Back

Identification of the receptor-binding protein of Clostridium difficile phage CDHS-1 reveals a new class of receptor-binding domains.

Dowah, A. S. A.; Xia, G.; Thanki, A. M.; Ali, A. A. K.; Shan, J.; Wallis, R.; Clokie, M. R. J.

2021-07-05 molecular biology
10.1101/2021.07.05.451159 bioRxiv
Show abstract

As natural bacterial predators, bacteriophages have the potential to be developed to tackle antimicrobial resistance, but our exploitation of them is limited by understanding their vast uncharacterised genetic diversity1,2. Fascinatingly, this genetic diversity reflects many ways that phages can make proteins, performing similar functions that together form the familiar phage particle. Critical to infection are phage receptor-binding proteins (RBPs) that bind bacterial receptors and initiate bacterial entry3. Here we identified and characterised Gp22, a novel RBP for phage CDHS-1 that infects pathogenic C. difficile, but that had no recognisable RBPs. We showed that Gp22 antibodies neutralised CDHS-1 infection and used immunogold-labelling and transmission electron microscopy to identify their location on the capsid. The Gp22 three-dimensional structure was resolved by X-ray crystallography revealing a new RBP class with an N-terminal L-shaped -helical superhelix domain and a C-terminal Mg2+-binding domain. The findings provide novel insights into C. difficile phage biology and phage-host interactions. This will facilitate optimal phage development and future engineering strategies4,5. Furthermore, the AlphaFold2-predicted Gp22 structure, which was strikingly accurate, paves the way for a structurome based transformation and guidance of future phage studies where many proteins lack sequence homology but have recognisable protein structures.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.