Dismantling superbugs: engineering third-generation lysins as enzybiotics for burn wound infections

Burn wounds remain frequently occurring injuries around the world. The World Health Organization (WHO) estimates 11 million burn injuries occur annually. For these complicated injuries to heal properly, a delicate balance between a pro- and anti-inflammatory response is required, making them challenging to treat. This delicate balance is frequently disrupted by opportunistic pathogens, and can lead to life-threatening infections. Prominent burn wound infecting pathogens, including Staphylococcus aureus and Acinetobacter baumannii, are known for their capacity to persist in (burn) wounds, reflected in biofilm formation. Moreover, these organisms are shortlisted by the WHO as priority pathogens for which new antibacterial approaches are urgently needed in the light of the antibacterial resistance crisis. Policy makers state that this crisis can be solved by reducing the use of the current antibiotics on the one hand and by supplying new antibacterial strategies on the other hand.

Bacteriophages, viruses specifically infecting bacteria, have been engaging in this evolutionary arms race for ages. From this perspective, bacteriophages encode a broad range of co-evolved antibacterial proteins, which could be used as a source of bacteriophage-inspired antibacterials. One class of promising antibacterial proteins are found in endolysins. These peptidoglycan-degrading enzymes are expressed at the end of the lytic infection cycle, where they induce host lysis to release newly formed bacteriophage particles. Exogenous application of endolysins to Gram-positive bacteria resulted in rapid lysis of a bacterial culture, revealing its potential as alternative antibiotic. Two decades later, this first generation of natural endolysins has entered a phase III clinical trial targeting S. aureus. In Gram-negative pathogens, protein engineering has enabled targeting pathogens, including Pseudomonas aeruginosa and A. baumannii. These domain swapping and other engineering efforts resulted in lysins with superior antibacterial activity or improved biochemical properties that have been defined as a second generation of lysins.

However, several challenges in lysin engineering remain. Therefore, it might be interesting to tailor the lysin to address challenges in its final clinical application. Engineering strategies to tackle such hurdles are labeled as third-generation engineering strategies. Therefore, reliable models including these hurdles are of potential interest as well. As a result, the first part of this dissertation describes an ex vivo burn wound model to assess experimental antibacterial therapies. Where most of these models rely on time-consuming CFU-counting or qPCR methods with data points on discrete time points, we integrate the use of bioluminescent strains in this model. This implementation increased throughput and time resolution, allowing insights into bacterial repopulation of the explant following antibacterial treatment. Moreover, hands-on time was reduced significantly. To expand this model for any Gram-negative, clinical isolate of interest, a Tn7-based tagging system was developed, enabling site-specific integration of a reporter gene in the chromosome of the desired strain. This system was used to tag A. baumannii NCTC 13423 with the luxCDABE operon. Site-specific integration was validated using Nanopore sequencing and de novo genome assembly. This strain was subsequently validated in the ex vivo model, representing a new tool for the characterization of novel (phage-inspired) antibacterials targeting burn wound infections.

The second part of this dissertation explores a novel third-generation concept in engineering lysins for the treatment of burn wound infections. As these types of infections are often characterized by an impaired immune response and biofilm formation, a lysin fusion with peptides modulating inflammation and the ability to disrupt biofilms was pursued. To illustrate this concept, targeted VersaTile libraries were created using 14 different peptides and one previously characterized, engineered lysin LysRODI∆amidase that targets S. aureus. Despite the identification of four hits using three independent muralytic assays, fragmentation of these fusion proteins was observed to be a major hurdle in this engineering concept. Screening of a different proof-of-concept library starting from the previously engineered lysin 1D10, targeting A. baumannii, revealed three clear hits of which one was considered to be a stable lysin without major fragmentation.

Characterization of this lead variant, BZAb1, resulted in superior bacterial killing in a time-kill assay with respect to its parental lysin, 1D10. On the other hand, no superiority was observed when studied in the ex vivo model of burn wound infection. Therefore, it might be interesting to screen libraries in this model in the future to select for hits outperforming the parental lysin in these specific settings. Moreover, this lysin was characterized as a narrow-spectrum lysin as only growth of A. baumannii strains could be inhibited at low concentrations. Concerning third-generation properties, BZAb1, did not eradicate 24h old in vitro biofilms, and also no anti-inflammatory effect could be observed in RAW264.7 macrophages. However, an anti-inflammatory clue could be confirmed as this lead variant was able to bind LPS in vitro. Lastly, no cytotoxicity was observed in HeLa cells. As a result, the first third-generation lysin targeting a Gram-negative pathogen is described in this dissertation.

This second part illustrates the possibility to include additional properties into a single fusion protein. On the other hand, these proof-of-concept libraries involving synthetic linkers also revealed that these fusion proteins might be very fragile. As a result, one could opt not to use linkers or to use naturally occurring linkers to reduce autoproteolysis. Additionally, innovative coformulation strategies of these active compounds with anti-inflammatory compounds might result in a similar outcome.

In conclusion, third-generation lysin engineering might be more successful in the future when choosing less toxic fusion partners increasing protein yield. Together with innovative approaches in other developmental areas, lysins will further fuel the transition to sustainable antibacterial therapies in the future to overcome the silent pandemic of antibacterial resistance.