- Email: [email protected]
Unacylated tridecaptin A1 acts as an effective sensitiser of Gram-negative bacteria to other antibiotics
International Journal of Antimicrobial Agents 44 (2014) 493–499
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Unacylated tridecaptin A1 acts as an effective sensitiser of Gram-negative bacteria to other antibiotics Stephen A. Cochrane, John C. Vederas ∗ Department of Chemistry, University of Alberta, 30 University Campus NW, Edmonton, AB, Canada T6G 2G2
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 14 July 2014 Accepted 13 August 2014 Keywords: Tridecaptin Lipopeptide Antibiotic Gram-negative bacteria Rifampicin Vancomycin
a b s t r a c t A derivative of the linear cationic lipopeptide tridecaptin A1 missing the N-terminal lipophilic acyl group, termed H-TriA1 , is devoid of antimicrobial activity but is extremely effective at sensitising Gram-negative bacteria to certain antibiotics. H-TriA1 has low cytotoxicity compared with the natural peptide and in low concentrations it can substantially lower the minimum inhibitory concentration (MIC) of some antibiotics against strains of Escherichia coli, Campylobacter jejuni and Klebsiella pneumoniae. In particular, the MIC of rifampicin was lowered 256–512-fold against K. pneumoniae strains using low concentrations of HTriA1 . H-TriA1 does not exert its synergistic effect through partial membrane lysis, but does bind to model bacterial membranes in a manner akin to the natural peptide. Formation of this stable secondary structure on the outer membrane may account for the observed synergistic activity. © 2014 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction The continued emergence of multidrug-resistant (MDR) bacteria is a growing concern worldwide. MDR bacterial infections are twice as likely to result in death and place a significant physical and financial strain on healthcare systems [1]. Pharmaceutical companies have been quite effective at producing antibiotics for the treatment of Gram-positive bacterial infections, however new antibiotics for Gram-negative bacterial infections are still needed [2]. There has also been a lack of new structural classes of antibiotics, with only four new types entering the market in the last 40 years, namely linezolid, daptomycin, fidaxomicin and bedaquiline [2,3]. Extremely multidrug-resistant (XMR) strains of Gram-negative bacteria are of particular concern [4]. Carbapenem-resistant Klebsiella pneumoniae (CRKP) strains are resistant to multiple classes of antibiotics, including fluoroquinolones, aminoglycosides and all available -lactams [5]. Treatment of CRKP-associated infections often meets with failure, with an associated mortality rate of ≥50% [6]. XMR strains of Acinetobacter baumannii are attributed to hospital-acquired infections, especially among immunocompromised patients, and show a tendency for the development of antibiotic resistance [7]. Even XMR strains of Escherichia coli have been detected, which could have serious implications if the
∗ Corresponding author. Tel.: +1 780 492 5475; fax: +1 780 492 8231. E-mail address: [email protected] (J.C. Vederas).
need for new antibiotics targeting Gram-negative bacteria is not met [8]. The polymyxins, a class of cyclic cationic lipopeptides, are used as a last-resort treatment for XMR Gram-negative bacterial infections, and colistin (polymyxin E) is one of the few antibiotics effective against XMR strains. Recent work by Vaara and co-workers has led to the development of novel polymyxin derivatives with decreased toxicity profiles that are active against a variety of Gram-negative bacteria [9–13]. Lipopeptides generally exert their bactericidal effect through membrane disruption. As it is difficult for bacteria to alter their phospholipid bilayer, the development of resistance is often limited [14]. There are many different classes of lipopeptides and it has been suggested that they could be useful candidates for new antibiotics [15]. Tridecaptin A1 (TriA1 ) (Fig. 1) is a linear cationic lipopeptide produced by Paenibacillus spp. that shows strong activity against Gram-negative bacteria, including K. pneumoniae and A. baumannii [16–18]. A synthetic derivative of TriA1 , octyl-tridecaptin A1 (Oct-TriA1 ), also displays potent activity against clinically isolated strains of CRKP and XMR Enterobacter cloacae [18]. An alternative approach to the treatment of Gram-negative bacteria is to sensitise the outer membrane so that they are more susceptible to antibiotics normally reserved for the treatment of Gram-positive bacterial infections. Treatment of polymyxin B with ficin or related enzymes results in removal of the lipid tail and the N-terminal amino acid, yielding the less toxic polymyxin B nonapeptide (PMBN). PMBN is devoid of antimicrobial activity but acts synergistically with antibiotics that are normally blocked by the
http://dx.doi.org/10.1016/j.ijantimicag.2014.08.008 0924-8579/© 2014 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
494
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Fig. 1. Structure of tridecaptin A1 (TriA1 ) analogues.
outer membrane [19,20]. Making Gram-negative bacteria susceptible to antibiotics typically reserved for Gram-positive bacteria is an extremely attractive approach to tackling XMR Gram-negative bacterial infections, and recent work on the polymyxins has produced new derivatives for this purpose [9,21]. In this study, we show that an unacylated analogue of TriA1 , termed H-TriA1 , can act synergistically with certain classes of Gram-positive-targeting antibiotics, namely rifampicin and vancomycin, against Gram-negative bacteria. H-TriA1 has low cytotoxicity and haemolytic activity and is readily accessible by chemical synthesis. Our studies suggest that H-TriA1 does not fully disrupt the outer or inner membranes of Gram-negative bacteria, but appears to selectively permeabilise them to certain antibiotics. It does adopt a defined secondary structure similar to TriA1 in the presence of model membranes, which is a signature for antimicrobial activity of this class of lipopeptides and may be implicated in the synergistic effect.
˚ 5 m) (Phenomenex, Torrance, CA) was used for preparative 100 A, scale purification.
2. Materials and methods
2.4. Broth dilution assay
2.1. General analytical information
All minimum inhibitory concentrations (MICs) were determined according to Clinical and Standards Laboratory Institute (CLSI) guidelines [22].
All chemicals were purchased from Sigma-Aldrich (Oakville, ON, Canada) unless otherwise stated. Media, broth, bacterial strains and human embryonic kidney 293 (HEK 293) cells were purchased from ATCC (Manassas, VA). Nuclear magnetic resonance (NMR) spectra were recorded on a Varian Inova 600 MHz spectrometer (Agilent Technologies, Santa Clara, CA). For 1 H NMR spectra, ı values were referenced to D2 O (4.79 ppm). High-resolution mass spectrometry spectra were recorded on a Bruker 9.4T Apex-Qe Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (Bruker Daltonics, Billerica, MA). Circular dichroism (CD) spectra were recorded on an Olis DSM 17 CD spectrophotometer (Olis Inc., Bogart, GA). The fluorescence spectra of samples were measured in quartz cuvettes on an MP1 Fluorescence System (Photon Technology International, Edison, NJ). Ultraviolet and visible light (UV/vis) absorbances were measured on a SpectraMax® 340PC spectrophotometer (Molecular Devices, Sunnyvale, CA). Peptides were purified on a preparative high-performance liquid chromatography (HPLC) system (Gilson Inc., Middleton, WI) equipped with a model 322 HPLC pump, GX-271 liquid handler, 156 UV/vis detector and a 10 mL sample loop. A Phenomenex C18 column (21.2 × 250 mm,
2.2. Peptide synthesis See Supplementary material. 2.3. Bacterial growth conditions All cultures were grown from glycerol stocks, including E. coli DH5␣, E. coli ATCC 25922, Salmonella enterica ATCC 13311, Campylobacter jejuni NCTC 11168, K. pneumoniae ATCC 13883, A. baumannii ATCC 19606, Pseudomonas aeruginosa ATCC 27853, Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212 and Listeria monocytogenes ATCC 15313. All organisms (excluding C. jejuni) were grown in Mueller–Hinton (MH) broth at 37 ◦ C at 225 rpm. C. jejuni was grown on MH agar plates at 37 ◦ C under 10% CO2 , 5% O2 and 85% N2 .
2.5. Synergy assay A modified broth dilution assay in a 96-well plate (Corning Inc., Corning, NJ) was used to test for synergy between H-TriA1 and a panel of 10 antibiotics, where row (A) is a sterility control, (B) is antibiotic alone, (C) is antibiotic in 12.5 g/mL H-TriA1 , (D) is antibiotic in 6.25 g/mL H-TriA1 , (E) is antibiotic in 3.13 g/mL H-TriA1 , (F) is antibiotic in 1.56 g/mL H-TriA1 , (G) is 12.5 g/mL H-TriA1 growth control and (H) is a growth control. MH broth (50 L) was added to all wells of a 96-well plate. A solution of the desired antibiotic (4× desired starting concentration) in MH broth (50 L) was added to B1, C1, D1, E1 and F1, and serial dilutions were made across these rows. MH broth (50 L) was added to rows A, B and H. A 25, 12.5, 6.25 and 3.13 g/mL solution (50 L) of H-TriA1 in MH broth was added to rows C, D, E and F, respectively. A 25 g/mL solution of H-TriA1 in MH broth (50 L) was added to row G. Rows B–H were inoculated with 5 L of bacterial suspension (see CLSI guidelines) so that the final concentration in each well was 5 × 105 CFU/mL.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
The MIC was measured as the lowest concentration of antibiotic required to prevent visible growth. 2.6. Cytotoxicity assay with HEK 293 cells A frozen dimethyl sulphoxide (DMSO) stock of HEK 293 cells was thawed and diluted with 10 mL of HyCloneTM DMEM/High Glucose (ATCC) + 10% foetal bovine serum (ATCC) medium. Cells were centrifuged and re-suspended in medium (10 mL) in a 75-mL culture flask and were grown at 37 ◦ C for 48 h in a 5% CO2 incubator. The medium was removed, cells were washed with phosphate-buffered saline (PBS) solution (10 mL) and fresh medium was added (10 mL). After 48 h, the medium was removed and cells were washed with PBS, followed by trypsinisation by incubating with 0.0625% trypsin in medium (3 mL) for 5 min. The trypsinised suspension was added to medium (8 mL) and centrifuged. The cell pellet was re-suspended in medium (1 mL) and 100 L was used to start a new culture or for a 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (MTT) assay. The re-suspended cell pellet (100 L) was diluted with medium (10 mL). Cells were counted using a haemocytometer and the concentration was adjusted to 2.5 × 105 cells/mL. The MTT assay was performed in a 96-well plate, where row (A) is a sterility control, (B–D) are peptide and (E) is a growth control. The suspension of HEK 293 cells (100 L) was added to rows B–E, and medium (100 L) was added to row A. The cells were incubated at 37 ◦ C for 24 h. The medium was replaced in all wells. A 16 mg/mL peptide solution (100 L) was added to B1, C1 and D1 and serial dilutions were performed across each row. The plate was incubated for 48 h at 37 ◦ C. A 2.5 mg/mL solution of thiazolyl blue tetrazolium bromide in PBS (20 L) was added to all wells and the plate was incubated for 3 h. The medium was carefully removed from all wells and was replaced with DMSO (100 L). After 30 min, the absorbance at 570 nm was measured in each well. A graph of the absorbance versus concentration was plotted and the MIC was determined as the lowest concentration before absorbance increased from background values. 2.7. Outer-membrane permeability assay The ability of peptides to permeate the outer membrane of E. coli ATCC 25922 cells was determined using the 8-anilinonaphthalene1-sulphonic acid (ANS) uptake assay [23]. A fully-grown overnight culture in MH broth was used to inoculate fresh broth to an optical density at 600 nm (OD600 ) of 0.05. Cells were grown to mid-log phase (OD600 ∼ 0.3–0.8) and were pelleted by centrifugation. Cells were washed with buffer [10 mM Tris (pH 7.4), 150 mM NaCl] and were pelleted by centrifugation. Buffer was added to obtain at least 20 mL of a suspension of cells with an OD600 of 0.065. A constant concentration of 5 M ANS in bacterial buffer suspension (2 mL) was used for all experiments and the concentration of TriA1 and H-TriA1 was varied. The peptide was added and the fluorescence emission was measured after 10 min between 450 nm and 600 nm with excitation at 380 nm. The increase in fluorescence from blank readings was used to assess the penetration of the outer membrane. 2.8. Inner membrane permeability The ability of peptides to penetrate the inner membrane of E. coli ML-35 cells was determined using an o-nitrophenyl--dgalactopyranoside (ONPG) assay [24]. An overnight culture in MH broth was grown from a glycerol stock and 100 L was added to fresh broth (10 mL). Cells were grown for 3 h with shaking at 37 ◦ C to mid-log phase (ca. 2 × 1011 CFU/mL) and the suspension was cooled to 0 ◦ C. A 96-well plate was used for the ONPG assay, where row (A) is a blank, (B) is a positive control (melittin) and (C) and (D) are peptide. ONPG buffer [70 L; 10 mM NaH2 PO4 (pH 7.4), 100 mM NaCl
495
and 1.5 mM ONPG] and cell suspension (20 L) were added to the first six wells in rows A–D. A 64 g/mL solution of melittin (10 L) was added to row B. Next, 10 L of 40×, 20×, 10×, 5×, 2.5× and 1.25× MIC peptide solutions was added to C1, C2, C3, C4, C5 and C6, respectively. The experiment was run in duplicate using D1–D6. A multichannel pipette was used to add the melittin and peptide solutions as close in time as possible. After 15 min, the absorbance at 405 nm of each well was measured. The percentage of leakage was calculated for each peptide concentration taking the positive control as 100%. 2.9. Large unilamellar vesicle (LUV) preparation 1-Palmitoyl-2-oleoylphosphatidylcholine (POPC) and 1palmitoyl-2-oleoylphosphatidylglycerol (POPG) were purchased as 25 mg/mL chloroform solutions from Avanti Polar Lipids (Alabaster, AL). The mini-extruder and membranes were also purchased from Avanti Polar Lipids. POPC (228 L, 7.5 mol) and POPG (77 L, 2.5 mol) were mixed and dried under an argon stream and were placed under high vacuum for 2 h. The film was hydrated with phosphate buffer [1 mL, 5 mM KH2 PO4 (pH 7.0)] and was vortexed for 30 s. The mixture was frozen using a dry ice/acetone bath, thawed with warm water and vortexed for 30 s. The freeze–thaw–vortex cycle was repeated twice. Passing the milky suspension through a 50-nm polycarbonate membrane 10 times using a mini-extruder formed 50-nm LUVs. The phosphate concentration of the LUV solution was determined using the Stewart method [25]. 2.10. Circular dichroism spectroscopy Samples were added to a 0.4-mL quartz cuvette with a 0.1 cm path length, measured from 190–250 nm at 20 ◦ C and averaged over five scans. The peptide concentration remained constant at 20 M for all experiments. The CD spectrum of phosphate buffer was subtracted from all spectra. Ten lipid equivalents (by phosphate concentration) were mixed thoroughly with the peptide solution before measuring. A digital filter of 15 was applied to CD spectra, which were converted to molar ellipticity units. 3. Results and discussion 3.1. Preliminary synergistic testing of three tridecaptin A1 derivatives Our previous structure–activity relationship studies on Nterminal derivatives of TriA1 identified three analogues with significantly lower antimicrobial activities than the natural peptide [18]. H-TriA1 is not acylated at the N-terminus with a lipid tail, PEG-tridecaptin A1 (PEG-TriA1 ) is acylated with a triethylene glycol derivative, and biotin-tridecaptin A1 (Bio-TriA1 ) is acylated with biotin (Fig. 1). The antimicrobial activities of these analogues were determined against a panel of Gram-negative bacteria (Table 1). PEG-TriA1 was inactive at all concentrations tested (>100 g/mL), H-TriA1 required 100–200 g/mL to inhibit the growth of bacteria, and Bio-TriA1 was slightly more active with a MIC of 50 g/mL against most bacteria tested. These analogues were screened for synergistic activity with rifampicin against E. coli ATCC 25922 at sub-MIC concentrations of 12.5, 6.25, 3.13 and 1.56 g/mL (Table 2). Enterobacteriaceae, Acinetobacter and Pseudomonas spp. are intrinsically resistant to rifampicin, but disruption of the outer membrane increases their susceptibility to this antibiotic. PEGTriA1 showed no synergistic activity, whilst Bio-TriA1 caused a small decrease in the MIC of rifampicin (8×). H-TriA1 displayed excellent synergy with rifampicin, lowering the MIC by 512× at 12.5 g/mL and by 64× at 6.25 g/mL. H-TriA1 was therefore used
496
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Table 1 Minimum inhibitory concentrations (MICs) of tridecaptin A1 (TriA1 ) analogues against Gram-negative bacteria. MIC (g/mL)
Organism
TriA1 Escherichia coli ATCC 25922 E. coli DH5␣ Salmonella enterica ATCC 13311 Campylobacter jejuni NCTC 11168 Pseudomonas aeruginosa ATCC 27853 Klebsiella pneumoniae ATCC 13883 Acinetobacter baumannii ATCC 19606
H-TriA1
3.13 3.13 6.25 3.13 50 3.13 12.5
100 100 100 100 >100 100 200
Bio-TriA1
PEG-TriA1
50 50 50 50 >100 50 >100
>100 >100 >100 >100 >100 >100 >100
H-TriA1 , unacylated analogue of tridecaptin A1 ; Bio-TriA1 , biotin-tridecaptin A1 ; PEG-TriA1 , PEG-tridecaptin A1 .
for subsequent synergistic studies. It is interesting to note that capping the N-terminus significantly lowers the ability of TriA1 analogues to sensitise Gram-negative bacteria. This may be due to the decreased electrostatic attraction between the positively charged peptide and the anionic bacterial outer membrane. 3.2. Cytotoxicity and plasma stability studies Our previous work has shown that H-TriA1 has low haemolytic activity, causing only 0.5% haemolysis at a concentration of 83 g/mL [18]. To further assess the toxicity profile of H-TriA1 , toxicity against HEK 293 cells was determined. The natural peptide inhibits all growth of HEK 293 cells at a concentration of 100 g/mL, whereas H-TriA1 is significantly less toxic, requiring 4000 g/mL for the same effect. The cytotoxic concentration of H-TriA1 against HEK 293 cells is 320× larger than the highest concentration used in the synergy assay (12.5 g/mL) and suggests that H-TriA1 could be part of a clinically safe antibiotic combination. The stability of H-TriA1 in human plasma was also assessed. Incubation of the peptide in human plasma at 37 ◦ C for 30 min showed minimal degradation of the peptide (<1%; see Supplementary material). Furthermore, the MIC of natural TriA1 against E. coli ATCC 25922 was determined in the presence of 4% bovine serum albumin and was found to be unaffected by protein binding. Therefore, it is unlikely that the mechanism by which H-TriA1 exerts its synergistic effect will be affected by protein binding. In addition to its favourable toxicity profile and plasma stability, H-TriA1 can also be synthesised easily and efficiently by standard peptide synthesis protocols. In contrast to the cyclic polymyxins, no orthogonal protection of 2,4-diaminobutyric acid residues or complex cyclisation steps are required to synthesise the tridecaptins. 3.3. Synergy of H-TriA1 with other antibiotics The synergistic activity of H-TriA1 with ten other antibacterial agents was assessed against a panel of Gram-positive and Gramnegative bacteria (Table 3). The lantibiotics nisin and gallidermin sequester lipid II and cause pore formation in Gram-positive bacteria [26] but cannot cross the outer membrane of Gram-negative bacteria. Vancomycin also binds to lipid II and inhibits peptidoglycan formation [27] but is similarly blocked by the outer membrane.
Daptomycin exerts its bactericidal effect against Gram-positive bacteria through membrane disruption [28]. The other antibiotics tested (penicillin G, ampicillin, ciprofloxacin, tetracycline and streptomycin) have activity both against Gram-positive and Gramnegative bacteria. However, resistance mechanisms are common for these antibiotics. Natural TriA1 has weak activity against Gram-positive strains as well as the Gram-negative Pseudomonas spp. It is therefore not surprising that no notable synergistic activity with antimicrobial agents was observed when P. aeruginosa, S. aureus, E. faecalis and L. monocytogenes were treated with H-TriA1 . However, synergistic effects with certain antimicrobial agents were observed for the other Gram-negative strains tested. The largest synergistic effects were observed with rifampicin, showing a decrease in MIC of 512× against E. coli strains and K. pneumoniae ATCC 13883, and 128× against C. jejuni. Moderate decreases of 32× and 16× were found against S. enterica and A. baumannii, respectively. Strong synergistic effects were also found between vancomycin and H-TriA1 , with a large MIC decrease of 256× against K. pneumoniae ATCC 13883, 32–64× against E. coli strains and 16× against the other Gramnegative strains. Of particular note is the large decrease in MIC observed for vancomycin against K. pneumoniae, which normally has a high tolerance to this antibiotic (MIC = 100 g/mL). This combination could therefore have potential for the treatment of CRKP infections, against which Oct-TriA1 is active [18]. Surprisingly, moderate decreases in the MIC (2–16×) of the lantibiotics gallidermin and nisin were found against the E. coli strains, S. enterica, C. jejuni and K. pneumoniae, which was unexpected given the large size of lantibiotics (Mw > 2 kDa). Although lantibiotics are not yet commonly used in the clinical setting, they are excellent food preservatives [29], where the use of antimicrobial peptides is often preferred over conventional antibiotics. Given the synergy observed against these common foodborne pathogens and the fact that H-TriA1 is also a peptide, there may be an application for HTriA1 as a food preservative. No synergistic effects were observed with daptomycin, which was not active at the highest concentration tested (500 g/mL). The synergistic effects of the remaining antimicrobial agents (penicillin G, ampicillin, ciprofloxacin, tetracycline and streptomycin) ranged from small to moderate. Ciprofloxacin and ampicillin had similar MIC decreases (8–16×) to vancomycin and rifampicin against A.
Table 2 Synergy between rifampicin and peptides against Escherichia coli ATCC 25922. Analogue
PEG-TriA1 Bio-TriA1 H-TriA1
Rifampicin MICa at peptide concentrationsa of 12.5
6.25
3.13
1.56
0
1.56 0.4 0.012
1.56 0.78 0.1
3.13 1.56 0.78
3.13 3.13 0.78
3.13 3.13 6.25
MMD
FICI
2× 8× 512×
0.514 0.132 0.002
MIC, minimum inhibitory concentration; MMD, maximum MIC decrease; FICI, fractional inhibitory concentration index [FICI = (MICpeptide+antibiotic )/(MICpeptide ) + (MICpeptide+antibiotic )/(MICantibiotic )]; PEG-TriA1 , PEG-tridecaptin A1 ; Bio-TriA1 , biotin-tridecaptin A1 ; H-TriA1 , unacylated analogue of tridecaptin A1 . a All values reported in g/mL.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
497
Table 3 Synergy between H-tridecaptin A1 (H-TriA1 ), an unacylated analogue of tridecaptin A1 , and antimicrobial agents. MICa at H-TriA1 concentrationsa of
MMD
FICI
25 100 >500 50 6.25 31.3 0.39 3.13 0.003 6.25
8× 16× N/D 64× 512× 4× 2× 2× 0 2×
0.157 0.125 N/D 0.023 0.002 0.327 0.515 0.514 1.000 0.532
6.25 6.25 >500 6.25 0.39 12.5 0.1 0.1 0.003 0.2
12.5 50 >500 50 6.25 12.5 0.1 0.1 0.003 1.56
4× 16× N/D 32× 512× 2× 0 0 0 8×
0.282 0.094 N/D 0.047 0.002 0.563 1.000 1.000 1.000 0.130
25 50 >500 12.5 0.78 3.13 0.39 6.25 0.003 0.78
25 50 >500 50 1.56 3.13 0.39 6.25 0.003 1.56
50 100 >500 100 3.13 12.5 0.39 6.25 0.006 3.13
4× 4× N/D 16× 32× 8× 2× 2× 2× 4×
0.375 0.500 N/D 0.125 0.033 0.140 0.515 0.532 0.500 0.257
6.25 6.25 >500 12.5 1.56 6.25 0.024 1.56 0.024 6.25
6.25 6.25 >500 50 12.5 12.5 0.024 1.56 0.024 6.25
12.5 6.25 >500 100 12.5 12.5 0.024 1.56 0.024 12.5
12.5 12.5 >500 200 12.5 12.5 0.024 1.56 0.024 12.5
2× 4× N/D 16× 128× 2× 0 0 4× 2×
0.563 0.281 N/D 0.188 0.009 0.563 1.000 1.016 0.250 0.563
Klebsiella pneumoniae ATCC 13883 0.78 Nisin 1.56 Gallidermin >500 Daptomycin Vancomycin 0.39 0.006 Rifampicin 3.13 Penicillin G 0.78 Tetracycline 0.1 Streptomycin 0.002 Ciprofloxacin 50 Ampicillin
3.13 3.13 >500 1.56 0.024 25 0.78 0.20 0.003 100
3.13 3.13 >500 3.13 0.024 50 0.78 0.39 0.003 100
6.25 6.25 >500 6.25 0.03 50 0.78 0.39 0.003 250
12.5 50 >500 100 3.13 200 1.56 0.39 0.012 250
16× 32× N/D 256× 512× 64× 2× 4× 4× 5×
0.070 0.047 N/D 0.008 0.002 0.047 0.508 0.257 0.167 0.700
K. pneumoniae ATCC 700603 Vancomycin Rifampicin
31.3 0.2
62.5 0.4
62.5 3.13
250 12.5
8× 256×
0.438 0.005
3.13 12.5 >500 12.5 0.20 125 0.20 25 0.10 62.5
6.25 12.5 >500 25 0.39 125 0.20 25 0.10 125
6.25 12.5 >500 50 0.39 250 0.20 50 0.20 250
6.25 12.5 >500 50 0.39 500 0.20 50 0.20 500
2× 0 N/D 16× 16× 16× 0 4× 8× 16×
0.516 1.063 N/D 0.078 0.062 0.375 1.001 0.313 0.120 0.219
Bacteria/antibiotic
12.5
6.25
3.13
1.56
0
Escherichia coli ATCC 25922 Nisin Gallidermin Daptomycin Vancomycin Rifampicin Penicillin G Tetracycline Streptomycin Ciprofloxacin Ampicillin
3.13 6.25 >500 0.78 0.012 7.8 0.2 1.56 0.003 3.13
3.13 6.25 >500 6.25 0.1 7.8 0.2 1.56 0.003 3.13
6.25 6.25 >500 25 0.78 7.8 0.2 1.56 0.003 3.13
6.25 12.5 >500 50 0.78 15.6 0.39 1.56 0.003 3.13
E. coli DH5␣ Nisin Gallidermin Daptomycin Vancomycin Rifampicin Penicillin G Tetracycline Streptomycin Ciprofloxacin Ampicillin
3.13 3.13 500 1.56 0.012 6.25 0.1 0.1 0.003 0.2
6.25 6.25 >500 6.25 0.2 6.25 0.1 0.1 0.003 0.2
6.25 6.25 >500 6.25 0.39 12.5 0.1 0.1 0.003 0.2
Salmonella enterica ATCC 13311 12.5 Nisin Gallidermin 25 >500 Daptomycin 6.25 Vancomycin 0.1 Rifampicin 1.56 Penicillin G 0.2 Tetracycline 3.13 Streptomycin Ciprofloxacin 0.003 0.78 Ampicillin
12.5 50 >500 25 0.39 1.56 0.2 3.13 0.003 0.78
Campylobacter jejuni NCTC 11168 Nisin 6.25 Gallidermin 3.13 Daptomycin >500 Vancomycin 12.5 0.1 Rifampicin Penicillin G 6.25 Tetracycline 0.024 Streptomycin 1.56 0.006 Ciprofloxacin 6.25 Ampicillin
31.3 0.05
Acinetobacter baumannii ATCC 19606 Nisin 3.13 12.5 Gallidermin >500 Daptomycin 3.13 Vancomycin 0.024 Rifampicin 31.25 Penicillin G 0.20 Tetracycline 12.5 Streptomycin 0.024 Ciprofloxacin 31.25 Ampicillin
498
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Table 3 (Continued) Bacteria/antibiotic
MICa at H-TriA1 concentrationsa of 12.5
6.25
3.13
1.56
0
A. baumannii ATCC BAA-1605 Vancomycin 250 0.05 Rifampicin
250 0.2
250 0.2
250 0.2
250 0.78
MMD
FICI
0 16×
1.625 0.064
MIC, minimum inhibitory concentration; MMD, maximum MIC decrease; FICI, fractional inhibitory concentration index [FICI = (MICpeptide+antibiotic )/(MICpeptide ) + (MICpeptide+antibiotic )/(MICantibiotic )]; N/D, not determined. No synergy was observed against Pseudomonas aeruginosa ATCC 27853, Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212 or Listeria monocytogenes ATCC 15313. a All values reported in g/mL.
baumannii, which appears to be less susceptible to the synergistic action of H-TriA1 than the other strains. The synergistic combinations of rifampicin/H-TriA1 and vancomycin/H-TriA1 were also screened against more virulent strains of K. pneumoniae and A. baumannii. K. pneumoniae ATCC 700603 produces -lactamase SHV-18 and is resistant to several -lactams, cephalosporins and tetracycline. A. baumannii ATCC BAA-1605 is resistant to several -lactams, cephalosporins, gentamicin and ciprofloxacin. A moderate MIC decrease of 8× was observed for vancomycin against the MDR Klebsiella strain, however excellent synergistic effects were observed with rifampicin, improving the MIC by 256×. No synergistic effect was observed against A. baumannii ATCC BAA-1605 with vancomycin and a 16× improvement was found with rifampicin. 3.4. Rationale for synergistic effect In an attempt to explain the synergistic effects of H-TriA1 on Gram-negative bacteria, a series of mechanistic studies were performed. An ANS uptake assay can be used to monitor the ability of antimicrobial peptides to disrupt the outer membrane. Fluorescence of the hydrophobic fluorophore ANS increases when it relocates from bulk solution to the membrane interior upon membrane disruption [23]. An ANS uptake assay using E. coli ATCC 25922 cells revealed that although natural TriA1 caused an increase in ANS fluorescence at MIC concentrations and higher, H-TriA1 had no effect on fluorescence (Fig. 2). This may explain the difference
Fig. 2. Increase in fluorescence observed when Escherichia coli ATCC 25922 cells in 8-anilinonaphthalene-1-sulfonic acid (ANS) buffer are exposed to increasing concentrations of tridecaptin A1 (TriA1 ) and to H-TriA1 (an unacylated analogue of TriA1 ): (1) 8× MIC TriA1 ; (2) 4× MIC TriA1 ; (3) 2× MIC TriA1 ; (4) 1× MIC TriA1 ; (5) H-TriA1 (25 g/mL); (6) no peptide. No effect is observed with H-TriA1 . MIC, minimum inhibitory concentration.
in antimicrobial activities and cytotoxicity between the natural and unacylated peptides, however it is surprising that H-TriA1 allows passage of large antibiotics such as vancomycin and rifampicin through the outer membrane but not a small hydrophobic fluorophore. Like the polymyxins, the N-terminal lipid tail is likely required for permeation of the outer membrane and would explain why H-TriA1 may not permeate the outer membrane. The ability of H-TriA1 to disrupt the inner membrane of E. coli ML-35 cells was assessed using an ONPG assay. Disruption of the inner membrane releases a -galactosidase that can cleave the o-nitrophenyl group from ONPG in the buffer, whose absorbance is monitored at 405 nm. No -galactosidase activity was detected in cells treated with H-TriA1 (60 g/mL), whereas natural TriA1 caused 95%, 69% and 25% leakage at 4×, 2× and 1× MIC, respectively. These results suggest that the synergistic effect of H-TriA1 with other antibiotics is not due to partial membrane lysis. Previous studies have shown that Oct-TriA1 is unstructured in water but adopts a stable secondary structure in the presence of negatively charged phospholipid LUVs that mimic the bacterial membrane [30]. Replacement of key amino acids with alanine resulted in the loss of this secondary structure and a concomitant loss of antimicrobial activity. The CD spectrum of HTriA1 in the presence of 10 equivalents (by phosphate content) of 3:1 POPC:POPG LUVs was obtained (Fig. 3). The CD spectrum of H-TriA1 in the model membrane environment is very similar to Oct-TriA1 , suggesting that H-TriA1 adopts a similar secondary structure on the outer membrane of Gram-negative bacteria, which may account for the observed synergistic activity. The magnitude of the negative band at 196 nm can be correlated with antimicrobial activity and would also explain the significantly lower activity of H-TriA1 relative to the natural peptide.
Fig. 3. Circular dichroism spectra of octyl-tridecaptin A1 (Oct-TriA1 ) and HTriA1 (an unacylated analogue of tridecaptin A1 ) in the presence of 10 equivalents (by phosphate content) of 3:1 POPC:POPG large unilamellar vesicles: (1) H-TriA1 ; (2) Oct-TriA1 . = wavelength (nm), [] = molar ellipticity (cm2 /dmol). POPC, 1-palmitoyl-2-oleoylphosphatidylcholine; POPG, 1-palmitoyl-2oleoylphosphatidylglycerol.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
4. Conclusions An unacylated analogue of the cationic lipopeptide TriA1 , termed H-TriA1 , was found to be an excellent sensitiser of Gramnegative bacteria to certain antimicrobial agents normally reserved for the treatment of Gram-positive bacterial infections. In particular, H-TriA1 increased the potency of the hydrophobic antibiotic rifampicin 512× and vancomycin 256× against one of the K. pneumoniae strains that is normally resistant to these compounds. H-TriA1 has an excellent toxicity profile, only showing a cytotoxic effect against human embryonic kidney cells at 320× the highest synergistic concentration. Mechanistic studies revealed that HTriA1 does not exert its synergistic effect through membrane lysis, but may sensitise Gram-negative bacteria by forming a stable secondary structure on the outer membrane. With the continued rise in the number of XMR strains of Gram-negative bacteria, finding new antibiotics or methods to increase the effectiveness of existing antibiotics is vital. Development of resistance to lipopeptide-based antimicrobial agents is often limited, therefore H-TriA1 could be an ideal candidate for further biological studies. Funding: Natural Sciences and Engineering Research Council of Canada (NSERC), Griffith Laboratories Canada, and the Canada Research Chair in Bioorganic and Medicinal Chemistry. Competing interests: None declared. Ethical approval: Not required. Acknowledgments The authors are grateful to Shaun McKinnie for help with plasma stability assays, to Dr Brandon Findlay for his advice on antimicrobial testing, to Jing Zheng and Dr Randy Whittal for assistance with mass spectrometry, to Mark Miskolzie for assistance with NMR, to Bernadette Beadle for assistance with Campylobacter testing, and to Wayne Moffat for assistance with CD experiments.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijantimicag. 2014.08.008. References [1] Paphitou NI. Antimicrobial resistance: action to combat the rising microbial challenges. Int J Antimicrob Agents 2013;42:S25–8. [2] Butler MS, Blaskovich MA, Cooper MA. Antibiotics in the clinical pipeline in 2013. J Antibiot (Tokyo) 2013;66:571–91. [3] Roemer T, Boone C. Systems-level antimicrobial drug and drug synergy discovery. Nat Chem Biol 2013;9:222–31. [4] Tzouvelekis LS, Markogiannakis A, Psichogiou M, Tassios PT, Daikos GL. Carbapenemases in Klebsiella pneumoniae and other Enterobacteriaceae: an evolving crisis of global dimensions. Clin Microbiol Rev 2012;25:682–707. [5] Lee GC, Burgess DS. Treatment of Klebsiella pneumoniae carbapenemase (KPC) infections: a review of published case series and case reports. Ann Clin Microbiol Antimicrob 2012;11:32.
499
[6] Gasink LB, Edelstein PH, Lautenbach E, Synnestvedt M, Fishman NO. Risk factors and clinical impact of Klebsiella pneumoniae carbapenemase-producing K. pneumoniae. Infect Control Hosp Epidemiol 2009;30:1180–5. [7] Karah N, Sundsfjord A, Towner K, Samuelsen O. Insights into the global molecular epidemiology of carbapenem non-susceptible clones of Acinetobacter baumannii. Drug Resist Updat 2012;15:237–47. [8] Cantón R, Akóva M, Carmeli Y, Giske CG, Glupczynski Y, Gniadkowski M, et al. European Network on Carbapenemases. Rapid evolution and spread of carbapenemases among Enterobacteriaceae in Europe. Clin Microbiol Infect 2012;18:413–31. [9] Vaara M, Fox J, Loidl G, Siikanen O, Apajalahti J, Hansen F, et al. Novel polymyxin derivatives carrying only three positive charges are effective antibacterial agents. Antimicrob Agents Chemother 2008;52:3229–36. [10] Vingsbo Lundberg C, Vaara T, Frimødt-Moller N, Vaara M. Novel polymyxin derivatives are effective in treating experimental Escherichia coli peritoneal infection in mice. J Antimicrob Chemother 2010;65:981–5. [11] Mingeot-Leclercq MP, Tulkens PM, Denamur S, Vaara T, Vaara M. Novel polymyxin derivatives are less cytotoxic than polymyxin B to renal proximal tubular cells. Peptides 2012;35:248–52. [12] Vaara M, Sader HS, Rhomberg PR, Jones RN, Vaara T. Antimicrobial activity of the novel polymyxin derivative NAB739 tested against Gram-negative pathogens. J Antimicrob Chemother 2013;68:636–9. [13] Vaara M, Vaara T. The novel polymyxin derivative NAB739 is remarkably less cytotoxic than polymyxin B and colistin to human kidney proximal tubular cells. Int J Antimicrob Agents 2013;41:292–3. [14] Pirri G, Giuliani A, Nicoletto SF, Pizzuto L, Rinaldi AC. Lipopeptides as antiinfectives: a practical perspective. Cent Eur J Biol 2009;4:258–73. [15] Cochrane SA, Vederas JC. Lipopeptides from Bacillus and Paenibacillus spp.: a gold mine of antibiotic candidates. Med Res Rev 2014 (Epub ahead of print). doi: 10.1002/med.21321. [16] Kato T, Hinoo H, Shoji J. The structure of tridecaptin A (studies on antibiotics from the genus Bacillus. XXIV). J Antibiot (Tokyo) 1978;31:652–61. [17] Lohans CT, van Belkum MJ, Cochrane SA, Huang Z, Sit CS, McMullen LM, et al. Biochemical, structural, and genetic characterization of tridecaptin A1 , an antagonist of Campylobacter jejuni. ChemBioChem 2014;15:243–9. [18] Cochrane SA, Lohans CT, Brandelli JR, Mulvey G, Armstrong GD, Vederas JC. Synthesis and structure–activity relationship studies of N-terminal analogues of the antimicrobial peptide tridecaptin A1 . J Med Chem 2014;57: 1127–31. [19] Vaara M, Vaara T. Sensitization of Gram-negative bacteria to antibiotics and complement by a nontoxic oligopeptide. Nature 1983;303:526–8. [20] Viljanen P, Vaara M. Susceptibility of Gram-negative bacteria to polymyxin B nonapeptide. Antimicrob Agents Chemother 1984;25:701–5. [21] Vaara M, Siikanen O, Apajalahti J, Fox J, Frimodt-Møller N, He H, et al. A novel polymyxin derivative that lacks the fatty acid tail and carries only three positive charges has strong synergism with agents excluded by the intact outer membrane. Antimicrob Agents Chemother 2010;54:3341–6. [22] Wiegand I, Hilpert K, Hancock RE. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat Protoc 2008;3:163–75. [23] Thennarasu S, Lee DK, Tan A, Prasad Kari U, Ramamoorthy A. Antimicrobial activity and membrane selective interactions of a synthetic lipopeptide MSI843. Biochim Biophys Acta 2005;1711:49–58. [24] Pate M, Blazyk J. Methods for assessing the structure and function of cationic antimicrobial peptides. Methods Mol Med 2008;142:155–73. [25] Stewart JC. Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 1980;104:10–4. [26] Bonelli RR, Schneider T, Sahl H-G, Wiedemann I. Insights into in vitro activities of lantibiotics from gallidermin and epidermin mode-of-action studies. Antimicrob Agents Chemother 2006;50:1449–57. [27] Watanakkunakorn C. Mode of action and in-vitro activity of vancomycin. J Antimicrob Chemother 1984;14:7–18. [28] Capepari P, Boaretti M, Lleo MM, Satta G. Lipoteichoic acid as a new target for activity of antibiotics: mode of action of daptomycin (LY146032). Antimicrob Agents Chemother 1990;34:1220–6. [29] Hansen JN. Nisin as a model food preservative. Crit Rev Food Sci Nutr 1994;34:69–93. [30] Cochrane SA, Findlay B, Vederas JC, Ratemi ES. Key residues in octyl-tridecaptin A1 analogues linked to stable secondary structure in the membrane. ChemBioChem 2014;15:1295–9.
Contents lists available at ScienceDirect
International Journal of Antimicrobial Agents journal homepage: http://www.elsevier.com/locate/ijantimicag
Unacylated tridecaptin A1 acts as an effective sensitiser of Gram-negative bacteria to other antibiotics Stephen A. Cochrane, John C. Vederas ∗ Department of Chemistry, University of Alberta, 30 University Campus NW, Edmonton, AB, Canada T6G 2G2
a r t i c l e
i n f o
Article history: Received 8 May 2014 Received in revised form 14 July 2014 Accepted 13 August 2014 Keywords: Tridecaptin Lipopeptide Antibiotic Gram-negative bacteria Rifampicin Vancomycin
a b s t r a c t A derivative of the linear cationic lipopeptide tridecaptin A1 missing the N-terminal lipophilic acyl group, termed H-TriA1 , is devoid of antimicrobial activity but is extremely effective at sensitising Gram-negative bacteria to certain antibiotics. H-TriA1 has low cytotoxicity compared with the natural peptide and in low concentrations it can substantially lower the minimum inhibitory concentration (MIC) of some antibiotics against strains of Escherichia coli, Campylobacter jejuni and Klebsiella pneumoniae. In particular, the MIC of rifampicin was lowered 256–512-fold against K. pneumoniae strains using low concentrations of HTriA1 . H-TriA1 does not exert its synergistic effect through partial membrane lysis, but does bind to model bacterial membranes in a manner akin to the natural peptide. Formation of this stable secondary structure on the outer membrane may account for the observed synergistic activity. © 2014 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
1. Introduction The continued emergence of multidrug-resistant (MDR) bacteria is a growing concern worldwide. MDR bacterial infections are twice as likely to result in death and place a significant physical and financial strain on healthcare systems [1]. Pharmaceutical companies have been quite effective at producing antibiotics for the treatment of Gram-positive bacterial infections, however new antibiotics for Gram-negative bacterial infections are still needed [2]. There has also been a lack of new structural classes of antibiotics, with only four new types entering the market in the last 40 years, namely linezolid, daptomycin, fidaxomicin and bedaquiline [2,3]. Extremely multidrug-resistant (XMR) strains of Gram-negative bacteria are of particular concern [4]. Carbapenem-resistant Klebsiella pneumoniae (CRKP) strains are resistant to multiple classes of antibiotics, including fluoroquinolones, aminoglycosides and all available -lactams [5]. Treatment of CRKP-associated infections often meets with failure, with an associated mortality rate of ≥50% [6]. XMR strains of Acinetobacter baumannii are attributed to hospital-acquired infections, especially among immunocompromised patients, and show a tendency for the development of antibiotic resistance [7]. Even XMR strains of Escherichia coli have been detected, which could have serious implications if the
∗ Corresponding author. Tel.: +1 780 492 5475; fax: +1 780 492 8231. E-mail address: [email protected] (J.C. Vederas).
need for new antibiotics targeting Gram-negative bacteria is not met [8]. The polymyxins, a class of cyclic cationic lipopeptides, are used as a last-resort treatment for XMR Gram-negative bacterial infections, and colistin (polymyxin E) is one of the few antibiotics effective against XMR strains. Recent work by Vaara and co-workers has led to the development of novel polymyxin derivatives with decreased toxicity profiles that are active against a variety of Gram-negative bacteria [9–13]. Lipopeptides generally exert their bactericidal effect through membrane disruption. As it is difficult for bacteria to alter their phospholipid bilayer, the development of resistance is often limited [14]. There are many different classes of lipopeptides and it has been suggested that they could be useful candidates for new antibiotics [15]. Tridecaptin A1 (TriA1 ) (Fig. 1) is a linear cationic lipopeptide produced by Paenibacillus spp. that shows strong activity against Gram-negative bacteria, including K. pneumoniae and A. baumannii [16–18]. A synthetic derivative of TriA1 , octyl-tridecaptin A1 (Oct-TriA1 ), also displays potent activity against clinically isolated strains of CRKP and XMR Enterobacter cloacae [18]. An alternative approach to the treatment of Gram-negative bacteria is to sensitise the outer membrane so that they are more susceptible to antibiotics normally reserved for the treatment of Gram-positive bacterial infections. Treatment of polymyxin B with ficin or related enzymes results in removal of the lipid tail and the N-terminal amino acid, yielding the less toxic polymyxin B nonapeptide (PMBN). PMBN is devoid of antimicrobial activity but acts synergistically with antibiotics that are normally blocked by the
http://dx.doi.org/10.1016/j.ijantimicag.2014.08.008 0924-8579/© 2014 Elsevier B.V. and the International Society of Chemotherapy. All rights reserved.
494
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Fig. 1. Structure of tridecaptin A1 (TriA1 ) analogues.
outer membrane [19,20]. Making Gram-negative bacteria susceptible to antibiotics typically reserved for Gram-positive bacteria is an extremely attractive approach to tackling XMR Gram-negative bacterial infections, and recent work on the polymyxins has produced new derivatives for this purpose [9,21]. In this study, we show that an unacylated analogue of TriA1 , termed H-TriA1 , can act synergistically with certain classes of Gram-positive-targeting antibiotics, namely rifampicin and vancomycin, against Gram-negative bacteria. H-TriA1 has low cytotoxicity and haemolytic activity and is readily accessible by chemical synthesis. Our studies suggest that H-TriA1 does not fully disrupt the outer or inner membranes of Gram-negative bacteria, but appears to selectively permeabilise them to certain antibiotics. It does adopt a defined secondary structure similar to TriA1 in the presence of model membranes, which is a signature for antimicrobial activity of this class of lipopeptides and may be implicated in the synergistic effect.
˚ 5 m) (Phenomenex, Torrance, CA) was used for preparative 100 A, scale purification.
2. Materials and methods
2.4. Broth dilution assay
2.1. General analytical information
All minimum inhibitory concentrations (MICs) were determined according to Clinical and Standards Laboratory Institute (CLSI) guidelines [22].
All chemicals were purchased from Sigma-Aldrich (Oakville, ON, Canada) unless otherwise stated. Media, broth, bacterial strains and human embryonic kidney 293 (HEK 293) cells were purchased from ATCC (Manassas, VA). Nuclear magnetic resonance (NMR) spectra were recorded on a Varian Inova 600 MHz spectrometer (Agilent Technologies, Santa Clara, CA). For 1 H NMR spectra, ı values were referenced to D2 O (4.79 ppm). High-resolution mass spectrometry spectra were recorded on a Bruker 9.4T Apex-Qe Fourier transform ion cyclotron resonance (FT-ICR) mass spectrometer (Bruker Daltonics, Billerica, MA). Circular dichroism (CD) spectra were recorded on an Olis DSM 17 CD spectrophotometer (Olis Inc., Bogart, GA). The fluorescence spectra of samples were measured in quartz cuvettes on an MP1 Fluorescence System (Photon Technology International, Edison, NJ). Ultraviolet and visible light (UV/vis) absorbances were measured on a SpectraMax® 340PC spectrophotometer (Molecular Devices, Sunnyvale, CA). Peptides were purified on a preparative high-performance liquid chromatography (HPLC) system (Gilson Inc., Middleton, WI) equipped with a model 322 HPLC pump, GX-271 liquid handler, 156 UV/vis detector and a 10 mL sample loop. A Phenomenex C18 column (21.2 × 250 mm,
2.2. Peptide synthesis See Supplementary material. 2.3. Bacterial growth conditions All cultures were grown from glycerol stocks, including E. coli DH5␣, E. coli ATCC 25922, Salmonella enterica ATCC 13311, Campylobacter jejuni NCTC 11168, K. pneumoniae ATCC 13883, A. baumannii ATCC 19606, Pseudomonas aeruginosa ATCC 27853, Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212 and Listeria monocytogenes ATCC 15313. All organisms (excluding C. jejuni) were grown in Mueller–Hinton (MH) broth at 37 ◦ C at 225 rpm. C. jejuni was grown on MH agar plates at 37 ◦ C under 10% CO2 , 5% O2 and 85% N2 .
2.5. Synergy assay A modified broth dilution assay in a 96-well plate (Corning Inc., Corning, NJ) was used to test for synergy between H-TriA1 and a panel of 10 antibiotics, where row (A) is a sterility control, (B) is antibiotic alone, (C) is antibiotic in 12.5 g/mL H-TriA1 , (D) is antibiotic in 6.25 g/mL H-TriA1 , (E) is antibiotic in 3.13 g/mL H-TriA1 , (F) is antibiotic in 1.56 g/mL H-TriA1 , (G) is 12.5 g/mL H-TriA1 growth control and (H) is a growth control. MH broth (50 L) was added to all wells of a 96-well plate. A solution of the desired antibiotic (4× desired starting concentration) in MH broth (50 L) was added to B1, C1, D1, E1 and F1, and serial dilutions were made across these rows. MH broth (50 L) was added to rows A, B and H. A 25, 12.5, 6.25 and 3.13 g/mL solution (50 L) of H-TriA1 in MH broth was added to rows C, D, E and F, respectively. A 25 g/mL solution of H-TriA1 in MH broth (50 L) was added to row G. Rows B–H were inoculated with 5 L of bacterial suspension (see CLSI guidelines) so that the final concentration in each well was 5 × 105 CFU/mL.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
The MIC was measured as the lowest concentration of antibiotic required to prevent visible growth. 2.6. Cytotoxicity assay with HEK 293 cells A frozen dimethyl sulphoxide (DMSO) stock of HEK 293 cells was thawed and diluted with 10 mL of HyCloneTM DMEM/High Glucose (ATCC) + 10% foetal bovine serum (ATCC) medium. Cells were centrifuged and re-suspended in medium (10 mL) in a 75-mL culture flask and were grown at 37 ◦ C for 48 h in a 5% CO2 incubator. The medium was removed, cells were washed with phosphate-buffered saline (PBS) solution (10 mL) and fresh medium was added (10 mL). After 48 h, the medium was removed and cells were washed with PBS, followed by trypsinisation by incubating with 0.0625% trypsin in medium (3 mL) for 5 min. The trypsinised suspension was added to medium (8 mL) and centrifuged. The cell pellet was re-suspended in medium (1 mL) and 100 L was used to start a new culture or for a 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide (MTT) assay. The re-suspended cell pellet (100 L) was diluted with medium (10 mL). Cells were counted using a haemocytometer and the concentration was adjusted to 2.5 × 105 cells/mL. The MTT assay was performed in a 96-well plate, where row (A) is a sterility control, (B–D) are peptide and (E) is a growth control. The suspension of HEK 293 cells (100 L) was added to rows B–E, and medium (100 L) was added to row A. The cells were incubated at 37 ◦ C for 24 h. The medium was replaced in all wells. A 16 mg/mL peptide solution (100 L) was added to B1, C1 and D1 and serial dilutions were performed across each row. The plate was incubated for 48 h at 37 ◦ C. A 2.5 mg/mL solution of thiazolyl blue tetrazolium bromide in PBS (20 L) was added to all wells and the plate was incubated for 3 h. The medium was carefully removed from all wells and was replaced with DMSO (100 L). After 30 min, the absorbance at 570 nm was measured in each well. A graph of the absorbance versus concentration was plotted and the MIC was determined as the lowest concentration before absorbance increased from background values. 2.7. Outer-membrane permeability assay The ability of peptides to permeate the outer membrane of E. coli ATCC 25922 cells was determined using the 8-anilinonaphthalene1-sulphonic acid (ANS) uptake assay [23]. A fully-grown overnight culture in MH broth was used to inoculate fresh broth to an optical density at 600 nm (OD600 ) of 0.05. Cells were grown to mid-log phase (OD600 ∼ 0.3–0.8) and were pelleted by centrifugation. Cells were washed with buffer [10 mM Tris (pH 7.4), 150 mM NaCl] and were pelleted by centrifugation. Buffer was added to obtain at least 20 mL of a suspension of cells with an OD600 of 0.065. A constant concentration of 5 M ANS in bacterial buffer suspension (2 mL) was used for all experiments and the concentration of TriA1 and H-TriA1 was varied. The peptide was added and the fluorescence emission was measured after 10 min between 450 nm and 600 nm with excitation at 380 nm. The increase in fluorescence from blank readings was used to assess the penetration of the outer membrane. 2.8. Inner membrane permeability The ability of peptides to penetrate the inner membrane of E. coli ML-35 cells was determined using an o-nitrophenyl--dgalactopyranoside (ONPG) assay [24]. An overnight culture in MH broth was grown from a glycerol stock and 100 L was added to fresh broth (10 mL). Cells were grown for 3 h with shaking at 37 ◦ C to mid-log phase (ca. 2 × 1011 CFU/mL) and the suspension was cooled to 0 ◦ C. A 96-well plate was used for the ONPG assay, where row (A) is a blank, (B) is a positive control (melittin) and (C) and (D) are peptide. ONPG buffer [70 L; 10 mM NaH2 PO4 (pH 7.4), 100 mM NaCl
495
and 1.5 mM ONPG] and cell suspension (20 L) were added to the first six wells in rows A–D. A 64 g/mL solution of melittin (10 L) was added to row B. Next, 10 L of 40×, 20×, 10×, 5×, 2.5× and 1.25× MIC peptide solutions was added to C1, C2, C3, C4, C5 and C6, respectively. The experiment was run in duplicate using D1–D6. A multichannel pipette was used to add the melittin and peptide solutions as close in time as possible. After 15 min, the absorbance at 405 nm of each well was measured. The percentage of leakage was calculated for each peptide concentration taking the positive control as 100%. 2.9. Large unilamellar vesicle (LUV) preparation 1-Palmitoyl-2-oleoylphosphatidylcholine (POPC) and 1palmitoyl-2-oleoylphosphatidylglycerol (POPG) were purchased as 25 mg/mL chloroform solutions from Avanti Polar Lipids (Alabaster, AL). The mini-extruder and membranes were also purchased from Avanti Polar Lipids. POPC (228 L, 7.5 mol) and POPG (77 L, 2.5 mol) were mixed and dried under an argon stream and were placed under high vacuum for 2 h. The film was hydrated with phosphate buffer [1 mL, 5 mM KH2 PO4 (pH 7.0)] and was vortexed for 30 s. The mixture was frozen using a dry ice/acetone bath, thawed with warm water and vortexed for 30 s. The freeze–thaw–vortex cycle was repeated twice. Passing the milky suspension through a 50-nm polycarbonate membrane 10 times using a mini-extruder formed 50-nm LUVs. The phosphate concentration of the LUV solution was determined using the Stewart method [25]. 2.10. Circular dichroism spectroscopy Samples were added to a 0.4-mL quartz cuvette with a 0.1 cm path length, measured from 190–250 nm at 20 ◦ C and averaged over five scans. The peptide concentration remained constant at 20 M for all experiments. The CD spectrum of phosphate buffer was subtracted from all spectra. Ten lipid equivalents (by phosphate concentration) were mixed thoroughly with the peptide solution before measuring. A digital filter of 15 was applied to CD spectra, which were converted to molar ellipticity units. 3. Results and discussion 3.1. Preliminary synergistic testing of three tridecaptin A1 derivatives Our previous structure–activity relationship studies on Nterminal derivatives of TriA1 identified three analogues with significantly lower antimicrobial activities than the natural peptide [18]. H-TriA1 is not acylated at the N-terminus with a lipid tail, PEG-tridecaptin A1 (PEG-TriA1 ) is acylated with a triethylene glycol derivative, and biotin-tridecaptin A1 (Bio-TriA1 ) is acylated with biotin (Fig. 1). The antimicrobial activities of these analogues were determined against a panel of Gram-negative bacteria (Table 1). PEG-TriA1 was inactive at all concentrations tested (>100 g/mL), H-TriA1 required 100–200 g/mL to inhibit the growth of bacteria, and Bio-TriA1 was slightly more active with a MIC of 50 g/mL against most bacteria tested. These analogues were screened for synergistic activity with rifampicin against E. coli ATCC 25922 at sub-MIC concentrations of 12.5, 6.25, 3.13 and 1.56 g/mL (Table 2). Enterobacteriaceae, Acinetobacter and Pseudomonas spp. are intrinsically resistant to rifampicin, but disruption of the outer membrane increases their susceptibility to this antibiotic. PEGTriA1 showed no synergistic activity, whilst Bio-TriA1 caused a small decrease in the MIC of rifampicin (8×). H-TriA1 displayed excellent synergy with rifampicin, lowering the MIC by 512× at 12.5 g/mL and by 64× at 6.25 g/mL. H-TriA1 was therefore used
496
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Table 1 Minimum inhibitory concentrations (MICs) of tridecaptin A1 (TriA1 ) analogues against Gram-negative bacteria. MIC (g/mL)
Organism
TriA1 Escherichia coli ATCC 25922 E. coli DH5␣ Salmonella enterica ATCC 13311 Campylobacter jejuni NCTC 11168 Pseudomonas aeruginosa ATCC 27853 Klebsiella pneumoniae ATCC 13883 Acinetobacter baumannii ATCC 19606
H-TriA1
3.13 3.13 6.25 3.13 50 3.13 12.5
100 100 100 100 >100 100 200
Bio-TriA1
PEG-TriA1
50 50 50 50 >100 50 >100
>100 >100 >100 >100 >100 >100 >100
H-TriA1 , unacylated analogue of tridecaptin A1 ; Bio-TriA1 , biotin-tridecaptin A1 ; PEG-TriA1 , PEG-tridecaptin A1 .
for subsequent synergistic studies. It is interesting to note that capping the N-terminus significantly lowers the ability of TriA1 analogues to sensitise Gram-negative bacteria. This may be due to the decreased electrostatic attraction between the positively charged peptide and the anionic bacterial outer membrane. 3.2. Cytotoxicity and plasma stability studies Our previous work has shown that H-TriA1 has low haemolytic activity, causing only 0.5% haemolysis at a concentration of 83 g/mL [18]. To further assess the toxicity profile of H-TriA1 , toxicity against HEK 293 cells was determined. The natural peptide inhibits all growth of HEK 293 cells at a concentration of 100 g/mL, whereas H-TriA1 is significantly less toxic, requiring 4000 g/mL for the same effect. The cytotoxic concentration of H-TriA1 against HEK 293 cells is 320× larger than the highest concentration used in the synergy assay (12.5 g/mL) and suggests that H-TriA1 could be part of a clinically safe antibiotic combination. The stability of H-TriA1 in human plasma was also assessed. Incubation of the peptide in human plasma at 37 ◦ C for 30 min showed minimal degradation of the peptide (<1%; see Supplementary material). Furthermore, the MIC of natural TriA1 against E. coli ATCC 25922 was determined in the presence of 4% bovine serum albumin and was found to be unaffected by protein binding. Therefore, it is unlikely that the mechanism by which H-TriA1 exerts its synergistic effect will be affected by protein binding. In addition to its favourable toxicity profile and plasma stability, H-TriA1 can also be synthesised easily and efficiently by standard peptide synthesis protocols. In contrast to the cyclic polymyxins, no orthogonal protection of 2,4-diaminobutyric acid residues or complex cyclisation steps are required to synthesise the tridecaptins. 3.3. Synergy of H-TriA1 with other antibiotics The synergistic activity of H-TriA1 with ten other antibacterial agents was assessed against a panel of Gram-positive and Gramnegative bacteria (Table 3). The lantibiotics nisin and gallidermin sequester lipid II and cause pore formation in Gram-positive bacteria [26] but cannot cross the outer membrane of Gram-negative bacteria. Vancomycin also binds to lipid II and inhibits peptidoglycan formation [27] but is similarly blocked by the outer membrane.
Daptomycin exerts its bactericidal effect against Gram-positive bacteria through membrane disruption [28]. The other antibiotics tested (penicillin G, ampicillin, ciprofloxacin, tetracycline and streptomycin) have activity both against Gram-positive and Gramnegative bacteria. However, resistance mechanisms are common for these antibiotics. Natural TriA1 has weak activity against Gram-positive strains as well as the Gram-negative Pseudomonas spp. It is therefore not surprising that no notable synergistic activity with antimicrobial agents was observed when P. aeruginosa, S. aureus, E. faecalis and L. monocytogenes were treated with H-TriA1 . However, synergistic effects with certain antimicrobial agents were observed for the other Gram-negative strains tested. The largest synergistic effects were observed with rifampicin, showing a decrease in MIC of 512× against E. coli strains and K. pneumoniae ATCC 13883, and 128× against C. jejuni. Moderate decreases of 32× and 16× were found against S. enterica and A. baumannii, respectively. Strong synergistic effects were also found between vancomycin and H-TriA1 , with a large MIC decrease of 256× against K. pneumoniae ATCC 13883, 32–64× against E. coli strains and 16× against the other Gramnegative strains. Of particular note is the large decrease in MIC observed for vancomycin against K. pneumoniae, which normally has a high tolerance to this antibiotic (MIC = 100 g/mL). This combination could therefore have potential for the treatment of CRKP infections, against which Oct-TriA1 is active [18]. Surprisingly, moderate decreases in the MIC (2–16×) of the lantibiotics gallidermin and nisin were found against the E. coli strains, S. enterica, C. jejuni and K. pneumoniae, which was unexpected given the large size of lantibiotics (Mw > 2 kDa). Although lantibiotics are not yet commonly used in the clinical setting, they are excellent food preservatives [29], where the use of antimicrobial peptides is often preferred over conventional antibiotics. Given the synergy observed against these common foodborne pathogens and the fact that H-TriA1 is also a peptide, there may be an application for HTriA1 as a food preservative. No synergistic effects were observed with daptomycin, which was not active at the highest concentration tested (500 g/mL). The synergistic effects of the remaining antimicrobial agents (penicillin G, ampicillin, ciprofloxacin, tetracycline and streptomycin) ranged from small to moderate. Ciprofloxacin and ampicillin had similar MIC decreases (8–16×) to vancomycin and rifampicin against A.
Table 2 Synergy between rifampicin and peptides against Escherichia coli ATCC 25922. Analogue
PEG-TriA1 Bio-TriA1 H-TriA1
Rifampicin MICa at peptide concentrationsa of 12.5
6.25
3.13
1.56
0
1.56 0.4 0.012
1.56 0.78 0.1
3.13 1.56 0.78
3.13 3.13 0.78
3.13 3.13 6.25
MMD
FICI
2× 8× 512×
0.514 0.132 0.002
MIC, minimum inhibitory concentration; MMD, maximum MIC decrease; FICI, fractional inhibitory concentration index [FICI = (MICpeptide+antibiotic )/(MICpeptide ) + (MICpeptide+antibiotic )/(MICantibiotic )]; PEG-TriA1 , PEG-tridecaptin A1 ; Bio-TriA1 , biotin-tridecaptin A1 ; H-TriA1 , unacylated analogue of tridecaptin A1 . a All values reported in g/mL.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
497
Table 3 Synergy between H-tridecaptin A1 (H-TriA1 ), an unacylated analogue of tridecaptin A1 , and antimicrobial agents. MICa at H-TriA1 concentrationsa of
MMD
FICI
25 100 >500 50 6.25 31.3 0.39 3.13 0.003 6.25
8× 16× N/D 64× 512× 4× 2× 2× 0 2×
0.157 0.125 N/D 0.023 0.002 0.327 0.515 0.514 1.000 0.532
6.25 6.25 >500 6.25 0.39 12.5 0.1 0.1 0.003 0.2
12.5 50 >500 50 6.25 12.5 0.1 0.1 0.003 1.56
4× 16× N/D 32× 512× 2× 0 0 0 8×
0.282 0.094 N/D 0.047 0.002 0.563 1.000 1.000 1.000 0.130
25 50 >500 12.5 0.78 3.13 0.39 6.25 0.003 0.78
25 50 >500 50 1.56 3.13 0.39 6.25 0.003 1.56
50 100 >500 100 3.13 12.5 0.39 6.25 0.006 3.13
4× 4× N/D 16× 32× 8× 2× 2× 2× 4×
0.375 0.500 N/D 0.125 0.033 0.140 0.515 0.532 0.500 0.257
6.25 6.25 >500 12.5 1.56 6.25 0.024 1.56 0.024 6.25
6.25 6.25 >500 50 12.5 12.5 0.024 1.56 0.024 6.25
12.5 6.25 >500 100 12.5 12.5 0.024 1.56 0.024 12.5
12.5 12.5 >500 200 12.5 12.5 0.024 1.56 0.024 12.5
2× 4× N/D 16× 128× 2× 0 0 4× 2×
0.563 0.281 N/D 0.188 0.009 0.563 1.000 1.016 0.250 0.563
Klebsiella pneumoniae ATCC 13883 0.78 Nisin 1.56 Gallidermin >500 Daptomycin Vancomycin 0.39 0.006 Rifampicin 3.13 Penicillin G 0.78 Tetracycline 0.1 Streptomycin 0.002 Ciprofloxacin 50 Ampicillin
3.13 3.13 >500 1.56 0.024 25 0.78 0.20 0.003 100
3.13 3.13 >500 3.13 0.024 50 0.78 0.39 0.003 100
6.25 6.25 >500 6.25 0.03 50 0.78 0.39 0.003 250
12.5 50 >500 100 3.13 200 1.56 0.39 0.012 250
16× 32× N/D 256× 512× 64× 2× 4× 4× 5×
0.070 0.047 N/D 0.008 0.002 0.047 0.508 0.257 0.167 0.700
K. pneumoniae ATCC 700603 Vancomycin Rifampicin
31.3 0.2
62.5 0.4
62.5 3.13
250 12.5
8× 256×
0.438 0.005
3.13 12.5 >500 12.5 0.20 125 0.20 25 0.10 62.5
6.25 12.5 >500 25 0.39 125 0.20 25 0.10 125
6.25 12.5 >500 50 0.39 250 0.20 50 0.20 250
6.25 12.5 >500 50 0.39 500 0.20 50 0.20 500
2× 0 N/D 16× 16× 16× 0 4× 8× 16×
0.516 1.063 N/D 0.078 0.062 0.375 1.001 0.313 0.120 0.219
Bacteria/antibiotic
12.5
6.25
3.13
1.56
0
Escherichia coli ATCC 25922 Nisin Gallidermin Daptomycin Vancomycin Rifampicin Penicillin G Tetracycline Streptomycin Ciprofloxacin Ampicillin
3.13 6.25 >500 0.78 0.012 7.8 0.2 1.56 0.003 3.13
3.13 6.25 >500 6.25 0.1 7.8 0.2 1.56 0.003 3.13
6.25 6.25 >500 25 0.78 7.8 0.2 1.56 0.003 3.13
6.25 12.5 >500 50 0.78 15.6 0.39 1.56 0.003 3.13
E. coli DH5␣ Nisin Gallidermin Daptomycin Vancomycin Rifampicin Penicillin G Tetracycline Streptomycin Ciprofloxacin Ampicillin
3.13 3.13 500 1.56 0.012 6.25 0.1 0.1 0.003 0.2
6.25 6.25 >500 6.25 0.2 6.25 0.1 0.1 0.003 0.2
6.25 6.25 >500 6.25 0.39 12.5 0.1 0.1 0.003 0.2
Salmonella enterica ATCC 13311 12.5 Nisin Gallidermin 25 >500 Daptomycin 6.25 Vancomycin 0.1 Rifampicin 1.56 Penicillin G 0.2 Tetracycline 3.13 Streptomycin Ciprofloxacin 0.003 0.78 Ampicillin
12.5 50 >500 25 0.39 1.56 0.2 3.13 0.003 0.78
Campylobacter jejuni NCTC 11168 Nisin 6.25 Gallidermin 3.13 Daptomycin >500 Vancomycin 12.5 0.1 Rifampicin Penicillin G 6.25 Tetracycline 0.024 Streptomycin 1.56 0.006 Ciprofloxacin 6.25 Ampicillin
31.3 0.05
Acinetobacter baumannii ATCC 19606 Nisin 3.13 12.5 Gallidermin >500 Daptomycin 3.13 Vancomycin 0.024 Rifampicin 31.25 Penicillin G 0.20 Tetracycline 12.5 Streptomycin 0.024 Ciprofloxacin 31.25 Ampicillin
498
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
Table 3 (Continued) Bacteria/antibiotic
MICa at H-TriA1 concentrationsa of 12.5
6.25
3.13
1.56
0
A. baumannii ATCC BAA-1605 Vancomycin 250 0.05 Rifampicin
250 0.2
250 0.2
250 0.2
250 0.78
MMD
FICI
0 16×
1.625 0.064
MIC, minimum inhibitory concentration; MMD, maximum MIC decrease; FICI, fractional inhibitory concentration index [FICI = (MICpeptide+antibiotic )/(MICpeptide ) + (MICpeptide+antibiotic )/(MICantibiotic )]; N/D, not determined. No synergy was observed against Pseudomonas aeruginosa ATCC 27853, Staphylococcus aureus ATCC 29213, Enterococcus faecalis ATCC 29212 or Listeria monocytogenes ATCC 15313. a All values reported in g/mL.
baumannii, which appears to be less susceptible to the synergistic action of H-TriA1 than the other strains. The synergistic combinations of rifampicin/H-TriA1 and vancomycin/H-TriA1 were also screened against more virulent strains of K. pneumoniae and A. baumannii. K. pneumoniae ATCC 700603 produces -lactamase SHV-18 and is resistant to several -lactams, cephalosporins and tetracycline. A. baumannii ATCC BAA-1605 is resistant to several -lactams, cephalosporins, gentamicin and ciprofloxacin. A moderate MIC decrease of 8× was observed for vancomycin against the MDR Klebsiella strain, however excellent synergistic effects were observed with rifampicin, improving the MIC by 256×. No synergistic effect was observed against A. baumannii ATCC BAA-1605 with vancomycin and a 16× improvement was found with rifampicin. 3.4. Rationale for synergistic effect In an attempt to explain the synergistic effects of H-TriA1 on Gram-negative bacteria, a series of mechanistic studies were performed. An ANS uptake assay can be used to monitor the ability of antimicrobial peptides to disrupt the outer membrane. Fluorescence of the hydrophobic fluorophore ANS increases when it relocates from bulk solution to the membrane interior upon membrane disruption [23]. An ANS uptake assay using E. coli ATCC 25922 cells revealed that although natural TriA1 caused an increase in ANS fluorescence at MIC concentrations and higher, H-TriA1 had no effect on fluorescence (Fig. 2). This may explain the difference
Fig. 2. Increase in fluorescence observed when Escherichia coli ATCC 25922 cells in 8-anilinonaphthalene-1-sulfonic acid (ANS) buffer are exposed to increasing concentrations of tridecaptin A1 (TriA1 ) and to H-TriA1 (an unacylated analogue of TriA1 ): (1) 8× MIC TriA1 ; (2) 4× MIC TriA1 ; (3) 2× MIC TriA1 ; (4) 1× MIC TriA1 ; (5) H-TriA1 (25 g/mL); (6) no peptide. No effect is observed with H-TriA1 . MIC, minimum inhibitory concentration.
in antimicrobial activities and cytotoxicity between the natural and unacylated peptides, however it is surprising that H-TriA1 allows passage of large antibiotics such as vancomycin and rifampicin through the outer membrane but not a small hydrophobic fluorophore. Like the polymyxins, the N-terminal lipid tail is likely required for permeation of the outer membrane and would explain why H-TriA1 may not permeate the outer membrane. The ability of H-TriA1 to disrupt the inner membrane of E. coli ML-35 cells was assessed using an ONPG assay. Disruption of the inner membrane releases a -galactosidase that can cleave the o-nitrophenyl group from ONPG in the buffer, whose absorbance is monitored at 405 nm. No -galactosidase activity was detected in cells treated with H-TriA1 (60 g/mL), whereas natural TriA1 caused 95%, 69% and 25% leakage at 4×, 2× and 1× MIC, respectively. These results suggest that the synergistic effect of H-TriA1 with other antibiotics is not due to partial membrane lysis. Previous studies have shown that Oct-TriA1 is unstructured in water but adopts a stable secondary structure in the presence of negatively charged phospholipid LUVs that mimic the bacterial membrane [30]. Replacement of key amino acids with alanine resulted in the loss of this secondary structure and a concomitant loss of antimicrobial activity. The CD spectrum of HTriA1 in the presence of 10 equivalents (by phosphate content) of 3:1 POPC:POPG LUVs was obtained (Fig. 3). The CD spectrum of H-TriA1 in the model membrane environment is very similar to Oct-TriA1 , suggesting that H-TriA1 adopts a similar secondary structure on the outer membrane of Gram-negative bacteria, which may account for the observed synergistic activity. The magnitude of the negative band at 196 nm can be correlated with antimicrobial activity and would also explain the significantly lower activity of H-TriA1 relative to the natural peptide.
Fig. 3. Circular dichroism spectra of octyl-tridecaptin A1 (Oct-TriA1 ) and HTriA1 (an unacylated analogue of tridecaptin A1 ) in the presence of 10 equivalents (by phosphate content) of 3:1 POPC:POPG large unilamellar vesicles: (1) H-TriA1 ; (2) Oct-TriA1 . = wavelength (nm), [] = molar ellipticity (cm2 /dmol). POPC, 1-palmitoyl-2-oleoylphosphatidylcholine; POPG, 1-palmitoyl-2oleoylphosphatidylglycerol.
S.A. Cochrane, J.C. Vederas / International Journal of Antimicrobial Agents 44 (2014) 493–499
4. Conclusions An unacylated analogue of the cationic lipopeptide TriA1 , termed H-TriA1 , was found to be an excellent sensitiser of Gramnegative bacteria to certain antimicrobial agents normally reserved for the treatment of Gram-positive bacterial infections. In particular, H-TriA1 increased the potency of the hydrophobic antibiotic rifampicin 512× and vancomycin 256× against one of the K. pneumoniae strains that is normally resistant to these compounds. H-TriA1 has an excellent toxicity profile, only showing a cytotoxic effect against human embryonic kidney cells at 320× the highest synergistic concentration. Mechanistic studies revealed that HTriA1 does not exert its synergistic effect through membrane lysis, but may sensitise Gram-negative bacteria by forming a stable secondary structure on the outer membrane. With the continued rise in the number of XMR strains of Gram-negative bacteria, finding new antibiotics or methods to increase the effectiveness of existing antibiotics is vital. Development of resistance to lipopeptide-based antimicrobial agents is often limited, therefore H-TriA1 could be an ideal candidate for further biological studies. Funding: Natural Sciences and Engineering Research Council of Canada (NSERC), Griffith Laboratories Canada, and the Canada Research Chair in Bioorganic and Medicinal Chemistry. Competing interests: None declared. Ethical approval: Not required. Acknowledgments The authors are grateful to Shaun McKinnie for help with plasma stability assays, to Dr Brandon Findlay for his advice on antimicrobial testing, to Jing Zheng and Dr Randy Whittal for assistance with mass spectrometry, to Mark Miskolzie for assistance with NMR, to Bernadette Beadle for assistance with Campylobacter testing, and to Wayne Moffat for assistance with CD experiments.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijantimicag. 2014.08.008. References [1] Paphitou NI. Antimicrobial resistance: action to combat the rising microbial challenges. Int J Antimicrob Agents 2013;42:S25–8. [2] Butler MS, Blaskovich MA, Cooper MA. Antibiotics in the clinical pipeline in 2013. J Antibiot (Tokyo) 2013;66:571–91. [3] Roemer T, Boone C. Systems-level antimicrobial drug and drug synergy discovery. Nat Chem Biol 2013;9:222–31. [4] Tzouvelekis LS, Markogiannakis A, Psichogiou M, Tassios PT, Daikos GL. Carbapenemases in Klebsiella pneumoniae and other Enterobacteriaceae: an evolving crisis of global dimensions. Clin Microbiol Rev 2012;25:682–707. [5] Lee GC, Burgess DS. Treatment of Klebsiella pneumoniae carbapenemase (KPC) infections: a review of published case series and case reports. Ann Clin Microbiol Antimicrob 2012;11:32.
499
[6] Gasink LB, Edelstein PH, Lautenbach E, Synnestvedt M, Fishman NO. Risk factors and clinical impact of Klebsiella pneumoniae carbapenemase-producing K. pneumoniae. Infect Control Hosp Epidemiol 2009;30:1180–5. [7] Karah N, Sundsfjord A, Towner K, Samuelsen O. Insights into the global molecular epidemiology of carbapenem non-susceptible clones of Acinetobacter baumannii. Drug Resist Updat 2012;15:237–47. [8] Cantón R, Akóva M, Carmeli Y, Giske CG, Glupczynski Y, Gniadkowski M, et al. European Network on Carbapenemases. Rapid evolution and spread of carbapenemases among Enterobacteriaceae in Europe. Clin Microbiol Infect 2012;18:413–31. [9] Vaara M, Fox J, Loidl G, Siikanen O, Apajalahti J, Hansen F, et al. Novel polymyxin derivatives carrying only three positive charges are effective antibacterial agents. Antimicrob Agents Chemother 2008;52:3229–36. [10] Vingsbo Lundberg C, Vaara T, Frimødt-Moller N, Vaara M. Novel polymyxin derivatives are effective in treating experimental Escherichia coli peritoneal infection in mice. J Antimicrob Chemother 2010;65:981–5. [11] Mingeot-Leclercq MP, Tulkens PM, Denamur S, Vaara T, Vaara M. Novel polymyxin derivatives are less cytotoxic than polymyxin B to renal proximal tubular cells. Peptides 2012;35:248–52. [12] Vaara M, Sader HS, Rhomberg PR, Jones RN, Vaara T. Antimicrobial activity of the novel polymyxin derivative NAB739 tested against Gram-negative pathogens. J Antimicrob Chemother 2013;68:636–9. [13] Vaara M, Vaara T. The novel polymyxin derivative NAB739 is remarkably less cytotoxic than polymyxin B and colistin to human kidney proximal tubular cells. Int J Antimicrob Agents 2013;41:292–3. [14] Pirri G, Giuliani A, Nicoletto SF, Pizzuto L, Rinaldi AC. Lipopeptides as antiinfectives: a practical perspective. Cent Eur J Biol 2009;4:258–73. [15] Cochrane SA, Vederas JC. Lipopeptides from Bacillus and Paenibacillus spp.: a gold mine of antibiotic candidates. Med Res Rev 2014 (Epub ahead of print). doi: 10.1002/med.21321. [16] Kato T, Hinoo H, Shoji J. The structure of tridecaptin A (studies on antibiotics from the genus Bacillus. XXIV). J Antibiot (Tokyo) 1978;31:652–61. [17] Lohans CT, van Belkum MJ, Cochrane SA, Huang Z, Sit CS, McMullen LM, et al. Biochemical, structural, and genetic characterization of tridecaptin A1 , an antagonist of Campylobacter jejuni. ChemBioChem 2014;15:243–9. [18] Cochrane SA, Lohans CT, Brandelli JR, Mulvey G, Armstrong GD, Vederas JC. Synthesis and structure–activity relationship studies of N-terminal analogues of the antimicrobial peptide tridecaptin A1 . J Med Chem 2014;57: 1127–31. [19] Vaara M, Vaara T. Sensitization of Gram-negative bacteria to antibiotics and complement by a nontoxic oligopeptide. Nature 1983;303:526–8. [20] Viljanen P, Vaara M. Susceptibility of Gram-negative bacteria to polymyxin B nonapeptide. Antimicrob Agents Chemother 1984;25:701–5. [21] Vaara M, Siikanen O, Apajalahti J, Fox J, Frimodt-Møller N, He H, et al. A novel polymyxin derivative that lacks the fatty acid tail and carries only three positive charges has strong synergism with agents excluded by the intact outer membrane. Antimicrob Agents Chemother 2010;54:3341–6. [22] Wiegand I, Hilpert K, Hancock RE. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nat Protoc 2008;3:163–75. [23] Thennarasu S, Lee DK, Tan A, Prasad Kari U, Ramamoorthy A. Antimicrobial activity and membrane selective interactions of a synthetic lipopeptide MSI843. Biochim Biophys Acta 2005;1711:49–58. [24] Pate M, Blazyk J. Methods for assessing the structure and function of cationic antimicrobial peptides. Methods Mol Med 2008;142:155–73. [25] Stewart JC. Colorimetric determination of phospholipids with ammonium ferrothiocyanate. Anal Biochem 1980;104:10–4. [26] Bonelli RR, Schneider T, Sahl H-G, Wiedemann I. Insights into in vitro activities of lantibiotics from gallidermin and epidermin mode-of-action studies. Antimicrob Agents Chemother 2006;50:1449–57. [27] Watanakkunakorn C. Mode of action and in-vitro activity of vancomycin. J Antimicrob Chemother 1984;14:7–18. [28] Capepari P, Boaretti M, Lleo MM, Satta G. Lipoteichoic acid as a new target for activity of antibiotics: mode of action of daptomycin (LY146032). Antimicrob Agents Chemother 1990;34:1220–6. [29] Hansen JN. Nisin as a model food preservative. Crit Rev Food Sci Nutr 1994;34:69–93. [30] Cochrane SA, Findlay B, Vederas JC, Ratemi ES. Key residues in octyl-tridecaptin A1 analogues linked to stable secondary structure in the membrane. ChemBioChem 2014;15:1295–9.