Resumen de Structure and function of salmonella virus epsilon15 and campylobacter virus F358 tailspikes
Bacteriophages are the most numerous organisms on Earth. They infect bacteria, including pathogenic ones, and have a huge influence on ecological processes. In the future, they can help to treat diseases caused by antibiotic-resistant bacteria. Phages use their receptor-binding proteins to adsorb to their receptors in bacteria. Consequently, receptor-binding proteins control the phage host range.
In my thesis, I focused on the Salmonella virus epsilon15 receptor-binding proteins, gp20, and on one of the four Campylobacter virus F358 receptor-binding proteins, RBP3. These phages infect Salmonella enterica serovar Anatum and Campylobacter jejuni, respectively. They are the two bacteria causing most of the food-borne diseases to humans in Europe. The epsilon15 receptor is the O-antigen of the lipopolysaccharide (LPS). It consists of repetitions of the trisaccharide D-Galp[6Ac]-α-1 – 6-D-Manp-β-1 – 4-L-Rhap-α-1 – ROH linked by α-1 – 3 bonds.
During my thesis, I determined the structure of gp20, which is a trimeric tailspike, with and without its receptor. Gp20ΔN is composed of three domains. From the N- to the C-termini, the β-helix domain, the β-sandwich domain and the petal domain. The β-helix domain is a right-handed 12-rung helix made of β-strands. Each rung has three β-strands linked by turns. Some of the turns have loops that create a groove facing outwards. The β-sandwich domain is composed of a six-stranded and a five-stranded antiparallel β-sheets. This is the most distal domain in gp20. A backward-running linker connects the β-sandwich with the petal domain. The petal domain has an α/β hydrolase fold with a β-barrel inserted in one of its loops. These two subdomains form a groove parallel to the one of the β-helix domain.
Salmonella enterica serovar Anatum O-antigen oligosaccharides bind to four regions of gp20; one to the petal domain, inside the groove; another to the β-sandwich domain and two to the β-helix domain. The two β-helix domain oligosaccharides are separated by the endorhamnosidase site. It is located in a negatively charged area, inside the β-helix groove. It is formed by two aspartic acids that probably act following the retaining mechanism, common to other glycosidases. This activity clears space to allow the virus to approach the bacterial membrane. The petal domain contains a putative esterase site, although it is unknown if this putative esterase site has activity. Removing the acetyl group bound to the galactose might reduce the interactions between uncut lipopolysaccharide chains and also help the virus approach the membrane.
I also determined the structure of the Campylobacter virus F358 receptor-binding protein RBP3. It crystallised both as a monomer and as a trimer. Each RBP3 monomer is composed of a β-helix domain. It is shorter than homologous trimeric tailspikes and monomeric β-helical bacterial enzymes. This feature might be related to the existence of monomeric and trimeric states. The trimeric RBP3 has a negatively charged area in its interchain grooves. It resembles the catalytic sites of other tailspikes.
In summary, I present structural and functional findings of Salmonella and Campylobacter bacteriophage receptor-binding proteins. This knowledge will be useful to design engineered bacteriophages for the application of tailor-made phage therapy and other phage-based applications
