The terms “spatial” or “spatially-resolved data” have been increasingly thrown around in the field of single-cell studies in the past few years – with spatial transcriptomics being celebrated as Method of the Year 2020 by Nature Methods – and for excellent reasons.
The spatial dimension added to single-cell resolution methods is a game-changer
The spatial localization of individual cells and particular cell subtypes within tissues or even the construction of whole organs uncovers vital relationships between histology and molecular functions. Going further, we can already observe the compartmentalization of specific molecules, from lipids to metabolites and mRNA, across every single cell in samples counting tens of thousands of the latter. We have reached the point where we can build an accurate model of an adult mouse brain and all its individual inner tissues purely based on individual cells’ mRNA content and completely detached from anatomical data. With spatial data reaching the single-cell and even single-molecule resolutions, how far are we in early 2021 and what kind of data can we currently generate? Here is a highlight of the available techniques to potentially include in your research study.
Have a look at what it took to build a model of a mouse brain only from sequencing data!
Single-cell spatial transcriptomics has bloomed thanks to high-throughput technologies
Spatial scRNA-seq, linking the spatial coordinates of a cell within a tissue layer to its individual mRNA content, was definitely in the spotlight of the recent progress in single-cell spatial data. Newest technologies require a tissue section to be layered on top of a glass slide with barcoded spots (Visium technology) or monolayered barcoded beads (Slide-seq). While they currently have technical limitations – for Visium, the current size of a barcoded spot can cover several cells and for Slide-seq the low amount of barcodes on a bead limits the RNA capture – small innovative increments have brought their resolution closer to scRNA-seq achieved through classic microfluidics. High-résolution histological imaging of the tissue layer on top of the slide complements the data provided by RNA-sequencing to increase the resolution and correctly maps transcripts to their individual cells. Already newer, more performant, instruments have been announced for release later this year.
Access to those high-throughput methods enabled large-scale spatial RNA-sequencing projects, with the creation of atlases covering the single-cell RNA analysis of entire organs. Following the spatial mapping of heart and brain tissues, the Human Cell Atlas project is currently aiming at building 3D models of 22 organs and systems based on scRNA-seq. Beyond tissues in normal physiological conditions, researchers use now spatial mapping to observe the three-dimensional progress of diseases in parallel with the identification of the molecular mechanisms and the differential gene expression in the affected cells.
Single-cell spatial transcriptomics by in situ imaging provides a snapshot of individual mRNAs molecule directly in tissues
If you are searching to localize known RNA transcripts within individual cells, Fluorescence in Situ Hybridization (FISH) has made tremendous progress to reach single-cell resolution. In principle, multiple fluorescent hybridization probes binding to a single specific RNA sequence of interest generate sufficient fluorescent signal to be picked up, but the technology was initially restrained to a low number of different transcript sequences due to limited color channels.
The recent introduction of MERFISH, using multiple rounds of hybridization (aka multiplexing) now enables the identification, localization, and quantification of thousands of sequences at the same time. In short, the addition of secondary readout probes binding to the primary encoding probes and successive rounds of excitement at different wavelengths permits the creation of complex barcodes for each transcript sequence, thus covering a much larger section of the transcriptome. Using the latest advancements with MERFISH, scientists demonstrated the simultaneous measurements of RNA transcripts from ∼10,000 genes in individual cells with ∼80% detection efficiency and ∼4% misidentification rate. MERFISH has also now also been applied to DNA spatial imaging, monitoring chromatin-dependent transcriptional activity.
Another alternative for in situ hybridization is Seq-FISH. Seq-FISH relies on hybridization probes each with their unique color sequence, which can be decoded through multiples round of imaging with the need to for secondary readout probes. Like MERFISH, Seq-FISH can also cover up to 10,000 mRNAs molecules for transcriptome-wide studies directly in tissue samples bypassing the need for cell separation and isolation.
Spatial data from Mass Spectrometry Imaging provides previous insights in single-cell proteomics and metabolomics
Using mass spectrometry at the single-cell resolution has been ongoing for decades, but the output was limited mainly due to the size sensitivity of the instruments and techniques. The wide range of sizes from small mRNA strands to massive proteins meant that researchers had to make choices for the targets to detect. Moreover, there are – still – issues linked to the respective abundances of molecules present in the cell of interest. Lipids and metabolites might be present in quantities large enough to be easily detected, but peptides or small RNAs are much trickier.
Nevertheless, with recent advances in hardware (such as detectors) and advanced protocols (multiplexing once again), single-cell Mass Spectrometry Imaging (MSI) was successfully used to reveal dynamic cellular processes, from biosynthesis to the activation or inhibition of signaling pathways or the impact of diseases on metabolic processes.
We made lists of suggestions for open-source visualisation software for your spatial sequencing and in situ visualization data!
Overview of single-cell and single-molecule resolution spatial technologies
A non-exhaustive list of existing methods to gather spatial data, depending on the target molecule, technologies, resolution and coverage. Adapted from Allam et al.
Accurate spatial visualization of multimodal single-cell data is an upcoming challenge
Filtering, normalization, processing, and interpretation of single-cell data is already a complex task due to the sheer size of the datasets and the requirement to allocate tens, if not hundreds, of thousands of entries to their original cells. The spatial dimension adds an additional complexity to the dataset, which is even multiplied for multi-modal / multi-omics studies. While the combination of approaches provides a much more complete picture to understand complex physiological processes, there is an obvious need to be able to visualize the multimodal data from all angles without losing the depth of information.
Spatial visualization tools already exists: some are proprietary of the companies selling spatial-oriented instruments, some are recent inclusions into single-cell analysis dedicated software (such as Seurat), and some such as Vitessce or GLISS are being developed concomitantly by academic researchers. However, looking at the ever-increasing scales of the current projects, a “gold standard” platform has yet to emerge to harmonize the integration and interpretation of such complex, multi-modal datasets.
Bibliography
- Allam, S. Cai, A. Coskun, Multiplex bioimaging of single-cell spatial profiles for precision cancer diagnostics and therapeutics npj Precision Oncology 4, 11 (2020).
- Larsson, J. Frisèn, J. Lundeberg, Spatially resolved transcriptomics adds a new dimension to genomics. Nature Methods 18, 15–25 (2021).
- The Human Cell Atlas Project, https://www.humancellatlas.org/ (accessed March 2021).
- Taylor, J. K. Lukowski, C. R. Anderton, Spatially Resolved Mass Spectrometry at the Single Cell: Recent Innovations in Proteomics and Metabolomics Journal of the American Society for Mass Spectrometry, ahead of print (2021).
- Xia, J. Fan, G. Emanuel, J. Hao, X. Zhuang, Spatial transcriptome profiling by MERFISH reveals subcellular RNA compartmentalization and cell cycle-dependent gene expression PNAS 116 (39), 19490-19499 (2019).
- Zhuang, Spatially resolved single-cell genomics and transcriptomics by imaging. Nature Methods 18, 15–25 (2021).
