A short artificial antimicrobial peptide shows potential to prevent or treat bone infections

Abstract

Infection of bone is a severe complication due to the variety of bacteria causing it, their resistance against classical antibiotics, the formation of a biofilm and the difficulty to eradicate it. Antimicrobial peptides (AMPs) are naturally occurring peptides and promising candidates for treatment of joint infections. This study aimed to analyze the effect of short artificial peptides derived from an optimized library regarding (1) antimicrobial effect on different bacterial species, (2) efficacy on biofilms, and (3) effect on osteoblast‑like cells. Culturing the AMP-modifications with Escherichia coli, Enterococcus faecalis, Pseudomonas aeruginosa, Staphylococcus aureus (including clinical isolates of MRSA and MSSA) and Staphylococcus epidermidis identified one candidate that was most effective against all bacteria. This AMP was also able to reduce biofilm as demonstrated by FISH and microcalorimetry. Osteoblast viability and differentiation were not negatively affected by the AMP. A cation concentration comparable to that physiologically occurring in blood had almost no negative effect on AMP activity and even with 10% serum bacterial growth was inhibited. Bacteria internalized into osteoblasts were reduced by the AMP. Taken together the results demonstrate a high antimicrobial activity of the AMP even against bacteria incorporated in a biofilm or internalized into cells without harming human osteoblasts.

Figure 1

Minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) in µg/ml for gentamicin (Genta) and for five different AMPs. MH⁺⁺: Mueller-Hinton bouillon with cations (n = 4–8), MH⁻: Mueller-Hinton bouillon without cations (n = 4). Color code: green low concentration, dark orange high concentration needed for antimicrobial effect.

Figure 2

Human primary osteoblast-like cells were cultured with different concentrations of AMP and gentamicin for three days. Viability (A) and alkaline phosphatase (AP) activity (B) were determined. Values for AP were normalized to 10⁴ cells and compared to control (untreated cells, 100%). Results are means of n = 6, cells from two donors, triplicates.). Data are shown as means with standard deviation. Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm post-hoc test. Significant inhibition compared to the control group, viability: AMP1 ≥ 100 µg/ml, AMP2 ≥ 267 µg/ml, AMP3 ≥ 68 µg/ml, AMP4 ≥ 251 µg/ml, AMP5 ≥ 63 µg/ml; Alkaline phosphatase activity: AMP1 all concentrations, AMP2–4 ≥ 125 µg/ml, AMP5 ≥ 63 µg/ml.

Figure 3

The differentiation experiment of the osteoblast-like cells was done with cultivation medium (CM), differentiation medium (DM) and DM with low (17 µg/ml) and high (133 µg/ml) AMP2. After 21 days alkaline phosphatase (AP) activity assay and staining (A) and alizarin red staining and quantification (B) were done (n = 6, cells from two donors, triplicates). Results were normalized to 10⁴ cells and expressed in percentage to control (DM) which was set 100%. Data are shown as median with 25 and 75 percentiles. Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm post-hoc test: *Significant difference compared to the DM group, p ≤ 0.05. (C) Gene expression at days 0, 7, 14 and 21. Results are calculated with the ΔΔCt method with efficiency correction and normalization to the control (day 0 of differentiation) and 18 S ribosomal RNA used as a reference gene (n = 2). Data are shown as means with standard deviation.

Figure 4

Scratch assay with human primary osteoblast-like cells and different AMP2 concentrations (0–133 µg/ml). Measurement of the gap was directly done after performing the scratch and after 24 and 48 hours incubation (n = 3–9, cells from one donor). (A) The cell-free area after 48 hours is shown in percentage of the gap compared to time point 0 hours ( = 100%). Positive control: CM with 10% FCS. Data are shown as median with 25 and 75 percentiles. Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm post-hoc test: *Significant difference compared to group treated with 0 µg/ml AMP, p ≤ 0.05. (B) Exemplary pictures of positive control and 67 µg/ml AMP2 after 24 and 48 hours. The dotted lines mark the scratch. Scale bar 500 µm.

Figure 5

S. aureus internalized in human primary osteoblast-like cells were treated with 33 or 67 µg/ml AMP2 and 7.5 µg/ml rifampicin (rifam) for 4 hours and 24 hours (n = 7, cells from one donor). Data are shown as median with 25 and 75 percentiles. Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm post-hoc test: *Significant compared to control group (0 µg/ml AMP), p ≤ 0.05.

Figure 6

Fluorescence in situ hybridization (FISH) of S. epidermidis biofilms grown on polyurethane (PU) waver treated with 0, 4 or 8 µg/ml AMP2. (A) On the left hand, the panel shows the nucleic acid stain DAPI (blue) overlaid with the Cy3-labeled Staphylococcus-specific FISH-probe STAPHY (orange) for each group. The middle and right panels show either the separate DAPI or FISH Cy3 images in black and white of the respective colored image. AMP2 treatment led to a dramatic reduction of total biofilm area and fraction of FISH positive bacteria. Note also the altered fragmented appearance of the AMP treated biofilms. Scale bar = 10 µm. Pictures of 10 random areas per group were taken and results of digital images analysis are shown. (B) Total biofilm area (stained with DAPI). (C) Area of the FISH-positive bacteria (STAPHY-probe labelled with Cy3). n = 10 pictures per group. Data are shown as median with 25 and 75 percentiles.

Figure 7

S.aureus biofilm was grown on glass beads and incubated with 133–533 µg/ml AMP2 and heat flow of the bacteria was measured over 48 hours. (A) Total heat over time. red: growth control (GC), gray: negative control (NC) - beads without bacteria, green: 133 µg/ml AMP2, purple: 267 µg/ml AMP2, black: 533 µg/ml µg/ml AMP2 (n = 6). Kruskal-Wallis, Mann-Whitney, Bonferroni-Holm post-hoc test: *Significant compared to growth control, p ≤ 0.05. (B) After calorimetry measurement, samples and glass beads were plated on one half of a CA plate, respectively. Shown are exemplarily plates with medium sample on the left and three glas beads (circles) of 133, 267, 533 µg/ml AMP2 and 128 µg/ml daptomycin on the right side.

Figure 8

Zone of inhibition (ZOI) test of the AMP2 or gentamicin incorporated at different concentrations into a poly(D,L-lactide) coating (n = 3). Gentamicin was used in 10% w/w PDLLA, AMP2 ranged from 10, 20 to 30% w/w PDLLA. (A) ZOI after 1 day, (B) ZOI after 3 days, (C) Quantification of the ZOI, n = 3. Data are shown as means with standard deviation.

Figure 9

Two dimensional drawing of the five selected peptides using the PepDraw online software. (A) AMP1, (B) AMP2, (C) AMP3, (D) AMP4 and (E) AMP5.