1 History and development
Spatial transcriptomics emerged from the broader effort to connect molecular measurements with tissue architecture. Early work focused on visualizing RNA in fixed samples, then progressed toward methods that could profile many transcripts at once while retaining positional information. Over time, improvements in chemistry, imaging, sequencing, and computational analysis transformed the field from a set of specialized assays into a versatile toolbox for tissue biology.
1.1 Early tissue imaging approaches
Before transcriptome-wide mapping became possible, researchers used histological stains and targeted nucleic acid probes to study where biological molecules were located in cells and tissues. In situ hybridization was especially important because it linked gene expression to visible tissue structure. These approaches were usually limited to one or a few targets at a time, but they established the central idea that location matters in biological interpretation.
1.2 Emergence of spatial gene expression profiling
The move from single-gene visualization to broader profiling came with methods that combined microscopic tissue views with molecular readouts from defined locations. Early spatial profiling systems used arrays, barcodes, or segmented tissue regions to associate RNA abundance with position. This transition allowed researchers to compare gene expression patterns across neighboring tissue areas and to study local specialization within organs.
1.3 Advances in high-throughput technologies
High-throughput sequencing and improved probe chemistry made it possible to measure many transcripts simultaneously in spatially resolved samples. Platforms became faster, more sensitive, and more standardized, enabling larger tissue maps and more complex experiments. These advances also encouraged the development of software for image alignment, count matrix generation, and pattern discovery.
1.4 Integration with single-cell genomics
Spatial transcriptomics expanded rapidly after the rise of single-cell RNA sequencing. Single-cell methods offered detailed cell-state classification, while spatial assays provided the missing anatomical context. Together, these approaches made it possible to infer which cell types occupied a region, how they interacted, and how local environments influenced gene expression.
2 Principles of spatial transcriptomics
Spatial transcriptomics measures RNA or derived gene-expression signals while preserving the spatial arrangement of the tissue. The core goal is not only to identify which genes are active, but also to determine where those signals arise relative to cells, tissue layers, and structural landmarks. This combination of molecular and anatomical information supports analyses that are difficult with bulk RNA sequencing alone.
2.1 Transcriptome measurement
Most spatial methods detect messenger RNA through hybridization, reverse transcription, or sequencing. The resulting data can be summarized as counts for many genes across discrete locations or image-derived coordinates. Depending on the platform, the output may represent whole cells, groups of cells, or very small subcellular regions.
2.2 Spatial localization in tissue
Spatial information is preserved by fixing tissue on a slide, labeling molecules in place, or recording the position of captured RNA. Tissue structure can then be aligned with expression data, allowing comparison across anatomical regions. This positional context is central to the method, because similar cell types may behave differently depending on their neighborhood.
2.3 Resolution and coverage trade-offs
Spatial assays often balance three factors: how many genes are measured, how precisely location is defined, and how much tissue can be analyzed at once. Methods with broad transcript coverage may have lower spatial resolution, while highly resolved imaging approaches may detect fewer genes per assay cycle. Choosing a platform usually depends on the biological question being asked.
2.4 Relationship to histology
Histology provides the structural framework onto which molecular data are mapped. Spatial transcriptomics often uses stained sections or microscopy images to identify anatomy, pathology, or tissue boundaries. This link to morphology helps researchers interpret expression patterns in the context of recognizable tissue organization.
3 Experimental platforms
Spatial transcriptomics includes multiple platform families, each using a different strategy to preserve or reconstruct position. Some methods capture RNA on patterned surfaces, others visualize transcripts directly in tissue, and still others combine localized indexing with sequencing. The choice of platform affects resolution, throughput, and the type of downstream analysis required.
3.1 Capture-based methods
Capture-based systems use a physical surface to collect transcripts from defined positions in a tissue section. The surface contains spatial identifiers that make it possible to assign RNA molecules to their original location after sequencing or readout. These methods are widely used because they scale well and can cover relatively large tissue areas.
3.1.1 Slide-based barcoding systems
Slide-based barcoding places arrays of spatially encoded capture probes on a microscope slide. When a tissue section is mounted on the slide, released RNA binds to the probes and is labeled with position-specific information. After sequencing, expression data can be reconstructed as a map across the tissue.
3.1.2 Array-based capture surfaces
Array-based capture surfaces use patterned spots or features that each represent a known coordinate. The spots may vary in size, density, or chemistry depending on the platform. These systems are useful for profiling tissue architecture at a level finer than bulk assays, though often not at single-cell resolution.
3.2 Imaging-based methods
Imaging-based methods detect RNA molecules directly within fixed tissue by repeated rounds of labeling and microscopy. Because the signals are observed in place, these techniques can achieve very high spatial precision. They are especially useful when detailed localization is more important than capturing the entire transcriptome.
3.2.1 In situ hybridization approaches
In situ hybridization uses labeled probes that bind to specific RNA sequences inside cells. Modern versions may include multiplexed probe sets and amplified signals to improve detection. These assays can reveal gene expression patterns with high anatomical clarity.
3.2.2 Sequential imaging and decoding
Sequential imaging methods identify many RNA targets by performing multiple cycles of hybridization, imaging, and signal removal or color decoding. Each cycle adds new information, allowing a large panel of genes to be read from the same tissue section. This strategy supports rich spatial maps, though the workflow can be technically demanding.
3.3 Sequencing-based methods
Sequencing-based methods combine spatial indexing with next-generation sequencing to identify transcripts and their locations. Instead of reading molecules only by microscopy, they attach location-specific tags that are later recovered by sequencing. These approaches are flexible and can support genome-wide assays.
3.3.1 Spatial barcoding and indexing
Spatial barcoding assigns a unique identifier to each tissue position or capture site. RNA molecules collected from that site inherit the barcode during library preparation, making it possible to match gene counts with coordinates. This principle underlies many widely used spatial platforms.
3.3.2 Tissue section processing
Successful sequencing-based workflows depend on careful handling of tissue sections before molecular capture. Fixation, mounting, and enzymatic treatment must be balanced to preserve RNA integrity while allowing access to transcripts. Small differences in processing can strongly affect data quality.
3.4 Single-cell and subcellular resolution systems
Some newer systems aim to localize gene expression at the scale of individual cells or even smaller compartments within cells. These platforms may use highly dense barcodes, advanced imaging, or computational reconstruction to sharpen spatial detail. They are especially valuable for tissues with tightly packed cells or complex microanatomy.
4 Sample preparation and workflow
Spatial transcriptomics experiments require a workflow that preserves tissue structure and RNA quality from collection through sequencing or imaging. Each step can influence the final data, so protocols are often optimized for the tissue type and platform being used. Careful preparation is especially important for fragile samples and archived material.
4.1 Tissue collection and preservation
Tissue collection aims to maintain both morphology and RNA stability. Fresh, frozen, and fixed samples are used depending on the method and downstream goals. Preservation conditions affect transcript recovery, background signal, and compatibility with staining or imaging.
4.2 Sectioning and mounting
After collection, tissue is cut into thin sections and mounted onto a slide or compatible surface. Section thickness influences diffusion, capture efficiency, and image quality. Accurate mounting is necessary to keep spatial relationships intact during later processing.
4.3 Permeabilization and RNA capture
Permeabilization exposes RNA molecules so they can be detected or captured without destroying the tissue’s structure. The timing of this step is critical: insufficient treatment reduces capture, while excessive treatment can blur spatial detail. Optimized conditions improve both signal and localization.
4.4 Library preparation and sequencing
For sequencing-based assays, captured RNA is converted into libraries that can be read on sequencing instruments. The process usually includes reverse transcription, amplification, and adapter addition. In imaging-based workflows, this stage is replaced by iterative probe hybridization and optical readout.
4.5 Quality control and normalization
Quality control evaluates RNA yield, tissue morphology, background contamination, and coverage across the section. Normalization methods help adjust for differences in capture efficiency, sequencing depth, or imaging intensity. These steps are essential for reliable comparisons across regions or samples.
5 Data analysis
Spatial transcriptomics generates data that combine molecular counts with images or coordinates, so analysis requires both computational biology and image processing. The goal is to translate raw assay output into biologically meaningful patterns such as cell neighborhoods, tissue domains, or disease-associated zones. Because platforms differ, analysis pipelines are often customized.
5.1 Image processing and tissue registration
Microscopy images are commonly aligned with expression data to determine where each signal originated. Registration methods correct for section distortion, rotation, and local tissue deformation. Accurate alignment is important for linking molecular measurements to histological features.
5.2 Spot or cell segmentation
Depending on the platform, data may be assigned to capture spots, whole cells, or smaller compartments. Segmentation separates meaningful units from surrounding background and can improve interpretation of mixed signals. In highly resolved assays, this step may require sophisticated image-based boundary detection.
5.3 Gene expression quantification
After spatial assignment, transcripts are counted and organized into matrices for statistical analysis. Researchers may examine individual genes, gene sets, or pathway-level signatures. Quantification can reveal gradients, localized activation, or region-specific expression programs.
5.4 Spatial clustering and domain detection
Spatial clustering groups nearby regions with similar expression profiles, often revealing tissue zones or microenvironments. Domain detection methods identify boundaries between functional areas without prior labeling. These analyses are useful in organs with layered organization or heterogeneous pathology.
5.5 Cell type deconvolution
When spatial spots contain more than one cell, computational deconvolution estimates the mixture of cell types contributing to each location. This often uses reference profiles from single-cell sequencing. The result helps reconstruct tissue composition while preserving the original map.
5.6 Integration with other omics data
Spatial transcriptomics is frequently combined with proteomics, epigenomics, genomics, or single-cell data. Integration can strengthen cell-type assignment and uncover relationships between expression, chromatin state, and tissue context. These multimodal studies are increasingly common in systems biology.
6 Applications
Spatial transcriptomics is used wherever tissue organization shapes biological function. It is particularly valuable for studies in which neighboring cells differ in behavior, or where small anatomical changes carry major functional consequences. The technique has become important in both basic research and translational studies.
6.1 Cancer biology
In cancer research, spatial transcriptomics helps characterize tumor heterogeneity, stromal neighborhoods, and invasive fronts. It can reveal how malignant cells relate to immune cells, blood vessels, and surrounding tissue. This makes it useful for studying tumor ecosystems and local gene programs.
6.2 Neuroscience
The nervous system has strong spatial organization, making it a natural setting for these methods. Spatial transcriptomics can distinguish layered structures, nuclei, circuits, and cell populations within the brain. Researchers use it to study regional specialization, development, and disease-related changes.
6.3 Developmental biology
During development, tissues change rapidly in both structure and gene expression. Spatial methods help track these changes across embryonic or regenerating tissues, showing where specific programs are activated. They are especially useful for mapping differentiation trajectories in anatomical context.
6.4 Immunology
Immune responses are often organized into localized cellular communities. Spatial transcriptomics can identify immune infiltrates, activation zones, and tissue-resident immune populations. It also helps clarify how immune signaling varies across microenvironments.
6.5 Plant and microbial research
In plants, spatial methods can reveal gene expression across leaves, roots, flowers, and vascular tissues. In microbiology, adapted approaches may examine structured communities or host-associated microbes. These studies highlight how spatial organization influences metabolism, defense, and interaction.
7 Advantages and limitations
Spatial transcriptomics offers a distinctive view of biology by combining molecular profiling with tissue geography. At the same time, the methods involve trade-offs in resolution, sensitivity, and experimental complexity. Understanding these strengths and constraints is important for choosing an appropriate platform.
7.1 Strengths of spatial context
The main advantage is preservation of anatomical information. Spatial data can distinguish adjacent regions with different functions, identify rare local populations, and support interaction analysis. This context often clarifies findings that would be ambiguous in bulk measurements.
7.2 Technical constraints
Spatial assays may require specialized equipment, carefully optimized protocols, and substantial computational support. Tissue damage, uneven capture, and imaging artifacts can affect results. Platform-specific limitations also make cross-study comparison challenging.
7.3 Resolution and sensitivity issues
Higher spatial resolution often comes at the cost of reduced transcript detection or lower throughput. Some methods capture broad regional patterns but cannot resolve individual cells, while others achieve fine detail but measure fewer genes. Sensitivity can also vary by tissue type and RNA quality.
7.4 Cost and throughput considerations
Many spatial workflows are more expensive than standard expression assays because they involve advanced instrumentation, sequencing, imaging, or custom reagents. Throughput depends on tissue size, assay design, and analysis capacity. These factors influence whether a method is practical for large studies or routine use.
8 Future directions
The field continues to move toward finer resolution, better multiplexing, and more routine use in research and clinical settings. Future developments are likely to reduce technical barriers and improve the ability to study tissue biology in three dimensions, across time, and in combination with other molecular layers.
8.1 Higher spatial resolution
Ongoing innovation aims to map gene expression closer to the level of single cells and subcellular compartments. Improved probe design, imaging optics, and barcoding density are likely to sharpen localization. Higher resolution should make it easier to study cell boundaries and local signaling.
8.2 Multi-omic integration
Future platforms will increasingly combine RNA measurements with protein, chromatin, metabolite, or lineage information. Multi-omic spatial data can provide a more complete view of cell state and tissue function. Such integration may also improve interpretation of complex tissues.
8.3 Clinical and diagnostic uses
Spatial transcriptomics has potential in pathology, biomarker discovery, and treatment stratification. In clinical contexts, it may help characterize tissue architecture in disease and reveal molecular features not visible by morphology alone. Broader adoption will depend on reproducibility, speed, and ease of analysis.
8.4 Automation and standardization
Automation could make sample handling, imaging, and data processing more consistent across laboratories. Standardized protocols and reference materials would support comparability and reduce technical variation. These changes would help move spatial transcriptomics from specialized studies toward broader routine use.