Discovery of hemocompatible bacterial biofilm-resistant copolymers

Abstract

Blood-contacting medical devices play an important role within healthcare and are required to be biocompatible, hemocompatible and resistant to microbial colonization. Here we describe a high throughput screen for copolymers with these specific properties. A series of weakly amphiphilic monomers are combinatorially polymerized with acrylate glycol monomers of varying chain lengths to create a library of 645 multi-functional candidate materials containing multiple chemical moieties that impart anti-biofilm, hemo- and immuno-compatible properties. These materials are screened in over 15,000 individual biological assays, targeting two bacterial species, one Gram negative (Pseudomonas aeruginosa) and one Gram positive (Staphylococcus aureus) commonly associated with central venous catheter infections, using 5 different measures of hemocompatibility and 6 measures of immunocompatibililty. Selected copolymers reduce platelet activation, platelet loss and leukocyte activation compared with the standard comparator PTFE as well as reducing bacterial biofilm formation in vitro by more than 82% compared with silicone. Poly(isobornyl acrylate-co-triethylene glycol methacrylate) (75:25) is identified as the optimal material across all these measures reducing P. aeruginosa biofilm formation by up to 86% in vivo in a murine foreign body infection model compared with uncoated silicone.

Keywords: Bacterial biofilm; Hemocompatiblility; High throughput screening; Polymer microarray; Pseudomonas aeruginosa; Staphylococcus aureus.

Fig. 1

An overview of polymer microarray screening and scale up of ‘hit’ material compositions for in vitro and in vivo testing. (a) Polymer microarray printing using a robotic contact printer on pHEMA coated round and rectangular slides. (b) Quasi-static and flow chambers used for human whole blood incubation. (c) (i) Quantification of platelet, leukocyte, fibrinogen, IgG, complement C3a binding/adsorption to each microarray polymer spot after 2 h incubation with whole blood; (ii) Incubation of microarrays conditioned in blood for 2 h with P. aeruginosa or S. aureus. (d) Confocal microscope images showing surface bacterial coverage of the lead ‘hit’ material coated onto silicone catheter compared to commercially available silver and uncoated silicone catheters; evaluation of haemostasis and inflammation markers on scaled up hit materials. (e) Coating of scaled up ‘hit’ compositions on silicone catheters (confirmed by SEM) and in-vivo testing of lead material in a murine foreign body infection model.

Fig. 2

High throughput biological screening of the polymer library and selection of hits. (a-b) Intensity map representations of fluorescence readouts (au) of biofilm formation on each polymer spot on the microarray for a) P. aeruginosa and b) S. aureus. c) Intensity scale of fluorescence values compiling a subset of biological measurements taken for 6 selected ‘hit’ copolymers (first 6 formulations listed), demonstrating the biofilm resistance and hemocompatibility of hit materials across the multiple parameters assessed (biofilm formation (a and b), leukocyte attachment, IgG adsorption, platelet adhesion and fibrinogen adsorption). Measurements for the entire polymer library are shown in Figure SI2. Values for control non-fouling materials and 2 materials with poor biological performance are also shown. d) Intensity scale used for intensity maps. e) The numerical high and low values (au) for each biological screen. f) The chemical structures of the monomers used to make the ‘hit’ copolymer formulations. The structures of all monomers used in the study are shown in Figure SI1.

Fig. 3

Blood clotting and immune system activation by scaled up selected copolymer formulations under flow () and quasi-static () conditions. a) List of materials. All monomer acronyms are listed in Figure SI1. (b-f) Blood clotting mediator assays: b) platelet activation measured by platelet factor 4 (PF4), c) fibrinogen adsorption, d) coagulation activation measured by prothrombin fragment 1 + 2 (F1+2), e) platelet decay assay, f) leukocyte-platelet conjugate assay, (g-l) immune component activation assays to assess: g) complement activation measured as complement C5a, h) leukocyte activation assay using granulocyte CD11b marker normalized to lipopolysaccharide, i) leukocyte loss assay, j) surface leukocyte density, k) IgG surface adsorption, l) complement C3b surface adsorption. Error bars equal ±1 standard deviation unit (n = 3). Statistical comparison of measurement shown in Fig. 4 and Figure SI5.

Fig. 4

Statistical analysis (student's t-test) of blood clotting cascade and immune system activation measurements on scaled up selected copolymer formulations under flow and quasi-static conditions. Individual datasets shown in Fig. 3. Squares coloured blue or red indicate samples where a significant (p < 0.05) decrease or increase was observed, respectively, compared to PTFE (comparison with other polymers is shown in Figure SI5). Grey squares indicate no significant difference. Squares coloured white were due to an error in the measurement acquisition for a particular sample. Flow or static conditions are indicated. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Murine foreign body (FB) infection with P. aeruginosa for testing the in vivo performance of the ‘hit’ copolymer. a) Luminescent images of the implantation site in live mice over 4 days for uncoated and iBnA-co-triEGMA (75:25) polymer coated silicone catheter segments inoculated with bioluminescent P. aeruginosa. The FBs were implanted subcutaneously. Light output from bacteria colonizing the implanted co-polymer coated segments in whole live mice was measured on days 0–4. After the mice were euthanized, the catheter segments were removed and both the surrounding tissues (day 4, ex vivo) and the implants (day 4, FB) imaged ex vivo. Inset: intensity scale (radiance) where red and blue refer to high and low light outputs respectively. Image dimensions = 16 × 16 mm. b) Quantification of light output (normalized radiance) from uncoated silicone (red) and poly(iBnA-co-triEGMA) polymer coated catheter segments (blue) for P. aeruginosa. Error bars show one standard deviation unit, N = 8. Significant differences (Student's unpaired t-test) are indicated as * = p < 0.1, ** = p < 0.05 and ***p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Histological analysis of tissue sections surrounding subcutaneously implanted silicone (a, c) and poly (iBnA-co-triEGMA) catheter segments (b, d) recovered from a-b) control (uninfected) mice and c-d) mice infected with P. aeruginosa. Tissue sections were stained from left to right with hematoxylin and eosin (general tissue morphology), Masson's trichome (collagen*), combined (all 5 stains), DNA (DAPI), lipids (FM1-163) and lectins (wheat germ lectin-Alexa 680 conjugate). Of particular note is the high level of lectin reactive staining (red) in the 2 lower righthand panels for c-d indicative of a strong cellular immune response due to the presence of bacteria compared with the sterile upper two right hand control panels (a–b). The insets in the ‘combined’ panel images show localized bacterial foci (c, white arrow) only in the infected mice with silicone implants. (d). There was reduced infiltration of reactive fibroblasts and immune cells to the sterile sites compared with the infection sites. Scale bar equals 50 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)