Röst, H. | bioRxiv Science

bioRxiv2025

Dynamic Proximal Interactomics and Chemical Genetic Screening Reveal CCR4-NOT Sequestration in Stress Granules as a Mechanism for Transcript Stabilization

Cells adapt to stress by rewiring their post-transcriptional gene regulation. Stress granules--biomolecular condensates composed of polyadenylated RNAs and RNA-binding proteins--are implicated in this process, yet their precise functional roles remain debated. To address this, we mapped the dynamic proteomic landscapes of stress granules formed under oxidative and hyperosmotic stress using multi-bait BioID proximity profiling coupled with quantitative mass spectrometry. This analysis revealed context-specific remodeling of proximal interaction networks and identified a conserved, stress-dependent shift in association with the CCR4-NOT deadenylase complex. A complementary genome-wide chemical genetic screen further implicated CCR4-NOT in stress granule biology, showing that reduced CCR4-NOT activity bypassed lipoamide-mediated inhibition of stress granule assembly. Microscopy showed sequestration of the CCR4-NOT complex into stress granules, and global transcriptomic analyses revealed that this relocalization promotes poly(A) tail lengthening and increased abundance of stress-induced survival transcripts. Together, integration of proteomics, chemical genetics, and transcriptomics uncovers a spatial mechanism by which stress granule assembly promotes cellular adaptation to stress through sequestration of CCR4-NOT from the cytosol and consequent remodeling of post-transcriptional regulation.

cell biology↗

bioRxiv2022

Comparing Machine Learning Architectures for the Prediction of Peptide Collisional Cross Section

1Mass spectrometry is the method of choice in large-scale proteomics studies. One common method is data-independent acquisition (DIA), which allows for high-throughput analysis of biological samples, but also produces complex data. Methods of peptide separation, in addition to retention time, improve data analysis and there has been increasing interest in separating peptides based on collisional cross section (CCS), which is a measure of the size of a peptide. However, existing libraries that are used during data analysis lack CCS measurements, and this data is expensive and time-consuming to acquire. This has led to the desire to predict library values for mass spectrometry analysis. Here we compare three deep learning architectures, LSTM, CNN, and transformer, for the tasks of retention time and collisional cross section prediction. We show that the LSTM and CNN models perform similarly and that the transformer has a lower performance than expected.

bioinformatics↗