Effects of antibiotic treatment and phagocyte infiltration on development of Pseudomonas aeruginosa biofilm-Insights from the application of a novel PF hydrogel model in vitro and in vivo

Abstract

Background and purpose: Bacterial biofilm infections are major health issues as the infections are highly tolerant to antibiotics and host immune defenses. Appropriate biofilm models are important to develop and improve to make progress in future biofilm research. Here, we investigated the ability of PF hydrogel material to facilitate the development and study of Pseudomonas aeruginosa biofilms in vitro and in vivo.

Methods: Wild-type P. aeruginosa PAO1 bacteria were embedded in PF hydrogel situated in vitro or in vivo, and the following aspects were investigated: 1) biofilm development; 2) host immune response and its effect on the bacteria; and 3) efficacy of antibiotic treatment.

Results: Microscopy demonstrated that P. aeruginosa developed typical biofilms inside the PF hydrogels in vitro and in mouse peritoneal cavities where the PF hydrogels were infiltrated excessively by polymorphonuclear leukocytes (PMNs). The bacteria remained at a level of ~10⁶ colony-forming unit (CFU)/hydrogel for 7 days, indicating that the PMNs could not eradicate the biofilm bacteria. β-Lactam or aminoglycoside mono treatment at 64× minimal inhibitory concentration (MIC) killed all bacteria in day 0 in vitro biofilms, but not in day 1 and older biofilms, even at a concentration of 256× MIC. Combination treatment with the antibiotics at 256× MIC completely killed the bacteria in day 1 in vitro biofilms, and combination treatment in most of the cases showed significantly better bactericidal effects than monotherapies. However, in the case of the established in vivo biofilms, the mono and combination antibiotic treatments did not efficiently kill the bacteria.

Conclusion: Our results indicate that the bacteria formed typical biofilms in PF hydrogel in vitro and in vivo and that the biofilm bacteria were tolerant against antibiotics and host immunity. The PF hydrogel biofilm model is simple and easy to fabricate and highly reproducible with various application possibilities. We conclude that the PF hydrogel biofilm model is a new platform that will facilitate progress in future biofilm investigations, as well as studies of the efficacy of new potential medicine against biofilm infections.

Keywords: PF hydrogel; Pseudomonas aeruginosa; antibiotic resistance; biofilm infection; in vivo model.

Figure 1

Development of P. aeruginosa biofilm in PF hydrogel in vitro (A) and in vivo (B). (A) Left panel: CLSM micrographs visualize clearly that P. aeruginosa PAO1 (labeled with GFP) develops biofilms progressively from single bacteria (almost no signal due to limited number of bacteria) at day 0, further to microcolonies at day 1 that increased in size at days 3 and 7. Right panel: Images taken by light microscopy of methylene stained day 3 biofilms show distinct biofilm structures and single bacterial cells inside the PF hydrogel. The red arrow points toward a microcolony in the upper image and the bacteria in the red rectangle area are visualized by higher magnification in the lower image. (B) Epifluorescence and merged epifluorescence/light images acquired in frozen sections of PF hydrogel implants harvested from mouse peritoneal cavities. PAO1 (labeled with GFP) formed first single microcolonies at day 1 and then developed biofilm in an extensive area at day 3.

Figure 2

Quantification of P. aeruginosa bacteria present in PF hydrogels harvested from mouse peritoneal cavities (A) and visualization of the harvested PF hydrogels (B). (A) Enumeration of the number of bacteria in the infections. All bar charts show the median ± standard deviation (n ≥ 6). Values are shown for bacteria in hydrogels (white bars), spleens (black bars), originally sterile hydrogels (dark gray bars), and peritoneal swabs (light gray-bars). (B) Gross view of PF hydrogels with or without P. aeruginosa harvested from mouse peritoneal cavities. (a) PF hydrogel implant before implantation (color induced by Resazurin sodium); (b, c) sterile PF hydrogel on day 3 after implantation was generally loose and easily collected from the mouse peritoneal cavity; (d, e) PF hydrogel with P. aeruginosa on day 3 after implantation was wrapped by peritoneal omentum and could still be carefully separated; (f, g) PF hydrogel with P. aeruginosa on day 14 after implantation was wrapped tightly by the peritoneal omentum with pus inside, and was impossible to separate from the peritoneum.

Figure 3

Infiltration of phagocytes in the PF hydrogels without (A) and with (B) P. aeruginosa PAO1 in vivo. Images were acquired in PF hydrogels harvested from mouse peritoneal cavities at days 1, 3, and 7 after implantation and were stained with live/dead stain or methylene blue. (A) Upper image: CLSM micrograph of live/dead stained sterile PF hydrogel indicates that it was penetrated by a thin layer (ca. 40 µm) of phagocytes. Lower images: Infiltration of phagocytes in the sterile PF hydrogels visualized by light microscopy at different magnifications (bar sizes: left, 200 µm; right, 20 µm). The phagocytes in the PF hydrogels were mostly peritoneal macrophages, and the infiltrations were remarkable only at day 1, reduced significantly at day 3, and disappeared completely at day 7. (B) Upper image: CLSM micrograph of live/dead stained bacteria-containing PF hydrogel indicates that it is penetrated by a thick layer (> 250 µm) of phagocytes. Lower images: Infiltration of phagocytes in the PF hydrogels visualized by light microscopy at different magnifications (bar sizes: left, 200 µm; right, 20 µm). The images show a high density of host immune cells in the PF hydrogels predominated by PMNs, and the infiltration of PMNs was continuously observed in the whole study period. PMNs migrated toward and stayed close to the bacterial biofilms. The PMNs dispersed widely following the development of biofilm at days 3 and 7.

Figure 4

Evaluation of mono and combination therapies of β-lactam (imipenem, IMP) and/or aminoglycoside (gentamycin, GM) in P. aeruginosa PAO1 biofilms in vitro (A) and in vivo (B). (A) Comparing the doses of 0×, 64×, 128×, and 256× MIC in monotherapy or combination therapy, it appears that higher concentrations of antibiotics exhibited stronger killing against the biofilm bacteria. However, significant difference was only seen in the combination treatment, p < 0.02. Increasing the dose from 64× to 128× MIC and 128× to 256× MIC in monotherapy did not show significant differences. (B) Bacteria in day 0 biofilms were 100% killed by IMP or GM alone at 64× MIC, which is consistent with the in vitro findings. In contrast, IMP monotherapy at 128× MIC killed <10% and <3% of the bacteria in day 1, 3, and 7 biofilms, respectively. GM alone at 128× MIC killed 85% of the biofilm bacteria and exhibited better bactericidal effect than IMP (p < 0.0001) at day 1 biofilms. Combination treatment at the same MIC killed 92.5%, 47.5%, and 38.5% of the bacteria in day 1, 3, and 7 biofilms, respectively, in which, killing rates were significantly higher than for either antibiotic monotherapy at all assessed time points (p < 0.03). The two sharply sloped parallelograms of the data distribution indicate that antibiotic treatment was more effective in young biofilm infections (days 0 and 1) compared with mature biofilm infections (days 3 and 7). CTR, control; IMP, imipenem; GM, gentamicin; IMP + GM, imipenem + gentamicin combination. MICs: IMP, 1.25 μg/ml; and GM, 0.5 μg/ml.