova et al., 2019).

In this work, we investigated the effectiveness of the combined use of human defensins HNP-1, hBD-1, hBD-3 and antibiotics (rifampicin and amikacin) against isolates of S. aureus and Escherichia coli collected from hospitalized patients.

Materials and Methods

Results

All S. aureus isolates tested (n = 27) were susceptible to AMPs and rifampicin (Table 2). The MIC of HNP-1 for the studied staphylococci was 4 (2–8) mg/L (hereinafter, MIC and FICI values are presented as median and interquartile range in brackets). The MIC of hBD-1 was 8 (4–8) mg/L; that of hBD-3—1 (0.5–4) mg/L; that of rifampicin—0.008 (0.004–0.016) mg/L. As can be seen, the highest anti-staphylococcal activity among defensins was demonstrated by hBD-3.

The results of MIC studies for E. coli isolates (n = 24) are presented in Table 3. In this case, the most effective AMP also was hBD-3 that had MIC of 4 (4–8) mg/L. The MIC of HNP-1 was 12 (4–32) mg/L. hBD-1 was found to be ineffective against 10 out of 24 E. coli isolates; its MIC against susceptible strains was 32 (14–32) mg/L. The MIC of amikacin was 3 (2–4) mg/L.

We showed that against S. aureus (including MRSA), the combinations of HNP-1 + rifampicin and hBD-3 + rifampicin in most cases demonstrated synergistic effects—the FICI values for both combinations were 0.5 (0.375–0.5) (Table 4). The combination of HNP-1 + rifampicin did not show a synergistic effect against only 3 out of 27 S. aureus isolates (SA-4, SA-6, SA-19). When the combination of hBD-3 + rifampicin was used, the FICI value exceeded 0.5 for three isolates of S. aureus (SA-4, SA-9, SA-21), which indicates the absence of interaction between these substances against these strains. As to the combination of hBD-1 + rifampicin, we showed that only in 3 cases out of 27 there is a synergism of action (against isolates SA-5, SA-13, SA-14, see Table 4), while the median FICI value was 0.75 (0.75–1.25).

The study of the combined antimicrobial action of defensins with amikacin against E. coli isolates produced similar results. The combinations of HNP-1 + amikacin and hBD-3 + amikacin in most cases demonstrated synergistic action—the FICI values were 0.375 (0.375–0.5) and 0.5 (0.375–0.5), respectively (Table 5). The combined use of HNP-1 and amikacin did not show synergy in only 3 cases out of 24 (EC-6, EC-11, EC-13), and the combination of hBD-3 + amikacin—only in 2 cases out of 24 (EC-5, EC-13). The combined use of hBD-1 and amikacin against E. coli isolates did not show a synergistic effect: FICI = 1 (0.75–1.5). Moreover, in 10 cases out of 24, it was not possible to calculate the FICI value of this combination of substances, since the MIC of hBD-1 for these 10 isolates was >64 mg/L (Table 5).

Discussion

The results obtained are of interest from several points of view. First, even though studies of the antimicrobial activity of HNP-1, hBD-1 and hBD-3 against S. aureus and E. coli have previously been conducted, a thorough analysis of their MIC and FICI values against a large number of clinical isolates with heterogeneous antibiotic resistance phenotypes has not been carried out. Second, the obtained data can be used to search for and develop new strategies for overcoming resistance to antimicrobial drugs used in clinical practice, since this work shows that, for instance, the MIC values of rifampicin and amikacin decrease by several times when they are used in combination with HNP-1 or hBD-3.

We showed that antibiotic resistance phenotype (including methicillin resistance) does not affect the sensitivity of the studied bacterial isolates to AMPs (Tables 2 and 3). It can be argued that the mechanism of antimicrobial action of the studied defensins is at least not associated with the targets against which conventional antibiotics are directed. The antimicrobial activity of positively charged defensins is realized when a certain threshold concentration of AMP molecules on the outer surface of the lipid membrane of a bacterial cell is reached, so that the tangential tension is compensated, which ultimately leads to the formation of pores of different structure (Sengupta et al., 2008). The differences in MIC values between different isolates of the same species may be due to differences in the structures of their cell walls.

In this work, we did not study the activity of defensins at concentrations above 64 mg/L; there is no need to carry out MIC analysis at such high concentrations, since the use of AMPs in practice has a number of limitations. Implementation of AMPs is complicated by the fact that they are rapidly degraded by peptidases (Steckbeck, Deslouches & Montelaro, 2014). This leads to their short duration of action and significantly affects the pharmacokinetics. Moreover, AMPs can have a cytotoxic effect on eukaryotic cells and cause hemolysis (Matsuzaki, 2009; Takahashi et al., 2010). Another problem with the large-scale use of AMPs is the high cost of their production (Pachón-Ibáñez et al., 2017). To solve these problems, several strategies have been proposed: modification of AMPs or creation of new peptides (with a short amino acid sequence) (Pachón-Ibáñez et al., 2017), stimulation of the production of endogenous AMPs by administration of low molecular weight compounds (Chen et al., 2020), implementation of low doses of AMPs to enhance the antimicrobial activity of conventional antibiotics (Zharkova et al., 2019).

The effectiveness of AMPs significantly varies in different studies and against different strains of the same species (Ganz et al., 1985; Turner et al., 1998; Yang et al., 1999; Sahly et al., 2003; Dürr, Sudheendra & Ramamoorthy, 2006; Wilmes et al., 2011; Xhindoli et al., 2016). It should be noted that there is still no standard that defines control points (criteria) of susceptibility or resistance of certain species of bacteria to a specific AMP. Therefore, it is necessary to conduct studies of the MIC and FICI values against a large number of clinical isolates.

Earlier studies by other groups have shown that some natural and novel synthetic AMPs can exhibit synergistic effects in combination with aminoglycosides or rifampicin (Pollini et al., 2017; Wu et al., 2017). The mechanism underlying the synergistic action of HNP-1/hBD-3 with rifampicin and amikacin is most likely to be related to the fact that AMPs facilitate the penetration of antibiotics into cells (Zharkova et al., 2019). Zharkova et al. (2019) have shown that often there is synergy between highly active AMPs targeting membranes (for example, protegrin 1, hBD-3) and antibiotics with intracellular targets (for example, gentamicin, rifampicin), which suggests an increase in bioavailability as the main model of such interaction.

The implementation of low doses of AMPs can reduce the MICs of some antibiotics, which has been shown in numerous studies. For instance, the combination of hBD-3 and methicillin demonstrates a synergistic effect against clinical strains of MRSA with FICI values in the range of 0.09-0.45 (Midorikawa et al., 2003). This is very interesting because the use of methicillin alone is not effective against MRSA (The European Committee on Antimicrobial Susceptibility Testing, 2020), thus, hBD-3 can help overcome the resistance of MRSA to beta-lactam antibiotics. Similar results can be obtained for other combinations of AMPs with antibiotics to which the bacteria have acquired resistance. This would require studies with reference strains, followed by verification with respect to a large number of appropriate clinical isolates.

It has previously been shown that hBD-3 can effectively combat MRSA biofilms by suppressing bacterial growth, regulation of inflammation and immune responses in vivo (Zhu et al., 2017). In general, the effects of AMPs in vivo are very diverse: from wound healing (Bolatchiev et al., 2020) to the ability to neutralize the lethal toxin of the anthrax pathogen (Kim et al., 2005). Defensins can be considered as an effective link between innate and adaptive immune responses (Colavita et al., 2015), since AMPs directly stimulate the migration of immune cells, promote the release of pro-inflammatory cytokines, and activate antigen-presenting cells through the Th1 immune response (Suarez-Carmona et al., 2015). Thus, it is obvious that synergistic effects should be assessed in in vivo experimental models.

Thinking about further strategies for using these defensins to solve the problem of antibiotic resistance, we suggest that one of the approaches to future clinical applications of AMPs may be the search for ways to produce endogenous AMPs (for instance, by introducing low molecular weight compounds or viral vectors encoding peptide sequences) in combination with conventional antibiotics. On the one hand, this strategy can help to increase the effectiveness of antibacterial drugs, and on the other hand, the stimulation of endogenous AMPs is a much cheaper way of application of AMPs.

Conclusions

Thus, in this work, we investigated the antimicrobial activity of human defensins HNP-1, hBD-1, and hBD-3 against clinical isolates of S. aureus (n = 27) and E. coli (n = 24). Among the studied defensins, HNP-1 and hBD-3 were the most effective. Moreover, these antimicrobial peptides showed a synergistic effect against most of the studied isolates when applied together with rifampicin and amikacin.

Supplemental Information