New Tools for Proteome-Wide Cross-Linking Mass Spectrometry

This project deals with the establishment of methods and techniques to investigate protein-protein interaction networks within living cells. Such complex networks are essential for all biochemical processes within a cell and their regulation. To deepen knowledge on protein crosstalk is therefore one of the most important questions to answer in basic research. To visualize such a crosstalk, chemical reagents called cross-linkers are used to selectively connect and stabilize proteins at positions in close proximity. After that, proteins can be isolated and cut into pieces using an enzyme. With the help of a mass- spectrometer the proteins can be identified, and the exact position of their interaction can be localized. This works by detection of a fingerprint resulting from the cut protein parts. Unfortunately, the chemical reaction efficiency of such cross-linkers is quite low, hampering a direct detection of those within the mass-spectrometer. To overcome this issue, we will design, develop and test novel cross-linker reagents with improved reactivity. Reactive sidechains attached to the cross-linkers will furthermore enable a selective enrichment. By combination of these strategies, we aim to obtain purified samples in sufficient amounts, in which the cross-link detection works effectively. In the next step, the recorded data has to be analyzed with the help of an algorithm. To do so, we already developed software that can identify more cross-links while being less prone to errors compared to competitors. During this project, the software will be further advanced, and a machine-learning algorithm will improve the exact localization of the cross-link on the protein. The functionality will be additionally expanded to several types of cross-linkers. In a final project step, we will apply the established holistic workflow to investigate differences in protein-protein interactions between areas of the cell, where the genetic material is condensed vs. better accessible. This research work will significantly deepen the knowledge about how genetic material is read within cells.

Our project has significantly advanced crosslinking mass spectrometry (XL-MS), a technology not only relevant to the field of structural biology but further revealing how proteins interact inside cells. These interactions form the molecular networks that govern life processes. Understanding them is key to tackling diseases like cancer and neurodegeneration. Traditional XL-MS methods struggle with efficiency, robustness, and coverage. We set out to overcome these limitations by creating new chemical tools, enrichment strategies, and computational solutions. Major achievements: Innovative chemical linkers: We developed and optimized novel reagents such as DISPASO, which rapidly enters cells and nuclei within minutes, enabling efficient in vivo crosslinking. We also introduced DSSO-carbamate, a more stable alternative to existing linkers, and initiated work on photocleavable enrichment sites. These innovations allow researchers to capture protein interactions previously inaccessible. Breakthrough in coverage: By refining our workflows, we increased the number of identified crosslinked sites from 1,000 in 2023 to over 5,000 in 2025, mapping interaction networks of nearly 400 proteins and discovering 56 new connections. This includes insights into RNA biology, such as the structural dynamics of DDX39 proteins, and supports (re-)modelling of AI predicted 3D protein structures. Low-input capability: We demonstrated successful crosslinking from as little as 25 nanograms of protein-comparable to results previously requiring 40 times more material-paving the way for studies on scarce samples. Computational advances: Our partner group released MS Annika 3.0, a next-generation search engine for XL-MS data. It introduces a novel algorithm for non-cleavable crosslinkers, enabling proteome-wide studies that were previously impossible. MS Annika 3.0 outperforms competing tools in accuracy and speed, reducing false positives by 75% and supporting GPU acceleration for high-throughput analysis. It is open source and integrated into widely used platforms, ensuring global accessibility. Collaborative impact: Together, we applied these tools to complex biological systems, such as the C. elegans Box C/D RNP complex, and published multiple open-access studies. Our work has set new standards for XL-MS, combining chemistry, biology, and AI-driven computation. These advances transform XL-MS into a powerful tool for exploring protein networks at unprecedented depth and scale. This is of high importance for understanding disease mechanisms and accelerating drug discovery. By making our workflows and software openly available, we empower researchers worldwide to push the boundaries of structural biology.