Elucidating the mechanisms of action and evolutionary relationships of phage anti-defence proteins

Abstract

Bacteriophages (phages) are viruses that infect bacteria. As such, phages and bacteria are locked in an ongoing evolutionary arms race, driving the emergence of diverse bacterial defence systems that protect bacteria against phages, and corresponding phage-encoded anti-defence proteins (ADPs) that overcome these systems. Understanding the mechanisms of this co-evolutionary dynamic has had profound implications across different domains of life. For example, many components of eukaryotic innate immunity evolved from bacterial defence systems, forming a pool of so-called “ancestral immunity”. While many ADPs are now known, the mechanisms of action (MoA) for the majority remain uncharacterised. Among ADPs with known mechanisms, direct binders are the most frequent category. In this thesis, I demonstrate that many ADPs with uncharacterised MoA are likely to also function via direct binding to bacterial defence proteins. I predicted direct interactions between 20 such ADPs and their corresponding bacterial defence proteins. Structural analysis of these predictions provided mechanistic insights and aligned well with available experimental data. For instance, AcrIB4 shares homology with the C-terminal region of Cas8b, initially suggesting it might substitute for Cas8b in the complex, but the experimental data showed it does not affect incorporation of Cas8b into the cascade. Instead, my predictions suggest that AcrIB4 binds to Cas8b at the site normally occupied by Cas11. As Cas11 is translated from the 3' end of the cas8b gene and is homologous to AcrIB4, it is likely that AcrIB4 replaces Cas11, thereby destabilising the Cascade complex. Interestingly, some ADPs appear to inhibit one bacterial defence system while activating another. I showed that such ADPs are enriched in DNA mimics, suggesting their broad-spectrum activity against various nucleic acid-binding defence proteins may exert evolutionary pressure on bacteria to evolve counter-counter-adaptations. This hypothesis is supported by protein structure predictions, which revealed numerous high-confidence interactions between these ADPs and bacterial defence systems. I identified structural and functional homologs of phage ADPs in eukaryotic viruses, targeting defence systems that are evolutionary related in bacterial and eukaryotes. This suggests that immune evasion strategies may be conserved across bacterial and eukaryotic viruses. For example, I found functional homologs of phage NTase enzymes that inhibit bacterial STING by producing competing oligonucleotides. Given that the eukaryotic cGAS–STING pathway senses viral DNA in the cytosol and activates immune responses via cyclic oligonucleotide synthesis, similar competitive mechanism could plausibly inhibit its activity in eukaryotes as well. Current ADP databases capture only a small fraction of the known diversity and often lack structural and mechanistic information. To address this gap, I developed EVADES, a comprehensive ADP resource that integrates experimental and predicted protein structures, family annotations, and mechanistic classifications. Together, these findings underscore the evolutionary depth and functional complexity of phage–host interactions, with far-reaching implications for both microbial ecology and immunity research.