Single-cell sequencing: A cutting edge tool in molecular medical research

Abstract

The rapid development of advanced high throughput technologies and introduction of high resolution “omics” data through analysis of biological molecules has revamped medical research. Single-cell sequencing in recent years, is in fact revolutionising the field by providing a deeper, spatio-temporal analyses of individual cells within tissues and their relevance to disease. Like conventional sequencing, the single-cell approach deciphers the sequence of nucleotides in a given Deoxyribose Nucleic Acid (DNA), Ribose Nucleic Acid (RNA), Micro Ribose Nucleic Acid (miRNA), epigenetically modified DNA or chromatin DNA; however, the unit of analyses is changed to single cells rather than the entire tissue. Further, a large number of single cells analysed from a single tissue generate a unique holistic perception capturing all kinds of perturbations across different cells in the tissue that increases the precision of data. Inherently, execution of the technique generates a large amount of data, which is required to be processed in a specific manner followed by customised bioinformatic analysis to produce meaningful results. The most crucial role of single-cell sequencing technique is in elucidating the inter-cell genetic, epigenetic, transcriptomic and proteomic heterogeneity in health and disease. The current review presents a brief overview of this cutting-edge technology and its applications in medical research.

Introduction

Over the past two decades, development of advanced high throughput technologies has revamped the medical research in molecular biology at cellular and subcellular levels. The term “omics” refers to the study of genome, transcriptome, epigenome, proteome or metabolome of a given sample using high throughput techniques. The essential premise of these approaches is that if a complex system is viewed as a whole, it can be better understood. Genomics is the study of an organism's entire set of Deoxyribose Nucleic Acid (DNA), including all of its genes, referred to as the ‘genome’. Transcriptomics includes information of all transcripts expressed in the cell/tissue. Epigenomics is the study of the whole set of epigenetic alterations on a cell's genetic material.¹ Wasinger et al. (1995) created the word ‘proteome’, which is defined as the characterisation and quantification of all sets of proteins in a cell, organ, or organism at a certain moment. Metabolomics is the analysis of all small-molecule metabolite profiles; it is the examination of chemical fingerprints that specific biological processes establish during their activity. Based on the approach of analysis, currently omics can be of three types:

• •Bulk omics: when the bulk of cells in a tissue are processed as whole to obtain average genetic variation, gene expression, epigenetic modifications, protein and metabolic profiles. Bulk omics captures numerous cells from given sample and considers it as same homogenous population. Hence, the output of bulk omics leads to loss of information regarding cellular heterogeneity of the population. But these methods are comparatively easy to perform due to low technical noise and simple protocols. Bulk techniques are also beneficial for dissecting ensemble signatures at the individual/tissue level, as well as undertaking comparative omics research.²
• •Single-cell omics: when the information of genome, transcriptome, epigenome, proteome, and metabolome is obtained from an individual cell of a tissue. Compared to conventional bulk omics single-cell omics have advantage over dissecting cellular heterogeneity. Also, differences arising due to different cell cycle stages and apoptosis processes can be resolved using single-cell omics.³
• •Spatial omics: This is an evolving technique wherein individual cells are analysed for various omics while maintaining their spatial orientation within the tissue. This allows examination of relation of cellular state with its neighbouring cells or surrounding extra cellular matrix.⁴

Antibody-based technologies including Western blot or flow and mass cytometry were the first to allow detection of proteins in single cells; however, these methods rely heavily on high-quality antibodies and are naturally limited in their multiplexing capacity. However, recent advances in gas chromatography and mass spectrometry have enabled the study of complete proteome at single-cell level.⁵ Similar techniques are also used for single-cell metabolomics which is a leading field in understanding phenotypic variations between cells. Of all these omics approaches, single-cell omics is currently proving to be a most promising technology in medical research. The current review is to outline the cutting-edge technology of single-cell sequencing method and its applications in medical research.

Single-cell sequencing

Sequencing refers to deciphering the sequence of nucleotides (adenine, guanine, cytosine, and thymine) in a given DNA, Ribose Nucleic Acid (RNA), Micro Ribose Nucleic Acid (miRNA), epigenetically modified DNA or chromatin DNA. Next generation sequencing, previously known as massively parallel sequencing is currently the most popular sequencing method. This method involves construction of library, amplification, sequencing of library and data analysis. NGS can be performed using various platforms such as sequencing by synthesis, sequencing by hybridisation, pyrosequencing, ion torrent, Illumina, and nanopore technology.⁶ This technique is called single-cell sequencing when applied on a single cell after isolating it from a tissue. The most significant benefit of single-cell sequencing technique is its ability to reveal inter- and intra-tissue heterogeneity. This method requires relatively lesser amounts of tissues that may be collected from different sites in the organ.

Single-cell genomic sequencing is also known as single-cell DNA sequencing (scDNA-seq). Whole genome sequencing (sequencing of an organism's whole genome),⁷ whole exome sequencing (sequencing of only the coding portions of the genome),⁸ and clinical exome sequencing (DNA tests for identification of the molecular basis of genetic disorder of a suspected patient)⁹ are examples of single-cell genomics. Single-cell transcriptome sequencing may dynamically reflect the total gene expression of a single cell at a specific functional stage. The first report of single-cell cDNA amplification was published in 1990. The concept evolved slowly and the first exhaustive technical report of high-throughput single-cell transcriptome sequencing (scRNA-seq) was published by Tang et al., in 2009.¹⁰ Single-cell epigenomics allows us to understand complete set of epigenetic modifications of a particular cell. Whole genome bisulfite sequencing enables quantitative assessment of the methylation status of all CpG sites in a genome for a thorough understanding of the significance of genome-wide DNA methylation patterns, often known as the methylome.¹¹ To determine single-cell chromatin accessibility and DNA-binding site by proteins, assays for transposase-accessible chromatin with sequencing (ATAC-Seq)¹² and chromatin immunoprecipitation sequencing (ChIP-Seq) are also available; single-cell epigenomics is performed through by DNA methylome sequencing (scDNA-Met-seq) (Fig. 1).

Fig. 1.

Next generation sequencing based omics and their approaches.

Single-cell sequencing: Workflow

The processing pipeline for single-cell sequencing is essentially same as that for bulk next generation sequencing except for an additional preparatory and highly critical step of single-cell isolation. Depending upon the downstream studies proposed to be undertaken, different approaches for single-cell isolation including cell sorting (fluorescence-activated or magnetic-activated cell sorting-FACS), microfluidics, micromanipulation, laser capture micro dissection, microdroplet, etc., can be integrated into the work flow (Fig. 2).

Fig. 2.

Single-cell sequencing workflow.

When the amounts of available tissues are restrictive, such as in early embryo stages, microscopic inspection and micromanipulation can be additionally performed to ensure the isolation of single cells. This procedure does not necessitate any special equipment, but it is labour intensive and time consuming when dealing with a large number of cells.^10,13 Cells expressing specific surface marker(s) can be isolated from a population using FACS. A downside of employing FACS is the requirement for a high number of cells as the beginning material. Furthermore, the high vibrations utilised in FACS to separate the stream into individual droplets may harm the cell, causing membrane leakage and mRNA or even DNA loss.¹⁴ Microfluidics, on the other hand, is more promising since it allows for high-throughput single-cell separation, pipetting error elimination, easy inclusion of subsequent DNA/RNA amplification, has high fidelity, and efficiency.¹⁵ Laser capture microscopy does not entail the suspension of cells. Using a laser, the cells can be separated from the tissue. This saves the information about the spatial location of each individual cell being analysed. On the flip side however, the probability of contamination of the sample with neighbouring cells exists.¹⁶ Microdroplet-based technologies may also allow for high-throughput reagent delivery to each droplet, either by droplet fusion or via instruments like the Pico injector. This will downscale reactions in a highly parallel manner to nanolitre and picolitre volumes, substantially lowering costs.¹⁷

The isolated cell is evaluated for the quality and integrity of its content. After single-cell isolation, and depending upon type of omics study, the DNA, RNA, miRNA, or chromatin DNA is extracted and processed accordingly. For single-cell DNA sequencing, genomic DNA is extracted and fragmented followed by amplification.¹⁸ After isolating RNA for scRNA seq, reverse transcription is carried out after poly-T enrichment. Polymerase Chain Reaction (PCR) is then used to amplify reverse transcribed cDNA.¹⁹ miRNA sequencing follows same protocols as RNA-seq except that an additional step for filtering shorter fragments is performed. In methylome sequencing, DNA is fragmented, adaptors are attached, bisulphite conversion of the DNA is executed followed by global amplification.²⁰

The development of a sequencing library by tagging sequences of interest with short DNA barcodes of approximately 08 nucleotides so that sequenced reads can be identified for their cell of origin is the cornerstone of any single-cell sequencing investigation. High-throughput sequencing technologies such as Illumina, nanopore, PacBio, AbSoLid, and 10X genomics are being used to sequence single-cell sequencing libraries. The sequenced reads from single cells are then analysed using standard bioinformatics tools.

Single-cell sequencing: Data analysis

There is no one-size-fits-all strategy to single-cell sequencing data analysis. Each type of genomics, transcriptomics, and epigenomics data has its own customised approaches for data analysis. The general pipeline, however, stays the same as is depicted schematically in Fig. 3.

Fig. 3.

Overview of the bioinformatics approaches to analyze single cell sequencing data.

Single-cell sequencing analysis can be broadly divided into two steps; preprocessing of sequencing data followed by downstream analysis of the data.²¹ The DNA library which is prepared by nucleotide barcoding for specific cells prior to sequencing is multiplexed for sequencing. These sequenced reads are then separated into different samples based on their cellular bar codes. This process is called demultiplexing. Thereafter, assessment of sequenced reads is carried out for both quality and number of reads. Once the reads are aligned with reference databases, they are subjected to a quality check. Low-quality reads are removed from the data, which is then normalised to remove any changes caused by low input material from single cells, as well as numerous other biases and disturbances inherent in the sequencing process.²² Feature selection is a further step which identifies features which are more informative and can provide meaning to data. Dimensionality reduction of data provides reduced but efficient sequencing data using several methods like principal component analysis (PCA), linear discriminant analysis (LDA), generalized discriminant analysis (GDA) etc.²³

Clustering is a critical step in downstream analysis, where individual cell types are recognised based on molecular properties analysed.²⁴ Differential gene expression patterns in scRNA-seq, copy number variations (CNVs) and single nucleotide variations (SNVs) in scDNA-seq, differential DNA methylation in scDNA-Met-seq, and chromatin DNA analysis and accessibility Single-cell chromatin immunoprecipitation sequencing (scChIP-Seq) and Single-cell assay for transposase-accessible chromatin with sequencing (scATAC-Seq) are examined.²⁵ Finally, to obtain relevant data, a tailored statistical analysis is performed utilising tools and pipelines built specifically for each single-cell omics.²⁶

Single-cell sequencing: Applications

Due to its unprecedented utility, single-cell sequencing technology has revolutionised medical research in recent years. Integrative analysis of these single-cell multi-omics data can provide unrivalled information on developmental phases or underlying disorders. Single-cell gene expression profiling and epigenetic variation may help us understand the evolution of diverse cell lineages during the embryonic development, when relatively fewer cells are available as in liquid biopsy of circulating tumour cells and foetal cells in maternal circulation, which can provide a wealth of development, disease or physiology-related data.¹⁵ Novel mutations can be found in small populations of cells at primary and metastatic sites, revealing the molecular pathways as well as genetic and epigenetic heterogeneity of the disease. Single-cell transcriptomics establishes gene expression of individual cells amongst heterogeneous populations. Regulation of gene expression may now be examined at a single-cell resolution through single-cell miRNA–mRNA co-sequencing technologies,²⁷ while single-cell methylation sequencing can reveal epigenetic variations in seemingly uniform populations of cells. Even the role of unusual cell types in disease processes can be discovered (Fig. 4).

Fig. 4.

Applications of single-cell sequencing.

These approaches in single-cell sequencing have found applications in medical research. This includes advances in clinical diagnosis, prenatal-diagnosis and germline transmission.²⁸ Currently, it has proved to be a most promising tool to create human cell atlas which would have cataloguing of all human cell types. A cell atlas would essentially be a collection of cellular reference maps that describe each of the tens of thousands of cell types that make up the human body and their locations.²⁹ The single-cell multi-omics has transformed the concept of precision medicine in both genetic and acquired diseases. The cancer research in terms of diagnosis, prognosis, treatment, monitoring, cancer progression, tumor heterogeneity, tumor immunology, metastasis and chemoresistance is unthinkable without the single-cell sequencing.^30,31 The relatively newer omics the ‘microbiomics’ too has exclusively evolved due to single-cell sequencing.³²

Single-cell sequencing: Limitations

Single-cell sequencing technology has some drawbacks. Because the initial material from a single cell is so little, any degradation, sample loss, or contamination would have a big impact on the outcome. For single-cell sequencing, restrictive amounts of starting material necessitate considerable amplification, which is prone to biases and errors and inaccuracy of sequencing results. In addition, if the cellular bar codes are misaligned, demultiplexing will not occur, resulting in data loss.

Conclusion and future perspectives

Single-cell sequencing is a next-generation sequencing technique that has revealed additional information on the molecular pathways that underpin health and illness. This cutting-edge technique is now most well-established in the field of transcriptomics; with procedures and analysis pipelines for genomics and epigenomics rapidly expanding as well. It has the potential to play a significant role in medical research in the future by overcoming the hurdles of understanding complex biological diversity and heterogeneity.³³ Future technologies should focus on correlating phenotypes and genotypes in single cells using a combination of live-cell imaging and sc-seq approaches concentrating on integrative sequencing analyses from a single cell in parallel (viz. DNA, RNA, protein, and epigenomic changes). In situ single-cell sequencing technologies that can measure genetic data on single cells while retaining their spatial context in tissues should be the focus of future technology development initiatives.¹⁵ Reduction in DNA/RNA amplification may further increase the efficiency and accuracy of single-cell omics. However, the ability to directly sequence unamplified DNA and RNA produced from single cells will require more innovation.²⁸ However, it is a certainty that single sequencing has and will continue to revolutionise the field of medical research in the near future.

Disclosure of competing interest

The authors have none to declare.