An Algorithm to Quantify Inducible Protein Condensates In Eukaryotic Cells

Reorganization of cellular proteins into subcellular compartments, such as the rearrangement of RNA-binding proteins into cytoplasmic stress granules and P-bodies, is a well-recognized, widely studied physiological process currently under intense investigation. Using the assembly of a novel, inducible, nuclear granule formed from the yeast RNA-binding transcription termination factors Nab3 and Nrd1, we present a freely-accessible, high-throughput and unbiased algorithm written in MATLAB that detects and measures protein distribution, partitioning, and sequestration into subcellular compartments captured by fluorescence microscopy; an invaluable advancement to current image analysis methods which utilize experiment-specific custom scripts or subjective manual counting. Employing our algorithm, we quantified thousands of cells, ensuring rigorous examination of Nab3 granule formation across strains with reproducible statistical analyses. We document strain differences in Nab3 granule formation and an associated growth defect. Additionally, we applied our algorithm to immunofluorescent images of the inducible polymerization into filaments of an enzyme in human cells, demonstrating the algorithms versatility and adaptability. SUMMARY STATEMENTWe describe a computational tool that enables the quantification of protein condensation during assembly of a subnuclear compartment. The algorithm scores assembly of fluorescently tagged proteins in yeast or human cells.