Signal Attrition in Whole Cell Cross-Linking Mass Spectrometry

Cross-linking mass spectrometry (XL-MS) could replace
traditional
techniques for sampling the cellular interactome, such as affinity
pulldown MS. It generates superior interaction data that can be used
to better model cellular structure and function. However, the sampling
depth of current XL-MS techniques is disappointing. Poor sampling
is often blamed on a low-yielding cross-linking reaction, estimated
to be ≤0.1% based on relative ion abundance in the mass spectrometer.
Here, using a new two-step cross-linker installation process, we demonstrate
that the yield of the chemical reaction is not the limiting factor
in sampling the interactome. Low cross-link detection levels persist
even when cross-link yields approach 30% of total peptide. We show
that cross-linked peptides are not preferentially lost during sample
workup; they enter the mass spectrometer at least as efficiently as
linear peptides. Low detection rates arise mostly from severe signal
splitting that reduces the S/N of cross-linked peptides, which is
likely exacerbated by low-sensitivity database search tools and compounded
by ion suppression.