Rationale and development of a rapid, quantitative antimicrobial flow cytometry method

We reasoned that as AMPs disrupt the cytoplasmic membrane we should be able to monitor this disruption using membrane permeable/impermeable fluorescent intercalating DNA dyes that are used to determine the viability of a microbial culture. We initially synthesised melittin (peptide sequence shown in S1 Table) a well characterised, broad-spectrum and potent AMP to investigate whether viability dyes, Syto9 (membrane permeable) and propidium iodide (PI, membrane impermeable), could be used to determine AMP activity, against the melittin-susceptible bacterium Fusobacterium nucleatum. Initially, MIC and MBC (using standard CLSI methodology) of melittin for F. nucleatum were determined by harvesting late exponentially growing cells (to ensure ≥95% viable bacteria) and incubating (2.5 × 10⁶ cells/mL) with serial dilutions of melittin for 90 minutes, followed by bacterial growth assays (Fig 1a). Table 1 shows that F. nucleatum is highly melittin-sensitive with low MIC and MBC activities observed. Using the average MIC/MBC value of 4.0 μM we incubated F. nucleatum with melittin at 0.5× and 10× the MIC/MBC and then added varying concentrations of Syto 9/PI to evaluate whether peptide concentrations resulted in variant and observable membrane disrupted/non-disrupted cell populations. Staining cells with 0.1% v/v Syto 9 in 0.9% v/v saline (3.34mM stock solution) and 0.1% v/v PI in saline (50μg/mL stock solution) provided optimal concentrations that gave consistent and distinct cell populations; a membrane intact (Syto9+/PI-) and two membrane disrupted populations (Syto9+/PI+ and Syto9-/PI+; S1 Fig). At the optimal Syto 9/PI concentrations, incubating F. nucleatum with 2 μM and 40 μM melittin resulted in 48% and >97.5% of the bacteria being PI+ (i.e. Syto9+/PI+ and Syto9-/PI+), respectively (S1 Fig). Following from this we then incubated F. nucleatum with variant melittin concentrations and determined the proportion of membrane intact (Syto9+/PI-) and membrane disrupted (PI+ i.e. Syto9+/PI+ and Syto9-/PI+) populations. Fig 1b shows that increasing melittin concentration resulted in a corresponding increase in membrane disrupted cells. At melittin concentrations of 3.13 and 6.25 μM and higher; 87% and >98% of the bacteria were PI+, respectively, indicating that the melittin concentration that induces 98%+ membrane disrupted bacteria is between 3.13–6.25 μM (Fig 1c). Linear regression of the PI+ values for the peptide concentrations prior to full membrane disruption (% PI+ cells ≤ 99% and ≥ 1–5%) gave an R² value of >0.9 and from extrapolation the minimum peptide concentration that induced 100% membrane disruption (MDC) could be determined (Fig 1c, insert). For melittin the MDC was 3.3±0.3 μM, which was in agreement with the melittin MIC and MBC values (Table 1). Using another lytic AMP, magainin II, the membrane disruption of F. nucleatum displayed a similar positive correlation with AMP concentration. The MDC obtained was 9.5±2.4 μM which was again in agreement with the MIC and MBC for that AMP (S2 Fig and S2 Table). Melittin and magainin II are known to have high and low activities against the Gram positive bacterium Streptococcus mutans, respectively. Table 1 and S2 Table, shows that the MDC assay discriminated between the two peptides where melittin was found to be 25 fold more potent than magainin II against S. mutans and that the MDC was comparable with the MIC and MBC of these peptides. Although the standard procedure is a 90 minute incubation of bacteria and AMP, we investigated whether the incubation time could be shortened. Table 1 shows that shortening the incubation time affected the MDC for F. nucleatum but not for S. mutans when incubated with melittin, indicating that the AMP/bacteria incubation could be shortened but this is AMP/species dependant.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Increasing concentration of the antimicrobial peptide Melittin increases bacterial membrane permeability to propidium iodide.

(a) Schematic representation of the work flow for determining MIC, MBC and MDC. (b) Flow cytometry dot blots of F. nucleatum incubated with increasing concentrations of Melittin and stained with Syto9 dye (membrane permeable) and propidium iodide dye (membrane impermeable). (c) Percent propidium iodide positive F. nucleatum cells correlates with increasing peptide concentration. Insert shows the reciprocal plot of percent membrane disrupted cells (PI+) and peptide concentration to determine MDC.

https://doi.org/10.1371/journal.pone.0151694.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. MIC and MBC of melittin for F. nucleatum and S. mutans and comparison with the membrane disrupting concentration (MDC) determined at different peptide/bacteria incubation times.

https://doi.org/10.1371/journal.pone.0151694.t001

Applicability of the flow cytometry MDC assay

To validate the flow cytometry MDC assay further we synthesised three antimicrobial peptides with defined modes of antibacterial killing [11]; ovispirin (carpet mode), magainin II (toroidal pore mode) and alamethicin (barrel-stave mode) along with a potent lytic/pore forming AMP; caerin 1.1 [25] (S1 Table). The MIC, MBC and MDC of these peptides were then determined against each of the ESKAPE pathogens, MRSA and E. coli. Table 2 shows that each of the AMPs had different antimicrobial activities against each of the eight species tested. In comparing the activity of any one AMP, the flow cytometry MDC value was found to be similar/comparable to the MBC value and in the majority also had similar MIC values. Where MDC or MBC differed from MIC, the MIC was always significantly lower than the MDC or MBC reflecting that growth can be inhibited at a lower concentration than that required for cell lysis. The MIC value was observed to be significantly lower than the MDC or MBC values for; ovispirin with E. aerogenes and K. pnuemoniae; magainin II with P. aeruginosa and E. coli; alamethicin with A. baumannii and caerin 1.1 with S. aureus (MSSA and MRSA). In other experiments we found that S. mutans and F. nucleatum were susceptible to alamethicin and ovispirin and that the MDC value for each AMP was consistent with the MBC and MIC values (S2 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Comparison of the antimicrobial activity of three defined pore forming peptides and a membrane lytic peptide determined by microdilution growth assay (MIC), colony count assay (MBC) and the flow cytometry assay (MDC) for Gram positive and Gram negative bacteria.

https://doi.org/10.1371/journal.pone.0151694.t002

A potential advantage of the flow cytometry assay is the speed at which AMP analogues can be screened in a semi or fully quantitative assay. To investigate this we synthesised five magainin II analogues with two critical properties for AMP activity altered (cationicity and α-helicity). Cationic properties were altered by substituting all of the lysine residues (residues 4, 10 11, 14) with either; arginine (Arg), ornithine (Orn), diaminobutyric acid (Dab) or diaminopropionic acid (Dap), and α-helicity increased by replacing residues 8 (Ser), 13 and 18 (Gly) with alanine. The magainin II sensitive bacteria, S. mutans and F. nucleatum were then screened in the flow cytometry assay with each of the analogues at a peptide concentration equivalent to MDC/MBC50 which equates to a 50:50 membrane intact:disrupted cell population (Fig 2). For F. nucleatum the Orn, Dab and Arg analogues did not alter the intact:disrupted ratio of 50:50, however the Dap analogue induced less disruption, while the Ala analogue increased disruption. For S. mutans the Orn, Dab and Dap analogues induced less disruption, while the Arg and Ala analogues increased disruption. To validate this semi-quantitative screening assay we then determined the MIC, MBC and MDC for each analogue (Table 3). The MIC, MBC and MDC values for each analogue reflected the screening assay, in that where membrane disruption was found to be similar, increased or decreased, the MIC, MBC and MDC values were either similar, lower or higher respectively when compared with that of native Magainin II (Table 3). Although the Arg and Ala analogues induced greater cell lysis compared with native magainin II they also induced greater hemolysis of red blood cells and had a lower therapeutic index (S3 Table). Interestingly for F. nucleatum the Dab analogue induced less hemolysis resulting in a higher therapeutic index compared with native magainin II (S3 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Flow cytometry dot plots of F. nucleatum and S. mutans incubated with magainin analogues at a concentration equivalent to the MIC50 concentration of magainin.

Cells stained with Syto9 dye (membrane permeable) and propidium iodide dye (membrane impermeable).

https://doi.org/10.1371/journal.pone.0151694.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Antimicrobial activity of Magainin analogues against S. mutans, and F. nucleatum.

https://doi.org/10.1371/journal.pone.0151694.t003

Bacterial strains and growth conditions

The bacterial strains; Fusobacterium nucleatum ATCC 25586, Streptococcus mutans Ingbritt, Staphylococcus aureus (MSSA) ATCC 29213; Staphylococcus aureus (MRSA) ATCC 43300; Enterococcus faecalis ATCC 29212; Enterobacter aerogenes ATCC 13048; Pseudomonas aeruginosa strain PAO1-LAC, ATCC 47085; Klebsiella pneumoniae ATCC 13883; Acinetobacter baumannii strain 2208, ATCC 19606; Escherichia coli ATCC 25922 lyophilised or glycerol stocks were obtained from the culture collection of the Oral Health Cooperative Research Centre, The Melbourne Dental School, University of Melbourne, Australia. These strains were selected for this study as they are widely used in antimicrobial studies and are considered to be important pathogens and major targets for antibiotic resistant studies. Bacteria were grown aerobically and maintained by passage at ambient temperature (E. coli, K. pneumoniae, A. baumannii and S. aureus strains) or at 37°C (P. aeruginosa, E. faecalis and E. aerogenes) on horse blood agar (10% v/v defibrinated horse blood, 4.4% w/v Oxoid Blood Agar Base No. 2). S. mutans and F. nucleatum were grown anaerobically (MK3 anaerobic workstation, Don Whitley Scientific Limited, England) and maintained by passage at 37°C on Todd Hewitt agar (3.6% w/v Oxoid Todd-Hewitt Broth, 1.5% w/v sucrose, 1.5% w/v BactoTM Agar, 0.8% w/v Oxoid Yeast Extract) or horse blood agar, respectively. For the antimicrobial peptide assays a 20 mL starter culture was produced by taking single colonies from blood or Todd Hewitt agar plates, respectively, to inoculate either: Luria Broth (LB; 1% w/v BactoTM Tryptone, 1% w/v NaCl, 0.5% w/v Oxoid Yeast Extract, pH 7.5, Thermo Scientific Pty, Ltd, Sydney, Australia) for bacteria; E. coli, K. pneumoniae, P. aeruginosa, A. baumannii, E. faecalis, E. aerogenes and S. aureus which were grown aerobically at 37°C, or Todd Hewitt broth (3.6% w/v Oxoid Todd-Hewitt Broth, 1.5% w/v sucrose, 0.8% w/v Oxoid Yeast 9 Extract) for S. mutans, or Brain Heart Infusion broth (BHI; 3.7% w/v Oxoid BHI, 0.05% w/v L-cysteine) for F. nucleatum which were both grown anaerobically at 37°C. After overnight incubation, 0.5–2.0 mL of the starter culture was used to inoculate the appropriate fresh broth (200 mL) and growth monitored at 650 nm using a spectrophotometer (model 275E; Perkin-Elmer, Sydney, Australia) with culture purity checked by microscopic examination and culture. Bacteria were harvested during late exponential growth phase and viability and bacteria/mL determined using a BacLight viability kit (Invitrogen, Sydney, Australia) and a Quanta SC-MPL flow cytometer (Beckman coulter Pty, Ltd, Sydney, Australia).

Synthesis of Antimicrobial Peptide (AMPs)

O-Benzotriazole-N,N,N',N'-tetramethyluronium hexafluorophosphate (HBTU), 1hydroxybenzotriazole (HOBt), diisopropylethylamine (DIPEA), N,N-dimethylformamide (DMF), piperidine, trifluoroacetic acid (TFA) and 9-fluorenylmethoxycarbonyl (Fmoc) amino acids were obtained from Auspep Pty Ltd (Melbourne, Australia). Fmoc-Phenylalaninol was obtained from ChemPep Inc (FL). Triisopropylsilane (TIPS) was obtained from Aldrich (New South Wales, Australia). 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), 4-dimethylamino-pyridine (DMAP) and N,N’-diisopropyl-carbodiimide (DPCIDI) were obtained from Sigma Chemical Company (New South Wales, Australia). Diethyl ether and dichloromethane (DCM) were obtained from BDH (Poole, UK). diethylaminosulfur trifluoride was obtained from Alfa Aesar (WA). Unless otherwise stated chemicals were of peptide synthesis grade or its equivalent.

The antimicrobial peptides; caerin 1.1, melittin, ovispirin, magainin II and magainin II analogues (S2 Table) were chemically synthesized on a CEM Liberty microwave peptide synthesizer (Ai Scientific, Victoria, Australia). The peptide-resins were assembled from Fmoc-Rink-AM SURETM Resin or Fmoc-Gly-TGA resin (Merck Millipore Pty Ltd, NSW Australia) to produce the C-terminal carboxyamide or carboxylic acid AMPs, respectively, in the Fmoc/tBu mode of synthesis. For a 0.1 mmol reaction scale, Fmoc-deprotection was performed in two stages by initial treatment with 20% piperidine/0.1 M HOBt/DMF (v/v, 7 ml) under microwave radiation for 30 s (40 W, 40°C), followed by filtration and a second addition of the above solution (45 W, 75°C; 3 min). The peptide-resins were then rinsed with DMF (4 × 7 ml). Acylation, where required, was achieved by the addition of a solution containing amino acid (5 eq, relative to reaction scale), HBTU (5 eq) and DIEA (10 eq) in DMF/NMP (7:1, v/v; 4 ml) to the Nα-deprotected peptide-resin and the mixture agitated under microwave radiation for 10 min (30 W, 75°C, vessel under external chilled air flow). Dichloromethane (DCM) (5 × 2 min) was used to rinse the peptide-resins prior to the cleavage step. The peptide was cleaved from the resin support by the addition of TFA/TIPS/thioanisole/phenol/water (90:2.5:2.5:2.5:2.5, % v/v/v/v/v; 5 ml) for 2.5 h or 4 h for Arg containing AMPs, after which the cleavage filtrates were evaporated under nitrogen flow and the crude product was isolated by precipitation in cold ether (4 × 30 ml).

The crude peptides were purified using an Agilent 1200 series liquid chromatograph instrument (Agilent, NSW, Australia) equipped with a UV detector (model G1316A) and a Zorbax 300 SB-C18 reversed phase column (9.4 mm × 25 cm). Crude peptide analysis was achieved using a linear acetonitrile gradient in 0.1% TFA at a flow rate of 4 mL/min (linear gradient of 0 to 54% CH₃CN over 15 min). Analysis of the purified peptide was performed using an Esquire HCT electrospray ionization-mass spectrometry system (Bruker Daltronics, NSW, Australia), each of the purified peptides gave the expected masses (S2 Table).

For the synthesis of Alamethicin, Fmoc-Phenylalaninol-Resin was produced by suspending 2-Chlorotrityl resin (750 mg; loading 1.36 mmol/g) in a solution of DCM and DMF [6ml; 1:3 (v/v)] followed by the addition of Fmoc-Phenylalaninol (545mg, 1.46 mmol) and pyridine (231 mg, 2.92 mmol) and the mixture gently stirred for 16 h, after which time methanol (4 ml) was added, the resin stirred for a further 2 h, filtered, rinsed with DCM and dried under reduced pressure overnight. Resin substitution was determined by incubating a resin sample (39.2 mg) in 2% DBU/DMF (2 mL) for 30 min. Acetonitrile was then added to achieve a final volume of 10 mL and the amount of Fmoc in solution determined by absorbance at 490nm and the resin substitution determined using control and reference solutions, which gave a resin loading of 0.31 mmol/g. The addition of Fmoc-aminoisobutyric acid (Fmoc-Aib) into the peptide chain was accomplished using Fmoc-Aib acid fluoride (Fmoc-Aib-F), whereby, Fmoc-Aib-F was synthesised by diethylaminosulfur trifluoride (DAST, 5.3 mL; 40.0 mmol) was added to a suspension containing Fmoc-Aib-OH (7.5 g; 30.7 mmol) in anhydrous DCM (30 mL) under nitrogen. After 1 h agitation, the reaction mixture was washed with cold water (2 ×40 mL) and the DCM layer collected and dried with Mg₂SO₄. The solution was then filtered, the solvent evaporated and the resultant light yellow powder re-dissolved in DCM (20 mL) and n-hexane (20 mL) and placed at 4°C for 48 h. The recrystallized product was filtered, washed with cold n-hexane and dried under reduced pressure. Three crystallizations were performed to give a final yield of 7.3g. Fmoc-Aib-F purity was determine by converting Fmoc-Aib-F to the methyl ester (Fmoc-Aib-OMe) by dissolving the crystalline product in methanol (4.75 mL) and pyridine (0.25 mL). After 10 min, DCM (10 mL) was added, the solution washed with 10% citric acid (2 × 10 mL) and the DCM layer evaporated. The resultant white product was dissolved in 60% buffer B containing a small amount of isopropanol and analyzed by RP-HPLC and ESI-MS which showed the presence of Fmoc-Aib-OMe in > 95% purity [m/z 340.0; (M+H)+]. The synthesis of alamethicin was performed using Fmoc-Phenylalaninol-Resin (359 mg; 0.1 mmol) and a CEM Liberty Microwave Synthesizer as described above for the other AMPs, with following modifications: Fmoc-Aib-F was coupled by the addition of the amino acid fluoride in DMF (5 eq, 2.5 mL) and DIEA in NMP (9 eq, 0.5 mL) to the peptide-resin followed by microwave heating (10 min, 30 W, 75°C). The peptide-resin was then filtered and the coupling step repeated once, followed by rinsing with DMF (3 × 7 mL). Upon completion of the final Fmoc deprotection step, the peptide-resin was rinsed with DMF (3 × 7 mL), DCM (3 × 10 mL) and dried under reduced pressure. A solution containing acetic anhydride (0.26 mL) and DIEA (0.98 mL) in DMF (4.0 mL) was then added and the mixture agitated for 30 min. Following final rinsing of the peptide-resin with DCM (5 × 10 mL), the resin-bound peptide was dried under reduced pressure and cleaved by the addition of TFA/DCM/H2O/TIPS (47:47:4:2; v/v; 2 h). The solvent was evaporated and the crude peptide precipitated in cold diethyl ether. Crude peptide was dissolved in 25% Buffer B/75% Buffer A and purified as described above using a linear gradient of 50–90% Buffer B in 40 min. The desired fraction was identified by MALDI-TOF MS using an Ultraflex TOF/TOF instrument (Bruker) in negative ion and reflectron mode.

Antimicrobial assays

Antibacterial assays were undertaken to determinate the minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC) and minimum membrane disruptive concentration (MDC) of each of the antimicrobial peptides (AMP) and AMP analogues (S1 Table). For each bacterim a stock solution (2.5 × 10⁶ cells/mL) in the respective media was made and an aliquot incubated with AMPs within 15 minutes from viable cell count and stock preparation. All AMPs were dissolved in DMSO and a 500 μM stock solution prepared by adding the respective media and serial dilutions (250–0.244 μM) of the AMP in media (100 μL/well) made just prior addition of bacteria. DMSO was found to be necessary for solvating the AMPs in media and the final assay concentration of DMSO was ≤ 2.5% v/v, which did not affect bacteria viability or susceptibility to AMPs. One hundred microliter aliquots of the bacterial stock solution (2.5 × 10⁵ cells/well) were added to the AMP serial dilutions and incubated at 37°C for 90 min. Assay preparation and incubations were conducted in aerobic or anaerobic conditions dependent on the bacteria. Bacteria were also incubated in the absence of AMP to serve as a growth control for the assay. After the 90 minute incubation period the antimicrobial activity (MIC, MBC and MDC) was determined as follows;

For MIC the CLSI broth microdilution assay [16] was followed as we have previously described [33]. Briefly, after the 90 minute incubation, bacterial growth was monitored at 10 min intervals over a 12 or 24 hour period (bacteria dependent) at OD₆₅₀ using an iEMS microplate reader (Pathtech Pty Ltd, Melbourne, Australia) which incubated the cultures at 37°C. The MIC was calculated using the Lambert and Pearson growth curve analysis method [38], by plotting the relative growth at each peptide concentration compared to maximal growth (determined as the point when bacteria incubated in media alone entered stationary phase of growth, 100% growth), the MIC was determined as the lowest peptide concentration (μM) required to completely inhibit the growth of the bacteria i.e the intersection of the linear curve with the x-axis.

For determination of MBC, the CLSI protocol [16] was followed, with the modifications being that due to the rapid action of the AMPs the assay incubation time was shortened from 16–20 hours to 90 minutes. Following incubation cells were pelleted by centrifugation (7,000g, 30 mins) and resuspended in 10 assay volumes (2 mL) of appropriate media and aliquots plated on agar and incubated for 24 to 48 h at 37°C (aerobically or anaerobically, bacteria dependent), and colony forming units (CFU) quantified. The MBC was thus determined as the lowest concentration (μM) of AMP yielding ≥ 99.9% reduction in the initial colony count after incubation.

For determination of MDC a 50 μL aliquot of the bacteria/AMP was mixed with 100 μL of 0.9% w/v saline containing 0.1% v/v of SYTO^® 9 (3.34mM stock solution) and 0.1% v/v of PI (50μg/mL stock solution) and incubated for 10 minutes at room temperature in the dark. Following incubation the bacterial cell samples were analysed using a Cell Lab Quanta SC MPL flow cytometer (Beckman Coulter) equipped with a 100 W stabilized mercury arc lamp with wavelengths of 365, 404, and 435 nm, and a 488 nm diode laser. The fluorescence from SYTO^® 9 was measured through a 525-nm band-pass filter (Fluorescent Channel 1, FL-1), and the red emission of PI was measured with a 670-nm long pass filter (Fluorescent Channel 3, FL-3). A minimum of 100,000 bacterial events were collected and the multi-parametric data analyzed using the Cell Lab Quanta SC software. NB for flow cytometers not fitted with an electrical volume detector for cell counting, the addition of microbeads to the samples can be used to determine cell numbers. MDC was calculated by plotting the percent PI+ cells for the each peptide concentration and fitting a linear curve to the data that is ≤ 99% PI+ cells and ≥ 1–5% PI+ cells (minimum of 4 data points). Inclusion of data points outside of these PI+ cell percentages will result in a positive or negative skewing of the MDC value. By solving the line equation (y = mx + C) for x (peptide concentration) when y = 1 (where 1 = 100% of PI+ cells) the MDC (x) can be determined i.e. MDC = x = (y − C)/m. The MDC was thus determined as the lowest concentration (μM) of AMP yielding ≥ 99.0% PI+ cells.

Hemolysis assay

Fresh sheep red blood cells in Alsever’s solution (RBCs, Equicell, Victoria, Australia) were diluted 1 in 20 in PBS (pH 7.4), pelleted by centrifugation, and washed three times in PBS (1000 g, 10 min). The RBCs were counted using a cell counter (Coulter Particle Counter Z series, Beckman Coulter) and diluted to a final concentration of 2 × 10⁷ cells/mL. 100 μL aliquots of the RBC solution were seeded into a V-bottomed 96-well plate containing 100 μL of serial dilutions (250–0.244 μM) of the AMP in PBS and incubated in a humidified atmosphere containing 5% CO₂ at 37°C for 2 h. Following incubation, the RBCs were pelleted centrifuged (1000 g, 10 min) and the amount of haemoglobin in 100 μL aliquots determined by measuring the absorbance at 405 nm using a microplate reader (PerkinElmer 1420 Multilabel Counter VICTOR3). Positive and negative controls for hemolysis were taken as RBC lysed with 0.5% Triton X-100 (1:1 v/v) and RBC suspension in PBS, respectively. The percentage of hemolysis was calculated using the following formula:

The percentage hemolysis was plotted against peptide concentration and linear regression analysis was used to determine the hemolytic concentration needed to lyse 50% (HC50) of RBCs.

Inhibition of cell proliferation assay

To determine the inhibition of cell proliferation, 100 μl of HEK-293 (ATCC CRL-1573TM) cells (5×10³) in media (Eagle's Minimum Essential Medium (EMEM) supplemented with 10% v/v foetal bovine serum) were seeded into 96-well plates containing 100 μL serial dilutions (250–0.244 μM) of the AMP in media and cultured at 37°C, 5% CO2 incubator. After 28 h, 20 μL of 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) solution (CellTiter 96 AQueous Non-Radioactive Cell Proliferation Assay kit, Promega) were added to each well and the plates were incubated for a further 1–2 h at 37°C, 5% CO2 incubator. Cell proliferation was determined by measuring absorbance at 490 nm using a microplate reader (PerkinElmer 1420 Multilabel Counter VICTOR3). Positive and negative controls for inhibition were taken as HEK-293 cells incubated with mytomycin C (5 μg/mL, Sigma Pty Ltd) and HEK-293 cells incubated in media alone, respectively. The percentage of inhibition was calculated using the following formula:

The percentage of inhibition was plotted against peptide concentration and linear regression analysis was used to determine the AMP concentration needed to inhibit 50% (IC50) of the HEK-293 cells.