Chapter 2 Single-Cell RNA-seq Technologies

FIGURE 2.1: Explosion of scRNA-seq Throughput (2018)

Technological advances increase the efficiency of single-cell sequencing methods while the cost decreases day by day. In this chapter, we will briefly go through some of the single-cell RNA-seq technologies.

2.1 Plate-Based SMART-seq

SMART-seq is a plate-based single-cell RNA-seq method that captures full-length transcripts using oligo(dT) priming and template-switching reverse transcriptase. Each cell is sorted (e.g., via FACS) into individual wells of a 96- or 384-well plate.

The cDNA is amplified using PCR, and sequencing libraries are prepared from this amplified cDNA. SMART-seq and its improved version SMART-seq2 are well-suited for detecting isoforms and lowly expressed genes.

Advantages of SMART-seq:

• Full-length transcript coverage — useful for isoform and splicing analysis
  
• High sensitivity, low dropout rate
  
• Compatible with rare/low-throughput samples

Limitations of SMART-seq:

• Lower throughput than droplet-based methods
  
• Higher cost and hands-on time per cell
  
• Not UMI-based — less accurate quantification

2.2 MARS-seq

MARS-seq (Massively Parallel Single-Cell RNA-Seq) is another plate-based scRNA-seq method that focuses on 3’ end-tagged sequencing. Cells are sorted into 384-well plates containing barcoded primers for mRNA capture. Unique molecular identifiers (UMIs) and well-specific barcodes are introduced during reverse transcription.

After cDNA synthesis, all reactions are pooled for linear amplification by IVT (in vitro transcription), followed by library preparation.

Advantages of MARS-seq:

• Well-suited for profiling many cells in parallel
  
• Supports UMIs — accurate transcript quantification
  
• Suitable for low-input samples and immune profiling

Limitations of MARS-seq:

• 3′ bias — cannot resolve isoforms
  
• Requires customized primers and plate setup
  
• Slightly lower sensitivity compared to SMART-seq

2.3 Droplet Microfluidics (DROP-seq)

FIGURE 2.2: Extraction and Processing of Single-Cell Transcriptomes by Drop-Seq (2015)

DROP-seq enables high throughput profiling of thousands of individual cells in a short period of time by encapsulating each cell individually with a unique barcode in a droplet. The process takes place within a microfluidic device.

FIGURE 2.3: Drop-Seq barcoding scheme

After dissociating the complex tissue sample into individual cells, each cell is suspended in droplets with microparticle beads which deliver barcoded primers within a microfluidic device. When cells within the droplet are lysed, the mRNA of these cells is attached to microparticles.A set of beads are generated with the reverse transcription of mRNAs into cDNAs. These sets of beads are called STAMPs, “single-cell transcriptomes attached to microparticles”. Finally, an RNA-seq library with about 10,000 single-cell transcriptomes was constructed by barcode STAMPs amplifying.

FIGURE 2.4: Sequence of primers on the microparticle

Each barcoded primer bead contains an unique cell barcode and, an unique molecular identifiers (UMI)s along with the common PCR handle. The common PCR handle sequences used for PCR amplification after STAMPs formation. The cell barcodes are unique for each droplets, therefore, each microparticles contains primers with the same cell barcode. UMIs are unique barcodes to each transcript, UMIs are counted digitially to count unique number of mRNA transcripts. And, counting UMIs also removes the uncertainty of PCR duplicates from transcript copy numbers.

Limitations of DROP-seq:

• Most of the droplets are empty and, very few of the droplets contains both a cell and microparticle. Thus, it is not an efficient method.

2.4 10X Genomics Chromium

FIGURE 2.5: 10x Genomics Chromium scheme (2017)

The 10X Genomics method is very similar to DROP-seq and, it is first described in (2017). First, the barcoded beads are mixed with the cell suspension. Then, the mixture is capsulated with the emulsion generated by oil. The microfluidic device and the barcoded beads of the 10X Genomics are engineered to increase the efficiency of barcoding cells. Therefore almost 90% of the droplets contain a cell and a barcoded bead. The main difference between the 10X Genomics Chromium and DROP-seq is the reverse transcription occurs within the droplets. After the amplification process, the transcriptomics library is constructed. For each lanes of a 10X Genomics Chromium chip, approximately 20,000 cells can be barcoded and sequenced simultaneously.

Advantages of 10X Genomics Chromium:

• 10X Genomics is a commercial platform
• Data obtained is high-throughput
• Almost 10,000s reads per cell
• Fast, inexpensive
• 3’ transcript sequencing
• Single-nucleus sequencing possible
  • Less bias by cell morphology
  • Only captures nuclear RNA

2.5 Nanowell Platforms

Nanowell-based scRNA-seq platforms isolate single cells into thousands of tiny wells patterned on a microchip. Each well contains one cell, and barcoded beads or primers are then added to each well. After cell lysis, mRNAs are captured by the barcoded oligos on beads, similar to Drop-seq, followed by reverse transcription and cDNA amplification.

Common nanowell-based methods include Seq-Well and BD Rhapsody, which allow moderate throughput and compatibility with fragile cell types. These platforms combine physical cell isolation with scalable processing.

Advantages of Nanowell Platforms: - Gentle on cells — suitable for fragile or rare populations - More control over cell-bead pairing than droplet systems - Lower equipment complexity vs. droplet microfluidics

Limitations of Nanowell Platforms: - Lower throughput than 10X or Drop-seq - More manual steps — potential for variability - Limited capture efficiency compared to high-end droplet systems

2.6 Sample Indexing (Illumina)

Sample indexing involves tagging different samples or individual cells with unique barcode sequences (index tags) during library preparation. These tags are incorporated either via adapter ligation or during PCR amplification. After sequencing, reads are demultiplexed based on these indexes, enabling multiplexing of multiple samples in a single sequencing run.

This method is widely used in plate-based protocols like SMART-seq2 and is foundational in many combinatorial indexing strategies.

Advantages of Sample Indexing:

• Cost-efficient multiplexing of samples or conditions
• Increases throughput without added sequencing lanes
• Compatible with standard Illumina sequencing workflows

Limitations of Sample Indexing:

• Barcode swapping or misassignment can affect data accuracy
• Limited by the number of unique index combinations
• Cannot handle extremely large numbers of cells without additional combinatorial strategies

2.7 Single-cell Combinatorial Indexing: sciRNA-seq

sci-RNA-seq (single-cell combinatorial indexing RNA-seq) enables profiling of thousands to millions of cells without physically isolating each cell. Instead, it uses combinatorial barcoding in multiple rounds of pooling and splitting cells or nuclei, introducing unique barcode combinations in each round (e.g., via transposase or reverse transcription primers).

The method begins with cells or nuclei distributed across multiple wells, where the first set of barcodes is introduced. The contents are then pooled and redistributed, and another set of barcodes is applied. This process repeats across multiple rounds, allowing millions of uniquely labeled single-cell transcriptomes.

Advantages of sci-RNA-seq: - Ultra-high throughput — millions of cells can be profiled - No need for droplet microfluidics or physical single-cell isolation - Can be applied to fixed or frozen cells and nuclei - Compatible with multi-omic profiling (e.g., sci-ATAC-seq)

Limitations of sci-RNA-seq: - Lower transcript capture efficiency than droplet-based methods - Complex data processing and barcode decoding - May require deep sequencing to resolve sparse expression profiles

References

Macosko, Evan Z., Anindita Basu, Rahul Satija, James Nemesh, Karthik Shekhar, Melissa Goldman, Itay Tirosh, et al. 2015. “Highly Parallel Genome-Wide Expression Profiling of Individual Cells Using Nanoliter Droplets.” Cell 161 (5): 1202–14. https://doi.org/https://doi.org/10.1016/j.cell.2015.05.002.

Svensson, Valentine, Roser Vento-Tormo, and Sarah A. Teichmann. 2018. “Exponential scaling of single-cell RNA-seq in the past decade.” Nature Publishing Group. https://doi.org/10.1038/nprot.2017.149.

Zheng, Grace, Jessica Terry, Phillip Belgrader, Paul Ryvkin, Zachary Bent, Ryan Wilson, Solongo Ziraldo, et al. 2017. “Massively Parallel Digital Transcriptional Profiling of Single Cells.” Nature Communications 8 (January): 14049. https://doi.org/10.1038/ncomms14049.