The heart is a highly structured organ whose normal functions are highly reliant on coordinated activities of different subcellular organelles. Dysregulated subcellular proteome and perturbed inter-organelle communications have been well-documented in different cardiomyopathies and attracted tremendous interest in mass spectrometry (MS)-based proteomics profiling. To date, biochemical enrichment or labelling of single organelles have provided critical insights into subcellular proteomes. However, a comprehensive view of cardiac protein subcellular localization with high spatial resolution has remained an unmet challenge in the field.
Here, we optimised a reproducible and high-resolution proteomics pipeline for profiling the spatial distribution of subcellular proteome from as little as 40 mg of mouse heart tissue using differential centrifugation-based fractionation, label-free data independent acquisition (DIA) MS-based profiling and machine learning (ML)-assisted prediction of protein subcellular locations. Based on protein fractionation patterns, our pipeline has resolved cardiac proteins to 9 subcellular features including cytosol, endoplasmic reticulum, plasma membrane, ECM/cytoskeleton, nucleus, mitochondria, ribonucleoprotein complex, Golgi apparatus/endo-lysosome, peroxisomes. Importantly, we have revealed different protein compositions of cardiac-specific sub-organellar structures such as transverse-tubule, intercalated disc, and junctional versus longitudinal sarcoplasmic reticulum.
The pipeline optimised in this proof-of-concept study will facilitate future investigations into inter-organelle communication and dynamic protein localization with high-resolution spatial information, providing critical molecular insights into cardiovascular physiology and disease development.