Targeted Antimicrobial Photodynamic Therapy of Biofilm-Embedded and Intracellular Staphylococci with a Phage Endolysin's Cell Binding Domain

Abstract

Bacterial pathogens are progressively adapting to current antimicrobial therapies with severe consequences for patients and global health care systems. This is critically underscored by the rise of methicillin resistant Staphylococcus aureus (MRSA) and other biofilm-forming staphylococci. Accordingly, alternative strategies have been explored to fight such highly multidrug resistant microorganisms, including antimicrobial photodynamic therapy (aPDT) and phage therapy. aPDT has the great advantage that it does not elicit resistance, while phage therapy allows targeting of specific pathogens. In the present study, we aimed to merge these benefits by conjugating the cell-binding domain (CBD3) of a Staphylococcus aureus phage endolysin to a photoactivatable silicon phthalocyanine (IRDye 700DX) for the development of a Staphylococcus-targeted aPDT approach. We show that, upon red-light activation, the resulting CBD3-700DX conjugate generates reactive oxygen species that effectively kill high loads of planktonic and biofilm-resident staphylococci, including MRSA. Furthermore, CBD3-700DX is readily internalized by mammalian cells, where it allows the targeted killing of intracellular MRSA upon photoactivation. Intriguingly, aPDT with CBD3-700DX also affects mammalian cells with internalized MRSA, but it has no detectable side effects on uninfected cells. Altogether, we conclude that CBD3 represents an attractive targeting agent for Staphylococcus-specific aPDT, irrespective of planktonic, biofilm-embedded, or intracellular states of the bacterium. IMPORTANCE Antimicrobial resistance is among the biggest threats to mankind today. There are two alternative antimicrobial therapies that may help to control multidrug-resistant bacteria. In phage therapy, natural antagonists of bacteria, lytic phages, are harnessed to fight pathogens. In antimicrobial photodynamic therapy (aPDT), a photosensitizer, molecular oxygen, and light are used to produce reactive oxygen species (ROS) that inflict lethal damage on pathogens. Since aPDT destroys multiple essential components in targeted pathogens, aPDT resistance is unlikely. However, the challenge in aPDT is to maximize target specificity and minimize collateral oxidative damage to host cells. We now present an antimicrobial approach that combines the best features of both alternative therapies, namely, the high target specificity of phages and the efficacy of aPDT. This is achieved by conjugating the specific cell-binding domain from a phage protein to a near-infrared photosensitizer. aPDT with the resulting conjugate shows high target specificity toward MRSA with minimal side effects.

Keywords: Staphylococcus aureus; Staphylococcus epidermidis; antimicrobial photodynamic therapy; biofilms; cell-binding domain; endolysin; intracellular Staphylococcus aureus; intracellular bacteria.

FIG 1

(A) Co-localization of the GFP-CBD3 fusion protein (green) with bacteria of the CA-MRSA strain AH4807 and the S. epidermidis ATCC 35984 strain stained with DAPI (blue). A total of 1 μM GFP-CBD3 was incubated with diluted bacterial overnight cultures (OD₆₀₀ = 0.5) and imaged by confocal laser scanning microscopy. (B) Binding intensity of 0.64 μM CBD3-700DX to bacteria of the CA-MRSA-AH4807 strain and the S. epidermidis ATCC strain 35984. Fluorescence was measured with an Amersham Typhoon Biomolecular Imager and analyzed with ImageJ. Data are presented as the mean ± standard error of the mean (SEM) of two experiments performed in triplicate. Welch’s test was used for statistical analysis. Significant differences are marked (**, P < 0.002).

FIG 2

Photoactivated killing of S. aureus CA-MRSA-AH4807 (A) and S. epidermidis ATCC 35984 (B) grown to exponential phase. Approximately 1 × 10⁷ CFU/mL of the bacteria were incubated with stepwise increasing concentrations of CBD3-700DX (0.02 to 2.6 μM) or without photosensitizer. Bacteria were irradiated with red light at a radiance exposure of 30 J · cm⁻² (+) or kept in the dark (–). (C) H₂O₂ production upon aPDT of CA-MRSA-AH4807 with 5 μM CBD3-700DX or without photosensitizer. H₂O₂ was detected with 10 μM of an AquaSpark Peroxide Probe. In all experiments, irradiation was performed with a LED system that emits red light. Data are presented as the mean ± SEM of three experiments performed in triplicate (panels A and B) and two experiments performed in triplicate (panel C). The Kruskal-Wallis test with subsequent Dunn’s multiple-comparison tests was used for statistical analysis. Significant differences compared with the negative-control group (no CBD3-700DX and no light) are shown as follows: *, P < 0.03; **, P < 0.002; ***, P < 0.0002; ****, P < 0.0001.

FIG 3

aPDT of S. epidermidis biofilms with CBD3-700DX. Biofilms formed by S. epidermidis ATCC strain 35984 were incubated with either 8 μM CBD3-700DX (P+) or PBS (P–), and they were either kept in the dark (L–) or treated with red-light LEDs at a radiance exposure of 30 J · cm⁻² (L+). To assess bacterial viability, biofilms were stained with BacLight LIVE/DEAD stain and imaged by confocal laser scanning microscopy. Green fluorescence (Syto9) marks living bacteria and red fluorescence (propidium iodide) marks dead bacteria. Video S2 in the supplemental material shows a three-dimensional reconstruction from stacks of 2-dimensional confocal microscopy images recorded upon aPDT with CBD3-700DX (P+L+). Fig. S2 shows the unmerged images of the Syto9 and propidium iodide fluorescence.

FIG 4

(Photo)cytotoxicity of CBD3-700DX toward HeLa cells. HeLa cells were incubated with different concentrations of CBD3-700DX for 15 min, and the unbound conjugate was removed by washing with DPBS prior to treatment with red light (+) at a radiance exposure of 30 J · cm⁻². (A) (Photo)cytotoxicity was assessed using the colorimetric MTT assay 24 h after treatment. The percentage of cell viability, expressed as MTT reduction, was calculated relative to that of viable control cells that were mock-treated with DPBS in the dark. Cells treated with 1% SDS were used as a control for cell killing. (B) The H₂DCFDA fluorescence fold increase per mg protein (y axis) was determined by fluorescence spectroscopy immediately after aPDT as a measure for cellular ROS levels. Data are presented as the mean ± SEM of three experiments performed in triplicate. Notably, in panel B, two outlier data points of 2.17 × 10⁷ and 3.90 × 10⁷ fluorescence fold increases per mg protein were removed from the condition where HeLa cells were treated with 0.64 μM CBD3-700DX in the presence of red light. An ordinary one-way ANOVA with a subsequent Holm-Sidak’s multiple-comparison test was used for statistical analysis. Significant differences compared with the control group (no photosensitizer and no light) are shown as follows: *, P < 0.03; **, P < 0.002; ***, P < 0.0002; ****, P < 0.0001.

FIG 5

CBD3 internalization by HeLa cells and binding to intracellular S. aureus. (A) Uninfected HeLa cells incubated with 0.2 μM of the GFP-CBD3 fusion protein (false color represented in red). (B) HeLa cells were infected with CA-MRSA D15-GFP (MOI = 10) and subsequently incubated with lysostaphin to eliminate extracellular bacteria. The HeLa cells with internalized bacteria (bottom left image; green in the overlay) were then incubated with 0.5 μM CBD3-Alexa Fluor 647 (top right image; red in the overlay). Nuclei were stained with DAPI (top left image; blue in the overlay). Fluorescence images were acquired by confocal laser scanning microscopy. Note that in the overlay (bottom right image), the co-localizing fluorescence signals of the GFP (green) and CBD3-Alexa Fluor 647 (red) appear yellow.

FIG 6

aPDT with CBD3-700DX kills intracellular S. aureus and infected HeLa cells. HeLa cells were incubated overnight without (P–) or with (P+) 0.2 μM CBD3-700DX. The following day, the cells were either infected with CA-MRSA-AH4807 at an MOI of 10 for 2 h, or they remained uninfected. Cells were then incubated with lysostaphin to eliminate extracellular bacteria. Both the uninfected (A) and the infected (B) HeLa cells were irradiated with red light (L+) at a radiance exposure of 30 J · cm⁻² or were kept in the dark (L–). To assess bacterial and HeLa cell viability, BacLight LIVE/DEAD staining was performed followed by confocal laser scanning microscopy. Green fluorescence (Syto9) marks living cells and bacteria, while red fluorescence (propidium iodide) marks dead cells and bacteria (the red arrows mark dead bacteria). Fig. S4 and S5 in the supplemental material show unmerged images of the Syto9 and propidium iodide fluorescence for panels A and B, respectively.