Spatial Epigenomics: Mapping Modifications In Situ

Chromatin state is usually measured on dissociated cells, which removes the one thing that defines a tissue: where each cell sits. In a tumour, a developing organ, or a fibrotic lesion, cells a few tens of micrometres apart can occupy different regulatory states while looking identical after dissociation.

Spatial epigenomics attempts to keep the coordinate attached to the mark. The field is young, and the honest framing is that methods currently deliver a coarse-to-moderate spatial picture of a subset of marks — useful for discovery, insufficient for mechanism on its own.

The fourth row is unglamorous and frequently the best choice: laser-capture microdissection of regions, followed by conventional ATAC or CUT&RUN, gives high-quality mark data at the cost of region-level rather than single-cell resolution.

Spatial ATAC produces, per binned area, an accessibility profile. In practice the useful readout is often a module score — for example, the aggregate accessibility of marker-defined enhancer sets — rather than per-element calls, because per-element signal at these resolutions is sparse.

Spatial CUT&Tag gives one histone mark at a time across a section. It is powerful for questions of the form "where in this tissue is the repressive domain", and weak for questions about which specific element changes, because coverage per bin is limited.

Imaging-based methods resolve marks at optical resolution and can multiplex a few marks with RNA. They report abundance and localization of a mark, not genome-wide distribution; the readout is a structure, not a sequence.

Microdissection plus sequencing remains the reference against which the others are best validated.

Every spatial method sits somewhere on a triangle: resolution, genome coverage, and mark multiplexing. Improving one typically costs another.

Practical consequence for design: decide which vertex your question needs.

• • "Which regions of this tumour have a repressed chromatin state?" → low resolution, one mark is enough.
• • "Which cells at the invasive edge lose an enhancer mark?" → moderate resolution, one mark, careful registration to histology.
• • "Does this specific element change accessibility in the fibrosis zone?" → microdissection plus bulk, or spatial ATAC with deep sequencing and region-level statistics.

A common failure is asking a per-element question of a per-region dataset; the result is underpowered and the conclusion is usually "we observed a trend".

• • Registration error. Aligning the barcode array to the histological image dominates spatial error; a few hundred micrometres of misalignment can reverse a conclusion about which region changed.
• • Bin size as a hidden parameter. Conclusions must be shown to be stable across reasonable bin sizes.
• • Normalization across regions. Differences in cell density, nuclei capture, and RNA/chromatin content between regions mimic biological differences.
• • Pseudoreplication. Bins within a section are not independent replicates; the replicate unit is the section or the donor.
• • Mark abundance versus mark distribution. An imaging signal for H3K27me3 intensity does not tell you which loci carry it.

Three applications where spatial epigenomics already earns its place:

• • Tumour architecture. Mapping accessibility or repressive marks across a section to define niches that differ in regulatory state, then testing candidate states in matched cell models.
• • Developmental anatomy. Regional chromatin differences in an organ primordium where dissociation loses positional information.
• • Fibrosis and inflammation. Localized chromatin changes in a lesion that bulk tissue would average with adjacent normal regions.

In each case, the spatial dataset is the discovery layer, and mechanism comes from perturbation in a model system.

The last row is increasingly expected: a spatial claim supported by a second modality is far more robust than one resting on a single barcode array.

EDITGENE builds the cell models that follow spatial discovery: region-matched primary-like or iPSC-derived cells, CRISPRi/a-ready lines for element tests, and epigenome-editing lines for mark-level experiments.

Share the tissue context and the candidate regions. Contact EDITGENE to request a project consultation or quote.

Spatial epigenomic methods differ substantially in resolution, coverage, and mark multiplexing; reported performance depends on tissue type, section quality, and analysis choices.

Spatial patterns describe where a mark or accessibility state is enriched, not which loci carry it or what it does. Mechanistic claims require locus-level data and perturbation.

• • Deng Y, et al. Spatial-CUT&Tag: spatially resolved chromatin modification profiling at the cellular level. Science. 2022;375:681–686.
• • Deng Y, et al. Spatial profiling of chromatin accessibility in mouse and human tissues. Nature. 2022;609:375–383.
• • Llorens-Bobadilla E, et al. Solid-phase capture and profiling of open chromatin by spatial ATAC. Nature Biotechnology. 2023;41:1085–1088.
• • Takei Y, et al. Integrated spatial genomics reveals global architecture of single nuclei. Nature. 2021;590:344–350.
• • Moses L, Pachter L. Museum of spatial transcriptomics. Nature Methods. 2022;19:534–546.