Crosslink examples
Disulfide bridges
ProteomeXchange dataset PXD009460 contains data from a study of natural disulfide linkages. One file of data from Chicken Lysozyme (Lysozyme_2p2mJ_1minute.raw) was downloaded and processed in Mascot Distiller 2.7. Fragment masses were de-charged to 1+. SwissProt was searched using Mascot Server 2.7, with these settings:
Crosslinking : Disulfide bridge in Lysozyme Enzyme : Trypsin/P Variable modifications : Nethylmaleimide (C), Oxidation (M) Peptide mass tolerance : ± 20 ppm Fragment mass tolerance : ± 20 ppm Max missed cleavages : 2 Instrument type : ESI-TRAP
The crosslinking method, in XML format, looked like this. A reference section can be found here, but most items will be self evident.
<mxm:method description="" name="Disulfide bridge in Lysozyme" strategy="Brute-force">
<mxm:linkers>
<mxm:linker ModFileName="Xlink:Disulfide (C)"/>
</mxm:linkers>
<mxm:accessions>
<mxm:accession>LYSC_CHICK</mxm:accession>
</mxm:accessions>
<mxm:scope>
<mxm:parameter name="IntraLink">True</mxm:parameter>
<mxm:parameter name="LoopLink">True</mxm:parameter>
</mxm:scope>
<mxm:filters>
<mxm:parameter name="MinLen">2</mxm:parameter>
</mxm:filters>
</mxm:method>
You can view the result report here. Four disulfide bridges are annotated in the SwissProt entry for LYSC_CHICK. All are detected, although one is a weak match:
- FT DISULFID 24 145: Query 550, score 24 (weak match)
- FT DISULFID 48 133: Query 1118, score 101
- FT DISULFID 82 98: Query 1495, score 91 (bridge 4 reduced and alkylated)
- FT DISULFID 94 112: Query 1101, score 65 (bridge 3 reduced and alkylated)
The last two disulfides overlap one another, so that one must be broken to characterise the other. The study used a time course experiment with TCEP reduction and NEM alkylation to achieve this. This example uses just the first time point, where some spectra suggest that there may be a degree of scrambling. For example, the looplink in the match to query 1101 is clearly between 94 and 98, rather than the expected 94 and 112.
EDC
EDC is a carbodiimide crosslinker that forms links between carboxyl groups and primary amines. The ‘zero-length’ link involves elimination of water, so that the delta mass of the intact crosslink is -18 Da, and monolinks do not occur. PRIDE project PXD001538 contains EDC data from a study of the interaction of a pair of proteins: the HOP2-MND1 heterodimer in Arabidopsis thaliana.
The proteins were expressed in E. coli, alkylated with iodoacetamide, and digested with trypsin. Samples were analysed using HCD on a Thermo Q-Exactive plus. 20140603_QEx3_RSLC4_Rampler_Mechtler_IMP_shotgun_EDCMES12.raw was downloaded from PRIDE PXD001538. Mascot Distiller 2.7 was used for peak picking. Fragment masses were de-charged to 1+. The sequences of the heterodimer, a contaminants Fasta, and the E. coli sequences in SwissProt were searched using Mascot Server 2.7, as follows:
Crosslinking : PXD001538 EDC Enzyme : Trypsin/P Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Peptide mass tolerance : ± 10 ppm Fragment mass tolerance : ± 0.03 Da Max missed cleavages : 4 Instrument type : ESI-TRAP
The crosslinking method, in XML format, looked like this. A reference section can be found here, but most items will be self evident.
<mxm:method description="EDC zero-length crosslinks in PXD001538"
name="EDC MND1_ARATH+HOP2_ARATH" strategy="Brute-force">
<mxm:linkers>
<mxm:linker ModFileName="Xlink:EDC (K)">
<mxm:does_not_pair_with ModFileName="Xlink:EDC (K)" />
<mxm:does_not_pair_with ModFileName="Xlink:EDC (Protein N-term)" />
</mxm:linker>
<mxm:linker ModFileName="Xlink:EDC (Protein N-term)">
<mxm:does_not_pair_with ModFileName="Xlink:EDC (K)" />
<mxm:does_not_pair_with ModFileName="Xlink:EDC (Protein N-term)" />
</mxm:linker>
<mxm:linker ModFileName="Xlink:EDC (DE)">
<mxm:does_not_pair_with ModFileName="Xlink:EDC (DE)" />
<mxm:does_not_pair_with ModFileName="Xlink:EDC (Protein C-term)" />
</mxm:linker>
<mxm:linker ModFileName="Xlink:EDC (Protein C-term)">
<mxm:does_not_pair_with ModFileName="Xlink:EDC (DE)" />
<mxm:does_not_pair_with ModFileName="Xlink:EDC (Protein C-term)" />
</mxm:linker>
</mxm:linkers>
<mxm:accessions>
<mxm:accession>MND1_ARATH</mxm:accession>
<mxm:accession>HOP2_ARATH</mxm:accession>
</mxm:accessions>
<mxm:scope>
<mxm:parameter name="InterLink">True</mxm:parameter>
<mxm:parameter name="IntraLink">True</mxm:parameter>
<mxm:parameter name="LoopLink">True</mxm:parameter>
</mxm:scope>
</mxm:method>
The Linkers element for this particular chemistry is complicated because it is hetero-functional. EDC links carboxyl to amine but not carboxyl to carboxyl or amine to amine, and this has to be defined explicitly in the method. The Accessions element specifies that we will only look for links within and between the listed proteins. The Scope element specifies that we want to see interlinks, intralinks, and looplinks. The “Brute-force” strategy is an N^2 search for all possible pairs of peptides from the two Arabidopsis proteins.
You can view the result report here. The two Arabidopsis proteins are at the top of the report with large numbers of crossslink matches. Amongst the high scoring matches, take a look at the Peptide View report for query 22552, an interlinked pair. Fragmentation is relatively complete, giving a good level of confidence in both linkage sites.
DSS/BS3
Pride project PXD006131 contains data for DSS/BS3 crosslinking of some standard proteins, analysed using different fragmentation methods on an Orbitrap Fusion Lumos. EThcD data for Human serum albumin (B160326_14_Lumos_IN_LK_HSA-BS3_EThcd_Rep1.raw) was downloaded and peak picked using Mascot Distiller 2.7. Fragment masses were de-charged to 1+. SwissProt was searched using Mascot Server 2.7, and the following settings:
Crosslinking : HSA Xlink:DSS Enzyme : Trypsin/P Fixed modifications : Carbamidomethyl (C) Variable modifications : Oxidation (M) Peptide mass tolerance : ± 10 ppm Fragment mass tolerance : ± 20 ppm Max missed cleavages : 4 Instrument type : EThcD
The crosslinking method, in XML format, looked like this. A reference section can be found here, but most items will be self evident.
<mxm:method description="" name="HSA Xlink:DSS" strategy="Brute-force">
<mxm:linkers>
<mxm:linker ModFileName="Xlink:DSS (K)">
<mxm:monolink>A</mxm:monolink>
<mxm:monolink>W</mxm:monolink>
</mxm:linker>
<mxm:linker ModFileName="Xlink:DSS (Protein N-term)">
<mxm:monolink>A</mxm:monolink>
<mxm:monolink>W</mxm:monolink>
</mxm:linker>
</mxm:linkers>
<mxm:accessions>
<mxm:accession>ALBU_HUMAN</mxm:accession>
</mxm:accessions>
<mxm:scope>
<mxm:parameter name="InterLink">True</mxm:parameter>
<mxm:parameter name="IntraLink">True</mxm:parameter>
<mxm:parameter name="LoopLink">True</mxm:parameter>
</mxm:scope>
<mxm:filters>
<mxm:parameter name="MinLen">2</mxm:parameter>
</mxm:filters>
</mxm:method>
You can view the result report here. Amongst the high scoring matches for crosslinked pairs of peptides, take a look at the Peptide View report for query 28141. Fragmentation is relatively complete, giving a good level of confidence in both linkage sites.
