Correspondence: Viswanatha Krishnakumari, CSIR – Centre for Cellular and Molecular Biology, Uppal Road, Hyderabad- 500 007, India. Tel.: +91 40 27192586; fax: +91 40 27160591; e-mail: firstname.lastname@example.org
Human-β-defensins 1-3 (HBD-1-3) and their C-terminal analogs Phd-1-3 do not show antibacterial activity against Escherichia coli in the presence of mono- and divalent cations. Activity of peptides was examined against E. coli pretreated with carbonyl cyanide m-chlorophenylhydrazone (CCCP) and salt remedial Escherichia coli ftsEX, a deletion mutant of FtsEX complex [an ATP-binding cassette (ABC) transporter protein], in the presence of Na+, Ca2+, and Mg2+. Activity was observed in the presence of Na+ and Ca2+, although not in the presence of Mg2+ against E. coli, when proton motive force (PMF) was dissipated by CCCP. The peptides exhibited antibacterial activity against E. coli ftsEX even in the presence of Na+ and Ca2+. Our results indicate that HBD-1-3 and Phd-1-3 do not require PMF for their antibacterial activity. The absence of activity against E. coli in the presence of Na+ and Ca2+ ions is due to not only weakened electrostatic interactions with anionic membrane components, but also involvement of electrochemical gradients. However, Mg2+ prevents electrostatic interaction of the peptides with the outer membrane resulting in loss of activity.
Antimicrobial human-β-defensin peptides (HBD-1-4) are important components of innate and adaptive immunity (Goldman et al., 1997; Bals et al., 1998; Harder et al., 2001; Lehrer & Ganz, 2002; Selsted & Ouellette, 2005; Agerberth & Gudmundsson, 2006). They exhibit broad-spectrum antibacterial activity (Goldman et al., 1997; Bals et al., 1998; Harder et al., 2001; Lehrer & Ganz, 2002; Selsted & Ouellette, 2005; Agerberth & Gudmundsson, 2006). NaCl inhibits the antibacterial activity of HBD-1, HBD-2, and HBD-4, although not of HBD-3 (Goldman et al., 1997; Bals et al., 1998; Garcia et al., 2001; Harder et al., 2001). At physiological concentrations, divalent cations attenuate the activity of HBD-1-3 (Goldman et al., 1997; Bals et al., 1998; Hoover et al., 2000, 2001; Tomita et al., 2000; Bowdish et al., 2005; Selsted & Ouellette, 2005; Maisetta et al., 2008). Extensive studies have been carried out to understand structure–function relationships in human-β-defensins (Hoover et al., 2003; Krishnakumari et al., 2003, 2006; Wu et al., 2003; Kluver et al., 2005; Varkey & Nagaraj, 2005). The activity of several human-β-defensins analogs was sensitive to the presence of mono- and divalent cations (Krishnakumari et al., 2003, 2006; Kluver et al., 2005). The mechanism(s) of action and biophysical basis for sensitivity to cations have not yet been established unequivocally.
It has been shown that E. coli can grow in the presence of the protonophore carbonyl cyanide m-chlorophenylhydrazone (CCCP), which dissipates PMF (Ohyama et al., 1992). CCCP has been used extensively in studies related to understanding resistance mechanisms of microbial pathogens that use energy-dependent efflux system against antibacterial peptides (Shafer et al., 1998; Bengoechea & Skurnik, 2000). Our earlier studies on β-defensin C-terminal analogs have indicated that the initial site of action is the membrane of bacteria and fungi (Krishnakumari et al., 2003, 2006, 2009). In this article, we show that PMF is not required for antibacterial activity of HBD-1-3 and their C-terminal analogs against E. coli. In the absence of PMF, it is observed that Na+ and Ca2+ does not attenuate their antibacterial activity. However, inhibition of activity is observed with Mg2+ both in the presence and absence of PMF. Supporting evidence is further obtained from studies with E. coli ftsEX, a deletion mutant of FtsEX complex that belongs to the ABC transporter family of proteins that is salt remedial (Reddy, 2007).
Materials and methods
HBD-1, HBD-2, and HBD-3 were purchased from Peptides International (Louisville, KY). The C-terminal analogs that possess a single disulfide bond: Phd-1, ACPIFTKIQGTYRGKAKCK; Phd-2, FCPRRYKQIGTGLPGTKCK; and Phd-3, SCLPKEEQIGKSTRGRKCRRKK were synthesized as described earlier using 4-(hydroxymethyl) phenoxy acetamidomethyl resin and 9-fluorenylmethoxycarbonyl chemistry (Krishnakumari et al., 2006). The formation of disulfide bonds was accomplished by air oxidation at a peptide concentration of 0.5 mg mL−1 for 24 h at room temperature. Purified peptides were characterized by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry on an ABI Voyager DE STR using recrystallized α-cyano-4-hydroxycinnamic acid as the matrix (Krishnakumari et al., 2006). CCCP and reagents for peptide synthesis were purchased from Sigma Chemical Co. (St. Louis, MO).
Bacterial strains and growth conditions
Wild-type E. coli K-12 strain MG1655 (F-, λ−, rph-1; CGSC-7740) referred to as E. coli subsequently (Nikaido, 1996; Blattner et al., 1997) and E. coli ftsEX (MR10; MG1655 ΔlacI ΔftsEX210::Kan), a derivative of MG1655 lacking FtsEX complex (FtsE and FtsX proteins are associated with inner membrane of bacteria) described earlier (de Leeuw et al., 1999; Reddy, 2007), were used in this study. Luria–Bertani (LB) medium that contains 1% NaCl was used for E. coli MG1655, and 0.2 M glucose was added to the LB medium for the growth of E. coli ftsEX (Reddy, 2007; Sezonov et al., 2007). The growth temperature was 37 °C for both strains.
The antibacterial activity of peptides HBD-1-3 and Phd-1-3 against E. coli and E. coli ftsEX was examined in sterile 96-well plates in a volume of 100 µL as follows (Krishnakumari et al., 2006): E. coli was grown overnight in LB medium at 37 °C. After 20 h, 0.2 mL from this suspension was subcultured for 3 h in 20 mL of LB broth to obtain a mid-log-phase culture. Cells were harvested by centrifugation, washed thoroughly with 10 mM phosphate buffer, pH 7.4 (PB), to remove LB medium. Cells were resuspended in the same buffer, and concentration was adjusted to 106 colony-forming units (CFUs) per mL. These cells were incubated with different concentrations of peptides for 2 h at 37 °C. Aliquots were spread on LB agar plates and incubated for 18 h at 37 °C. The CFUs were counted, and percentage killing of bacteria relative to the CFU counts in untreated controls was calculated. The concentration of the peptides at which no viable colonies formed was taken as lethal concentration (LC). The results were expressed as average of LCs obtained from three independent experiments carried out in duplicate. Mono- or divalent cations (150 mM NaCl, 2.5 mM CaCl2, or 2.5 mM MgCl2) were added in the incubation buffer along with peptides at their LC of the test peptide to determine their effect on the activity. No visible precipitation was observed. The activity of peptides (HBD-1-3 and Phd-1-3) against E. coli ftsEX grown in LB supplemented with 0.2 M glucose, was determined in a similar manner.
Mid-log-phase culture of E. coli, diluted to 106 CFUs mL−1 in PB, was pretreated with 50 μM CCCP for 20 min at 37 °C. CCCP was dissolved in methanol as a stock solution at 25 mM concentration. The methanol concentration in the medium was 0.2% that did not affect cell viability. CCCP-treated cells in the presence of Na+, Ca2+, or Mg2+ were incubated with peptides at their LCs for 2 h at 37 °C to determine their effect on the antibacterial activity. Aliquots were spread on LB agar plates and incubated at 37 °C for 18 h. The CFUs were counted, and percentage killing of bacteria relative to the CFU counts in untreated controls was calculated.
The antibacterial activities of HBD-1-3 and Phd-1-3 against E. coli and E. coli ftsEX are summarized in Table 1. The data indicate that the peptides show equivalent activity against both strains of bacteria. The assays were also performed with CCCP-treated E. coli to evaluate the requirement of PMF for activity. No difference in viability was observed between CCCP-treated and untreated E. coli control cells. The peptides showed no variation in LC against CCCP-treated and untreated E. coli. Phd-1-3 showed lower activity as compared to HBD-1-3.
Table 1. Antibacterial activities of human β-defensins HBD-1-3 and C-terminal analogs Phd-1-3
The values reported are average of LCs obtained from three independent experiments carried out in duplicate.
The antibacterial activities of HBD-1-3 and Phd-1-3 in the presence of Na+ against E. coli with and without CCCP treatment and E. coli ftsEX are shown in Fig. 1a–d. All the peptides showed loss in activity against E. coli in the presence of Na+ (indicated by * in Fig. 1a and b). When cells were pretreated with CCCP, in the presence of Na+, 40% killing of bacteria was observed for HBD-1. The activities of HBD-2 and HBD-3 were not attenuated to a great extent (Fig. 1a). No loss in activity was observed for Phd-1-3 (Fig. 1b), in the presence of Na+. Attenuation of activity was considerably less against E. coli ftsEX in the presence of Na+ for all the peptides in the absence of CCCP treatment as shown in Fig. 1c and d.
The antibacterial activities of HBD-1-3 and Phd-1-3 in the presence of divalent cations Ca2+ and Mg2+ against E. coli with and without CCCP treatment and E. coli ftsEX are shown in Fig. 2a–d. All the peptides were inactive in the presence of Ca2+ and Mg2+ at their LCs against E. coli (indicated by * (in Fig. 2a and b). When E. coli cells were pretreated with CCCP, no loss in activity was observed for HBD-3 and Phd-1-3 in the presence of Ca2+ (Fig. 2a and b). HBD-1 and HBD-2 showed 40% and 85% killing of bacteria, respectively (Fig. 2a). Low activity was observed in the presence of Mg2+ for both HBD-1-3 and Phd-1-3 (Fig. 2a and b). Ca2+ did not inhibit the activity of HBD-1-3 and Phd-1-3 against E. coli ftsEX, as observed in the case of E. coli without CCCP treatment (Fig. 2c and d). However, activity was significantly inhibited by Mg2+ (Fig. 2c and D).
The antibacterial activity of human-β-defensins (HBD-1-4) is sensitive to the presence of mono- and divalent cations (Goldman et al., 1997; Tomita et al., 2000; Garcia et al., 2001; Harder et al., 2001; Maisetta et al., 2008). Because defensins are highly cationic, binding to the negatively charged bacterial cell surface is conceivably reduced in the presence of cations. There have been several attempts to design analogs with improved salt-resistant antibacterial activity (Kluver et al., 2005; Scudiero et al., 2010; Jung et al., 2011). Based on these studies, different domains or segments in HBD-1-3 that are critical for antibacterial activity and salt sensitivity have been identified. The antibacterial activity of these analogs in the presence of salt has been attributed to their compact structure (Scudiero et al., 2010). In this article, the effect of dissipation of PMF on the antibacterial activity of HBD-1-3 and Phd-1-3 in the presence of Na+, Ca2+, and Mg2+ has been investigated.
Our results indicate that HBD-1-3 and analogs Phd-1-3 exert antibacterial activity against E. coli in the absence of PMF. It has been reported that HNP-1 (an alpha-defensin) and its analogs exhibited antibacterial activities against E. coli, S. aureus, and P. aeruginosa in the presence of CCCP (Varkey & Nagaraj, 2005). HNP-2 showed activity against Neisseria gonorrhoeae in the presence of CCCP (Shafer et al., 1998). We show that HBD-1-3 and analogs Phd-1-3 exhibit activity in the presence of Na+ and Ca2+, when PMF is dissipated in E. coli by CCCP. Phd-1 was more effective in killing CCCP-treated bacteria in the presence of Na+ and Ca2+ as compared to HBD-1. The net charge at neutral pH is +5 in Phd-1 and +4 in HBD-1. The net charge in Phd-1 and the other peptides (in parenthesis) HBD-2 (+6), HBD-3 (+11), Phd-2 (+5), and Phd-3 (+7) is higher than that in HBD-1, suggesting that in defensins, a minimal net charge (at neutral pH) may be required for electrostatic interactions with the bacterial membranes, which is reflected in their higher antibacterial activity.
In E. coli, both inwardly directed Na+ gradient and constant intracellular pH are maintained by Na+/H+ antiporter, coupled to PMF (Ohyama et al., 1992; Padan & Schuldiner, 1994). The development and maintenance of proton motive force are accompanied by rapid Na+ extrusion and the maintenance of a substantial inwardly directed Na+ gradient. Earlier, NMR studies have indicated that when external concentration of NaCl is ∼100 mM, the Na+ extrusion system ensures that lower concentration of Na+ (4 mM) is maintained in the cytoplasm (Castle et al., 1986). This possibly prevents effective binding of HBD-1-3 and C-terminal analogs to the inner membrane via electrostatic interactions. In the presence of CCCP, the Na+ extrusion system is no longer functional and this results in equilibration of Na+. This enhances the effective interaction of peptides with the inner membrane. Similar mechanism exists for Ca2+ in E. coli in which Ca2+ extrusion is also coupled to PMF via secondary Ca2+–proton exchange (Tsujibo & Rosen, 1983). Supporting data were obtained with E. coli ftsEX mutant. The mutant is salt remedial as well as osmoremedial, which renders it conditionally dependent on high osmolarity for its growth (Ukai et al., 1998; de Leeuw et al., 1999; Reddy, 2007). It has been suggested that optimal osmotic strength may be required for proper assembly of cell division machinery (Reddy, 2007). Peptides showed activity with Na+ and Ca2+ in the absence of CCCP with this salt remedial strain. The results indicate that inhibition of activity in the presence of Na+ and Ca2+ against E. coli is due to not only weakened electrostatic interactions with anionic phospholipid head groups, but also involvement of electrochemical gradients. The inhibition of activity in the presence of Mg2+, which is not reversed even with CCCP-treated E. coli and E. coli ftsEX, could be due to inability of the peptides in binding to the polyanionic LPS in the outer membrane.
The inner membrane is responsible for crucial metabolic processes in E. coli, and any destabilization of the membrane would lead to cell death. It has been shown that HBD-2-3 and analogs (Phd-1-3) associate with negatively charged lipids of model membranes, causing membrane destabilization (Bohling et al., 2006; Bai et al., 2009; Schmidt et al., 2011; Krishnakumari & Nagaraj, 2012). Electrostatic interactions with the bacterial surface are important for the activity of HBD-1 (Pazgier et al., 2007). We conclude that the loss of activity in the presence of Na+ or Ca2+ ions is due to involvement of electrochemical gradients that prevents effective electrostatic interactions with the bacterial inner membrane, and Mg2+ inhibits activity by preventing interaction of the peptides with LPS. This study provides newer insights into mechanism of action of human-β- defensins particularly attenuation of activity in the presence of mono- and divalent cations.
Funding from CSIR Network project NWP-05 is gratefully acknowledged. RN is the recipient of JC Bose Fellowship from the DST India. We acknowledge Dr. Manjula Reddy for providing us with wild-type and E. coli ftsEX mutant strains.