Development of 3D MERFISH in Drosophila

Abstract

In multicellular organisms, biological processes ranging from development to homeostasis and disease rely on both the composition and the precise spatial organisation of cells. To understand these processes at the tissue or organ level, it is crucial to study cellular activities and interactions within their native spatial context. While single-cell RNA sequencing provides high-dimensional transcriptomic profiles, it lacks the critical spatial information. Multiplexed error-robust fluorescence in situ hybridisation (MERFISH) is an advanced imaging-based spatial transcriptomic technique that enables high-throughput, single-cell-resolution measurement of hundreds to thousands of RNA species within intact tissues. This report evaluates an improved Indirect-BarCoding (IBC) design for 3D MERFISH to enhance library flexibility and sensitivity. The IBC approach successfully targets transcripts, but its current implementations produce a signal intensity of only ~70% of classical smFISH and increasing to 3-level amplification resulted in non-specific background and only a marginal brightness increase. Most likely reason behind is probe cross-hybridisation, thus requiring further refinement in terms of hybridisation kinetics of IBC. Furthermore, we optimised the Cellpose-SAM (CP-SAM) pipeline for cell segmentation, implementing an anisotropy correction of 3.07 and fine-tuning scaling parameters. This provides a robust foundation for 3D cell segmentation to assign transcripts to precise cellular boundaries.