Fig 1.
Schematic showing the overall PAN-INTACT workflow. Mouse hearts (or other tissues) are harvested and minced extensively before douncing. Homogenized tissue is passed over sequential filters, and a crude nuclear pellet is obtained following centrifugation. The crude pellet is layered on a sucrose cushion, and ultracentrifugation yields a pure nuclear pellet, which is resuspended in buffer supplemented with 10% glycerol to avoid clumping. Pure nuclei are incubated with an appropriate antibody and purified further by magnet-assisted nuclei immunoprecipitation (MAN-IP) using anti-rabbit IgG Microbeads. The protocol is broadly split into nuclei isolation and MAN-IP segments, and the dashed boxes designate optimization steps that are shown in the indicated figures.
Fig 2.
Optimization of nuclei extraction from mouse hearts by dounce homogenization.
A) Diagram detailing the PAN-INTACT steps involved in nuclei extraction from heart tissue with the dashed box highlighting the douncing step that underwent optimization. B) Hearts derived from mice of varying postnatal age were systematically assessed for the ideal number of douncing strokes to release nuclei (stained with DAPI). For P28 hearts, 18 dounces were optimal. Under-douncing (8 strokes) led to incomplete nuclei release, while over-douncing (28 strokes) led to nuclei fragmentation. Scale bar: 200 μm. C) The optimal range of douncing strokes required to release nuclei at various postnatal ages is shown as a table (i) and graph (ii) with Standard Deviation (S.D.).
Fig 3.
Altering the density of the sucrose cushion enriches for CM or heterogenous cardiac nuclei.
A) Diagram detailing the PAN-INTACT steps involved in nuclei purification using two alternative sucrose cushions. B) Following nuclei purification over a 2.2M sucrose cushion, qRT-PCR was performed. Enrichment of CM nuclei was demonstrated by increased Troponin T (Tnnt2) expression in purified versus input (crude nuclear pellet). In contrast, epicardial (Wt1/Upk1b), fibroblast (Col1a1), and endothelial (Pecam1) markers were not enriched. * indicates p-value < 0.05 compared with input (crude nuclear pellet). C) Nuclei purified over a 2.2M sucrose cushion were immunolabeled with PCM1 (green) and Phospholamban (PLN; red) antibodies with DAPI counterstaining (blue). Nearly every DAPI+ nucleus is PLN+, while PCM1 labels only a subset of purified nuclei. Zoomed images are provided for detailed visualization (solid arrow, PCM1+/PLN+ nuclei; dashed arrow, PCM1-/PLN+ nuclei). Scale bar: 100μm. D) Cardiac nuclei were isolated over a 2.2M sucrose cushion. Half of the purified nuclei were then subjected to PCM1-based immunoaffinity purification for 30 minutes. PCM1+ nuclei were directly compared to input nuclei to determine cardiac enrichment by performing qRT-PCR with lineage-specific markers: Tnnt2 (CM), Wt1/Upk1b (epicardial cell), Col1a1 (fibroblast), and Pecam1 (endothelial cell). * and # indicate p-value < 0.05 for Pure CM nuclei 2.2M and PCM1+ CM nuclei, respectively, compared with input. E) Following nuclei purification over a 1.8M sucrose cushion, qRT-PCR was performed. Enrichment of cardiomyocyte, epicardial, fibroblast, and endothelial markers was observed. * indicates p-value < 0.05 compared with input (crude nuclear pellet). (B, D, and E) Y-axis scale: Log2; Gapdh was used as a reference gene to normalize qRT-PCR data, which is represented as average fold enrichment ± S.D. for reactions done in triplicate.
Fig 4.
Efficient isolation of Nkx2-5 lineage positive cardiac nuclei using PAN-INTACT.
A) Schematic of the breeding strategy used for these experiments. Crossing of R26Sun1-2xsf-GFP-6xmyc/+ and Nkx2-5Cre/+ mice led to Nkx2-5Cre/+; R26Sun1-2xsf-GFP-6xmyc/+ offspring that drive expression of the SUN1 fusion protein in the nuclear membrane of Nkx2-5 lineage positive cells. The major Nkx2-5 positive lineages in the heart include CMs, smooth muscle cells (SMCs), and endocardial cells (ECs). B) Experimental workflow for magnet-assisted nuclear immunoprecipitation (MAN-IP). C-E) Heterogeneous cardiac nuclei were purified from P1 mouse hearts using a 1.8M sucrose cushion followed by Myc MAN-IP. Confocal images of DAPI-stained nuclei showing native sfGFP expression in input nuclei (C), MAN-IP eluate (D), and MAN-IP flow-through (E). Magnification 100μm. F) Quantification of labeled nuclei from four independent experiments (n = 100 nuclei per experiment) for Myc MAN-IP of Nkx2-5 lineage positive cardiac nuclei was used to calculate specificity and sensitivity (range in percentage with S.D. in parentheses). G) qRT-PCR was performed on purified Nkx2-5 lineage positive nuclei and pure mixed nuclei compared with input. Gapdh served as an internal standard for qPCR, and the data are represented as average fold enrichment ± S.D. of triplicate reactions for each marker gene over input. * and # indicate p-value < 0.05 compared with input. Y-axis scale: Log2.
Fig 5.
PAN-INTACT enables generation of high-resolution chromatin accessibility maps.
A) Schematic diagram of experimental setup. CM nuclei were isolated from WT and Nkx2-5Cre/+; R26Sun1-2xsf-GFP-6xmyc/+ P1 mouse hearts using 2.2M sucrose-cushion, an aliquot of which was used as the input sample. Remaining nuclei were subjected to MAN-IP using a Myc or PCM1 antibody. Chromatin accessibility maps were generated for each sample individually using the OMNI-ATAC-Seq protocol. B) Similarity matrix demonstrating high pairwise Pearson correlations for ATAC-Seq biological replicates and amongst different samples. C) Heatmap showing a high degree of overlap for accessible sites within input, PCM1, and Myc samples. As an additional comparison, DHS-seq data is shown for P1 mouse heart from the ENCODE dataset. Kb (Kilobases) D) Genome browser tracks for canonical markers of CMs (i, Myh6) and fibroblasts (ii, Col3a1) with ATAC-seq peaks for each sample and DHS-seq peaks for comparison. E) Gene Ontology (GO) analyses for annotated nearest neighboring genes surrounding ATAC-seq peaks from all 3 samples (i-iii) display terms related to cardiac growth and function. Displayed results represent GO terms only with False Discovery Rate (FDR) < 0.05.
Fig 6.
PAN-INTACT enriches for rare cell type specific nuclei from adult mouse kidney.
A) R26Sun1-2xsf-GFP-6xmyc/+ female mice were bred with Wt1eGFPCre/+ males. Upon Cre-mediated recombination of the floxed STOP cassette, the SUN1 fusion protein is expressed on the nuclear membrane in Wt1 lineage positive cells. At P28, Wt1eGFPCre/+;R26RSun1-2XsfGFP-6-myc/+ mouse kidneys were harvested. B) Mixed kidney nuclei were purified using a 1.8M sucrose cushion followed by immuno-purification with Myc MAN-IP to isolate podocyte nuclei. C) qRT-PCR analysis demonstrates enrichment of podocyte markers Actn4, Synpo, Nphs2, and Nphs1. The CM marker Tnnt2 was not enriched. 1.8M sucrose-cushion purified nuclei was used as the input for qRT-PCR to calculate fold enrichment, and Gapdh served as an internal standard. Data is represented as average fold enrichment ± S.D. of triplicate reactions for each marker gene over input. Y-axis scale: Linear. * indicates p-value < 0.05 compared with input. Acronyms: Proximal Convoluted Tubule (PCT), Glomerulus (G), Bowman’s Capsule (BC), Renal podocytes (P).
Table 1.
Systematic optimization of PAN-INTACT.