Addition of L-cysteine to the N- or C-terminus of the all-D-enantiomer [_D(KLAKLAK)₂] increases antimicrobial activities against multidrug-resistant Pseudomonas aeruginosa, Acinetobacter baumannii and Escherichia coli

Peptide design and synthesis

Peptides [(D-Lys-D-Leu-D-Ala-D-Lys-D-Leu-D-Ala-D-Lys)₂] (DP) (McGrath et al., 2013), DP to which L-cysteine was added at the N-terminus, [L-Cys-(D-Lys-D-Leu-D-Ala-D-Lys-D-Leu-D-Ala-D-Lys)₂] (C-DP), and DP to which L-cysteine was added at the C-terminus [(D-Lys-D-Leu-D-Ala-D-Lys-D-Leu-D-Ala-D-Lys)₂-L-Cys] (DP-C), were synthesized with >95% purity (Scrum Inc., Tokyo, Japan). DP-C was dimerized by heating at 60 °C for 30 min (DP-C dimer) to convert cysteine to cystine. The purity of the synthetic peptides was checked on a SunFire C18 column (100 Å, 5 µm, 4.6 mm inner diameter × 250 mm; Waters, Milford, MA, USA). Each peptides were gradiently eluted with solution A (water containing 0.1% trifluoroacetic acid) and solution B (acetonitrile containing 0.1% trifluoroacetic acid) at a flow rate of 1.0 mL/min. The elution program for DP and C-DP was as follows: at 0 min, 10% of B; at 20 min, 60% of B. The elution program for DP-C was as follows: at 0 min, 0% of B; at 20 min, 100% of B. The separated components were detected at 220 nm. The DP-C dimer was separated on a COSMOSIL 5C18-AR-300 reversed phase column (4.6 mm inner diameter × 250 mm; Nacalai Tesque, Kyoto, Japan), using an automated HPLC system (LC-2010AHT; Shimadzu, Kyoto, Japan). The reaction products were gradiently eluted with solution A (water containing 0.086% trifluoroacetic acid) and solution B (acetonitrile containing 0.086% trifluoroacetic acid) at a flow rate of 1.0 mL/min. The elution program for DP-C dimer was as follows: at 0 min, 20% of B; at 20 min, 50% of B. The peptide masses were determined by MALDI-TOF MS on a microflex (Bruker, Billerica, MA, USA).

Bacterial strains

MDR P. aeruginosa NCGM2.S1 (Miyoshi-Akiyama et al., 2011) and drug-susceptible P. aeruginosa PAO1 (CLSI, 2018), P. aeruginosa OCR1 (Poole et al., 1996), and P. aeruginosa PAO4290 (Yoneyama et al., 1997) were grown in Luria Bertani (LB) broth (BD Japan, Tokyo, Japan) or on LB plates containing 15 g/L agar at 37 °C. Drug-susceptible E. coli ATCC 25922, a clinical isolate of E.coli NCCHD1261-5 (Uchida et al., 2018), and S. aureus ATCC 25923 were grown at 37 °C in tryptic soy broth (BD). Drug-susceptible A. baumannii ATCC 15308, a clinical isolate of MDR A. baumannii IOMTU433 (Tada et al., 2015) (GenBank accession no. AP014649), a clinical isolate of MDR A. baumannii NCGM237 (Tada et al., 2015) (GenBank accession no. AP013357), a clinical isolate of MDR A. baumannii NCGM253 (Tada et al., 2015) (GenBank accession no. AB823544), K. pneumoniae ATCC-BAA-2146, K. pneumoniae ATCC15380 (Reading & Cole, 1977), and Serratia marcescens NBRC102204^T were grown at 37 °C in Difco™ Nutrient broth (BD).

Drug susceptibility testing

The minimum inhibitory concentrations (MICs) of peptides and antibiotics, including meropenem, amikacin, ofloxacin, and colistin, were determined by the broth microdilution method according to Clinical Laboratory Standards Institute (CLSI) guidelines (CLSI, 2018). Bacterial strains were inoculated at 5  × 10⁵ CFU/mL per well into 96-well round-bottom microtiter plates (Watson Bio Lab, Kobe, Japan) containing an equal volume of serially diluted peptides or antibiotics. Three independent experiments were performed to confirm reproducibility.

The synergistic effects of the peptides and the antibiotics amikacin, colistin, meropenem, ofloxacin, and rifampicin against P. aeruginosa NCGM2.S1 were evaluated using the checkerboard dilution method. The peptides were two-fold serially diluted to final concentrations ranging from 0.125- to 2-times the MIC longitudinally in 96-well round-bottom microtiter plates. Subsequently, antibiotic was two-fold serially diluted to final concentrations ranging from 0.125- to 2-times the MIC transversely into the plates. NCGM2.S1 was inoculated at 5  × 10⁵ CFU/mL per well at a volume equal to that of the diluted peptide and antibiotic. Three independent experiments were performed to confirm reproducibility. The synergistic effect of the peptides and antibiotics was assessed by determining the fractional inhibitory concentration (FIC) index (Berenbaum, 1978), using the formula: ${\displaystyle \mathrm{FIC}=\frac{\text{MIC of peptide}\in \text{combination}}{\text{MIC of peptide alone}}+\frac{\text{MIC of antimicrobial agent}\in \text{combination}}{\text{MIC of antimicrobial agent alone}}.}$

An FIC index ≤0.5 was defined as synergistic, an FIC index >0.5 to 4.0 was defined as additive or unrelated, and an FIC index >4.0 was defined as antagonistic.

Cytotoxicity tests

The human hepatoblastoma cell line, HepG2 (ATCC HB-8065), was obtained from American Type Culture Collection and cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum (FBS). HepG2 cells were seeded at 3,000 cells/well in 96-well cell culture-treated flat bottom microtiter plates (Falcon, Corning, NY, USA). The cells were incubated at 37 °C for 48 h in an atmosphere containing 5% CO₂, followed by the addition of peptides at a final concentration of 0–256 µg/mL, or colistin, at a final concentration of 0–3,000 µg/mL, by serial dilution. The plates were incubated with 0.2% FBS in 5% CO₂ at 37 °C for 48 h; under these conditions, HepG2 cells were alive, but they did not grow. Cell viability was determined using a Cell Counting Kit-8 (Dojin, Tokyo, Japan), and colorimetric changes were determined at OD₄₅₀ with a microplate reader (Corona Electric Co. Ltd., Ibaraki, Japan). LD₅₀ was defined as the concentration of peptides or colistin that resulted in 50% cell viability. Three independent experiments were performed to confirm reproducibility.

Statistical analysis

The Mann–Whitney U test was used to compare the MIC values of C-DP, DP-C, and DP-C dimer with DP in P. aeruginosa and A. baumannii. P-values less than 0.05 were considered statistically significant. Cell survival was expressed as a percentage of the control were obtained as mean  ± standard deviation of three independent experiments done in three replicates for each treatment. Significant differences of cell survival rate between each concentration and the control were statistically evaluated by Student’s t-test.

Addition of L-cysteine to the N- or C- terminus enhanced the antimicrobial activity of the original peptide

HPLC analysis indicated that the purities of synthetic DP, C-DP, and DP-C were 100, 95.52, and 98.68%, respectively (Figs. S1–S3). Additionally, the masses of DP, C-DP, and DP-C by MALDI-TOF MS analysis, were 1525.444, 1626.595, and 1627.151, respectively that matched well with the theoretical molecular weights (1524.0, 1627.1, and 1627.1) (Figs. S1–S3). Similarly, the formation of DP-C dimer was assessed by HPLC and MALDI-TOF MS. The HPLC analysis of heat treated DP-C showed one large peak estimated as DP-C dimer and one small peak estimated as DP-C monomer the peak area ratio was 6.3:1 (Fig. S4). The MALDI-TOF MS result of heat treated DP-C was 3252.7, which was consistent with the estimated molecular weight of DP-C dimer (Fig. S4). The grand average hydropathy (GRAVY) values were −0.07 for DP and 0.1 each for C-DP and DP-C, indicating that the addition of L-cysteine affected the hydrophobicity of DP.

Assessment of antibiotic susceptibility showed that drug-susceptible P. aeruginosa PAO-1 (CLSI, 2018) and PAO4290 expressing a normal level of MexAB-OprM (Yoneyama et al., 1997), were susceptible to all antibiotics tested; MDR P. aeruginosa NCGM2.S1 (Miyoshi-Akiyama et al., 2011) was susceptible to colistin, but resistant to amikacin, meropenem and ofloxacin; and OCR1, a nalB MDR mutant that overproduces the outer membrane protein OprM (Poole et al., 1996) was susceptible to amikacin and colistin, intermediately susceptible to ofloxacin, but resistant to meropenem (Table 1). DP showed antimicrobial activity against PAO-1, with an MIC of 300 µg/mL, consistent with previous findings (McGrath et al., 2013). DP also had antimicrobial activity against NCGM2.S1 and PAO4290, with MICs of 300 µg/mL, but DP showed no antimicrobial activity against OCR1 within the tested concentration range. C-DP had greater antimicrobial activities than DP against all of these strains (The Mann–Whitney U test, p < 0.05), with MICs of 64–128 µg/mL, and DP-C had greater antimicrobial activities than C-DP and DP (The Mann–Whitney U test, p < 0.05), with MICs 16–32 µg/mL (Table 1). The DP-C dimer also had antimicrobial activities and showed MICs identical to DP-C against all the strains tested. These results indicate that the addition of L-cysteine to the N- or C-terminus of [_D(KLAKLAK)₂] increased its antimicrobial activity.

DP showed antimicrobial activity against four A. baumannii strains, with MICs of 64 to 300 µg/mL, and against a clinical isolate of colistin- and carbapenem-resistant E. coli NCCHD1261-5 co-harboring mcr-1 and bla_NDM-5 genes (Uchida et al., 2018), with an MIC of 64 µg/mL. In contrast, DP was inactive against drug-susceptible E. coli ATCC 25922, two K. pneumoniae strains, S marcescens NBRC102204 and S. aureus ATCC 25923 (Table 2). C-DP and DP-C showed higher antimicrobial activities than DP (The Mann–Whitney U test, p < 0.05), with MICs of 4–8 µg/mL against the four A. baumannii strains and MICs of 4 to 16 µg/mL against the two E. coli strains. C-DP and DP-C showed antimicrobial activity against carbapenem-resistant K. pneumonia e ATCC BAA-2146 harboring bla_NDM-1, with both having MICs of 16 µg/mL, but not against penicillin-resistant, β-lactamase-producing K. pneumonia e ATCC 15380, S. marcescens NBRC102204 and S. aureus ATCC 25923 (Table 2).

Synergistic effects of peptides and antibiotics

The combinations of DP, C-DP, and DP-C with colistin had synergistic effects on antimicrobial activity (Table 3). For example, the growth of P. aeruginosa NCGM2.S1 was inhibited by a combination of one-sixteenth the MIC of DP (19 µg/mL), and one-fourth the MIC of colistin (0.25 µg/mL), by a combination of one-fourth the MIC of C-DP (32 µg/mL) and one-eighth the MIC of colistin (0.125 µg/mL), and by a combination of one-eighth the MIC of DP-C (4.0 µg/mL) and one-fourth the MIC of colistin (0.25 µg/mL). Although the AMPs that induced susceptibility to rifampicin were reported in clinical MDR isolates of P. aeruginosa (Baker et al., 2019), DP-C only slightly enhanced the susceptibility to rifampicin. The combination of C-DP with meropenem or ofloxacin had additive effect on antimicrobial activity. Synergistic effects were not observed when DP-C was combined with amikacin.

Cytotoxicity of antimicrobial peptides to HepG2 cells

The LD₅₀ values of each peptide in HepG2 cells were >256, >256, 192, and 2100 µg/mL for DP, C-DP, DP-C, and colistin, respectively (Figs. S5 and S6). The raw data of Figs. S5 and S6 are available in Files S2 and S3, respectively. Since the antimicrobial activity of these peptides against P. aeruginosa were synergistic with colistin, we examined whether a combination of these peptides and colistin was more toxic than the peptide alone. The combination of DP-C and colistin dose-dependently induced the death of HepG2 cells (Fig. 1), with more than 50% of the cells dying at 256 µg/ml DP-C and 25.6 µg/mL colistin. The cytotoxicity of the peptide was not enhanced by combination of low doses of colistin. There was no cytotoxicity to HepG2 at 4 µg/mL DP-C and 0.4 µg/mL colistin, which concentrations showed synergistic effects of colistin and DP-C. The raw data of Fig. 1 are available in File S2.