University of Groningen Breaking walls: combined peptidic activities against Gram-negative human pathogens Li, Qian

(1)

University of Groningen

Breaking walls: combined peptidic activities against Gram-negative human pathogens

Li, Qian

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.

Document Version

Publisher's PDF, also known as Version of record

Publication date: 2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Li, Q. (2019). Breaking walls: combined peptidic activities against Gram-negative human pathogens. University of Groningen.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

(2)

Chapter

5

Efficient killing

of Gram-negative pathogens

by highly synergistic action of

GNP-D8 and vancomycin or nisin:

biological and pharmaceutical

characterization

Qian Li1, Manuel Montalban-Lopez1,2, Oscar P. Kuipers1

1Department of Molecular Genetics, Groningen Biomolecular Sciences and Biotechnology Institute, University of Groningen, Groningen, the Netherlands 2 Department of Microbiology, Faculty of Sciences, University of Granada, Spain

(3)

Abstract

The intrinsic resistance of Gram-negative pathogenic bacteria is a major threat to human health. It has been shown that synergism be-tween combinations of antibiotics or antibiotics with other compounds can be used to target specific problematic pathogens. The aim of this work is to optimize and characterize the in vitro synergism of GNPs (designed peptides with outer-membrane penetrating capability) with either vancomycin or nisin. Several variants of GNP-6 and GNP-8 were designed, synthesized and tested against selected Gram-negative patho-gens either alone or combined with vancomycin or nisin. The combi-nations of D-amino acid isomers of either GNP-D8 and vancomycin, or GNP-D6 and vancomycin, displayed strong in vitro activity against a panel of relevant Gram-negative clinical isolates. GNP-D8 alone and combined with vancomycin was non-toxic against HEK-293 (human embryonic kidney) cells at the measured concentrations. Moreover, no hemolytic effect towards human red blood cells (hRBCs) at the measured concentrations up to at least 500 µM could be detected. The IC50 (half maximal inhibitory concentration) of GNP-D6 alone against HEK-293 was 33 µM and 50 % haemolysis of GNP-D6 alone against hRBCs was 170 µM. We conclude that vancomycin, GNP-D6 and GNP-D8 are safe to use at their MIC concentrations, as far as tested with respect to hemolysis and HEK cell viability. Overall, GNP-D8 and vancomycin exerted a highly synergistic action against Gram-negative bacteria and provides a very promising combination for further clinical characterization towards Phase I clinical trials.

Key words:

Gram-negative bacteria, pathogens, outer-membrane-penetrating pep-tides, vancomycin, nisin, synergism, toxicity, hemolysis

(4)

CH APTER 5: I nt ro duc tio n

5 Introduction

The World Health Organization (WHO) has published a report in 2017 providing a global priority list of antibiotic-resistant bacteria to guide research, discovery and development of antibiotics [1]. 12 bacterial species have been listed, the top 3 making up the category “criti-cal”, while 6 were classified as “high” and the other 3 were defined as “ medium” priority. Remarkably, 9 of these 12 “superbugs” are Gram-negative pathogens, while the 3 “critical” bacterial species are the Gram-negative pathogens Acinetobacter baumannii (carbapenem- resistant), Pseudomonas aeruginosa (carbapenem-resistant) and

En-terobacteriaceae (carbapenem-resistant, 3rd generation cephalosporin-

resistant).

The emergency and increasing spread of antibiotic resistance in Gram-negative pathogens is a major global medical challenge and highlights the urgent need for new therapeutic strategies. Bacteria can be intrinsically resistant to certain antibiotics, which can be explained by specific structural and functional characteristics [2, 3]. The outer- membrane (OM) of Gram-negative bacteria is composed of glycolip-ids, principally lipopolysaccharide (LPS) [4]. The well-packed sugar chains of LPS on the cell face create an extremely ordered network and form a well-assembled and low fluidity surface, which is directly responsible for the low permeability of the OM for some compounds [5, 6]. The protective OM of Gram-negative bacteria forms an efficient barrier to prevent several antibiotics (vancomycin, teicoplanin, nisin, gallidermin, epidermin, mersacidin and other lantibiotics) from reach-ing their targets at the cytoplasmic membrane and/or the cytoplasm, which complicates treatments towards (multidrug-resistance (MDR)) Gram-negative pathogens [7, 8]. Thus, the main bottleneck for these antibiotics to be active against Gram-negative bacteria relies on their ability to pass the OM.

Notably, it has been reported that in the presence of chelating agents (EDTA, citrate monohydrate, or trisodium orthophosphate), nisin can inhibit the growth of Gram-negative bacteria more efficiently [9–11]. However, these chelating agents are not applicable for clinical use. Some synergistic effects between antibiotics and membrane penetrating agents against Gram-negative pathogens have been reported [12, 13]

(5)

CH APTER 5: Efficien t k illin g o f G ra m-n ega tiv e p at hog en s b y hig hl y sy ner gi stic ac tio n o f GNP -D8 a nd va nco m ycin o r ni sin: b io log ic al a nd p ha rm aceu tic al c ha rac ter iza tio n

5

and this strategy was proven to be efficient as well as attractive in clinics (e.g. to reduce the toxicity of colistin).

Our previous work described in Chapter 4 showed that GNP-6 and GNP-8 are promising for further development and characterization as helper molecules against Gram negatives. In order to improve the activ-ities of GNP-6 and GNP-8, either alone or combined with vancomycin or nisin, amino acids substitutions, replacement of L- by D-amino acids and reversed sequences were employed. The synthesized GNPs with either L-amino acids or D-amino acids are listed in Table 1. Their efficacy when combined with either nisin or vancomycin was studied. Furthermore their safety with respect to cytotoxicity and hemolytic activity was assessed.

1. Materials and methods

1.1. Bacterial strains, cell-line and growth conditions

The bacteria used in this study are listed in Table 2. All the bacteria used were obtained from the Belgian Co-ordinated Collections of Microorganisms (BCCM) or isolated by Fidelta Ltd. in Croatia.

Table 1. List of GNPs used in this work.

Name Sequences Name in literature Reference

GNP-6 RRLFRRILRWL-NH2 RW-BP100 [14]

GNP-8 RIVQRIKKWLR-NH2 This work

GNP8–1 RIVQRIKKWL-NH2 This work

GNP8–2 KIVQRIKKWLR-NH2 This work

GNP8–3 AIVQRIKKWLR-NH2 This work

GNP8–4 RIRKRIKKWLR-NH2 This work

GNP8–5 RIKRRIKKWLR-NH2 This work

GNP8–6 RIVQRIKKWR-NH2 This work

GNP-D6 rrlfrrilrwl-NH2 This work

D6-rev lwrlirrflrr-NH2 This work

GNP-D8 rivqrikkwlr-NH2 This work

D8-rev rlwkkirqvir-NH2 This work

Note: Capital letters are used to represent L-amino acids-containing peptides; lower case letters are used to represent all-D-amino acids-containing peptides.

(6)

5

CH APTER 5: M at er ia ls a nd m et ho ds

L. lactis MG1363 was cultured in M17 broth supplemented with

0.5 % (w/v) glucose (GM17) or GM17 agar at 30 °C. Escherichia coli,

Klebsiella pneumoniae, Pseudomonas aeruginosa, Acinetobacter bau-mannii, and Enterobacter aerogenes were grown in Luria-Bertani (LB)

broth shaken (200 rpm) or on LB agar at 37 °C. All the strains were used to test the minimum inhibitory concentration (MIC) of nisin, vancomycin, synthesized peptides (GNPs) and combinations thereof.

The human embryonic kidney cell line HEK-293 was from the Amer-ican Type Culture Collection (ATCC).

Table 2. Strains used in this work.

Strains Charateristics Purpose References

Escherichia coli LMG15862 beta lactamase Indicator strain Lab collection (BCCM)

Klebsiella pneumoniae

LMG20218 beta lactamase Indicator strain Lab collection(BCCM)

Pseudomonas aeruginosa LMG

6395 Indicator strain Lab collection(BCCM)

Acinetobacter baumannii LMG

Enterobacter aerogenes LMG

Escherichia coli ATCC

25922 Indicator strain Lab collection(Fidelta Ltd )

Escherichia coli ATCC

BAA-2452 Clinical isolate,multidrug-resistant (MDR) Indicator strain Lab collection(Fidelta Ltd )

Escherichia coli

B1927 Clinical isolate, MDR Indicator strain Lab collection(Fidelta Ltd )

ATCC 700603 Indicator strain Lab collection(Fidelta Ltd )

ATCC BAA-2524 Clinical isolate, MDR Indicator strain Lab collection(Fidelta Ltd )

Pseudomonas aeruginosa

ATCC 27853 Indicator strain Lab collection(Fidelta Ltd )

ATCC BAA-2108 Clinical isolate, MDR Indicator strain Lab collection(Fidelta Ltd )

Acinetobacter baumannii ATCC

17978 Indicator strain Lab collection(Fidelta Ltd )

Acinetobacter baumannii ATCC

BAA-1605 Clinical isolate, MDR Indicator strain Lab collection(Fidelta Ltd )

Acinetobacter baumannii B2026 Clinical isolate, MDR Indicator strain Lab collection (Fidelta Ltd.)

Lactococcus lactis

(7)

5

1.2. Materials and chemicals

The purification and quantification of nisin was achieved with an Agilent 1260 series HPLC as described previously [16]. Vancomy-cin, MgCl2, lipopolysaccharides (LPS) from Escherichia coli O55:B5, Mueller- Hinton broth (MHB) and fetal bovine serum (FBS) were purchased from Sigma-Aldrich (Oakville, Ontario, Canada). Sodium phosphate tablets, Dulbecco’s modified Eagle’s medium (DMEM), non-essential amino acids (NEAA) were purchased from Thermo Fisher Scientific (Waltham, Massachusetts, America). CellTiter-glo reagent was purchased from Promega (Leiden, the Netherlands). Human plasma EDTA K2 Mixed-Gender was purchased from Sera Laboratories International, Ltd (West Sussex, United Kingdom). Syn-thesized peptides (purity > 99 %) were supplied by Pepscan (Lelystad, the Netherlands).

1.3. Minimum inhibitory concentration (MIC) and synergism test

MIC tests were performed in triplicate by liquid growth inhibition mi-crodilution assays in sterile polypropylene microtiter plates according to Wiegand et al. [17]. The lowest concentration of the antimicrobials that inhibits visible growth of the indicator is identified as the MIC. In some experiments, 21 mM MgCl2 and 1.0 mg/mL LPS purified from

E. coli O55:B5 were added to MHB to check their inhibitory effects.

We conducted standard checkerboard broth microdilution assays to test the synergistic effect of combined antimicrobials [13, 18]. Exoge-nous LPS purified from E. coli O55:B5 and 21 mM Mg2+_{were applied}

in the synergy inhibition experiments. Human plasma (10 % or 30 %) was applied both in the MIC tests and synergy inhibition experiments, when indicated.

The fractional inhibitory concentration (FIC) indices [19] were cal-culated, using the formula FICI = MICac/MICa + MICbc/MICb, to determine whether the combination is additive, synergistic or antag-onistic [13, 19]. The MICa and MICb is the MIC of compound A or B alone, respectively. MICac is the MIC of compound A in combination with compound B and MICbc is the MIC of compound B when it was combined with compound A. The FIC corresponds to the MIC of a compound in combination with the other compound divided by the

(8)

5

MIC of the compound alone. FICa is FIC of compound A while FICb is FIC of compound B. The FICI is interpreted as follows: synergistic, FICI ≤ 0.5; additive, 0.5 < FICI ≤ 1; indifferent, 1 < FICI < 2; antagonistic, FICI > 2.

1.4. Effect of antimicrobial agents on bacterial growth

The bacteria were incubated in 96-wells microplates at 37 °C for 16~20 h in the presence of the antimicrobial, GNPs alone or their combination at varying final concentrations. The bacterial cells were used at a standard final concentration of 1 × 105 CFU/mL. Fresh MHB broth with or with-out a bacterial inoculum was used in parallel in the same conditions as growth and sterility controls, respectively. OD600 was measured every 2 hours during overnight incubation [20]. At intervals, 10 µl aliquots were taken from the wells and 10-fold serially diluted in PBS. 50 µL of the suspensions were immediately plated on LB agar plates. Bacterial colonies were counted after overnight incubation at 37 °C.

1.5. Hemolytic activity and cytotoxicity

Hemolytic activity was determined with fresh human red blood cells (hRBCs) [12]. The hRBCs were centrifuged (4000 rpm, 5 min) and the supernatant was discarded. The pellets were washed three times with PBS and diluted in PBS to make the cell stock solution with a cell density of 2 × 108 cells/mL. The hRBC suspensions were incubated with van-comycin, GNP-D6, GNP-D8 and the combinations thereof at various concentrations, as indicated. 1 % Triton X-100 and PBS were used as controls. The cells were incubated at 37 °C for 1 hour and centrifuged at 1000 g for 5 min. The supernatant was transferred to a new 96-well plate and hemolysis was monitored by measuring the absorbance at 450 nm. Hemolysis levels were calculated as percentage. % hemolyt-ic = 100 × (As-A0)/(At-A0), where As, A0 and At are the absorbance of the hRBCs suspension in PBS with antimicrobial agents (s), without antimicrobial agents (0), and with 1 % Triton X-100 (t).

Human embryonic kidney cell line HEK-293 was used to assess the cell viability as reported [21]. 96-well plates were seeded with HEK-293 cells at concentration of 3 × 105 cells per well in 100 µL of DMEM growth media supplemented with 1 % non-essential amino acids (NEAA) and 10 % fetal bovine serum (FBS). Border wells were

CH APTER 5: M at er ia ls a nd m et ho ds

(9)

5

filled with 100 µL of sterile PBS. The compounds were added to the cells the next day at various concentrations. Compounds were tested in duplicate. ATP levels were measured by adding 50 µL of CellTiter-Glo reagent to each well and after 5 minutes of incubation luminescence was measured with a SpectraMax i3 microplate reader. The potential effect of the tested compounds on cell viability was determined by com-paring the signal obtained in the presence of different concentrations of the compounds with those obtained in the control wells.

2. Results

2.1. MICs and synergism of GNP-8 and L-analogs with vancomycin or nisin

As previous results showed, GNP-8 alone was not very efficient and structural propensity against Gram negative pathogens and the com-bined concentrations required to inhibit the Gram-negative pathogens were still above 1 µM in some cases. Amino acid substitutions can en-hance the net cationic charge, hydrophobicity and molecular flexibility and therefore change the activity of peptides [13, 22]. Thus, aiming at further improving the synergistic activity, a set of GNP-8 mutants was designed, synthesized (L-analogs) and tested (Table 1).

The C-terminal arginine of GNP-8 was deleted to decrease the net charge of the peptide and became GNP8-1. GNP8-2 kept the net charge of the peptide but used lysine to replace arginine as the N-terminal residue. Alanine was used to replace arginine-1 to increase the hy-drophobicity of the peptide in GNP8-3. Valine and glutamine were replaced by arginine and lysine to increase the hydrophilicity and positive charge in GNP8-4 and GNP8-5. The order of valine and ar-ginine conversed in GNP8-4 and GNP8-5. The alteration of GNP8-6 with respect to GNP-8 was the deletion of leucine between tryptophan and arginine in the penultimate position. GNP8-6 had the same net charge as GNP-8 but one hydrophobic amino acid less. The positively charged amino acids lysine and arginine at the C-terminal half were also closer than in GNP-8 (Table 1).

However, the analysis of GNP-8 analogs failed to identify any sin-gle or double amino acid alteration that could lead to improved

(10)

5

CH

APTER 5: R

es

ul

ts

Table 3. Combined activity of GNP-8 and GNP-8 analogs and vancomycin or nisin against 5 Gram-negative pathogens.

Pathogens Antibiotic GNP MICa

(µM) MICb (µM) MICac(µM) MICbc (µM) FICI

E. coli LMG15862 Vancomycin GNP-8 64 12 4 1.5 0.188 GNP8-1 64 32 8 4 0.25 GNP8-2 64 16 4 2 0.188 GNP8-3 64 12 4 3 0.313 GNP8-4 64 16 4 2 0.188 GNP8-5 64 24 4 3 0.188 GNP8-6 64 256 8 16 0.188 Nisin GNP-8 12 12 1.5 1.5 0.25 GNP8-1 12 32 3 2 0.313 GNP8-2 12 16 1.5 2 0.25 GNP8-3 12 12 1.5 1.5 0.25 GNP8-4 12 16 3 2 0.375 GNP8-5 12 24 3 3 0.375 GNP8-6 12 256 6 64 0.75 K. pneu-moniae LMG20218 Vancomycin GNP-8 128 32 4 2 0.094 GNP8-1 128 64 16 16 0.375 GNP8-2 128 32 16 8 0.375 GNP8-3 128 24 16 3 0.25 GNP8-4 128 >256 ND ND ND GNP8-5 128 >256 ND ND ND GNP8-6 128 >256 ND ND ND Nisin GNP-8 48 32 0.75 1 0.047 GNP8-1 48 64 3 1 0.078 GNP8-2 48 32 0.75 2 0.078 GNP8-3 48 24 1.5 3 0.156 GNP8-4 48 >256 ND ND ND GNP8-5 48 >256 ND ND ND GNP8-6 48 >256 ND ND ND P. aeruginosa LMG 6395 Vancomycin GNP-8 128 16 16 2 0.25 GNP8-1 128 16 16 2 0.25 GNP8-2 128 16 16 2 0.25 GNP8-3 128 16 4 8 0.531 GNP8-4 128 16 4 8 0.531 GNP8-5 128 32 16 8 0.5 GNP8-6 128 >256 ND ND ND Nisin GNP-8 36 16 4.5 1 0.188 GNP8-1 36 16 4.5 4 0.375 GNP8-2 36 16 4.5 4 0.375 GNP8-3 36 16 4.5 8 0.625 GNP8-4 36 16 2.25 4 0.313 GNP8-5 36 32 9 8 0.5 GNP8-6 36 >256 ND ND ND A. baumannii LMG01041 Vancomycin GNP-8 32 >64 2 1 <0.078 GNP8-1 32 128 1 16 0.156 GNP8-2 32 128 4 4 0.156 GNP8-3 32 128 4 8 0.188 GNP8-4 32 >256 ND ND ND GNP8-5 32 >256 ND ND ND GNP8-6 32 >256 ND ND ND Nisin GNP-8 6 >64 0.19 4 <0.094 GNP8-1 6 128 0.375 16 0.188 GNP8-2 6 128 0.375 16 0.188 GNP8-3 6 128 0.375 32 0.313 GNP8-4 6 >256 ND ND ND GNP8-5 6 >256 ND ND ND GNP8-6 6 >256 ND ND ND

(11)

5

antimicrobial activity alone (Supplementary Table 1). The MICs of GNP-8 mutants were quite different from GNP-8. The MICs of GNP8–4, GNP8–5 and GNP8–6 against K. pneumoniae, A. baumannii and E. aerogenes were all more than 256 µM. This might be caused by the change of the inter-amidine distance. These results emphasize the importance of “VQ” and “WLR” in the sequence of the peptide.

Since GNP-8 performed better against the test panel than the GNP-8 mutants tested, either alone or in combination with either vancomycin or nisin (Table 3), we did choose GNP-8 as a template for the next experiments.

2.2. Antimicrobial activity and synergy of D-form GNPs

GNP-6 and GNP-8 were synthesized with all D-amino acids as well as their reversed peptides with all D-amino acids. The activities of D-peptides against Gram-negative pathogens alone are listed in Ta-ble 4. The D-peptides had a MIC value equal or smaller than their L-counterparts. GNP-D6 MIC value was equal or 2-fold higher than GNP-6 in all the strains except E. aerogenes, where the MIC decreased 4 times. The impact of D-amino acid replacement in GNP-8 was more drastic, with a 4-fold MIC reduction against E. aerogenes and more than 16-fold against A. baumannii. The reverse D-peptides were in

E. aerogenes LMG 02094 Vancomycin GNP-8 192 128 12 8 0.125 GNP8-1 192 >256 ND ND ND GNP8-2 192 256 12 16 0.125 GNP8-3 192 128 12 16 0.188 GNP8-4 192 >256 ND ND ND GNP8-5 192 >256 ND ND ND GNP8-6 192 >256 ND ND ND Nisin GNP-8 32 128 4 8 0.188 GNP8-1 32 >256 ND ND ND GNP8-2 32 256 4 1 0.128 GNP8-3 32 128 4 8 0.188 GNP8-4 32 >256 ND ND ND GNP8-5 32 >256 ND ND ND GNP8-6 32 >256 ND ND ND

Green: the lowest FICI for the specific Gram-negative pathogen and antibiotic in this table; Red: FICI < 0.1. ND: not determined.

Note: MICa is the MIC of vancomycin or nisin alone; MICb corresponds to the MIC of GNPs when used alone; MICac is the MIC of vancomycin or nisin in combination with the GNPs at the MICbc concentration. MICbc is the MIC of GNPs when used with the MICac concentration of vancomycin or nisin.

(12)

5

Table 4. MIC values (µM) of GNPs and vancomycin alone and in the presence of LPS

(1 mg/mL) or Mg2+_{(21 mM).} Vancomycin GNP-6 GNP-D6 D6-rev GNP-8 GNP-D8 D8-rev LPS Mg2+ LPS Mg2+ LPS Mg2+ E. coli LMG15862 64 128 256 2 2 >32 16 4 12 4 >64 >64 8 K. pneumoniae LMG20218 128 256 >2048 4 2 >32 32 4 32 32 64 >512 32 P. aeruginosa LMG 6395 128 256 >2048 3 2 >32 >32 4 16 16 64 >256 16 A. baumannii LMG01041 32 32 256 2 2 >32 16 4 >64 8 128 >128 64 E. aerogenes LMG 02094 192 384 3072 8 2 >32 16 8 128 32 128 512 32 L. lactis MG1363 0.125 ND ND ND 2 ND ND ND ND 16 ND ND ND ND: not determined.

Table 5. Combined activity of peptide GNP-6 or GNP-8 with either vancomycin or nisin. Pathogens Antibiotic GNP MICa

E. coli LMG15862 Vancomycin GNP-6 64 2 16 1 0.75 GNP-D6 64 2 16 0.5 0.5 D6-rev 64 4 16 1 0.5 GNP-8 64 12 4 1.5 0.188 GNP-D8 64 4 4 0.25 0.125 D8-rev 64 8 8 0.5 0.188 Nisin GNP-6 12 2 1.5 0.5 0.375 GNP-D6 12 2 3 0.5 0.5 D6-rev 12 4 1.5 1 0.375 GNP-8 12 12 1.5 1.5 0.25 GNP-D8 12 4 1.5 1 0.375 D8-rev 12 8 1.5 2 0.375 K. pneumoniae LMG20218 Vancomycin GNP-6 128 4 32 1 0.5 GNP-D6 128 2 32 0.5 0.5 D6-rev 128 4 32 1 0.5 GNP-8 128 32 4 2 0.094 GNP-D8 128 32 8 1 0.094 D8-rev 128 32 16 1 0.188 Nisin GNP-6 48 4 3 0.5 0.188 GNP-D6 48 2 6 0.5 0.375 D6-rev 48 4 6 1 0.375 GNP-8 48 32 0.75 1 0.047 GNP-D8 48 32 0.75 2 0.078 D8-rev 48 32 3 2 0.125 CH APTER 5: R es ul ts

(13)

5

P. aeruginosa LMG 6395 Vancomycin GNP-6 128 3 32 1.5 0.75 GNP-D6 128 2 64 2 1.5 D6-rev 128 4 32 2 0.75 GNP-8 128 16 16 2 0.25 GNP-D8 128 16 32 4 0.5 D8-rev 128 16 32 4 0.5 Nisin GNP-6 36 3 4.5 0.75 0.375 GNP-D6 36 2 4.5 0.25 0.25 D6-rev 36 4 4.5 1 0.375 GNP-8 36 16 4.5 1 0.188 GNP-D8 36 16 4.5 0.5 0.156 D8-rev 36 16 4.5 1 0.188 A. baumannii LMG01041 Vancomycin GNP-6 32 2 8 0.5 0.5 GNP-D6 32 2 8 1 0.75 D6-rev 32 4 8 2 0.75 GNP-8 32 >64 2 1 <0.078 GNP-D8 32 8 2 1 0.188 D8-rev 32 64 2 2 0.094 Nisin GNP-6 6 2 0.75 0.5 0.375 GNP-D6 6 2 0.38 0.5 0.313 D6-rev 6 4 0.75 1 0.375 GNP-8 6 >64 0.19 4 <0.094 GNP-D8 6 8 0.19 0.5 0.094 D8-rev 6 64 0.75 2 0.188 E. aerogenes LMG 02094 Vancomycin GNP-6 192 8 64 2 0.583 GNP-D6 192 2 96 2 1.5 D6-rev 192 8 64 2 0.583 GNP-8 192 128 12 8 0.125 GNP-D8 192 32 24 4 0.25 D8-rev 192 32 24 4 0.25 Nisin GNP-6 32 8 2 1 0.188 GNP-D6 32 2 4 0.5 0.375 D6-rev 32 8 4 2 0.375 GNP-8 32 128 4 8 0.188 GNP-D8 32 32 4 2 0.188 D8-rev 32 32 2 4 0.188

Green: the lowest FICI for the specific Gram-negative pathogen and antibiotic in this table; Red: FICI < 0.1.

Note: MICa is the MIC of vancomycin or nisin alone; MICb corresponds to the MIC of GNPs when used alone; MICac is the MIC of vancomycin or nisin in combination with the GNPs at the MICbc concentration. MICbc is the MIC of GNPs when used with the MICac concentration of vancomycin or nisin.

(14)

5

general equally or 2- to 8-fold less active than the D-peptides against the whole panel.

The combined effect of either GNP-6, GNP-D6 or D6-rev with vancomycin showed in general a weak synergy (FICI = 0.5), or an additive or indifferent effect (Table 5). With nisin, the potentiation effect was even more clear and the synergy could be measured against all the strains in all cases. GNP-8, GNP-D8 and D8-rev showed a strong synergistic effect in all the combinations with either nisin or vancomycin. In general, the combination with nisin was more effective and achieved the lowest FICI values. Nevertheless, the combinations of GNP-D8 and vancomycin or nisin showed slightly less synergism than with the L-peptides with FICI ranging from 0.078 to 0.188. However, these values still suggest a very significant

in vitro synergy. P. aeruginosa and E. aerogenes were the strains

where the combined effect of either peptide and antimicrobial was less pronounced. Since the reversed peptide showed no big difference from their counterparts, and D-amino acids confer higher resistance against proteolysis, GNP-D6 and GNP-D8 were selected for further tests. Notably, the D-GNPs exhibit an enhanced activity against Gram-negative pathogens when compared with the L-counterparts. In all cases the concentration is equal to the MIC of L-GNPs or re-duced up to 4-fold. The reversal of peptides did not largely change the activities of the peptides against Gram-negative pathogens except for A. baumannii. These data indicate that the selected GNPs do not specifically interact with a receptor.

Vancomycin, GNP-D6 and GNP-D8 were tested against several multi-drug resistant (MDR) Gram-negative pathogens as well as the corresponding standard (i.e. antibiotic sensitive) bacteria. The MIC values and synergy activity tests are listed in the Table 6 and Sup-plementary Table 2. The obtained lack of activities for vancomycin against Gram-negative MDR strains was similar to the measured MIC against selected Gram-negative pathogens and it is consistent with the mode of action of vancomycin, which is inhibition of cell wall synthesis via binding to lipid II. The MIC for GNP-D6 against these pathogens remained between 1.26 and 5.05 µM, whereas GNP-D8 was less active (MIC between 2.68 and >85.68 µM) especially against strains of K. pneumoniae (Figure 1 and Table 6). However, the activity

CH

APTER 5: R

es

ul

(15)

5

of the combination of vancomycin and GNP-D8 was strain-specific and mostly synergistic, being P. aeruginosa the least sensitive bacteri-um in this test. Vancomycin and GNP-D8 showed a great synergism against E. coli ATCC BAA-2452, K. pneumoniae ATCC BAA-2524,

K. pneumoniae B1945 and A. baumannii ATCC BAA-1605 (Table 6).

The in vitro antimicrobial activity tests of vancomycin and GNP-D8, in the presence of human plasma, either alone or in combination, were performed against three selected MDR Gram-negative pathogens

Table 6. Combined activity of GNP-D6, GNP-D8 and vancomycin against MDR Gram-neg-ative pathogens.

E. coli

ATCC 25922 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 1.26_5.36 5.52_5.52 1.26_1.6 < 1.062_0.363

E. coli

ATCC BAA-2452 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 1.26_5.36 5.52_2.76 1.26_0.80 < 1.062_{< 0.181}

E. coli

B1927 Vancomycin GNP-D6_GNP-D8 88.32_88.32 1.26_2.68 5.52_5.52 1.26_1.6 1.062_0.659

K. pneumoniae

ATCC 700603 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 _>85.682.53 _11.045.52 _10.711.52 <0.663_0.25

K. pneumoniae

ATCC BAA-2524 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 _>85.682.53 _11.045.52 1.52_3.21 < 0.663_{< 0.163}

K. pneumoniae

B1945 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 _21.415.05 22.07_11.04 2.02_3.21 < 0.650_{< 0.275}

P. aeruginosa

ATCC 27853 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 _42.842.53 11.04_11.04 _10.712.53 <1.125_<0.375

P. aeruginosa

ATCC BAA-2108 Vancomycin GNP-D6_GNP-D8 >88.32_>88.32 _>85.685.05 88.31_44.15 _42.842.39 < 1.473_{< 1}

P. aeruginosa

B1954 Vancomycin GNP-D6_GNP-D8 88.32_88.32 _85.682.53 11.04_44.15 _42.843.03 1.323₁

A. baumannii

ATCC 17978 Vancomycin GNP-D6_GNP-D8 88.32_88.32 _10.711.27 11.04_5.52 1.01_1.61 0.925_0.22

A. baumannii

ATCC BAA-1605 Vancomycin GNP-D6_GNP-D8 44.16_44.16 _10.715.05 5.52_2.76 1.52_0.80 0.426_0.137

A. baumannii

B2026 Vancomycin GNP-D6_GNP-D8 22.08_22.08 _10.712.53 2.76_2.76 0.76_2.68 0.435_0.376 Highlighted: 0.1<FICI < 0.5.

Note: MICa is the MIC of vancomycin alone; MICb corresponds to the MIC of GNPs when used alone; MICac is the MIC of vancomycin in combination with the GNPs at the MICbc concentration. MICbc is the MIC of GNPs when used with the MICac concentration of vancomycin. Highlighted values indicate measured synergy.

(16)

5

(Table 7). In the presence of human plasma, all MIC values were proportionally decreased. In this case, the combined effect against

K. pneumoniae and P. aeruginosa is additive, whereas synergy is still

observed against A. baumannii. Remarkably, GNP-D8 is active in plasma concentrations up to 30 %.

2.3. Mechanistic aspects of the synergistic action against Gram-negative pathogens

We used the Gram-positive bacterial species L. lactis as a model Gram-positive organism with no outer-membrane and exposed it to GNP-D6, GNP-D8 and vancomycin, respectively (Table 4).

As expected, L. lactis MG1363 was sensitive to GNP-D6 (MIC = 2 µM) while the MIC value of GNP-D8 against L. lactis was 16 µM. The MIC value of GNP-D6 was 16 times higher, while the MIC value of GNP-D8 was more than 100 times higher than that of vancomycin (Table 4), thus suggesting that GNP-D6 (and GNP-D8 albeit less effi-ciently) can also target the cytoplasmic membrane. These data confirm the earlier observed additive effect between GNP-D6 with vancomycin. Moreover, the results clearly indicate that the main role of GNP-D8 is to assist either nisin or vancomycin to pass through the outer-membrane

Table 7. MIC value (µM) and combination effect of vancomycin, GNP-D8 and the combina-tions thereof against selected MDR Gram-negative pathogens without or with human plasma.

(µM) MICb (µM) MICac(µM) MICbc (µM) FICI Without human plasma

K. pneumoniae B1945 Vancomycin GNP-D8 >88.32 21.41 11.04 10.2 <0.625

P. aeruginosa ATCC

BAA-2108 Vancomycin GNP-D8 >88.32 >85.68 44.15 42.84 < 1

A. baumannii ATCC

BAA-1605 Vancomycin GNP-D8 44.16 10.71 2.76 2.68 0.188

With human plasma

K. pneumoniae B1945

(30% plasma) Vancomycin GNP-D8 22.08 2.68 2.76 2.68 1.125

P. aeruginosa ATCC

BAA-2108 ( 30% plasma) Vancomycin GNP-D8 22.08 1.34 1.38 1.34 1.06

A. baumannii ATCC

BAA-1605 (10% plasma) Vancomycin GNP-D8 11.04 5.36 0.69 0.67 0.188 Plasma concentration used is the highest one that enables bacterial growth. The ratio of the concen-trations of vancomycin and GNP-D8 is 1:1.

CH

APTER 5: R

es

ul

(17)

5

of Gram-negative bacteria, thus working as a gate-opener rather than effectively targeting the cytoplasmic membrane.

Next, we tested the role of the outer-membrane as the main target of these GNPs using the most exposed component, LPS, and magnesium ions as an LPS-stabilizing agent. Both vancomycin and GNPs exhibited a reduced antimicrobial effect against all the tested pathogens when the cells were grown in the presence of exogenous LPS or Mg2+_as

expected (Table 4). The exogenous LPS or Mg2+_{inhibited the activity}

of GNP-D6 and GNP-D8 more than that of vancomycin, probably because the target of vancomycin is lipid II which is located on the inner membrane. When there was 1 mg/mL LPS in the medium, the MICs of vancomycin against Gram-negative pathogens increased 2-fold except against A. baumannii. This might be put down to specific components of the outer-membrane of A. baumannii. Remarkably, at least 16-fold more GNP-D6 or GNP-D8 was needed to inhibit the growth of Gram- negative pathogens in most cases, in the presence of LPS or magnesium.

The presence of exogenous LPS considerably suppressed the ability of GNP-D6 and GNP-D8 to potentiate vancomycin against Gram- negative pathogens (Figure 1). 2 µM vancomycin and 1 µM GNP-D8 can inhibit the growth of K. pneumoniae without added LPS. However, at least 32 µM vancomycin and 16 µM GNP-D8 were needed for the inhibition of K. pneumoniae after the addition of excess LPS. When 21 mM Mg2+_{was added to the growth medium for the checkerboard}

broth microdilution assays, neither GNP-D6 nor GNP-D8 were able to potentiate vancomycin against any of the five Gram-negative pathogens, which grew at the tested concentrations (the concentrations for the test were the same as shown in Figure 1: the highest concentrations of vancomycin or GNP-D6 and GNP-D8 tested are their individual MICs) (data not shown). This is consistent with the results when van-comycin, GNP-D6 and GNP-D8 were tested alone in the presence of LPS and Mg2+_{(Table 4).}

These results suggest that the GNPs can disrupt the integrity of the outer-membrane of these Gram-negative pathogens via association with LPS, enhance the permeability and sensitize the Gram-negative pathogens to antibiotics that otherwise are restricted to Gram-positive bacteria.

(18)

5

In order to reveal the effect of the combination of vancomycin and GNP-D6 or GNP-D8, growth curve tests were performed. Pronounced inhibitory effects were detected when vancomycin was combined with either GNP-D6 or GNP-D8, compared to the untreated control or when each one of them was used alone (Figure 2). No growth was vis-ible when a GNP and vancomycin were added simultaneously during the assay. The concentrations of vancomycin, GNP-D6 and GNP-D8

Figure 1. GNP-D6, GNP-D8 potentiate vancomycin against Gram-negative pathogens.

Checkerboard broth microdilution assays using vancomycin and GNP-D6 or GNP-D8 against five Gram-negative pathogens either with or without exogenous 1 mg/mL LPS at 37 °C. Dark regions represent higher cell densities. Data are representative of at least duplicates. X-axis: µM concentration of vancomycin. Y-axis: µM concentration of GNP-D6 or GNP-D8 (as indicated on top). CH APTER 5: R es ul ts

(19)

5

(20)

5

when tested alone, or in a mixture, were the specific MICac or MICbc of them against different strains (according to Table 5).

The bactericidal effect of vancomycin, GNP-D6, GNP-D8 and the combinations thereof were determined. Thus, the final concentra-tion of the compounds used was 10-fold increased MIC when tested alone and 10-fold increased concentrations of previously measured MICac + MICbc (MICac and MICbc corresponse to the MICs of van-comycin or GNP-D6/GNP-D8 when tested in combination against five selected Gram-negative pathogens according to Table 5). Unlike the positive controls (untreated cells), no colonies of any of the five Gram-negative pathogens could grow after 3 hours treatment (Figure 3). There was no growth recovery after exposure to vancomycin, GNP-D6, GNP-D8 and the combinations thereof, so both the compounds alone and the combinations thereof were proven to be bactericidal.

Figure 2. Effect of vancomycin, GNP-D6, GNP-D8 and the combination thereof against Gram-negative pathogens. (A) ~ (E), Growth curve analysis. D6, D8 and Van represent GNP-D6,

GNP-D8 and Vancomycin, respectively. (A) E. coli in the presence of 1/4 * MIC (0.5 µM) GNP-D6, 1/16 * MIC (0.25 µM) GNP-D8, 1/4 * MIC (16 µM) vancomycin, 1/32 * MIC (2 µM) vancomycin, the combination of 1/4 * MIC GNP-D6 and 1/4 * MIC vancomycin, the combination of 1/16 * MIC GNP-D8 and 1/32 * MIC vancomycin and untreated control; (B), K. pneumoniae in the presence of 1/4 * MIC (0.5 µM) GNP-D6, 1/32 * MIC (1 µM) GNP-D8, 1/4 * MIC (32 µM) vancomycin, 1/16 * MIC (8 µM) vancomycin, the combination of 1/4 * MIC GNP-D6 and 1/4 * MIC vancomycin, the combination of 1/32 * MIC GNP-D8 and 1/16 * MIC vancomycin and untreated control; (C),

P. aeruginosa in the presence of 1* MIC (2 µM) GNP-D6, 1/4 * MIC (4 µM) GNP-D8, 1/2 * MIC

(64 µM) vancomycin, 1/4 * MIC (32 µM) vancomycin, the combination of 1 * MIC GNP-D6 and 1/2 * MIC vancomycin, the combination of 1/4 * MIC GNP-D8 and 1/4* MIC vancomycin and untreated control; (D), A. baumannii in the presence of 1/2 * MIC (1 µM) GNP-D6, 1/8 * MIC (1 µM) GNP-D8, 1/4 * MIC (8 µM) vancomycin, 1/16 * MIC (2 µM) vancomycin, the combina-tion of 1/2 * MIC GNP-D6 and 1/4 * MIC vancomycin, the combinacombina-tion of 1/8 * MIC GNP-D8 and 1/16 * MIC vancomycin and untreated control; (E), E. aerogenes in the presence of 1 * MIC (2 µM) GNP-D6, 1/8 * MIC (4 µM) GNP-D8, 1/2 * MIC (96 µM) vancomycin, 1/8 * MIC (24 µM) vancomycin, the combination of 1 * MIC GNP-D6 and 1/2 * MIC vancomycin, the combination of 1/8 * MIC GNP-D8 and 1/8 * MIC vancomycin and untreated control.

CH

APTER 5: R

es

ul

(21)

5

2.4. Cell toxicity and hemolytic activity of vancomycin and GNP-D6 or GNP-D8

Vancomycin is smaller than nisin and -in contrast to nisin- it has already been approved as an antibiotic in clinical therapy. Therefore, vancomycin and its combinations with GNP-D6 and GNP-D8 were tested with respect to cell toxicity and hemolysis. Human HEK-293 cells were used to assess cell viability (Figure 4). Vancomycin and GNP-D8 did not affect the ATP levels at the concentrations tested. Only GNP-D6 showed a reduction in the cell viability when it was tested alone (Figure 4). The IC50 of GNP-D6 was 33 µM while the working

Figure 3. Determination of viable cells after treatment with vancomycin, GNP-D6, GNP-D8 and the combinations thereof. The names of the strains are indicated on the left. An untreated

positive growth control (first column) and samples treated with different concentrations of antimicrobials (indicated on top) incubated for 3 h are displayed. The concentrations of antimi-crobials used are 10-fold the MIC indicated in table 5. Every plate was inoculated with 50 µL of a 103-fold-diluted sample from the different strains and treatments.

(22)

5

concentration to inhibit the growth of Gram-negative pathogens was always 2 µM (Table 5). When tested alone, in the presence of 20 µM GNP-D6, the HEK-293 retained 78 % of the activity. However, when the combination of 24 µM GNP-D6 and 88 µM vancomycin was added, the ATP levels of HEK-293 were not affected when compared with the control (Figure 4), suggesting a protective effect.

Fresh human red blood cells (hRBCs) were used for the hemolytic tests (Table 8, Supplementary Table 3 and Figure 5). Vancomycin and GNP-D8 did not cause hemolysis of hRBCs even at 500 µM and 600 µM, respectively, and the HC50 of GNP-D6 was 168.9 µM. Remarkably, 168.9 µM was 80-fold higher than the MIC of GNP-D6. When com-bined, vancomycin and GNP-D8 did not cause hemolysis at the highest concentration tested (360 µM vancomycin and 400 µM GNP-D8). The combination of vancomycin and GNP-D6 produced hemolysis at high concentrations that can be attributed to the effect of GNP-D6.

The cell selectivity of the peptides is identified as the therapeutic in-dex (TI) [23–25], which is a very important parameter used to estimate the toxicity and the effect of drugs. TI correlates the MHC and GM, in which MHC is the concentration of peptides needed to reach 10 %

Figure 4. Effect of vancomycin, GNP-D6, GNP-D8 and the combination thereof against HEK-293. CH APTER 5: R es ul ts

(23)

5

lysis of hRBCs, and GM is the geometric mean of MICs against all the tested bacteria (Table 8). A higher TI is preferable for an antimicrobial to be considered safe. The TI of GNP-D8 was the highest among the three agents, even higher than vancomycin. It is noteworthy that 88 µM vancomycin and GNP-D8 were used as the IC50 value, although at this concentration, the highest tested against HEK-293, there was no toxic effect measured.

Table 8. IC50, HC50, GMs, MHC and TI of vancomycin, GNP-D6 and GNP-D8.

IC50 (µM) a HC50 (µM)b GM (µM)c MHC (µM)d TI(MHC/GM)e

Vancomycin >88 >500 87.7 1000 11.40

GNP-D6 33 170 2.71 25 9.21

GNP-D8 >88 >200 30.75 1200 39.03

a, IC50 is the half maximal inhibitory concentration.

b, HC50 is the concentration that causes 50% hemolysis of hRBCs.

c, GM is the geometric mean of MIC value of all the Gram-negative strains tested.

d, MHC is the minimal hemolytic concentration that caused 10% hemolysis of hRBCs. If there was no detectable hemolytic activity observed at 500 µM, 1000 µM was used for calculation of the therapeutic index (TI).

e, Therapeutic index (TI) = MHC/GM. Larger values indicate greater cell specificity.

Figure 5. Effect of vancomycin, GNP-D6, GNP-D8 and combinations thereof on hemolytic activity. Hemolysis was observed in 96-well plates. The numbers shown on the figure represent

the concentrations (µM) of each compound. PC, positive control, hRBCs suspension with 1% TritonX-100; NC, negative control, hRBCs suspension in PBS.

(24)

5

CH APTER 5: Di sc us sio n

Neither vancomycin nor GNP-D8 caused lysis of human eryth-rocytes, nor showed significant toxicity against the human cell line HEK-293 up to 500 µM. GNP-D6 can cause hemolysis and toxicity at relatively high concentrations, which were 16- or 80-fold higher than the MICs against the Gram-negative pathogens assayed. The TI of vancomycin, GNP-D6 and GNP-D8 were 11.4, 9.21 and 39.03, re-spectively. The TI of GNP-D8 is ~3.4-fold higher than vancomycin and this indicates that GNP-D8 is quite safe and might have good potential for clinical use. In conclusion, the in vitro experiments indicate that (combinations of) vancomycin, with GNP-D8 or GNP-D6 are safe at therapeutic concentrations.

3. Discussion

In vitro synergism has been shown before for combinations of

anti-biotics with other compounds, which can be used to target specific problematic pathogens [12, 26–29]. Following this strategy, we have explored the combined activity of compounds with high potency against Gram-positive bacteria with natural and synthetic peptides (GNPs) that can facilitate their access to the cytoplasmic membrane. In this work, GNP-6 and GNP-8 were further engineered to alter the net charge, size, hydrophobicity, flexibility and stability to improve the activities of GNP-6 and GNP-8, either alone or combined with vanco-mycin or nisin. Amino acids substitutions, D-amino acids and reversed sequences were employed. Most amino acids in proteins of living cells contain L-amino acids. (Poly)peptides composed of D-amino acids are protected against peptidase and protease action [30, 31]. In addition, the introduction of D-amino acids may abolish receptor interaction thus allowing to establish whether or not a receptor is playing a role. Moreover, the reversed peptides (reversal of the peptides’ sequence) were reported to be interesting for peptidomimetics and can influence the activity of peptides [32, 33].

Overall, engineering and sequence reversal did not provide better molecules with the exception of the D-GNPs that retain the activity and are less prone to proteolytic degradation. The fact that both, the D- and L-amino acids-containing GNPs were similarly active indicated that

(25)

5

the activity did not rely on binding to a specific receptor but on their ability to interact with the outer membrane LPS. This is further con-firmed by the MIC data in the presence of LPS and magnesium, which clearly reduce the effect of any of the compounds or their combinations. As it was previously shown in chapter 4, a hypothetical mecha-nism of gate-opening activity of membrane disrupting peptides is proposed. LPS is a constitutive element of Gram-negative OM and crit-ical to maintain the permeability barrier function of the OM of Gram- negative pathogens and magnesium ions can stabilize the LPS layer of the OM [3, 34]. When LPS and Mg2+_{were added in the activity assays,}

the activity of vancomycin, GNP-D6 and GNP-D8 were significantly compromised either when they were tested alone or in combination. These results confirmed that GNPs, especially GNP-D8 was prone to

interact with LPS to improve the outer-membrane perturbation capac-ity, which potentiated the activity of Gram-positive-specific antibiotics (i.e. nisin and vancomycin) against Gram-negative pathogens. Human plasma is a complex mixture with lots of proteins and nutrients, which constitutes on one hand a rich medium for the bacteria, and on the other hand a source of host defense systems. As table 7 showed, the activities of vancomycin and GNP-D8, either alone or combination thereof, were increased in plasma. This might be because of the worse growth of bacteria in the presence of human plasma but also further pointed to the potential good activities of them in vivo. These data are encouraging since the in vivo application of antimicrobial peptides can fail due to the low activity in the presence of plasma [35].

Bacterial growth kinetics and bactericidal test were performed for a better understanding of the underlying mechanism. When combi-nations of vancomycin and GNPs were used, less compounds were needed to inhibit the growth of Gram-negative pathogens. There were no colonies growing after the treatment with vancomycin or/ and GNPs with 10-fold of MIC or 10-fold of MICac + MICbc, which verified that either the compounds alone or the combinations thereof are bactericidal.

The cell toxicity and hemolytic activity of vancomycin, GNP-D6 and GNP-D8 and the combinations thereof were assessed. Vancomycin and GNP-D8 neither caused lysis of hRBCs nor showed significant toxicity against the human kidney cell line HEK-239 when tested alone or in

(26)

5

combination. Some cationic compounds such as polymyxins exert a potent nefrotoxicity at concentrations similar to the therapeutic ones. The therapeutic window of these GNPs is much broader than that of polymyxins, in particular that of GNP-8. The concentration of GNP-D6 to cause a hemolysis and toxicity was extremely high compared to the concentration required to inhibit the growth of Gram-negative patho-gens, including MDR strains. Thus, GNP-D6 has still potential as a can-didate for therapeutic use, although further characterization is required.

Eight multi-drug resistant Gram-negative pathogens were utilized to test vancomycin, GNP-D6, GNP-D8 and the combinations thereof. The results were in line with the previous results with 5 selected pathogens. There is no doubt that synergistic effects of antimicrobials can enhance the therapeutic application of these compounds.

The synergy mechanism and in vivo activities in mice of GNPs and vancomycin/nisin is still under investigation. Here, we highlight the

Figure 6. Schematic overview of the hypothetical synergism of vancomycin/nisin and GNPs .

Left, vancomycin or nisin alone can hardly penetrate the outer-membrane; Middle, a gap will be formed on the OM in the presence of GNPs, and vancomycin or nisin can arrive at the inner membrane; Right, when exogenous LPS and Mg2+_{are added, it is much more difficult to form a}

the gap on the outer-membrane due to GNP sequestration or stabilization of the LPS, respectively.

CH APTER 5: Di sc us sio n

(27)

5

synergistic effect of GNP-D8 and vancomycin, which is also non-he-molytic and non-toxic against human cells in initial preclinical testing. Our preliminary results suggest that GNPs exert gate-opening activity on the OM (probably via LPS binding) and enable vancomycin or nisin to reach their targets. No specific receptors are involved in the actions of the selected GNP-D6 or GNP-D8. The hypothetical mechanisms of the action of these peptides are shown in Figure 6. In summary, the here described combinations constitute a new therapeutic approach to deal with (MDR) Gram-negative pathogens. The activities of vancomycin and GNP-D8 in the presence of human plasma further indicates the high potential of them for clinical use. In spite of the relatively low activity of GNP-D8 alone, the astonishing synergism of GNP-D8 with vancomycin and nisin raises the possibility of GNP-D8 to be registered as an adjuvant instead of antimicrobial itself. Mouse infection model experiments and preclinical development will be the next essential steps into further drug development.

Acknowledgements

Qian Li was supported by the Chinese Scholarship Council (NO 201306770012).

Manuel Montalban-Lopez was supported by a grant of EU FW7 project SynPeptide.

References:

[1] Organization WH. Global priority list of anti-biotic-resistant bacteria to guide research, discovery and development of new antibiot-ics. 2017. . https://www.who.int/medicines/ publications/global-priority-list-antibiot-ic-resistant-bacteria/en/.

[2] Santajit S, Indrawattana N. Mechanisms of an-timicrobial resistance in ESKAPE path-ogens. BioMed research international 2016;2016:2475067.

[3] Helander IM, Von Wright A, Mattila-Sandholm T. Potential of lactic acid bacteria and novel antimicrobials against Gram-negative bac-teria. Trends in food science & technology 1997;8:146–50.

[4] Kamio Y, Nikaido H. Outer membrane of

Salmonel-la typhimurium: accessibility of phospholipid

head groups to phospholipase c and cyanogen bromide activated dextran in the external me-dium. Biochemistry 1976;15:2561–70. [5] Okuda S, Sherman DJ, Silhavy TJ, Ruiz N,

Kah-ne D. Lipopolysaccharide transport and assembly at the outer membrane: the PEZ model. Nature reviews microbiology 2016;14:337.

[6] Delcour AH. Outer membrane permeability and antibiotic resistance. Biochimica et biophysica acta 2009;1794:808–16. [7] Blair JM, Webber MA, Baylay AJ, Ogbolu DO,

(28)

5

CH APTER 5: R ef er en ces

antibiotic resistance. Nature reviews mi-crobiology 2015;13:42–51.

[8] Breukink E, de Kruijff B. Lipid II as a target for antibiotics. Nature reviews drug discovery 2006;5:321–2.

[9] Boziaris I, Adams M. Effect of chelators and nisin produced in situ on inhibition and inactiva-tion of Gram negatives. Internainactiva-tional jour-nal of food microbiology 1999;53:105–13. [10] Stevens K, Sheldon B, Klapes N, Klaenhammer

T. Nisin treatment for inactivation of

Sal-monella species and other Gram-negative

bacteria. Applied and environmental mi-crobiology 1991;57:3613–5.

[11] Helander IM, Mattila-Sandholm T. Permeability barrier of the Gram-negative bacterial outer membrane with special reference to nisin. International journal of food microbiology 2000;60:153–61.

[12] Naghmouchi K, Baah J, Hober D, Jouy E, Ru-brecht C, Sané F, et al. Synergistic effect between colistin and bacteriocins in controlling Gram-negative pathogens and their potential to reduce antibiotic toxicity in mammalian epithelial cells. Antimicrobial agents and chemotherapy 2013;57:2719–25.

[13] Stokes JM, MacNair CR, Ilyas B, French S, Côté J-P, Bouwman C, et al. Pentamidine sensitizes Gram-negative pathogens to antibiotics and overcomes acquired colistin resistance. Nature microbiology 2017;2:17028.

[14] Torcato IM, Huang Y-H, Franquelim HG, Gaspar D, Craik DJ, Castanho MA, et al. Design and characterization of novel an-timicrobial peptides, R-BP100 and RW-BP100, with activity against Gram-negative and Gram-positive bacteria. Biochimica et biophysica acta 2013;1828:944–55. [15] Gasson MJ. Plasmid complements of

Strepto-coccus lactis NCDO 712 and other lactic

streptococci after protoplast-induced cur-ing. Journal of bacteriology 1983;154:1–9. [16] Zhou L, van Heel AJ, Kuipers OP. The length

of a lantibiotic hinge region has profound influence on antimicrobial activity and host specificity. Frontiers in microbiology 2015:6:100–7.

[17] Wiegand I, Hilpert K, Hancock RE. Agar and broth dilution methods to determine the minimal inhibitory concentration (MIC) of antimicrobial substances. Nature protocols 2008;3:163–75.

[18] Lehár J, Krueger AS, Avery W, Heilbut AM, Johansen LM, Price ER, et al. Synergistic drug combinations tend to improve ther-apeutically relevant selectivity. Nature bi-otechnology 2009;27:659–66.

[19] Odds F. Synergy, antagonism and what the chequerboard puts between them. Journal of antimicrobial chemotherapy 2003;52:1. [20] Naghmouchi K, Belguesmia Y, Baah J,

Teath-er R, DridTeath-er D. AntibactTeath-erial activity of class I and IIa bacteriocins combined with polymyxin E against resistant variants of

Listeria monocytogenes and Escherichia coli.

Research in microbiology 2011;162:99–107. [21] Peternel L, Kotnik M, Preželj A, Urleb U. Com-parison of 3 cytotoxicity screening assays and their application to the selection of novel antibacterial hits. Journal of biomo-lecular screening 2009;14:142–50. [22] de Breij A, Riool M, Cordfunke RA, Malanovic

N, de Boer L, Koning RI, et al. The antimi-crobial peptide SAAP-148 combats drug-re-sistant bacteria and biofilms. Science trans-lational medicine 2018;10:eaan4044. [23] Chen Y, Mant CT, Farmer SW, Hancock RE,

Vasil ML, Hodges RS. Rational design of α-helical antimicrobial peptides with en-hanced activities and specificity/therapeu-tic index. Journal of biological chemistry 2005;280:12316–29.

[24] Kim EY, Rajasekaran G, Shin SY. LL-37-derived short antimicrobial peptide KR-12-a5 and its D-amino acid substituted analogs with cell selectivity, anti-biofilm activity, synergistic effect with conventional an-tibiotics and anti-inflammatory activity. European journal of medicinal chemistry 2017;136:428–41.

[25] Muller PY, Milton MN. The determination and interpretation of the therapeutic index in drug development. Nature reviews drug discovery 2012;11:751.

[26] Field D, Seisling N, Cotter PD, Ross RP, Hill C. Synergistic nisin-polymyxin combina-tions for the control of Pseudomonas bio-film formation. Frontiers in microbiology 2016;7:1713.

[27] Hornsey M, Longshaw C, Phee L, Wareham DW. In-vitro activity of telavancin in com-bination with colistin versus Gram-negative bacterial pathogens. Antimicrobial agents and chemotherapy 2012;56:3080–5. [28] Barbee LA, Soge OO, Holmes KK, Golden MR.

In vitro synergy testing of novel

antimicro-bial combination therapies against

Neisse-ria gonorrhoeae. Journal of antimicrobial

chemotherapy 2014;69:1572–8. [29] Chen X, Niyonsaba F, Ushio H, Okuda D,

Na-gaoka I, Ikeda S, et al. Synergistic effect of antibacterial agents human β-defensins, cathelicidin LL-37 and lysozyme against

Staphylococcus aureus and Escherichia coli. Journal of dermatological science

(29)

5

[30] Güell I, Cabrefiga J, Badosa E, Ferre R, Talleda M, Bardají E, et al. Improvement of the efficacy of linear undecapeptides against plant-pathogenic bacteria by incorporation of D-amino acids. Applied and environ-mental microbiology 2011;77:2667–75. [31] Hong SY, Oh JE, Lee K-H. Effect of

D-ami-no acid substitution on the stability, the secondary structure and the activity of membrane-active peptide. Biochemical pharmacology 1999;58:1775–80. [32] Atzori A, Baker AE, Chiu M, Bryce RA, Bonnet

P. Effect of sequence and stereochemistry reversal on p53 peptide mimicry. PLoS one 2013;8:e68723.

[33] Vadukul DM, Gbajumo O, Marshall KE, Serpell LC. Amyloidogenicity and toxicity of the reverse and scrambled variants of amy-loid-β 1–42. FEBS letters 2017;591:822–30. [34] Zhang G, Meredith TC, Kahne D. On the essen-tiality of lipopolysaccharide to Gram-nega-tive bacteria. Current opinion in microbi-ology 2013;16:779–85.

[35] Marr AK, Gooderham WJ, Hancock RE. An-tibacterial peptides for therapeutic use: obstacles and realistic outlook. Current opinion in pharmacology 2006;6:468–72.

(30)

CH APTER 5: S up plem en ta ry M at er ia ls

5 Supplementary Materials

Supplementary Table 1. MIC value (µM) of GNP-8 and GNP-8 analogs alone against 5 Gram- negative pathogens.

E. coli

LMG15862 K. pneumoniae LMG20218 P. aeruginosaLMG 6395 A. baumannii LMG01041 E. aerogenesLMG 02094

GNP-8 12 32 16 >64 128 GNP8-1 32 64 16 128 >256 GNP8-2 16 32 16 128 256 GNP8-3 12 24 16 128 128 GNP8-4 16 >256 16 >256 >256 GNP8-5 24 >256 32 >256 >256 GNP8-6 256 >256 >256 >256 >256

Supplementary Table 2. MIC value (µM) of MDR Gram-negative pathogens.

Vancomycin GNP-D6 GNP-D8 E. coli ATCC 25922 >88.32 1.26 5.36 E. coli ATCC BAA-2452 >88.32 1.27 5.35 E. coli B1927 88.32 1.27 2.68 K. pneumoniae ATCC 700603 >88.32 2.53 >85.62 K. pneumoniae ATCC BAA-2524 >88.32 2.53 >85.62 K. pneumoniae B1945 >88.32 5.05 21.41 P. aeruginosa ATCC 27853 >88.32 2.53 42.84 P. aeruginosa ATCC BAA-2108 >88.32 5.05 >85.62 P. aeruginosa B1954 88.32 2.53 85.62 A. baumannii ATCC 17978 88.32 1.27 10.71 A. baumannii ATCC BAA-1605 44.16 5.05 10.7 A. baumannii B2026 22.08 2.53 10.7

(31)

5

Su pp le me nta ry T ab le 3. H emo ly tic ac tiv ity o f va nc om ycin, GNP -D6 a nd GNP -D8 a nd the c om bina tio ns the re of . Va nco m ycin GNP -D6 GNP -D8 Va nco m ycin-GNP -D6 Va nco m ycin-GNP -D8 C on a (µM) M HA b (%) ±S D C on a (µM) M HA b (%) ±S D C on a (µM) M HA b (%) ±S D C on a (µM) M HA b (%) ±S D C on a (µM) M HA b (%) ±S D 31 0 12.5 3.64±0.16 50 0 90/25 12.20±0.12 22/25 0 62.5 0 25 10.51±0.28 100 0 180/50 27.35±0.25 45/50 0 125 0 50 20.81±0.09 200 0 360/100 44.53±0.27 90/100 0 250 0 100 37.23±0.29 400 0 720/200 68.23±0.30 180/200 0 500 0 200 58.73±0.23 600 0 720/400 80.22±0.06 360/400 0 C on a, co ncen tra tio ns; MH A b, m ea n h em ol yt ic ac tiv ity ; S D , s ta nd ar d de vi at io n.

(32)

(33)