An in vitro collagen perfusion wound biofilm model; with applications for antimicrobial studies and microbial metabolomics

• Main Authors: Elisabeth A. Slade, Robin M. S. Thorn, Amber Young, Darren M. Reynolds
• Format: Article
• Language: English
• Published: BMC 2019-12-01
• Series: BMC Microbiology
• Subjects: Biofilm; Collagen; Wound; In vitro model; Volatile metabolite; Pseudomonas aeruginosa
• Online Access: https://doi.org/10.1186/s12866-019-1682-5

Abstract Background The majority of in vitro studies of medically relevant biofilms involve the development of biofilm on an inanimate solid surface. However, infection in vivo consists of biofilm growth on, or suspended within, the semi-solid matrix of the tissue, whereby current models do not effectively simulate the nature of the in vivo environment. This paper describes development of an in vitro method for culturing wound associated microorganisms in a system that combines a semi-solid collagen gel matrix with continuous flow of simulated wound fluid. This enables culture of wound associated reproducible steady state biofilms under conditions that more closely simulate the dynamic wound environment. To demonstrate the use of this model the antimicrobial kinetics of ceftazidime, against both mature and developing Pseudomonas aeruginosa biofilms, was assessed. In addition, we have shown the potential application of this model system for investigating microbial metabolomics by employing selected ion flow tube mass spectrometry (SIFT-MS) to monitor ammonia and hydrogen cyanide production by Pseudomonas aeruginosa biofilms in real-time. Results The collagen wound biofilm model facilitates growth of steady-state reproducible Pseudomonas aeruginosa biofilms under wound like conditions. A maximum biofilm density of 1010 cfu slide− 1 was achieved by 30 h of continuous culture and maintained throughout the remainder of the experiment. Treatment with ceftazidime at a clinically relevant dose resulted in a 1.2–1.6 log reduction in biofilm density at 72 h compared to untreated controls. Treatment resulted in loss of complex biofilm architecture and morphological changes to bacterial cells, visualised using confocal microscopy. When monitoring the biofilms using SIFT-MS, ammonia and hydrogen cyanide levels peaked at 12 h at 2273 ppb (±826.4) and 138 ppb (±49.1) respectively and were detectable throughout experimentation. Conclusions The collagen wound biofilm model has been developed to facilitate growth of reproducible biofilms under wound-like conditions. We have successfully used this method to: (1) evaluate antimicrobial efficacy and kinetics, clearly demonstrating the development of antimicrobial tolerance in biofilm cultures; (2) characterise volatile metabolite production by P. aeruginosa biofilms, demonstrating the potential use of this method in metabolomics studies.