DEMOCRATISING SPATIAL TRANSCRIPTOMICS WITH AUTOMATED IN SITU SEQUENCING

Spatial transcriptomics have revolutionised our ability to study the immense diversity of cell types in the brain within their tissue context. However, high running costs of commercial platforms limit the scalability of spatial transcriptomics methods. BARseq, a custom open-source approach, enables expression readout of hundreds of target genes along with arbitrary barcode sequences at low cost. BARseq uses sequencing by synthesis, whereby fluorophore-conjugated nucleotides are incorporated and imaged one base at a time. These reactions must be carried out by hand at every sequencing cycle, limiting the scalability and reproducibility of BARseq experiments. To address these limitations and make in situ sequencing more accessible, we developed a platform that performs tissue handling and imaging to read out the expression of target genes and barcode sequences in an automated manner.
To maximize throughput, we have designed custom flowcells that can hold 25 x 75 mm of tissue while minimizing aberrations during imaging. During sequencing, our system automatically alternates between nucleotide cleavage, incorporation reactions, and imaging on up to four flowcells in parallel. The reagents are delivered by microfluidic hardware, and the samples are maintained at the required temperature by a custom-made heating unit. To reduce imaging time, the four channels associated with the four DNA bases are acquired simultaneously using an epifluorescence microscope. This system lowers the barrier for entry into in situ sequencing, making it possible to investigate the diversity of cell types in the brain within their tissue context in a scalable and reproducible manner.