Bacterial biofilms in infective endocarditis: an in vitro model to investigate emerging technologies of antimicrobial cardiovascular device coatings

Abstract

Objective: In spite of the progress in antimicrobial and surgical therapy, infective endocarditis (IE) is still associated with a high morbidity and mortality. IE is characterized by bacterial biofilms of the endocardium, especially of the aortic and mitral valve leading to their destruction. About one quarter of patients with formal surgery indication cannot undergo surgery. This group of patients needs further options of therapy, but due to a lack of models for IE prospects of research are low. Therefore, the purpose of this project was to establish an in vitro model of infective endocarditis to allow growth of bacterial biofilms on porcine aortic valves, serving as baseline for further research.

Methods and results: A pulsatile two-chamber circulation model was constructed that kept native porcine aortic valves under sterile, physiologic hemodynamic and temperature conditions. To create biofilms on porcine aortic valves the system was inoculated with Staphylococcus epidermidis PIA 8400. Aortic roots were incubated in the model for increasing periods of time (24 h and 40 h) and bacterial titration (1.5 × 10⁴ CFU/mL and 1.5 × 10⁵ CFU/mL) with 5 L cardiac output per minute. After incubation, tissue sections were analysed by fluorescence in situ hybridization (FISH) for direct visualization of the biofilms. Pilot tests for biofilm growth showed monospecies colonization consisting of cocci with time- and inocula-dependent increase after 24 h and 40 h (n = 4). In n = 3 experiments for 24 h, with the same inocula, FISH visualized biofilms with ribosome-containing, and thus metabolic active cocci, tissue infiltration and similar colonization pattern as observed by the FISH in human IE heart valves infected by S. epidermidis.

Conclusion: These results demonstrate the establishment of a novel in vitro model for bacterial biofilm growth on porcine aortic roots mimicking IE. The model will allow to identify predilection sites of valves for bacterial adhesion and biofilm growth and it may serve as baseline for further research on IE therapy and prevention, e.g. the development of antimicrobial transcatheter approaches to IE.

Keywords: Biofilm; Bioreactor; Fluorescence in situ hybridization; In vitro model; Infective endocarditis; Staphylococcus epidermidis.

Fig. 1

Endocarditis bioreactor model. Pulsatile circulatory model implementing physiologic temperature and in vivo pulsatile flow conditions. Porcine valves are positioned between the aortic and ventricular compartment. The model allows bacterial growth on the heart valves and reproduces hydrodynamic factors and physical shear stress comparable to biofilm formation conditions in humans in vivo

Fig. 2

Localization sites of harvested samples after biofilm growth. (1) inter leaflet triangle, (2) commissure, (3) lateral leaflet cross-section, (4) middle leaflet cross-section

Fig. 3

FISH of tissue sections from ventricular surface of aortic valve leaflets. FISH analysis of tissue sections from ventricular surface of aortic valve leaflets incubated with S. epidermidis for 24 h in the bioreactor. FISH was performed using the pan-bacterial probe EUB338-Cy3 (yellow) and nucleic-acid-specific DAPI stain. Unspecific binding was excluded, using the nonsense probe NON338 (data not shown). (A1) A pattern of the aortic valve leaflet is included in the overview of location A at the beginning of the leaflet (bar = 100 µm). (A2) Higher resolution of the inset in A1. Overlay of all channels shows a small biofilm with EUB338 positive bacteria (yellow) (bar = 10 µm). (A3) Identical microscopic field of the DAPI channel in black (bar = 10 µm). (B1) Overview of location B at the middle of the leaflet (bar = 100 µm). (B2) At higher magnification FISH shows an active biofilm at the border of the tissue (bar = 10 µm). (B3) DAPI staining in black (bar = 10 µm). (C1): Overview of the location C in the endocardium (bar = 100 µm). (C2) Higher magnification of inset C showing massive infiltrating biofilms (bar = 10 µm). (C3) DAPI channel in black (bar = 10 µm)

Fig. 4

In vitro colonization corresponds to clinical human IE findings. FISH analysis of a human heart valve from a patient with S. epidermidis IE (a) compared to a porcine heart valve infected by S. epidermidis in the in vitro endocarditis bioreactor for 24 h (b). FISH using the pan-bacterial probe EUB338-Cy3 (yellow) and unspecific nucleic acid stain DAPI. Unspecific binding was excluded, using the nonsense probe NON338. (A1) Overview of the human heart valve tissue (bar = 100 µm). (A2) Higher magnification of inset (A1) showing an overlay of all channels, reveals a biofilm with EUB338-Cy3 positive cells (bar = 10 µm). (B1) Overview of the porcine leaflet tissue (bar = 100 µm). (B2) At higher magnification biofilms in the tissue appeared with strong EUB338-Cy3 FISH signals (bar = 10 µm). (A3, B3) Note the similar pattern in black and white images of the single channel DAPI, highlighting the biofilm formation (bar = 10 µm)