Journal: bioRxiv

Article Title: Cardiac-immune microniches programme macrophage states in the regenerating heart

doi: 10.64898/2026.03.05.709830

Figure Lengend Snippet: A, Schematic of the integrative analysis pipeline combining single cell RNA-seq, Visium spatial transcriptomics and ligand-receptor-target inference to identify niche-macrophage signalling programmes. B,C, Visium sections from homeostatic (B) and 5 dpi regenerating (C) hearts, with each white spot representing a 55µm capture area. D,D ′ , Spatial maps of key structural cell types identified (smooth muscle cells, cardiomyocytes, fibroblasts, epicardium and macrophages) in homeostatic (D) and regenerating (D ′ ) hearts according to inferred cell type composition after cell2location deconvolution. E, Representative MERFISH image illustrating single molecule transcript detection and cell segmentation. E ′ , Heat map summarising the biological categories included in designing the 500 gene MERFISH panel, grouped into structural cell markers, mpeg1.1 + subpopulation genes, injury-induced signatures, candidate niche-macrophage signalling mediators and other regeneration-related genes. F,G, Marker selection matrices used to distinguish structural cells (F) and mpeg1.1 + immune subpopulations (G) in MERFISH data. Detailed gene panel design included in Table S1. H-J, NicheNet-based prioritisation of ligand-receptor circuits, showing ligands upregulated after injury in structural “sender” populations (H), corresponding receptor activity in mpeg1.1 + “receiver” subsets (I) and predicted downstream target genes in mpeg1.1 + macrophages (J) that together define putative functional communication programmes. Detailed NicheNet output included in Supplementary Information 4.

Article Snippet: Using the integrated single cell transcriptomics dataset consisting of mpeg1.1 + cells as well as reference structural cells, a 500 gene MERFISH panel was designed for high-resolution spatial transcriptomics using the Merscope platform (Vizgen).

Techniques: Single Cell, RNA Sequencing, Spatial Transcriptomics, Marker, Selection, Activity Assay, Functional Assay

Journal: bioRxiv

Article Title: Cardiac-immune microniches programme macrophage states in the regenerating heart

doi: 10.64898/2026.03.05.709830

Figure Lengend Snippet: A,A ′ , MERFISH maps of homeostatic hearts with cells coloured by cell type (A) and with macrophages highlighted (A ′ ). A ′′ , MERFISH spatial expression of representative cell type markers (smooth muscle cells, endocardium/endothelium, cardiomyocytes, fibroblasts and epicardium) in homeostatic hearts. B,B ′ , Maps of 5 dpi regenerating hearts coloured by cell type (B, cell type key as in A), highlighting macrophage distribution (B ′ ). B ′′ , MERFISH spatial expression of cells type markers in regenerating hearts, illustrating depletion of cardiomyocytes and upregulation of fibroblast and macrophage signatures in the injury region. C, Proportion of structural and immune cell types per section (n=4 per condition) across biological replicates (n=2 per condition) in homeostatic and regenerating hearts, cell type key as in A. D,E, NicheCompass-derived multicellular niches in homeostatic (D) and regenerating (E) hearts, capturing anatomical domains including atrium, valves, myocardium, regenerating ventricle, border zone, injury core, injury epicardium, blood and outflow tract. F, Bar plots showing stereotyped cell type numbers and composition of each niche in homeostatic and regenerating conditions, cell type key as in A. Spatial distribution of macrophages in the homeostatic (G) and regenerating (H) hearts, with spatial expression of the macrophage markers mpeg1.1, marco, csf1ra, mrc1b and cxcr3.3 distributed along the injured heart, including the injury core and injury epicardium niches.

Article Snippet: Using the integrated single cell transcriptomics dataset consisting of mpeg1.1 + cells as well as reference structural cells, a 500 gene MERFISH panel was designed for high-resolution spatial transcriptomics using the Merscope platform (Vizgen).

Techniques: Expressing, Derivative Assay

Journal: bioRxiv

Article Title: Cardiac-immune microniches programme macrophage states in the regenerating heart

doi: 10.64898/2026.03.05.709830

Figure Lengend Snippet: A, Heat map of NicheCompass-derived gene programmes enriched across microniches in the regenerating injury territory, highlighting microniche-specific combinations of signalling modules. B, MERFISH-based spatial co-localisation of il34 expression in fibroblasts and egr1 expression in macrophages across the injury zone. B ′ , Assignment of egr1 expression to macrophage subpopulations at single cell resolution.

Article Snippet: Using the integrated single cell transcriptomics dataset consisting of mpeg1.1 + cells as well as reference structural cells, a 500 gene MERFISH panel was designed for high-resolution spatial transcriptomics using the Merscope platform (Vizgen).

Techniques: Derivative Assay, Expressing, Single Cell

Journal: bioRxiv

Article Title: Cardiac-immune microniches programme macrophage states in the regenerating heart

doi: 10.64898/2026.03.05.709830

Figure Lengend Snippet: A,A ′ , MERFISH maps of the injured heart of csf1ra j4e1/j4e1 mutants at 5 dpi, coloured by cell type (A) and highlighting macrophages (A ′ ). B, Comparison of cell type proportions in the injury territory between wild type and csf1ra mutants. C,C ′ , NicheCompass-defined niches in the mutant heart (C) and injury-restricted niches (C ′ ). D, Microniches within the mutant injury region. E, Volcano plot of differentially expressed genes in macrophages in wild type compared to mutant injury territories. F, Heatmap of genes upregulated in mutant macrophages and ploted across macrophage clusters, showing preferential enrichment of cluster 5 “phagolysosomal” and cluster 12 “stress-adapted inflammatory” signatures in mutants. Spatial module scores for clusters 5 (G) and 12 (H) projected onto wild type and mutant hearts, illustrating selective upregulation of programmes and increase of stress-adapted inflammatory states, particularly cluster 12, in mutant heart. I, Spatial expression of il34, csf1ra and egr1 in wild type and mutant hearts within the injury zone. J, Volcano plot of differentially expressed genes within the injury territory of wild type versus csf1ra mutants. K-N, MERFISH spatial expression maps of fibrosis-related genes tgfbr2a, acta2, ccn2b, serpine1 (K), ECM remodelling genes dcn4, timp4.3, ctnnbip1 (L), endothelial factors itgae.1, cdh5, plvapb (M) and epicardial factors sema3e, mdkb, tbx18 (N) in the wild type and mutant injury area. O, Model summarising how disruption of il34-csf1ra signalling prevents egr1 induction in resident and pro-resolving macrophages, shifting macrophages towards stress-adapted inflammatory states, which impacts endothelial and epicardial compartments and biases repair towards fibrosis.

Article Snippet: Using the integrated single cell transcriptomics dataset consisting of mpeg1.1 + cells as well as reference structural cells, a 500 gene MERFISH panel was designed for high-resolution spatial transcriptomics using the Merscope platform (Vizgen).

Techniques: Comparison, Mutagenesis, Expressing, Disruption

Journal: bioRxiv

Article Title: Cardiac-immune microniches programme macrophage states in the regenerating heart

doi: 10.64898/2026.03.05.709830

Figure Lengend Snippet: A,A ′ , MERFISH map of the homeostatic csf1raj4e1/j4e1 heart coloured by cell type (A) and highlighting macrophages (A ′ ). B, NicheCompass-derived niches in homeostatic csf1ra mutants, showing preserved global architecture. C, Change in microniche proportions within the injury territory of wild type versus csf1ra mutant. D,E, MERFISH spatial expression of col1a1a, il34, csf1ra and egr1 in wild type and csf1ra mutant hearts (D). Normalised fold change of captured col1a1a, il34, csf1ra and egr1 transcripts in mutant injury area compared to wild type. F,G, MERFISH whole-heart section of egr1 expression in wild type (F) and csf1ra mutant (G) hearts, showing that loss of egr1 is confined to the injury territory and not observed in remote myocardium. H, Normalised fold-change of egr1 transcripts in csf1ra mutants relative to wild type in regions outside the injury area.

Article Snippet: Using the integrated single cell transcriptomics dataset consisting of mpeg1.1 + cells as well as reference structural cells, a 500 gene MERFISH panel was designed for high-resolution spatial transcriptomics using the Merscope platform (Vizgen).

Techniques: Derivative Assay, Mutagenesis, Expressing

Journal: bioRxiv

Article Title: Spartan: Spatial Activation Aware Transcriptomic Analysis Network

doi: 10.64898/2026.02.18.706570

Figure Lengend Snippet: (A–B) Spatial domain identification for MERFISH 0.19 (SDMBench) and Vizgen MERFISH S2R3 samples. Ground-truth annotations (left) are compared with clustering results from Spartan (v1, v2), NichePCA, BANKSY, and SpatialLeiden (v1, v2). ARI values above each map quantify agreement with ground truth. (C) Spatially variable gene (SVG) examples for the MERFISH 0.19 sample. Spatial expression patterns align with anatomical boundaries and support the cluster assignments in (A). (D) Stereo-seq mouse embryo (E9.5E2S2). Ground-truth annotations (left) and spatial expression of representative embryo marker genes (right) show anatomically consistent localization. (E) SVGs for the Vizgen MERFISH mouse brain S2R3 sample, highlighting region- and cell-type–specific expression patterns. (F) 10x Visium human dorsolateral prefrontal cortex (DLPFC) sample 151673. Ground-truth cortical layers (left) and layer-enriched marker genes (right) demonstrate accurate laminar recovery.

Article Snippet: The data were acquired from the following websites or accession numbers: (1) Stereo-seq, BaristaSeq, MER-FISH, osmFISH, STARmap, STARmap*, 10x Visium datasets with regional annotations were downloaded from SDMBench; (2) Vizgen MERFISH dataset was downloaded from Vizgen MERFISH Mouse Brain Receptor Map; (3) The Visium HD dataset generated in this study is available in ArrayExpress under accession number E-MTAB-16671.

Techniques: Expressing, Marker

Journal: bioRxiv

Article Title: Spartan: Spatial Activation Aware Transcriptomic Analysis Network

doi: 10.64898/2026.02.18.706570

Figure Lengend Snippet: (A–B) Spatial domain identification for MERFISH 0.19 (SDMBench) and Vizgen MERFISH S2R3 samples. Ground-truth annotations (left) are compared with clustering results from Spartan (v1, v2), NichePCA, BANKSY, and SpatialLeiden (v1, v2). ARI values above each map quantify agreement with ground truth. (C) Spatially variable gene (SVG) examples for the MERFISH 0.19 sample. Spatial expression patterns align with anatomical boundaries and support the cluster assignments in (A). (D) Stereo-seq mouse embryo (E9.5E2S2). Ground-truth annotations (left) and spatial expression of representative embryo marker genes (right) show anatomically consistent localization. (E) SVGs for the Vizgen MERFISH mouse brain S2R3 sample, highlighting region- and cell-type–specific expression patterns. (F) 10x Visium human dorsolateral prefrontal cortex (DLPFC) sample 151673. Ground-truth cortical layers (left) and layer-enriched marker genes (right) demonstrate accurate laminar recovery.

Article Snippet: The Vizgen MERFISH mouse brain dataset further highlights Spartan’s ability to capture regionally specialized neuronal and glial populations across the adult brain ( ; Supplementary Fig. 3C).

Techniques: Expressing, Marker

Journal: bioRxiv

Article Title: Spartan: Spatial Activation Aware Transcriptomic Analysis Network

doi: 10.64898/2026.02.18.706570

Figure Lengend Snippet: (A–B) Spatial domain identification for MERFISH 0.19 (SDMBench) and Vizgen MERFISH S2R3 samples. Ground-truth annotations (left) are compared with clustering results from Spartan (v1, v2), NichePCA, BANKSY, and SpatialLeiden (v1, v2). ARI values above each map quantify agreement with ground truth. (C) Spatially variable gene (SVG) examples for the MERFISH 0.19 sample. Spatial expression patterns align with anatomical boundaries and support the cluster assignments in (A). (D) Stereo-seq mouse embryo (E9.5E2S2). Ground-truth annotations (left) and spatial expression of representative embryo marker genes (right) show anatomically consistent localization. (E) SVGs for the Vizgen MERFISH mouse brain S2R3 sample, highlighting region- and cell-type–specific expression patterns. (F) 10x Visium human dorsolateral prefrontal cortex (DLPFC) sample 151673. Ground-truth cortical layers (left) and layer-enriched marker genes (right) demonstrate accurate laminar recovery.

Article Snippet: The Vizgen MERFISH is imaging-based SRT dataset obtained by the MERFISH technology.

Techniques: Expressing, Marker