Biophysical evaluation of cardiolipin content as a regulator of the membrane lytic effect of antimicrobial peptides

Highlights

  • Cardiolipin is a tetra-acyl phospholipid synthesized in S. aureus as an adaptive response to external stress.

  • Increasing cardiolipin content in liposomes results in an increase in the level of lipid packing as measured by Laurdan GP.

  • Increasing content of cardiolipin reduces the lytic activity of the peptides LL-37 and ΔM2.

Abstract

Cardiolipin is an anionic tetra-acyl chained glycerophospholipid that increases lipid packing levels and induces intrinsic negative curvature in membranes. Cardiolipin is found in Staphylococcus aureus (S. aureus) membranes, where increased levels of this lipid are induced at the expense of diacyl phosphatidylglycerol in response to stress. We investigate cardiolipin as an inhibitor of the lytic activity of the cationic antimicrobial peptides LL-37 and ∆M2 in model systems with varying phosphatidylglycerol/cardiolipin ratios. Using HPTLC, we show that S. aureus (RN4220), under different growth conditions, has a phosphatidylglycerol/cardiolipin ratio of 80:20. From this, we chose three model systems to evaluate (100:0, 80:20, 60:40). ∆M2 presents higher binding affinity towards all mixtures compared to LL-37. This correlates with the higher antimicrobial activity of ∆M2 compared to LL-37 in S. aureus (MIC90 of 14 μM for ∆M2 and 57.7 μM for LL-37). Laurdan GP shows that Cardiolipin decreases lipid headgroup spacing. We find that cardiolipin does not affect ∆M2 or LL-37 binding to phosphatidylglycerol/cardiolipin liposomes. Instead, cardiolipin inhibits the ability of both peptides to induce calcein leakage in model liposomes. In conclusion, cardiolipin can reduce cAMP activity by inhibiting lysis but not binding.

Introduction

Staphylococcus aureus (S. aureus) is well recognized for being one of the leading pathogenic bacteria involved in health care-associated infections throughout the world [1]. In healthy individuals it is usually in a harmless form that can be found normally in the nose and on the skin of approximately 30% of adults [2]. However, this percentage is higher in patients and workers exposed to the bacteria within healthcare facilities. S. aureus becomes pathogenic in immuno-compromised surgery patients that are exposed to contaminated surgical instruments, and in patients with chronic skin lesions. When S. aureus travels through the body it can cause various diseases, from mild to severe, such as bacteremia, pneumonia, osteomyelitis, endocarditis and joint infections [1]. Also, patients may develop chronic S. aureus infections on exposed skin, and through indwelling medical devices due to S. aureus biofilm formation. Treatment of these infections has become more difficult due to the emergence of multidrug-resistant strains [[3], [4], [5]], like Methicillin-resistant S. aureus (MRSA). A growing number of cases of MRSA infections have been reported world-wide, increasing morbidity and putting a strain on health services [4,6].

Microorganisms, such as S. aureus, have the ability to modulate the lipid composition of their membranes as an adaptive response towards changing environmental conditions that can damage the membrane. Phospholipid modifications are triggered in response to the growth phase, [[7], [8], [9]], or medium stressors such as changes in pH, osmotic stress, the presence of organic solvents and high saline concentrations [10]. Additionally, this adaptive response can also be used to induce resistance to membrane-active antibiotics [11]. S. aureus is a Gram-positive bacteria characterized by a single bilayer membrane that acts as its plasma membrane. The predominant phospholipid constituents of the plasma membrane of S. aureus are phosphatidylglycerol (PG) and cardiolipin (CL). Cardiolipin contains four fatty acids and is synthesized in prokaryotes by a transphosphatidylation reaction through Cardiolipin Synthase that occurs between two phosphatidylglycerols that release glycerol and join the two phospholipids [12].

Phosphatidylglycerol contains one phosphate in its headgroup and no other groups with compensating positive charges and therefore has a net negative charge. Thus, it introduces a negative charge into the bacterial membrane surface and to the lipid–protein interface. It is believed that such properties allow the physical properties of the lipid phase to be regulated and therefore contribute to the optimal conditions of integral membrane proteins [13,14]. A change from a membrane composition rich in PG lipids to a membrane composition rich in CL may be expected to increase membrane packing due to an increase in the lateral density of fatty acids. However, the bacterial response that motivates the synthesis of higher amounts of CL and the physiological consequences of this increase in the CL content of S. aureus membranes are not yet fully understood. Some studies have suggested that an elevated CL content in the membrane may contribute to bacterial resistance to antibiotics. An increase in the CL level has been observed in S. aureus strains that have developed non-susceptibility to Daptomycin, suggesting that CL content is a possible response mechanism for adaptive resistance towards antimicrobial agents targeted onto the membrane [[15], [16], [17]]. Increasing concentrations of CL impose structural consequences upon the bilayer core of the S. aureus membrane and, accordingly, changes in the lipid phase behavior can occur that may affect physiologically relevant physical-chemical properties such as membrane permeability, fluidity and variation in lipid-protein interactions.

Cationic antimicrobial peptides (cAMPs) are short proteins, 20–40 amino acids long that preferentially bind to bacterial membranes and thereby affect lipid bilayer integrity, which finally leads to cell death. The initial binding of cAMPs with bacterial membranes is mediated through electrostatic interactions between the cationic residues on the peptide and anionic lipids (such as PG and CL) of the bacterial membrane. This implies that the presence of negative charges in membranes is a critical parameter for susceptibility to cAMPs. However, one of the resistance responses generally found in Gram-positive bacteria is a decrease in anionic charges on the membrane through the synthesis of lysyl-phospholipids [[18], [19], [20]]. After initial binding, cAMPs assume a secondary structure, and either compromise membrane integrity through detergent-like activity, or disrupt the membrane through multimeric pore formation. These action mechanisms are sensitive to the mechanical properties of the membrane, where a rigid membrane is likely to be more tolerant towards the presence of cAMPs on its surface. From this perspective, it is interesting to study how the condensing effects of CL formation may influence the membrane activity of antimicrobial peptides. With this aim, we evaluated two cAMPs active against S. aureus (LL-37 and ΔM2) in lipid model systems built of different ratios of PG and CL. Human cathelicidin LL-37 is a 37 residue peptide with a + 6 charge at pH 7.4, which after binding to the membrane assumes a α-helical structure [21]. It has several physiological roles in the body, being recognized for its antimicrobial, antifungal and antiviral activities [[22], [23], [24], [25], [26], [27]]. The cationic peptide ΔM2 is derived from Cecropin D-like from Galleria mellonella and it is composed of 39 amino acids with a + 9 charge at physiological pH [28].

Considering that both cAMPs are membrane-active, we specifically selected techniques based on the detection of membrane integrity. The effect of peptides on membrane disruption was evaluated by monitoring calcein leakage from lipid vesicles. Three lipid systems based on different compositions of CL were evaluated by monitoring the fluorescence intensity of the fluorophore. Since calcein trapped in the vesicles undergoes self-quenching, the disrupting mechanism of the peptides can be followed by inducing an increase in the fluorescence intensity of calcein released from liposomes. This model simplifies the cellular components present in the bacterial membrane, focusing on peptide-phospholipid interaction, and it is frequently used to evaluate the membrane permeabilization activity of cAMPs [29]. Generalized polarization (GP) experiments were performed on the model systems to understand the changes in the dynamics of the environment and monitor the phase state of lipid systems [30]. Laurdan provides information about hydration and molecular dynamics at the level of the glycerol backbone and can distinguish whether a membrane is in a gel or liquid-crystalline state [31]. Finally, the antimicrobial activity of peptides on S. aureus cell cultures was determined using a minimum inhibitory concentration (MIC) assay. The results were correlated to understand the role of CL in modulating the biophysical properties of the membrane.

Section snippets

Materials

The phospholipids used in this study were 1-palmitoyl-2-oleoyl-sn-glycerol-3-phosphoglycerol sodium salt (POPG, Lot. 160-181PG-135) and tetraoleoyl cardiolipin from heart bovine (CL, Lot. SLBL8560V; Fig. 1A and B). POPG and CL were purchased from Avanti Polar Lipids (Alabaster, AL, USA) and Sigma-Aldrich (St. Louis, MO, USA), respectively. 3,3-bis[N,N-bis(carboxymethyl)-aminomethyl] fluorescein (calcein), N-2-Hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES) and Sephadex G50-fine were

Results and discussion

First we focus on studying the activity of the two cationic antimicrobial peptides LL-37 and ∆M2 in Staphylococcus aureus, and compare this antimicrobial activity with their ability to bind and induce leakage in model POPG liposome. LL-37 is a cathelicidin derived cationic antimicrobial peptide (cAMP) that is part of the innate immune response in humans [34]. This peptide is 37 residues in length [35] and belongs to the class of amphipathic α-helical cAMPs that induce membrane leakage,

Conflict of interest

No potential conflict of interest was reported by the authors.

Acknowledgements

The authors are grateful to COLCIENCIAS for financial support (Project 120465843150, Contract 611-2014). This work was financially supported in part by the Statutory Funds of the Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian University. The Jagiellonian University is a partner of the Leading National Research Center (KNOW) supported by the Ministry of Science and Higher Education.