Single-cell approaches such as single-cell RNA sequencing (scRNA-seq) are essential to detect biologically relevant variability between cells or define the characteristics of rare cell types. Adapted methods such as in situ RNA-seq have tackled the challenge of capturing spatial context, but have yet to be applied to whole tissues owing to limitations such as low efficiency and scalability. A newly developed technology for 3D intact-tissue RNA-seq, named STARmap (spatially-resolved transcript amplicon readout mapping), promises to yield gene expression profiles while retaining 3D positional information at cellular resolution.

Recently developed in situ hybridization methods enable high-resolution imaging of RNAs within intact tissues by exploiting hydrogel-tissue chemistry (HTC) to link in situ-synthesized polymers with RNAs. Wang et al. hypothesized that exploiting HTC to convert a tissue into a hydrogel-embedded form might enable the application of in situ RNA-seq to the hydrogel-tissue formulation.

Credit: JG Photography/Alamy

In STARmap, to start, all cellular RNAs are labelled with two probes, one of which contains a five-base barcode, providing a gene-unique identifier for later multiplexed gene detection. To reduce noise, both probes must hybridize to the same RNA molecule for enzymatic amplification to occur, which generates a DNA nanoball (amplicon) that contains multiple copies of the cDNA probes. DNA amplicons are then anchored to an in situ-synthesized polymer network before removing proteins and lipids. This process transforms the tissue into a 3D hydrogel–DNA chip that can be used for sequencing. The identities of RNA transcripts, represented by DNA amplicons, are identified and quantified on the basis of the five-base barcode using a sequencing-by-ligation method with two-base encoding for error reduction, termed SEDAL (sequencing with error-reduction by dynamic annealing and ligation). SEDAL decodes DNA sequences into multi-coloured fluorescence signals that can be imaged. The authors validate their method by efficiently and reproducibly mapping 160 to 1,020 genes simultaneously in sections of mouse brain at single-cell resolution.

By combining HTC, targeted signal amplification and in situ sequencing, STARmap enables not only the quantification of gene expression in single cells but also the identification and mapping of cell types in 3D.