Crosslinking mass spectrometry (XL-MS) is a technique for identifying protein-protein interactions. It can be applied in vitro and in vivo. It uses chemically reactive, MS-cleavable crosslinkers to covalently link interacting proteins. Proceeding digestion and enrichment, crosslinked peptides are then subjected to LC-MS/MS. The major advantage of MS-cleavable crosslinkers is their capacity to be fragmented, yielding two peptides. On further fragmentation identification of the two peptides, and hence crosslinked proteins, is achieved.
To explore the requirements of XL-MS for investigating protein-protein interactions, two current XL-MS methods were compared. The first, from the lab of Albert Heck, which uses DSSO and XlinkX (Lui, 2015) and second the lab of Andrea Sinz using BuUrBu and MeroX (Arlt, 2015). Three aspects of each method were compared: fragmentation, coverage and data analysis. Subsequently, three experiments were undertaken. Firstly, fragmentation analysis at CID 10-70eV using two peptides of known sequence was performed. Secondly, crosslinking analysis was undertaken using two proteins known to interact, and thirdly a complex mixture of 100-300 proteins.
Results from the first experiments indicate relative fragmentation of the crosslinker preferentially occurs and starts at 25 eV, whereas relative peptide backbone fragmentation does not peak until 30-35 eV, depending on the peptide. These results were then applied to subsequent experiments. Two proteins, Npl3 and Hmt1, were successfully crosslinked, with greater than 100 crosslinks identified. A mixture of yeast proteins were then crosslinked and analysed by MeroX and XlinkX. This resulted in 109 and 30 crosslinks being identified, respectively.
Both methods successfully identified crosslinked proteins, however, the number and type of crosslinks discovered differed between the techniques. MeroX identifies inter-, intra- and dead-end crosslinks between serine, threonine and lysine, whereas XlinkX identifies inter-peptide links between lysines. Our results indicate that different scales of XL-MS experiments will influence the choice of crosslinker and analysis software.