Abstrakt
We have developed a de novo sequencing software tool (CYCLONE) and applied it for determination of cyclic peptides. The program uses a non-redundant database of 312 nonribosomal building blocks identified to date in bacteria and fungi (more than 230 additional residues in the database list were isobaric).
The software was used to fully characterize the tandem mass spectrum of several cyclic peptides and provide sequence tags. The general strategy of the script was based on fragment ion pre-characterization to accomplish unambiguous b-ion series assignments.
Showcase examples were a cyclic tetradepsipeptide beauverolide, a cyclic hexadepsipeptide roseotoxin A, a lasso-like hexapeptide pseudacyclin A, and a cyclic undecapeptide cyclosporin A. The extent of ion scrambling in smaller peptides was as low as 5 % of total ion current; this demonstrated the feasibility of CYCLONE de novo sequencing.
The robustness of the script was also tested against database sets of various sizes and isotope-containing data. It can be downloaded from the http://ms.biomed.cas.cz/MSTools/ website.