Spot-Based and Single-Cell-Resolution Methods in Spatial Transcriptomics

  • Spot-based and single-cell-resolution methods are two major approaches used in spatial transcriptomics to investigate gene expression while preserving the spatial organization of cells within tissues. Both methods combine molecular profiling with tissue location information, allowing researchers to understand not only which genes are expressed but also where their expression occurs. The main difference is the level of spatial resolution at which gene expression is measured. Spot-based methods capture RNA from defined tissue regions or spatial spots, which may contain multiple cells, whereas single-cell-resolution methods aim to identify gene expression at the level of individual cells or, in some technologies, subcellular structures.
  • Spot-based spatial transcriptomics uses spatially arranged capture areas containing barcoded probes to collect RNA from tissue sections. Each capture spot has a unique spatial barcode that identifies its position within the tissue. When a tissue section is placed on the capture surface, RNA molecules released from cells bind to the probes, undergo reverse transcription, and are converted into sequencing libraries. The spatial barcodes allow sequencing reads to be assigned to their original locations, producing a spatial gene-expression matrix that can be combined with tissue images. In many spot-based technologies, each spot captures transcripts from several neighbouring cells, so the resulting expression profile represents a mixture of cell types and cellular states.
  • Important spot-based technologies include 10x Genomics Visium, Visium HD, Slide-seq, Slide-seqV2, High-Definition Spatial Transcriptomics (HDST), and Open-ST. The original Visium platform uses relatively large capture spots, making it suitable for identifying gene-expression patterns across tissue regions but limiting the ability to distinguish individual cells within a spot. Visium HD provides much smaller capture areas and higher spatial resolution, allowing more detailed mapping of tissue organization. Slide-seq and Slide-seqV2 use densely packed, spatially barcoded beads to achieve finer spatial mapping, while HDST uses high-density barcoded beads for spatially resolved transcriptomic profiling. These technologies are particularly useful when researchers need broad transcriptome coverage across large tissue areas and want to identify regional gene-expression patterns, tissue compartments, and molecular differences between anatomical regions.
  • Single-cell-resolution spatial transcriptomics aims to measure gene expression in individual cells while retaining their positions within the tissue. Unlike conventional spot-based approaches, these methods attempt to separate the expression profiles of neighbouring cells through cell segmentation, spatially localized probes, or imaging-based molecular detection. Some platforms use fluorescent probes to detect selected RNA molecules directly inside intact tissue sections, while others use spatial barcodes or high-resolution capture areas combined with computational segmentation. This allows researchers to associate gene-expression profiles with individual cells and investigate cellular diversity, cell boundaries, tissue architecture, and interactions between neighbouring cells.
  • Imaging-based methods are important examples of single-cell-resolution spatial transcriptomics. Multiplexed Error-Robust Fluorescence In Situ Hybridization (MERFISH) uses fluorescently labelled probes and sequential imaging to identify many RNA species within individual cells. Sequential Fluorescence In Situ Hybridization (seqFISH and seqFISH+) uses repeated rounds of hybridization and imaging to detect multiple transcripts in their original cellular locations. 10x Genomics Xenium and NanoString CosMx Spatial Molecular Imager use targeted in situ hybridization and imaging to measure selected gene panels at cellular or subcellular resolution. These methods can provide detailed information about cell identity, gene-expression patterns, and the spatial relationships between cells. However, their measurements are generally restricted to predefined gene panels rather than providing unrestricted whole-transcriptome coverage.
  • Other spatial technologies combine high-resolution tissue imaging with molecular detection. STARmap uses in situ sequencing approaches to detect RNA molecules within intact tissues, while FISSEQ (Fluorescent In Situ Sequencing) identifies RNA sequences directly within cells through sequencing-based imaging. These approaches preserve the original tissue context and can provide spatially precise molecular information. RNAscope is another important in situ hybridization technique used to detect and validate selected RNA transcripts within tissues, although it is primarily a targeted detection method rather than a comprehensive transcriptome-wide profiling platform.
  • The computational analysis of spot-based and single-cell-resolution data requires different processing strategies. Spot-based datasets commonly contain spatial spots as measurement units, with each spot represented by a gene-expression profile and spatial coordinates. Software such as Space Ranger, Seurat, Scanpy, Squidpy, Giotto, STUtility, and SpatialExperiment supports data processing, visualization, clustering, spatial analysis, and integration with single-cell RNA sequencing datasets. Single-cell-resolution datasets require additional steps, including cell segmentation, transcript assignment to individual cells, removal of background signals, and quality assessment of cell-level expression profiles. Platforms such as Xenium and CosMx provide dedicated processing software, including Xenium Ranger and platform-specific analysis pipelines, while Seurat, Scanpy, Squidpy, and Giotto can support downstream analysis and integration.
  • A major challenge in spot-based spatial transcriptomics is the presence of multiple cells within a single spot. This creates mixed gene-expression profiles that make it difficult to determine which cell type produced a particular transcript. Computational deconvolution methods help estimate the cellular composition of each spot by integrating spatial transcriptomics with reference single-cell RNA sequencing data. Important tools include RCTD, cell2location, SPOTlight, stereoscope, DestVI, and Tangram. These methods use reference cell-type expression profiles to estimate the abundance or contribution of different cell populations within spatial spots. Their accuracy depends on the quality and biological relevance of the reference dataset, and estimated cell proportions should not automatically be interpreted as direct observations of individual cells.
  • Single-cell-resolution methods reduce some of the ambiguity associated with mixed spots, but they introduce other analytical challenges. Accurate cell segmentation is essential because incorrect identification of cell boundaries can assign RNA molecules to the wrong cells. Closely packed cells, overlapping structures, tissue damage, autofluorescence, background signals, and low transcript detection can affect the quality of cellular measurements. Computational methods are therefore needed to identify cell boundaries, assign detected transcripts to cells, remove background signals, and assess the reliability of each cellular profile. Even at high spatial resolution, measurements may not capture every transcript in a cell, and transcript detection does not necessarily indicate corresponding protein abundance or biological activity.
  • The choice between spot-based and single-cell-resolution methods depends on the research question, tissue type, required spatial resolution, gene coverage, sample availability, experimental resources, and cost. Spot-based methods are useful for investigating broad tissue organization, regional gene-expression patterns, and transcriptome-wide molecular differences across tissue sections. Single-cell-resolution methods are more appropriate when the objective is to identify individual cell populations, examine rare cell types, study cellular neighbourhoods, or investigate interactions between specific cells. High-density spot-based technologies may provide intermediate spatial resolution, although smaller capture areas do not always guarantee accurate single-cell identification.
  • Both approaches can be integrated with single-cell RNA sequencing, spatial proteomics, histological imaging, and other molecular datasets to provide a more complete understanding of tissue biology. Spatial analysis can identify tissue regions, detect spatially variable genes, characterize cellular neighbourhoods, and investigate potential ligand–receptor interactions between nearby cell populations. Tools such as SpatialDE, SPARK, and SPARK-X can identify genes with spatially structured expression patterns, while CellChat, CellPhoneDB, NicheNet, and LIANA can support the investigation of potential cell–cell communication when combined with suitable expression and spatial information.
  • Spot-based and single-cell-resolution methods have applications in cancer research, developmental biology, neuroscience, immunology, infectious disease research, and regenerative medicine. In cancer, they help identify tumour regions, immune-cell infiltration, stromal compartments, and spatial differences in gene expression. In neuroscience, they support the mapping of neuronal populations and molecular organization across brain regions. In developmental biology, they reveal spatial patterns of gene regulation during tissue formation and cellular differentiation. In animal genomics and breeding research, these methods can be used to investigate tissue-specific gene expression in livestock, identify cellular mechanisms underlying muscle development, study mammary gland biology and milk production, examine reproductive tissues, and characterize immune responses and disease resistance. They may also help connect genetic variation with molecular and cellular processes that influence economically important traits.
  • Despite their advantages, both approaches have limitations involving tissue preservation, RNA quality, spatial resolution, transcript detection efficiency, experimental cost, data complexity, and computational requirements. Spot-based measurements may combine signals from several cells, while imaging-based methods may be restricted to predefined gene panels and can be affected by segmentation errors or incomplete transcript detection. Results should therefore be interpreted alongside histology, immunofluorescence, immunohistochemistry, RNA in situ hybridization, and other independent validation methods. The integration of spot-based and single-cell-resolution technologies provides complementary information and is an important direction for understanding the molecular organization of complex tissues.

Author:

Leave a Reply

Your email address will not be published. Required fields are marked *