New Technique Enables Precise Location of gene-editing Events Inside Intact Tissue

Historically, there has been a trade-off between resolution (single cell) and context (spatial/tissue). Many gene-editing detection methods work on dissociated cells (losing spatial context) or bulk tissue (losing single-cell resolution). A recently published paper in Nature Biomedical Engineering (paid access) solves this problem using a technique called in situ sequencing (ISS) where they describe a time saving way to “read” genetic edits in single cells without ever separating cells from each other.

A Swiss team led by Sharan Janjuha and Tatjana Haenggi adapted ISS to read CRISPR-derived base and prime edits directly inside fixed tissue sections at single-cell resolution. Instead of dissociating tissue (losing spatial context) or microdissecting single cells (slow, low throughput), they use short targeting oligos + padlock probes that circularize only when the local genomic sequence matches an edited allele, amplify those circular products in situ, then sequence them by sequencing-by-synthesis chemistry inside the tissue. They validated the method in mouse brain (AAV-delivered editors) and in mouse + macaque liver (RNA-LNP delivered ABE), finding cell-type differences (e.g., neurons > astrocytes in brain) and broadly uniform hepatocyte editing across metabolic zones in macaque liver; repeated RNA-LNP dosing did not change editing distribution. This gives spatial maps of where edits actually occurred at a fraction of cost and time.

Detailed Comparison of the Different Approaches to Detect Base Edits

Feature	Bulk Sequencing (“Blender”)	Laser Microdissection	In Situ Sequencing (ISS, new method)
Sample prep	Whole organ homogenized; DNA extracted and sequenced	Tissue cut into thin slices; individual cells excised by laser	Tissue fixed on slide; cells remain in place
Resolution	Organ-level average (no spatial or cell-type resolution)	Single-cell, but only for selected cells (low throughput)	Single-cell resolution across entire tissue
Spatial context	Lost (cells mixed together)	Preserved (but only for dissected cells)	Fully preserved across intact tissue
Throughput	High (millions of reads from whole organ)	Very low (cell-by-cell cutting, slow)	High (entire tissue sections can be mapped in one run)
Quantitative accuracy	Accurate allele fractions at organ level	Accurate at single-cell level, but laborious	Accurate allele-specific detection in situ; quantitative with spatial maps
Targeting	Genome-wide (any edits detectable by sequencing)	Genome-wide for the cells isolated	Targeted to specific loci (padlock probes must be designed for sites of interest)
Labor intensity	Low (routine sequencing)	Very high (tedious microdissection, few cells)	Moderate (padlock probe design + ISS workflow, but scalable)
Use case	Benchmarking overall editing efficiency	Checking edits in selected cells / regions	Mapping editing efficiency & distribution across all cells in intact tissue
Limitations	No cell-type info; no spatial distribution	Extremely tedious; not scalable	Only detects pre-defined loci; needs probe design; not yet genome-wide
Cost	Low	High	Medium

Strengths of the approach

• Spatial + single-cell resolution of edits without single-cell dissociation. That directly addresses the “blender vs laser-microdissection” tradeoff described in the tweet. Nature
• Allele-level specificity: leverages padlock probe ligation to distinguish edited vs wild-type alleles (even single-nucleotide differences).
• Cross-species validation: applied to both mice and non-human primates (macaque), which strengthens translational relevance. PubMed+1
• Multiplex potential: in principle, multiple targets can be probed in the same tissue, enabling richer spatial perturbation maps.

Limitations & technical caveats

• Targeted (not whole-genome). ISS here is targeted — you must design padlock probes for specific loci. It won’t discover unknown off-target edits genome-wide; it quantifies predefined sites. Nature
• Sensitivity & false negatives. Padlock ligation and on-tissue amplification depend on probe access and tissue quality; some edited alleles may be missed if probe binding or ligation is inefficient.
• Signal interpretation. Detecting an edit in nuclear DNA inside a cell doesn’t necessarily indicate functional correction (e.g., low allelic fraction or mosaicism within a multinucleated cell may complicate interpretation).
• Tissue preservation constraints. Fixation / embedding protocols must preserve DNA accessibility for the padlock chemistry; not all clinical samples will be compatible.
• Quantitation vs sequencing depth. ISS reads are localized signals, not full-length molecule sequencing; they’re excellent for presence/absence and allele fraction estimation at the probed site but do not replace deep NGS for discovering complex indels or structural variants.

Practical implications for research & therapeutics

Regulatory relevance. Spatial evidence of editing distribution and repeat-dose behavior will be persuasive and relevant to safety discussions with regulators.

Better in vivo benchmarking. Researchers can now map which cells in an organ actually receive and execute editing, enabling more rigorous claims about biodistribution and efficacy. That matters for preclinical work and for safety/efficacy assessments in therapeutic pipelines. Nature

Optimizing delivery vehicles. Seeing where LNPs or AAVs deliver editors at cellular resolution helps improve targeting strategies (capsid selection, LNP formulation, dosing). The paper’s finding of uniform hepatocyte editing in macaques supports broad metabolic-liver-targeted therapies. ResearchGate

Assess off-target risk spatially. While targeted ISS doesn’t find unknown off-targets, it can be used to monitor known candidate off-target sites in situ, revealing whether potential off-targets appear in particular cell types or regions.

Check out the full article here.