Nuclear proteome map of mouse heart chambers

Figures

• 

  Fig. 1. Experimental and proteomic workflow for nucleocytoplasmic proteome enrichment in the mouse heart.A, optimized workflow of nucleocytoplasmic partitioning and nuclear enrichment. Fresh mouse heart regions (N = 4) were homogenized in cell lysis buffer and centrifuged at 100g; the supernatant was collected as the Global fraction. The remaining pellet underwent organelle lysis and centrifugation at 1000g to yield the Cytoplasmic fraction (supernatant). The final pellet was incubated with nuclear lysis buffer and centrifuged at 15,000g to obtain the nuclear fraction. Protein yield in each fraction across regions is shown (Created with BioRender.com). B, Western blot validation of nuclear and cytoplasmic markers comparing their intensities across methods and fractions. Cytoplasmic and nuclear fractions were probed for cytosolic/ER markers (Fh1, Canx, Gapdh) and nuclear markers (Lmna, Histone H3), demonstrating reduced cross-contamination and improved nuclear purity with the optimized protocol. C, proteomics workflow. Samples underwent reduction (DTT), alkylation (IAA), dual-enzyme digestion (Trypsin & LysC), SP3 cleanup, and LC–MS/MS acquisition in DIA mode, followed by data processing…

• 

  Fig. 2. Comparative analysis of subcellular fractions across mouse heart regions.A, heatmap of k-means clustering (k = 3) for significantly differentially abundant proteins (FDR<0.01) across Global, Cytoplasmic and Nuclear fractions in four heart regions (LA, RA, LV, RV). Columns are annotated by fraction and region; rows are grouped into clusters C1–C3, with Z-scored protein intensities shown. B, GO cellular component (GO:CC) enrichment analysis for each cluster. Enriched terms were redundancy-reduced using the simplify function. Dot size represents GeneRatio; color indicates Benjamini–Hochberg–adjusted p-value. Selected representative GO:CC terms are shown for clusters C1–C3. C, bar plots show VSN-normalized intensities for representative nuclear proteins (H1-2, Parp1, Hmgb2, Sfpq, Mecp2, Ubtf) across LV, RV, LA, and RA in matched Global, Cytoplasmic, and Nuclear fractions. Bars represent mean ± SEM, with points indicating individual samples. All proteins shown are significantly enriched in the nuclear fraction relative to the cytoplasmic fraction.

• 

  Fig. 3. Benchmarking heart nuclear proteome landscape at heart tissue, heart region, and nuclear heart levels.A, UpSet plot comparing heart nuclear proteome (current) with nuclear proteome of brain, liver, and kidney. Comparative analyses further included GO cellular component nuclear-annotated proteins for all mapped proteins, with 966 proteins uniquely identified in current heart nuclear proteome study relative to GO cellular component nucleus-annotated subset of each organ (brain, liver, kidney) nuclear proteome. B, Venn comparison of heart tissue enriched analysis (transcript) from Human Protein Atlas (HPA), with UniProt identifiers used in correlative analysis. C, UpSet plot comparing mouse chamber-resolved nuclear proteome from mouse right atrium (RA), left atrium (LA), right ventricle (RV), and left ventricle (LV). D, Venn comparison of heart nuclear proteome using global density-based fractionation of heart to resolve different subcellular niches of the mouse heart, including the nucleus. Validation against heart nuclear proteome revealed 365 co-identified proteins, 72 of which are identified across all anatomical regions of the mouse heart nuclear proteome.

• 

  Fig. 4. Regional and functional diversity of the cardiac nuclear proteome.A, average protein count in the nuclear fraction after filtering (≥67% valid values in at least one region) across LV, RV, LA, and RA. Bars represent mean values; dots indicate individual biological replicates. B, sample-to-sample correlation heatmap of nuclear proteomes showing clustering of biological replicates and separation of atrial (LA/RA) versus ventricular (LV/RV) profiles. C, heatmap of significantly regulated proteins (n = 973, FDR <0.05) identified by limma ANOVA. k-means clustering (k = 6) revealed six chamber-specific profiles (C1 = 111, C2 = 297, C3 = 183, C4 = 72, C5 = 188, C6 = 122). Columns are annotated by region (LV, RV, LA, RA); rows are Z-scored protein intensities ordered by cluster. D, KEGG pathway enrichment (Benjamini–Hochberg q < 0.05) for each cluster. Dot size indicates the number of mapped proteins (Count); color represents adjusted p-value. Numbers beneath each cluster indicate KEGG-mapped proteins relative to total cluster size (C1 = 67/111, C2 = 190/297, C3 = 107/183, C4 = 45/72, C5 = 114/188, C6 = 85/122). C1 was enriched for ferroptosis, peroxisome and cholesterol metabolism; C2…

• 

  Fig. 5. Hallmark nuclear programs show region-associated signatures across the cardiac nuclear proteome.A–I, heatmaps of VSN-normalized, Z-scored abundances for nucleus-annotated proteins selected from the 973 region-dependent nuclear proteins (limma one-way ANOVA across LV, RV, LA, RA; FDR <0.05). Proteins are grouped into functional nuclear programs: A, transcription factors and cofactors. B, RNA processing/splicing/export. C, chromatin organization and epigenetic regulation. D, nuclear envelope/lamina and nucleocytoplasmic transport. E, stress response and RNA granules. F, nuclear–cytoskeletal and sarcomere regulation. G, cell death and survival. H, cell cycle and mitosis, and I, proteostasis (UPS/SUMO/chaperones). Columns represent heart regions (LV, RV, LA, RA; color-coded as indicated) and rows represent individual proteins; values are displayed as Z-scores within each protein. Numbers in parentheses indicate proteins shown per program. Full protein lists are provided in Supplemental Table S9.

• 

  Fig. 6. Definition of a conserved core nuclear architecture across cardiac regions.A, frequency distribution of nuclear protein detection across samples, used to define a conserved set. Proteins that were non-significant across regions (ANOVA FDR ≥0.05) and detected in 15/16 or 16/16 nuclear samples were retained as the conserved set (958 proteins). Bars indicate the number of proteins per detection bin; colors denote whether proteins carry GO nucleus annotation (GO:0005634). B, volcano plot comparing matched nuclear versus cytoplasmic fractions for the nucleus-annotated subset, identifying proteins significantly enriched in the nuclear fraction (FDR <0.05; |fold-change| ≥ 1.5). The resulting 230 proteins (up in nucleus) were defined as the conserved nuclear core. C, unique transcription factor (TF) and co-factors identified in mouse heart nuclear core proteome relative to mouse heart tissue TFs and DNA-binding proteins network (29). D, GO-enrichment analysis of the conserved nuclear core. Left: molecular function (GO:MF) terms; right: cellular component (GO:CC) terms, highlighting canonical nuclear compartments and functions. E, protein–protein interaction network of conserved…

• 

  Fig. 7. Validation of region-dominant nuclear proteins by immunofluorescence.A–C, H2AC21, D–F, PTBP2, and G–I, SUN2. Nuclei are labeled with DAPI (red), cell membranes with WGA (blue), and target proteins are shown in green. A, D, and G, representative high-magnification images from LA, RA, RV, and LV showing nuclear-associated localization patterns. B, E, and H, whole-heart sections indicating the anatomical regions used for high-magnification imaging (dashed circles). C, F, and I, quantification of region-specific enrichment. Upper plots show mean nuclear IF intensity (fluorescence co-localized with DAPI; n = 3 hearts, three fields per chamber, mean ± SEM). Lower plots show corresponding VSN-normalized log2 proteomic intensities from nuclear fractions (non-imputed; n = 4 hearts, independent group), summarized by region (mean ± SEM). Asterisks mark the chamber with the highest IF signal. “Not detected” and “Detected in one out of four” indicate proteins not quantified in the proteomic dataset or quantified in only one biological replicate, respectively.