Introduction. Staphylococcus epidermidis is predominant in implant-associated infections due to its capability to form biofilms. It can deploy several strategies for biofilm development using either polysaccharide intercellular adhesin (PIA), extracellular DNA (eDNA) and/or proteins, such as the extracellular matrix-binding protein (Embp). Hypothesis/Gap Statement. We hypothesize that the dichotomic regulation of S. epidermidis adhesins is linked to whether it is inside a host or not, and that in vitro biofilm investigations in laboratory media may not reflect actual biofilms in vivo. Aim. We address the importance of PIA and Embp in biofilm grown in ‘humanized’ media to understand if these components play different roles in biofilm formation under conditions where bacteria can incorporate host proteins in the biofilm matrix. Methodology. S. epidermidis 1585 WT (deficient in icaADBC), and derivative strains that either lack embp, express embp from an inducible promotor, or express icaADBC from a plasmid, were cultivated in standard laboratory media, or in media with human plasma or serum. The amount, structure, elasticity and antimicrobial penetration of biofilms was quantified to describe structural differences caused by the different matrix components and growth conditions. Finally, we quantified the initiation of biofilms as suspended aggregates in response to host factors to determine how quickly the cells aggregate in response to the host environment and reach a size that protects them from phagocytosis. Results. S. epidermidis 1585 required polysaccharides to form biofilm in laboratory media. However, these observations were not representative of the biofilm phenotype in the presence of human plasma. If human plasma were present, polysaccharides and Embp were redundant for biofilm formation. Biofilms formed in human plasma were loosely attached and existed mostly as suspended aggregates. Aggregation occurred after 2 h of exposing cells to plasma or serum. Despite stark differences in the amount and composition of biofilms formed by polysaccharide-producing and Embp-producing strains in different media, there were no differences in vancomycin penetration or susceptibility. Conclusion. We suggest that the assumed importance of polysaccharides for biofilm formation is an artefact from studying biofilms in laboratory media void of human matrix components. The cell–cell aggregation of S. epidermidis can be activated by host factors without relying on either of the major adhesins, PIA and Embp, indicating a need to revisit the basic question of how S. epidermidis deploys self-produced and host-derived matrix components to form antibiotic-tolerant biofilms in vivo.

The treatment of implant‐associated bacterial infections and biofilms is an urgent medical need and a grand challenge because biofilms protect bacteria from the immune system and harbor antibiotic‐tolerant persister cells. This need is addressed herein through an engineering of antibody‐drug conjugates (ADCs) that contain an anti‐neoplastic drug mitomycin C, which is also a potent antimicrobial against biofilms. The ADCs designed herein release the conjugated drug without cell entry, via a novel mechanism of drug release which likely involves an interaction of ADC with the thiols on the bacterial cell surface. ADCs targeted toward bacteria are superior by the afforded antimicrobial effects compared to the non‐specific counterpart, in suspension and within biofilms, in vitro, and in an implant‐associated murine osteomyelitis model in vivo. The results are important in developing ADC for a new area of application with a significant translational potential, and in addressing an urgent medical need of designing a treatment of bacterial biofilms.

Implant-associated infections remain a grand unmet medical need because they involve biofilms that protect bacteria from the immune system and harbour antibiotic-tolerant persister cells. There is an urgent need for new biofilm-targeting therapies with antimicrobials, to treat these infections via a non-surgical way. In this work, we address this urgent medical need and engineer antibody-drug conjugates (ADC) that kill bacteria in suspension and in biofilms, in vitro and in vivo. The ADC contains an anti-neoplastic drug mitomycin C, which is also a potent antimicrobial against biofilms. While most ADCs are clinically validated as anti-cancer therapeutics where the drug is released after internalisation of the ADC in the target cell, the ADCs designed herein release the conjugated drug without cell entry. This is achieved with a novel mechanism of drug, which likely involves an interaction of ADC with thiols on the bacterial cell surface. ADC targeted towards bacteria were superior by the afforded antimicrobial effects compared to the non-specific counterpart, in suspension and within biofilms, in vitro and in vivo. An implant-associated murine osteomyelitis model was then used to demonstrate the ability of the antibody to reach the infection, and the superior antimicrobial efficacy compared to standard antibiotic treatment in vivo. Our results illustrate the development of ADCs into a new area of application with a significant translational potential.