Spatially exploring RNA biology in archival formalin-fixed paraffin-embedded tissues

Abstract

The capability to spatially explore RNA biology in formalin-fixed paraffin-embedded (FFPE) tissues holds transformative potential for histopathology research. Here, we present pathology-compatible deterministic barcoding in tissue (Patho-DBiT) by combining in situ polyadenylation and computational innovation for spatial whole transcriptome sequencing, tailored to probe the diverse RNA species in clinically archived FFPE samples. It permits spatial co-profiling of gene expression and RNA processing, unveiling region-specific splicing isoforms, and high-sensitivity transcriptomic mapping of clinical tumor FFPE tissues stored for 5 years. Furthermore, genome-wide single-nucleotide RNA variants can be captured to distinguish malignant subclones from non-malignant cells in human lymphomas. Patho-DBiT also maps microRNA regulatory networks and RNA splicing dynamics, decoding their roles in spatial tumorigenesis. Single-cell level Patho-DBiT dissects the spatiotemporal cellular dynamics driving tumor clonal architecture and progression. Patho-DBiT stands poised as a valuable platform to unravel rich RNA biology in FFPE tissues to aid in clinical pathology evaluation.

Keywords: RNA biology; clinical FFPE tissue; histopathology; microRNA; single-nucleotide RNA variants; spatial omics; spatiotemporal dynamics; splicing isoforms; whole transcriptome.

Figure 1.. Patho-DBiT workflow and spatial whole transcriptome mapping of mouse embryo

(A) Schematic workflow, molecular underpinnings, and technological spectrum of Patho-DBiT. (B) Patho-DBiT’s performance on an E13 mouse embryo section. Top left: H&E of an adjacent section. Top right: tissue scanning. Bottom: Unsupervised clustering. (C) Spatial pan-mRNA and UMI count maps. (D) Correlation analysis between replicates, with the Pearson coefficient indicated. (E) Read coverage along the gene body and the percentage of reads mapped to the 5’ UTR. (F) Proportion of reads mapped to different RNA categories. (G) Comparison of non-coding RNA read mapping ratios between Patho-DBiT and normal DBiT. (H) Spatial count maps of different non-coding RNAs. (I) Left: Cell patterning and embedding. Right: Relative miR-142 expression in KO and WT cells. (J) Spatial miR-142 expression in patterned cells (I) and the evaluation of detection sensitivity and specificity. (K) Detection of microRNAs in the E13 section. (L) Left: read coverage of miR-122 mapped to the genome location. Right: spatial distribution of miR-122. (M) Distribution of gene and UMI counts in different tissue types at varying spatial pixel sizes. See also Figure S1.

Figure 2.. Spatial co-mapping of gene expression and RNA processing in the mouse brain

(A) Patho-DBiT profiling of a mouse brain section. Left: H&E of an adjacent section. Middle: tissue scanning. Right: spatial pan-mRNA and UMI count maps. (B) Unsupervised clustering. The distribution aligned with the region annotation from the Allen Brain Atlas. (C) Integration of spatial data with scRNA-seq dataset. (D) Anatomical brain region labeling in the Patho-DBiT dataset. (E) Number of significant differentially spliced events and parental genes between brain region pairs under different read count thresholds. (F) Top-ranked 12 genes exhibiting significant regional differences in exon inclusion levels. (G and H) Junction read coverage of Myl6 (G) and Ppp3ca (H) splicing event in specific brain regions. (I) Left: spatial variations in A-to-I RNA editing. Right: distribution of editing ratio across all editing sites and the expression level of ADAR-encoding genes in different brain regions. (J) Left: spatial Adarb1 expression. Right: correlation between the Adarb1 expression and the average regional editing ratio across various brain regions, with the Spearman coefficient indicated. (K) Editing ratio correlation between 259 sites commonly detected by Patho-DBiT and long-read Nanopore sequencing, as reported in the reference literature, with the Pearson coefficient indicated. See also Figure S2 and Table S1.

Figure 3.. High-sensitivity spatial transcriptomics of a AITL sample stored for five years

(A) Spatial transcriptome mapping of an FFPE AITL tissue stored for five years. Left top: H&E of an adjacent section. Left bottom: tissue scanning. Right: unsupervised clustering. (B) Top ranked DEGs defining each cluster. (C) Spatial phenotyping of an adjacent section using CODEX technology. (D) Spatial distributions of B cells, T cells, and macrophages revealed by Patho-DBiT, exhibiting a strong Pearson correlation with the proteomic data from CODEX. (E) Top: CODEX data from the yellow square indicated area in (C). Bottom: DEGs in cluster 0 corresponding to the indicated region. (F) Ligand-receptor interactions within cluster 0. (G) Signaling pathways regulated by the DEGs in cluster 0. z score is computed and used to reflect the predicted activation level (z>0, activated; z<0, inhibited; z≥2 or z≤−2 can be considered significant). (H) Graphical network of canonical pathways, upstream regulators, and biological functions regulated by DEGs identified in cluster 0. See also Figure S3.

Figure 4.. High-resolution tissue architecture resolved by integrating Patho-DBiT with histology

(A) Spatial mapping of the MALT section using 20 μm Patho-DBiT device covering 100,000 spots. Left top: tissue scanning. Left bottom: H&E of an adjacent section. Right: unsupervised clustering. (B) Spatial identification of representative cell types. (C) IF staining of plasma cell marker (CD138) and macrophage marker (CD68) in the selected Region P and Region M in (B). (D) Representative transcriptomic neighborhoods revealed by Patho-DBiT, closely aligned with H&E tissue morphology. (E) Super-resolved high-resolution spatial clustering and representative gene expression enhanced with iStar. See also Figures S4 and S5.

Figure 5.. Genome-wide spatial variant profiling for differentiating malignant subclones

(A) Molecular underpinnings of detecting variations printed in pre-mRNA transcripts. (B) Comparison of genomic location coverage bandwidth between Patho-DBiT and scRNA-seq datasets. (C) Spatial expression map of accumulated SNVs burden. (D) IHC staining of BCL-2 and CD43 commonly used for immunophenotyping MALT tumor cells. (E) Unsupervised clustering of the spatial SNVs matrix and genome-wide distribution of somatic variations in clusters M1 and M3. Veen plot showing the pixel overlap between gene and SNVs tumor clusters. (F) WGS of the MALT section matched with a reference normal stomach sample from the same patient. (G) Read count of DNA-level SNVs and Patho-DBiT detected RNA-level SNVs, and their ratios. (H) Comparison of RNA variant allele frequencies in DNA-derived somatic mutation sites across gene clusters. (I) Evolutionary relationships among the identified gene clusters. (J) Spatial distribution of top-ranked SNV sites upregulated in tumor B-cell clusters. (K) InferCNV analysis of gene expression intensity across tumor genome positions in tumor B-cell clusters. See also Figure S6.

Figure 6.. Spatial microRNA regulations in the MALT section

(A) MicroRNAs detected by Patho-DBiT in the MALT section. The pie chart illustrates the percentage distribution of the detected count number per spatial pixel. (B) Spatial distribution of the Smooth muscle cell Score. (C) Spatial mapping of smooth muscle cell specific miR-143 and miR-145. (D) Differentially expressed microRNAs between the tumor and non-tumor regions. (E) Regulatory network between the top upregulated microRNAs and the gene expression in the tumor region. (F) Spatial expression, read coverage, and regulatory relationship of miR-21. (G) Spatial expression, read coverage, and expression comparison between tumor and non-tumor regions of miR-155. Significance level was calculated with two-tailed Mann-Whitney test, **** P < 0.0001. (H) Spatial interactions involving miR-155 and its upstream and downstream signaling pathways. The Pearson correlation was calculated across 1,366 spatial pixels within the tumor B-cell region. See also Figure S7 and Table S2.

Figure 7.. Cellular level spatial mapping of a DLBCL section elucidates tumor progression

(A) Spatial transcriptome mapping of the DLBCL biopsy. Sections from two different regions underwent 10 μm microfluidic barcoding. (B) Super-resolved high-resolution spatial clustering and representative gene expression enhanced with iStar. (C) Left: schematic illustration showing comparative analysis. Right: signaling pathways regulated by DEGs between tumor B cells in DLBCL vs. MALT biopsy. (D) Expression comparison of key genes involved in the NF-κB signaling between DLBCL vs. MALT biopsy. (E) IHC staining for Ki67 on adjacent sections from the two biopsies. (F) Spatial expression mapping of genes encoding plasma cell kappa and lambda chains in the two biopsies. (G) ISH staining for kappa and lambda chain mRNA in the designated area in (F). (H) Distance distribution between macrophages and tumor B cells in the two biopsies. Significance level was calculated with two-tailed Mann-Whitney test, **** P < 0.0001. (I) Signaling pathways regulated by DEGs between macrophages in DLBCL vs. MALT biopsy. (J) IHC staining for CD206 on an adjacent section from the DLBCL and MALT biopsy. (K) Ligand-receptor interactions between macrophage cluster 1 and tumor B cell clusters 2 and 5. The distinctive communication pattern between TGFB1 and the integrin family genes is indicated and spatially visualized. In (C) and (I), z score is computed and used to reflect the predicted activation level (z>0, activated; z<0, inhibited; z≥2 or z≤−2 can be considered significant). See also Figure S8.