Single Cell Technologies in Modern Core Facilities
Complementary Technologies Are Underpinning Single Cell Research
- Fig. 1: Looking at cell division using the Amnis ImageStream. Using a combination of Propidium iodide staining (for DNA; red) and mpm2 staining (for mitotic cells; green), it is possible to identify cells in prophase (upper panel), metaphase (middle panel) and anaphase (lower panel)
- Fig. 2: Data generated using the Hyperion Imaging mass cytometry system Human tonsil was stained with 35 differently labelled antibodies. False-color imaging of 4 of these is shown. Smooth muscle actin (red), CD31 (green), DNA (Blue) and CD45RA (Cyan). Left panel is slide overview, middle panel is zoomed in and right panel is after cell segmentation using Ilastik and Cell Profiler – this then allows quantitation of signal at the single cell level.
- Fig. 3: Dimensionality reduction of a 35-metal panel using withing the Phenograph package.
Being able to make measurements on a single cell level allows us to link genotype, phenotype and function. Several single cell technologies exist and there is now more of a need to integrate several of these into an analysis workflow. The complementary nature of flow cytometry, single cell sequencing and genomics platforms allows researchers to make more measurements than ever before. However, to do this knowledge of the advantages and disadvantages of each approach is vital especially as there is, of necessity, more of a link-up between these technological approaches.
The explosion of information that can be derived for specifically defined cells has great possibilities in biomarker discovery, therapeutic intervention and personalised medicine.
But to determine the function, genotype, and phenotype of an individual cell, it is essential to integrate several technological approaches. No one approach will fit all as it may depend on whether simple single cell information is needed (suspension-based methodologies) or whether there also needs to be a relational aspect i.e. the context of the cells in a tissue (tissue- or slide-based methodologies). This allows to gain a greater understanding of the complex nature of cells reactions to stimuli, cell to cell interactions and cell fate.
Cells are still identified based on some a priori knowledge of the cell system and its response to, for example, drug treatment. True discovery needs a high throughput mode that is not present in most of the single cell technologies discussed here. There are a number of techniques available and the one used will depend on the question asked and, more and more, researchers rely on integrated results from several approaches.
Flow cytometry is a well-established technique that uses fluorescence (labelled antibodies, fluorescent probes or proteins) to enable researchers to specifically identify sub-populations within a heterogeneous population. It has been established for almost 50 years and has been embraced particularly by the field of immunology. The ability to measure structural parts of a cell (protein, DNA, RNA) and cell function (pH, calcium flux, apoptosis) on a cell-by-cell basis makes it an extremely powerful technology.
Additionally, any sub-population identified by its fluorescence characteristics may be retrieved using a cell sorter. Sorted cells may be passaged, transplanted, put into functional assays or may be subject to downstream genetic, proteomic or metabolomic analysis.
Flow cytometry is still considered to be the gold standard of high parameter single cell analysis technologies . Throughput is high – 10k/second is relatively easy to achieve, but can be high cost given the amount of reagents needed. Currently, it is not easily possible to examine more than 30 parameters per cell which can be limiting and challenging in looking at highly heterogeneous populations. Cells have traditionally been sorted using electrostatic drop deflection but given the high pressures often used, low pressure microfluidics system such as the Miltenyi Biotec Tyto or the Cytonome microfluidic sorter can have advantages where subsequent functional studies are dependent on cells being as little stressed as possible. Single cell sorting for subsequent genomic, transcriptomic or proteomic analysis is also possible using an electrostatic drop deflection cell sorter.
Recently suspension mass cytometry (Fluidigm Helios) which uses metal-tagged antibodies to identify cells of interest has emerged. Although throughput here is slower, the number of analytes is increased compared with flow cytometry so allowing even greater sub-divisions of heterogenous populations.
Also, in recent years, imaging flow cytometry has allowed the positional information of fluorescence and the shape of cells to be captured as well as whole cell fluorescence. This is especially useful when looking for translocation or localisation of fluorescence, and because it is flow-based also allows thousands of cells per second to be analysed making it suitable for rare cell analysis.
Tissue Based Assays
Most single cell suspension approaches do not reveal spatial information either at the subcellular level or between cells as they need to be disaggregated. The Laser Scanning Cytometer was introduced in the late 1990s as a way of deriving subcellular information from single cells on a substrate, and recent iterations of this technique has been used to analysis tissue sections but have a limited number of targets. To address tissue cytometry needs, in recent years there has been an expansion of new techniques each with their own advantages, that are compatible with standard tissue sectioning techniques i.e. cryopreserved or formalin fixed paraffin embedded (FFPE), material which is critical for their adoption.
Nanostring Digital Spatial Profiling (DSP) can analysis tissue sections using a panel of either antibody or oligonucleotide probes that are marked with unique photo-cleavable oligonucleotide identifiers. Regions of interest down to a single cell are targeted with UV light releasing identifying oligonucleotides, which are collected and identified using either a multi-coloured fluorescent bar code system (up to 800 plex) or next generation sequencing (NGS) (kilo plex).
The Akoya Codex system can analysis tissue sections using a panel of antibodies that are labelled with an oligo duplex with unique 5’ overhangs. Antibody identity can be determined using iterations of single base extension to sequential identify oligo duplexes from the panel. Three antibodies can be identified per iteration by the addition of fluorescent nucleotides and their location within the section captured using standard fluorescence microscopy. Fluorescent dye molecules are then removed before the next iteration of extension and labelling. Antibody identity is determined by the iteration number and colour.
There are also techniques where the sequential labelling of tissue sections with fluorescent tagged antibodies (3 to 5 different colors) is performed in a single iteration. Images are captured using standard fluorescence microscopy followed by the removal or deactivation by photo-bleaching of fluorescence tags before the next labelling iteration. Zellkraftwerk ChipCytometry mounts a sample in a special flow cell that allows the exchange of solutions and Miltenyi have recently launched the MACSima system.
Hyperion mass cytometry system (Fluidigm) can be used to analyse sections with a panel of up to 40 different antibodies each coupled with a unique metal tag that are not common in nature giving very low background. A scanning laser beam is used to ablate the sample causing the release of particles that are carried by a stream of inert gas into a mass spectrometer for identification. This technique is capable of subcellular resolution down to 1 µm.
With Nanostring technology, conceptually thousands of different proteins or RNA can be labelled simultaneously, but not with subcellular detail. Whereas both Codex and ChipCytometry approaches can yield subcellular information, their iterative approaches requires significantly longer processing time, limiting the total number of targets and are dependent on image analysis software to identify cells and their labelling signatures. Tissue-based mass cytometry (Hyperion) is less susceptible to background noise but requires specialized equipment and currently is limited in the total number of targets. In many tissue-based techniques there is a need to identify individual cells using algorithms to separate cellular information from background allowing segmentation of single cells.
There is always a compromise – ideally, phenotype, function and positional information are wanted. No one technique does all of that to the same level of information/resolution, so prioritization may be necessary depending on the biological question being asked.
The challenges in the era of high-dimensional techniques are in data analysis as directed analysis is no longer possible. More and more operational pipelines have been published in general involving pre-processing of samples to remove unwanted and/or spurious events, this will be different with each methodology. Data analysis of highly multiplexed samples is a challenge and there is an increasing involvement of clustering or dimensionality reduction algorithms used in flow, imaging and mass cytometry data, although this has been common in sequencing for some time.
However, it is clear that emerging technologies such as those described here will be the driver in addressing tissue heterogeneity in health and disease.
Derek Davies1, Grant Calder2, Peter O’Toole2
1Derek Davies, The Francis Crick Institute, London UK
2Grant Calder, Bioscience Technology Facility, Department of Biology, University of York, UK
Director of the Bioscience Technology Facility
University of York
 Chattopadhyay PK, Winters AF, Lomas WE, Laino AS, Woods, DM. High-pararmeter single-cell analysis. Ann Rev Anal Chem 2019 12:2.1-2.20