Comprehensive identification of crosslinked peptides using a multi-crosslinker, fragmentation and data analysis approach

Smith DL1, Gotze M2, Hart-Smith G1 and Wilkins MR1

  1. Systems Biology Initiative, School of Biotechnology and Biomolecular Sciences, UNSW Sydney, NSW 2052, Australi.
  2. Institute of Biochemistry, Martin Luther University Halle-Wittenberg, Kurt- Mothes-Str. 3, D-06120 Halle (Saale), Germany.

Crosslinking mass spectrometry (XLMS) is a method to study protein-protein interactions. It combines chemical crosslinkers, protease digestion and tandem mass spectrometry, whereby fragmentation of the crosslinked peptides is used to discover intra- and inter-protein crosslinks. Early crosslinking studies used non-mass spectrometry cleavable crosslinkers such as BS3, BS2G and DTTSP. However, new mass spectrometry-cleavable linkers, such as DSSO, DSBU have helped address many of the prior limitations of crosslinking experiments, such as co-fragmentation of crosslinked peptides. They allow predictable generation of high-intensity reporter ions which new generation crosslink identification programs such as MeroX and XlinkX can use to increase speed, and confidence of identifications. Here, we have used a multi-crosslinker, fragmentation and program approach to address two major aims. Firstly, we investigated the enzyme/substrate interaction between Npl3p and its methyltransferase Hmt1p. Secondly, we used this combined analysis approach to understand how different crosslinkers (DSBU/DSSO), fragmentations (CID+ETD/SteppedHCD), programs (MeroX/XlinkX 2.0), and algorithms (Precursor and Reporter-Ion) impacted the results. From this study we have defined that the interaction between Npl3p and Hmt1p involves the intrinsically disordered “SRGG” region of Npl3p and the N-, C- termini and S-adenosyl methionine binding site of Hmt1p. For the first time, we have also shown direct evidence for Npl3p dimerization occurring at the “SRGG” region. We have also confirmed higher order multimerisation of Hmt1p in accordance to the known structure. Importantly, we demonstrate from our multi-crosslinker, fragmentation and program comparisons that one-type of analyses may not capture all identifiable crosslinks. This is attributed to a combination of biases inherent to either a MeroX/DSBU approach or an XlinkX/DSSO approach.