Back

Revisiting Stress Granule Transcriptomes Suggests Mitochondrial RNA Enrichment Despite Methodological Bias

Kaizeler, A.; Coutinho, D.; Vieira, D. V.; Morais, V. A.; Barbosa-Morais, N. L.

2026-09-11 cell biology
10.64898/2026.09.08.749846 bioRxiv
Show abstract

Stress granules (SGs) are dynamic, membraneless cytoplasmic condensates that form in response to diverse cellular stressors. Although proposed to modulate stress responses by selectively sequestering proteins and RNAs, their precise molecular composition and function remain unclear. Reported SG transcriptomes differ substantially due to methodological discrepancies, notably between differential centrifugation (DC) and proximity labeling (PL). Here, we reanalyse publicly available human SG transcriptomes across multiple stressors, cell types, and isolation strategies. DC-based profiles were strongly shaped by RNA length, consistent with a physical bias inherent to the sedimentation-based separation. Correcting for this effect reveals limited concordance between studies. Mitochondrially encoded RNAs nonetheless consistently stand out as a distinctively regulated transcript class, without a uniform direction of enrichment/depletion across datasets, and with a modest but significant enrichment in our consensus SG signature. Immunofluorescence experiments further support, without definitively demonstrating, sequestration of mitochondrial dsRNA by SGs upon stress. Since leakage of mitochondrial nucleic acids is a well-characterised damage-associated molecular pattern linked to inflammation, their sequestration in SGs suggests a potential role in modulating immune-related stress responses. These findings refine our understanding of SG composition, underscore the need to control for technical artifacts in SG isolation, and provide a framework to distinguish genuine biological signals from methodological noise in SG research.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.