Probing the differential stress granule proteome by mass spectrometry-based proteomics

Abstract

Stress granules (SG), as part of the biomolecular organization of the cell, are membraneless, predominantly formed by liquid-liquid phase separation (LLPS), and are essential for cellular homeostasis. Dysregulation of their assembly has been implicated in severe diseases such as cancer, amyotrophic lateral sclerosis (ALS), and frontotemporal dementia. Thus a comprehensive understanding of SG assembly, composition, and maintenance is of great interest. Their properties, such as small size, dynamic exchange with surrounding cytoplasm, and multivalent interactions, make their enrichment and analysis technically challenging. To overcome these obstacles, this study adapted an SG enrichment workflow based on green- fluorescent protein (GFP)-tagged G3BP1, the SG core protein. The approach combines chemical cross-linking for structural stabilization, enrichment via fluorescence-activated particle sorting (FAPS), and downstream quantitative mass spectrometry (MS) analysis. A screen for a suitable cross-linker on arsenate-stressed HeLa cells, considering the compatibility with FAPS, identified disuccinimidyl glutarate (DSG), with its short spacer arm, as the most effective. Alongside a gentle lysis procedure, involving syringe and cannula, and a short centrifugation protocol, yielded a clear and reproducible size-pre-enriched sample suitable for FAPS. FAPS discriminated the input according to ‘high’ or ‘medium’ intense particles, referring to the GFP-intensity. Subsequent label-free quantification (LFQ) in data-independent acquisition (DIA) MS analysis identified 362 statistically significant proteins in arsenate-stressed samples. Among these, 187 proteins were enriched in the stress high compared to the stress medium sorted fractions, indicating a subset of proteins whose association with SGs is highly stress dependent. Ratio- based filtering further refined the list, yielding 42 proteins confidently assigned as SG core components. Expanding the dataset to include samples subjected to heat stress and osmotic shock revealed stressor-specific protein sets alongside a shared cluster of proteins consistently present across all stress conditions. Using the chemical cross-linking and the fluorescent features of GFP tagged to G3BP1 for enrichment enables robust downstream data acquisition, yielding a highly reliable list of SG- associated and SG core proteins. This workflow is also suitable for comparative analysis across different stress conditions. Overall, it offers a specific approach for capturing in-depth quantitative proteomic information on SGs, and will help to unravel their composition, dynamics, and regulation.