Cellular exploration and examination are the foundation of biological research. However, the predominate amount of biological knowledge stems from the traditional approach of studying cell populations rather than the most fundamental unit of biology, a single cell. In this format cell populations are assumed to be homogenous in physical features, and genomic expression and mean averaged measurements can often obscure single-cell differences [2, 3]. Intrinsically, numerous critical areas of biology require studies that can only be addressed at the single-cell level such as stem cell differentiation and disease progression [4, 5]. It has been calculated that there are ~1.3 mutations per cell per division during the first five human postzygotic cleavages [6]. Somatic mutations increase during aging and neurodegeneration [7]. Improvements in biological and physical techniques have advanced single-cell studies and shed new insight into these areas and many others with potential for far greater impact [8]. At the single-cell level of exploration numerous differences can be seen and categorized. As the field develops these differences have been shown to be catalysts [9, 10], as well as regulators [11], of physiological and biological processes.

Advances in technology and analysis techniques have been a driving force in the single-cell field [12]. Previous theories are now testable in a timely and cost-efficient manner. Imaging techniques allow a much greater array of classification by precipitous nature of fluorescent secondary labeling and now commonplace transfection of GFP and other traits into cell lines. Such transfections can now be achieved on a single-cell level through virus stamping [13].

While most single-cell technologies require the isolation or separation of individual cells, combinatorial barcoding can work directly on a pool of cells [14]. Sometimes, nuclei are isolated in stead of cells and subjected to further analyses, such as single-nucleus RNA-seq [15-17], to minimize any contamination from other cells or degradation encountered with whole-neuron dissociation or laser caption micro-dissection [17, 18].

A primary limitation of single-cell studies is ensuring that the immense amount of generated data can be logically and successfully analyzed. This is especially true in single-cell proteomic and genomic analysis. As the techniques continue to be refined the ability to fully interpret data in a systematic and meaningful way will be of utmost importance. Monitoring single cells through image acquisition is rapidly increasing the experimental methods that are available to researchers. Open source and proprietary software can compile numerous images and generate comprehensive time-lapse images which lead to more intricate techniques becoming readily reproducible. Table 1 shows some of the key approaches being currently utilized along with an explanation of the relevant data produced and associated biological area of study.

Individual cells can often be picked up under microscope manually [19]. Beyond the manual approach, specific sorting methods are a critical component of single-cell strategies. The advancements of microfluidics have greatly increased the throughput of these methods, increasing the numbers of cells capable of being analyzed to relevant levels [20]. Fluorescent Activated Cell Sorting (FACS) is an established method and is effective for a broad range of applications [21]. However, new microprinted methods such as microraft arrays are rapidly contributing to the field and appear to be a current method of choice [22] (Table 2).

FACS is a specialized version of flow cytometry. It became a standard tool in biology for sorting live cells with a ready source of distinct, monoclonal antibodies [21]. Antibodies are linked to a fluorophore and desired markers are screened for in a cell suspension. The cells are sorted by passing through a laser to measure the fluorescent properties based on the binding of antibodies and diverted into respective categories [23]. The flow rate is critical in FACS to ensure that only one cell passes the laser at any given time. The number of distinct fluorophores available limits the number of parameters that can be screened for at any one time. Sorting NeuN-stained cells through FACS is a common approach for single neuron studies [24]. Multiple markers and dyes can be used to sort cells with unique marker expression. For example, B Shen et al isolated bone marrow haematopoietic cells and spleen cells through multiple markers [25]. Steichen JM et al sorted mouse splenocytes with a cocktail of monoclonal antibodies: "CD38 Alexa Fluor 700 (Clone 90, Invitrogen), CD45.2 PE (Clone 104, Biolegend), CD45.1 PerCP-Cy5.5 (Clone A20, Biolegend), B220 PB (Clone RA3-6B2, Biolegend), CD95 PE-Cy7 (Clone Jo2, BD Bioscience), CD4 APC-eFluor780 (Clone RM4-5, eBioscience), CD8a APC-eFluor780 (Clone 53-6.7, eBioscience), F4/80 APC-eFluor780 (Clone BM8, eBioscience), Ly-6G APC-eFluor780 (Clone RB6-8C5, eBioscience)" for single-cell PCR [26]. K Yoshida et al sorted each DAPI−CD45−CD31−EpCAM+ bronchial epithelial cell for somatic mutational analysis [27].

Microrafts are micrometer size wells that on average have zero or one cell present when cell isolates flow over them. The method of attachment can be magnetic, polystyrene, or biological [28] and holds the cell securely in place. Upon identification of a cell of interest the cell can be dislodged, along with the microraft, using microneedle or other similar technique. Sizing is typically in the thousands of wells but can be expanded into the millions and possibly more if the need arises.

Microraft arrays have been utilized to sort adherent cells with high viability for implantation [29]. Relative to FACS, microraft arrays do not require a high cell population or flow rate; both of which can cause variance in cell phenotypes [30]. Due to this it is a viable option for all cell types including fragile ones. Additionally, it is ideal for stem cells and primary cells lines since fewer passages are necessary to reach appropriate concentrations preventing irreversible alterations to pluripotency, gene expression, and phenotypic heterogeneity [31, 32]. After sorting, microraft arrays can also aid in the implantation for wound healing and regenerative applications by serving as microcarriers [33, 34]. The use of microcarriers has been shown to increase healing of liver, heart, and lung tissue and microrafts help screen for adherent cells, which are necessary for effective implantation [35-38]. For an example, Kaya-Okur HS et al used the SMARTer ICELL8 single-cell system from Takara Bio to select single K562 cells for CUT&TAG experiments [39].

Microfluidics, such as droplet microfluidics (inDrops) [40], Fluidigm C1 system or 10X Genomics Single-Cell 3’ system [18, 41], is also an important tool for isolating single cells, such as those used for hybridoma screening [42] or for single-cell RNA-seq [41, 43]. Packer JS et al captured C. elegans single cells with a 10X Genomics system to identify 502 terminal and pre-terminal cell types and resolve the lineage of individual cells during embryogenesis [44]. Single-cell RNA-seq is discussed in the article RNA-seq. In addition, Butler A et al devised an analytical strategy for integrating single-cell transcriptomic data across different conditions, technologies, and species to identify shared cell populations [45].

Cell migration is critical in many dynamic processes such as angiogenesis, embryonic development, wound healing, and disease progression [46-49]. As previously stated, much of the current body of knowledge was generated from studies done in aggregate. In regards to cell locomotion though, only select cells are able to engage in free movement and are often led by a single cell, acting as a guide [50]. It has also been shown that single-cell migration, mimics the movement of cells in 3D environments [51].

[enlarge]

Figure 1. 1D single-cell migration studies use microcontact printed tracks for the cells to attach to and move along (A). Another approach suspends a single fiber of fibrillar fibronectin (any fibrillar matrix protein would work) creating a pseudo-3D environment for cells to attach to and migrate (B).

Several methods exist to study single-cell migration. A popular approach is using microcontact printed regions to create “cell tracks” [51, 52]. This approach allows easy exploration of geometric limitations on migration. Printed regions can have a diverse array of shapes and patterns to tightly control and explore broad-ranging conditions (Figure 1A). 1D track studies have been completed that show cells take on a distinct morphology and phenotype that closely parallels cells in native 3D matrix [51]. To fully explore this phenomenon researchers created different thickness “tracks” by microprinting and then coating with matrix protein fibronectin (Fn) [51]. This observation has been recapitulated in other studies, including ones done on single fibers of Fn in a pseudo-3D environment [53] (Figure 1B). Single-fiber studies utilized endothelial cells and demonstrated distinct cellular attachment and migratory differences to fibers of different mechanical strain. Tip cells follow Fn preliminary matrix to create blood vasculature in angiogenesis and vasculogenesis. The ability to receive nutrients and other critical biological components is imperative in the dynamic processes of wound healing, disease progression, and development. This finding in single-cell migration presents a theory for directed migration relevant to all these processes and mechanistic insight into durotaxis due to ligand density in Fn [53].

A different variety of single-cell migration assay is run by utilizing a microfluidic platform. Systems of channels are created and cells are trapped in certain areas and then allowed to migrate across a portion [54]. These systems not only filter the cells but also allow migratory examination. Cancer studies have been conducted within this system to determine differences in migratory ability in cancer cells to examine metastasis root causes. Additionally, this system has the added benefit of being able to explore directed migration such as chemotaxis [54] and also further examine cells with the preferential migratory ability for other key differences [54]. It has been shown that metastasizing cells have a greater response to chemotactic gradients and also display higher levels of GTPase markers than non-metastasizing cells [54].

Traditional methods for tracking cell migration can also be applied to single cells. The key factor for single-cell migrations is reducing cell-to-cell contact initially. This is achieved by plating low densities of cells onto coated surfaces [55]. The molecule selected to coat a surface is dependent on study parameters as with bulk migration studies. Since cells have a preference for adhesion [56, 57] there must be a thorough screening process to ensure selected migrating cells have a relevant and representative phenotype. Advancements in stage mobility and time-lapse imaging capabilities, along with improved image analysis and tracking software make single-cell tracking achievable.

[enlarge]

Figure 2. FRET biosensors function by having a donor and acceptor fluorophore linked to molecules of interest. FRET occurs when the molecules interact and the donor is close enough to excite the acceptor fluorophore. This causes the intensity signal to shift from entirely donor to entirely acceptor.

3D migration is a technique that is being readily pursued as it combines the advantages of single-cell analysis while placing cells in an environment that closely parallels their native one. Collagen 3D matrices can now be readily produced in most labs [58] making these studies more accessible than previous ones conducted on matrix gels. Although there are similarities in movement between 2D and 3D migration and more thorough picture of cell movement is being generated. Cells migrate in 2D by actin polymerization creating a leading and lagging edge with lamellipodia. However, in 3D there are several modes of migration that are dependent on leading edge shape and cell-matrix adhesion [59]. Additionally, cells continue to respond to mechanical properties of the matrix even in 3D [60]. There are clear parallels between 2D and 3D migration but the environment is vastly more complex and dynamic leading to a great variety in migratory trends. One of the biggest challenges in 3D migration is visualizing the process with microscopy to gather relevant data but advanced imaging techniques and computer algorithms are making it achievable [61].

[enlarge]

Figure 3. Cell traction forces are calculated by cells attaching to and deforming pillars (A) or Fn-coated hydrogels (B).

Studying biological processes at the single-cell levels can expand on many of the previous knowledge bases by identifying key differences often obscured by analyzing large cell populations in mass. However, single-cell techniques are also applicable to studying the root biochemical activity that causes and drives these processes. Forester or fluorescent resonance energy transfer (FRET) is the radiationless transfer of energy from donor to acceptor fluorophore [62]. The transfer is proximity dependent and only occurs when the two respective fluorophores are between 1 and 10 nanometers apart [63]. Within these distances the excited (donor) fluorophore emits a virtual proton, which absorbed by the receiving (acceptor) fluorophore causing a corresponding excitation [62]. Therefore, results are interpreted as a function of the acceptor fluorescence (Figure 2). FRET allows the visualization of biological occurrences that are well below the resolution of traditional microscopy [64]. A wide array of molecules can be conjugated to donor and acceptor fluorophores giving FRET the ability to probe many biological processes. FRET has been successfully used to probe protein folding and conformations, DNA-DNA/RNA interaction, protein–protein interactions, and ligand-receptor binding [65]. FRET systems are commonly referred to as biosensors and are now commercially available in addition to being able to produce novel ones independently [66].

Biosensors were used to better understand the role of GTPases. The use of multiple FRET sensors, both intermolecular and intramolecular, produced spatiotemporal GTPase activity data not attainable by traditional biochemical methods [67, 68]. A better view of migration and the relation to GTPase Rho, Rac, and Cyc42 was revealed by through the information. It was determined that Rac is essential for cellular elongation of the leading edge in the initial phase of cellular migration, while Rho and Cyc42 are critical in retraction of the lagging end [67].

Cell traction forces (CTFs) are a driving force of many biological processes such as migration. As a regulator of cell motility they are critical in such dynamic processes as angiogenesis and vascular genesis, development, wound healing, and disease progression. The generation of mechanical force by cells also helps regulate cell shape and homeostasis through a feedback loop involving mechanical signaling [69]. Cells generate traction forces through attachment points to the extracellular matrix referred to as focal adhesions. As forces are applied it causes changes to the ECM, which is sensed through this adhesion, and causes a corresponding change in the cell [70]. Cells can also interact with one another through these forces [71]. single-cell measurements of traction forces are attained through multiple methods, but all of them utilize microcontact printing.

Measurements of CTFs require three steps: printing force arrays, null force array measurements, and force application array measurements [72]. The predominate method for CTF measurement is printing small pillars onto a flexible polyacrylamide gel (PG) and coating them with an extracellular matrix protein to form focal adhesions (Figure 3A). These pillars function as a cantilever and the force necessary to deform or bend them can be calculated through the known physical features of the pillars (Young’s modulus, height, and diameter). The parameters of the pillars and the pattern they are printed in are determined based on the relevant application. Variations in parameters have been used to determine the impact of the number of focal adhesions on CTFs [73].

The extracellular matrix proteins can be directly printed onto the flexible PG surface (Figure 3B). Spacing and patterns are controlled just as with the pillars. The advantage of this system is that multiple proteins can be printed, in different distribution and patterns, to explore differences in CTFs based on the adherent substrate. This method has already demonstrated that cells preferentially form focal adhesions with Fn rather than albumin or laminin [74]. Studies are planned to probe the binding preference with a more diverse set of proteins, both ECM and others. Additionally, this technique provides real-time analysis of the CTFs and can determine the mechanical properties of the cell simultaneously [75].

CTFs provide mechanistic insight into the process of adhesion and migration. Studies of endothelial cell CTFs affirmed the leading/lagging model of cell migration and also showed that cell motility is dependent on the density of cell attachment points [76-78]. Morphology is also impacted by cell adhesion and CTFs causing cells to take on different phenotypes relative to focal adhesion numbers [79]. A more thorough understanding of the mechanics of cell migration clearly has wide-ranging application in wound healing, metastasis, and development [80, 81].

The next frontier of single-cell exploration is the analysis of genetic diversity in cell populations. Continued advancements in throughput, cost and efficacy of sequencing techniques have made the analysis of large populations of individual cells feasible. All of the key molecules of the central dogma of biology can now be analyzed for individual cells including DNA, RNA (both in cell and in the nucleus), and the proteome [82-84]. Besides the improvement and adaption of current scientific methods for use with individual cells, new techniques are also being designed and effectively utilized.

Genes are regulated at the single-cell level being turned on and off in a stochastic process [85]. Due to this fact a homogenous population of cells have temporarily heterogenous genetic profiles [86]. The idea of single-cell genetic analysis has been around since the initial investigation of the genetic code and early development of the amplification technique PCR [87]. Early studies that proved the ability to analyze single cells at the genetic level was possible included studying the genomic DNA of single sperm cells [88] and genetic transcripts of single macrophages.

Three key steps are involved in the analysis of single-cell genetic information: isolation or sorting, lysis, and analysis. The sorting process can be conducted with the techniques discussed previously in this review. The choice of methods depends on the research area and topic. Additional concerns are cost and lab availability but these have been mitigated in recent years as the technology continues to advance. Sorting methods detailed here have all relied on cells already being dissociated; however, methods such as laser microdissection and glass capillary can also harvest cells from designated areas of tissue [87]. Laser microdissection has the advantage of being a microscopy technique and allows visual cell selection prior to isolation. The technique is also versatile in regards to cell source and is applicable to cell suspensions and tissue samples. A transfer film composed of a thermostable polymer is placed in contact with a cell source [89]. Target cells are identified with a microscope and a laser “cuts” the cells by close proximity, short bursts that activate the transfer film causing focal adhesions to form [90]. After cellular adhesion the film can be removed and additional processing can be completed.

After cell isolation access must be gained to the molecule of interest for further downstream processing such as amplification. Chemical lysis is the most common and preferred method for cells obtained using microfluidic approaches [91, 92]. The most important factor for reagent selection is ensuring that it does not interfere with the target molecules or negatively impact the enzymes or buffers for amplification and analysis stages. Methods for monitoring the interference that techniques create have been designed in order to minimize the impact of isolation and “noise” from various techniques [93]. Another consideration is that special instrumentations may be needed to ensure the minute amount of sample is not lost during manupilation. For example, Gkountela S et al used Corning DeckWork low binding barrier pipet tips (Sigma, Cat# CLS4135-4X960EA) for pipetting to avoid DNA loss during single cell whole-genome bisulfite sequencing [94].

Determination of gene expression profiles for single cells relies on reverse transcriptase quantitative PCR (RT–qPCR) and RNA fluorescent in situ hybridization (RNA FISH) [95, 96] (Table 3). These two methods provide great insight due to the discovery that two-thirds of the genome is transcribed to RNA but many of the transcripts do not code for proteins [97]. Non-coding transcripts are believed to play key regulatory roles and are thus of great interest. The ENCODE project that determined the predominance of transcription of non-coding RNA also discovered that the cellular location of RNA is critical to gene expression [97]. Studies conducted in several cellular compartments of one cell line concluded that most splicing is done during transcription [98]. Taking these factors into account single-cell studies focus on RNA detection and analysis methods to gain the most relevant information.

With regard to the studies of mRNA translation at the transcriptomic scale with spatial and single-cell resolution, the ribosome-bound mRNA mapping (RIBOmap), a three-dimensional in situ profiling method, has recently been developed [99]. RIBOmap-based analysis in HeLa cells showed cell cycle-dependent translational control and colocalized translation of functional gene modules. Mapping of 5413 genes in mouse brain tissues verified single-cell translatomic profiles for 119,173 cells and cell type-specific translational regulation, including translation remodeling in oligodendrocytes.

A novel single-cell RNA-sequencing method, which does not require microfluidic devices, expertise or hardware, applies particle-templated emulsification and allows single-cell encapsulation and barcoding of cDNA in uniform droplet emulsions using only a vortexer [100]. Particle-templated instant partition sequencing accommodates various emulsification formats, such as microwell plates and large-volume conical tubes, generates high-purity transcriptomes in mouse-human studies and can effectively characterize cell types in tissue compared to a commercial microfluidic method.

Single-cell techniques predominately suffer from a similar critique that they remove the cells from their native cellular environment. By removing cell contacts and interactions, as well as losing all spatial information, the techniques both potentially alter cell properties [101] and lose valuable data parameters respectively [102]. As the field matures these issues are being addressed. Transcriptome in vivo analysis (TIVA) is able to circumvent this issue (Table 3). The TIVA employs a photoactivatable double-stranded oligonucleotide linked to a cell penetrating peptide [103]. This system, referred to as a “TIVA tag” crosses the cell membrane and then loses its peptide, thus locking it inside the cell. Irradiation with a laser reveals a sequence that captures mRNAs in the designated cell. Cellular RNA is then isolated and amplified to quantify the gene expression for an entire cell, or specifically selected subcompartments [103]. This technique has already demonstrated that disassociated neuron cells have an increased number of transcripts as compared to ones in intact tissues [104]. This adds to and affirms previous theories of the importance of cell-cell interactions [105, 106].

Another major problem in droplet-based single-cell sequencing data is the presence of cell doublets. Y Chi et al applied an unsupervised machine learning classifier to identify sequencing data from cell doublets [107].

Stem cells are a focus area for single-cell genetic analysis. Identifying key differences in genetic profiles has a potentially revolutionary impact in developmental biology and regenerative medicine. Recent studies have shown the heterogeneous nature of stem cells with diverse genetic outputs [108-110]. Additionally, it has been shown that differences in gene expression do not automatically predict the difference in phenotypic expression [111]. Information is being gathered on cell fates throughout the developmental process. A study was conducted analyzing the gene expression of single cells from a mouse zygote to blastocyst. At the 64-stage blastocyst stage there are three distinct cell types present [112]. Analysis of cells during a four-day period leading to this stage showed relevant early stage markers for cell fate determination [113].

Single-cell split barcoding (SISBAR) has been developed to study neural cell lineages across developmental stages [114]. This method can be used for clonal tracking of single-cell transcriptomes in an in vitro model of ventral midbrain-hindbrain differentiation. The study showed that a transcriptome-defined cell type can develop from distinct lineages that provide molecular imprints on their progenies, and the multilineage fates of a progenitor cell-type are determined as the combined outcomes of distinct clonal fates of individual progenitors with specific molecular signatures. Furthermore, the results identified a ventral midbrain progenitor cluster as the clonal origin of midbrain dopaminergic neurons, midbrain glutamatergic neurons and vascular and leptomeningeal cells.

Proteomic studies are also conducted at the single-cell level to determine functional differences in cells and help identify key elements for expanded genetic exploration. High throughput proteomic screening is readily achievable with mass spectrometry [115], and potentially with nanopore technologies [116]. Unfortunately, issues arise with determining which cell is generating which data. This precision problem has been addressed by adding an additional microscopic localizing technique. Cells are dispersed on a slide and their nuclei are stained. A fluorescent scope then identifies the location of each cell and passes the coordinates to a laser for mass spectral analysis [117]. The expression of the specific gene in single cells can also be examined through single-cell western blot [118], for example, Milo single-cell Western Blot from ProteinSimple was used to estimate the percentage of enteroendocrine cells (neuropods) that expressed synapsin-1 [119].

Mass cytometry, a combination of mass spectrometry and flow cytometry, has been recently used to improve the research capabilities of the two independent techniques. Mass cytometers (CyTOFs) are inductively coupled plasma-mass spectrometers (ICP-MS) connected with time-of-flight detectors that together allow researchers to perform highly multiplexed single-cell assays.

Unlike traditional flow cytometers that use antibodies marked with fluorescent dyes, mass cytometers use antibodies tagged with heavy metals [120, 121]. The metal tags perform the same function as the fluorescent dyes but allow detection of a higher number of independent signals. At least 40-50 different tags can be concurrently detected, representing more than twice the number of non-overlapping fluorescent signals detected by flow cytometers [1, 122]. As conjugation chemistry catches up to the platform [123] this number will rapidly increase.

A CyTOF consists of three main functional units responsible for 1) sample introduction and preparation; 2) plasma and ion-cloud formation; and 3) ion acceleration and separation based on time of flight (Figure 4). The equipment and technology were described in detail in Badura et al [124] and in Tanner et al [125]. Briefly, the sample preparation step consists of cells of incubation (staining) with antibodies conjugated with polymeric chains able to chelate stable isotopes of heavy metals. Multiple samples can be mixed if they are first differentially labeled or “barcoded”. The staining step allows the affinity binding of the cellular parts of interest. Since the antibodies carry a tag, the expression level of each independent target can be monitored by tag detection. Cell suspensions are then introduced, either manually or with the help of an autosampler, into a nebulizer that transforms the cell suspension into fine water droplets. Once injected into the instrument, the droplets are heated up, water evaporates and cells are broken into atoms, which become charged in the process. The charged particles are accelerated and separated based on their mass-to-charge ratio in the time-of-flight mass spectrometer. Ions are detected and counted based on their flight time. Counts are converted in digital signals that are integrated on a per-cell basis, thus generating single-cell measurement data for further analysis. According to Atkuri et al [126] about 60-70% of the cells are lost through the process, making the efficiency of analysis of injected cells about 30-40%.

[enlarge]

Figure 4. Mass cytometry basic experimental setting. From Lauren et al., (2018) [1].

This technique has been successfully used in measuring levels of transmembrane, cytoplasmic or nuclear proteins. Membrane-bound proteins are often considered cellular markers, and cell populations are defined based on such markers expression [127, 128]. For example, Guo CJ et al measured the expression of IgA among the various cells from mouse small intestines [129]. Rosshart SP et al assessed immune responses in laboratory mice born to wild mice with mass cytometry [130]. This is often the case for immune cells populations, but not only [131]. Recent studies used mass cytometry to determine the lineage and the maturation state of each multiple cells simultaneously [1, 132] or to identify particular immune cells subsets [122]. Inside the cell, mass cytometry is able to detect the expression levels of individual transcription factors that control gene expression [133-135] or to measure the level of post-transcriptional modifications associated with the signaling pathways [136-138]. The intercellular communication mechanisms can also be analyzed by simultaneously monitoring the expression level of many cytokines or growth factors. By using this method, Newell et al [139] observed patterns of cytokine responses in T-cells that are specific for certain antigens, while Hartmann et al [140] noticed elevated inflammatory responses in patients with narcolepsy.

Other applications are the simultaneous investigation of gene transcription and translation in single cells by measurement of nucleic acids and proteins in the same experiment [141] and the determination of the epigenetic states of individual cells [142]. Mavropoulos et al [141] describe a novel assay called Metal In Situ Hybridization (MISH) for the simultaneous detection of three mRNA targets and cell surface markers in a cell in suspension. The assay involves hybridization of RNA-specific target probes, signal amplification and binding of amplifier-specific detector probes that are tagged with pure metal isotopes, which are further detected by mass cytometry. Recently, Cheung et al developed a mass cytometry-based system for the profiling of the global levels of chromatin modifications at the single-cell level [142, 143].

A new method, which enables direct capture of polyadenylated transcripts, in both single-cell RNA sequencing and single-cell Assay for Transposase Accessible Chromatin using sequencing assays, was introduced using CellTag-multi [144]. Also, core regulatory programs underlying on-target and off-target fates in the reprogramming of fibroblasts to endoderm progenitors were identified. Furthermore, the study revealed that the transcription factor Zfp281 is a regulator of reprogramming mechanisms directing cells toward an off-target mesenchymal fate.

The majority of mass cytometry applications use cell suspensions as samples for analysis, but analysis of homogenized tissues by this method has also been performed. Studies by Lavin et al and by Chevrier et al use mass cytometry to analyze tissue samples of immune cell populations, while Leelatian et al analyze human tissue and solid tumors [145-147]. Moreover, the analysis of samples from microscopy slides has also been made possible by mass cytometry [141, 148-150]. This is called imaging mass cytometry.

Mass cytometry can be used to address all biological questions previously answered with the help of flow cytometry with the great advantage of increasing the number of parameters that can be monitored and allowing the analysis of biological processes in different cell types simultaneously. However, the technique has its own limitations. The cells can only be analyzed after atomization and ionization, so there is no possibility of analyzing living cells. Because it is dependent on antibodies, the sensitivity and specificity of detection are limited to the characteristics of the antibodies used. High-quality low cross-reactivity antibodies, ideally pre-validated, are required. Also, the sensitivity of the ion detectors is lower than that of some fluorophores [151].

Single-cell techniques will continue to expand as technology advances by streamlining current applications and creating new ones. Studies of single cells have served to not only affirm previous population models and theories but also delve deeper into nuisances in cell diversity (Table 4). Techniques created to explore single cells are also providing valuable mechanistic information in many key biological practices, applicable in many key areas. Diagnostic capabilities are already focusing on single-cell detection methods as a potential candidate for rapid screening for pathogens and also applying it to food safety initiatives. A primary limitation of many of these techniques is an adequate analysis of all the data that is generated. More robust computational programs are currently under development and will serve to not only further the field fundamentally but also fully utilize and examine previously derived data sets.

Dr. Georgeta Basturea contributed the section on mass cytometry in December 2018.