The unique structure of the cell wall binding domain of phage endolysin

Ľubica Urbániková1, Martina Gerová1, Jiří Brynda2, Július Košťan3, Nora Halgašová1 and Gabriela Bukovská1

1Institute of Molecular Biology SAS, Dubravska cesta 21, Bratislava, Slovak Republic

2Institute of Organic Chemistry and Biochemistry CAS, Flemmingovo namesti 542/2, Prague, Czech Republic

3Max F. Perutz Laboratories, University of Vienna, Vienna, Austria

lubica.urbanikova@savba.sk

Bacteriophage-encoded endolysins, enzymes showing bacteriolytic activity, are of growing interest for their applications as enzybiotics in veterinary and/or human medicines and various field of biotechnology, e.g. food safety. Bacteriophage BFK20 is a lytic phage of Brevibacterium flavum CCM 251 (gram positive corynebacteria), industrial producer of L-lysine. The genome of the bacteriophage BFK20 has been sequenced and analyzed (EMBL accession no. AJ278322) [1]. The gene product of ORF24´ was identified as endolysin gp24´ (UniProt ID Q9MBI0), an enzyme necessary for cell lysis and release of mature phage particles from the infected bacterial cells. The protein is composed of two domains, a catalytic domain exhibiting N-acetylmuramoyl-L-alanine amidase activity, and a cell wall binding domain (gp24BD), which are connected by a proline-rich linker. The individual domains were cloned separately and the cell wall binding capability of the C-terminal region (81 aa) was proved experimentally [2]. The whole protein as well as individual domains were crystallized, but only crystals of individual gp24BD were obtained. Needle shaped crystals belonging to hexagonal space group P622 were grown overnight. Crystals diffracted to 3.2 Å resolution using synchrotron source of radiation, but were twinned and not suitable for structure determination. Later, the crystals recrystallized directly in the crystallization drop. The newly obtained crystals diffracted to 1.4 Å resolution using home diffractometer. They belonged to tetragonal space group P4212. The same protein samples stored for several weeks at 4 °C crystallized directly in the form of tetragonal crystals. Tetragonal crystals were dissolved and the protein analysis showed random proteolysis and protein shortening by 8 amino acid residues.

The 1.4 Å resolution data set was used for structure solution by the direct method using the program ARCIMBOLDO-LITE [3]. The overall structure revealed very loose bundle of three a-helices. The asymmetric unit contains one protein molecule. The crystal symmetry gives four molecules in the unit cell forming very compact tetramer. Based on the PISA prediction the tetramer is stable also in solution. The oligomers have also been experimentally detected, thus one can speculate about the oligomeric state as biologically active unit. Closer inspection revealed the amphipathic nature of the helices and the tetrameric coiled-coil structure. An electron density was found in the tetramer cavity to which metal and chlorine ions were modelled. The coiled-coil structure is interrupted by a loop (one from each molecule). Between each two neighbour molecules a molecule of glycerol was identified sitting mostly at the loop. Molecular docking experiments performed using a rhamnose molecule supported the idea that glycerol maps the binding site which may be large enough to accommodate a substrate composed of several monomers. The gp24BD is unrelated to any of the known cell wall binding domains of phage endolysins by the amino acid sequence and also by the structure; this is the first evidence of the endolysin binding domain showing the coiled-coil structure. BFK20 phage endolysin binds to the cell walls of corynebacteria in a highly specific manner, unfortunately, the specific composition and structure of their cell walls and the exact peptidoglycan substrate of gp24BD have not yet been determined. To answer the questions concerning the functioning of the binding domain and the whole endolysin, further work is needed, especially the structure of the whole molecule and the complexes with possible ligands should be solved.

 

1. G. Bukovska, L. Klucar, C. Vlcek, J. Adamovic, J. Turna, J. Timko, Virology, 348, (2006), 57-71.

2. M. Gerova, N. Halgasova, J. Ugorcakova, G. Bukovska, FEMS Microbiol Lett., 321, (2011), 83-91.

3. D. Rodríguez, M. Sammito, K. Meindl, I. M. de Ilarduya, M. Potratz, G. M. Sheldrick and I. Usón, Acta Crystallogr. D, D68, (2012), 336-343.