Download PDF

Introduction

Single-cell RNA sequencing (scRNA-seq) is a state of the art technology for the unbiased characterization of gene expression in complex cell and tissue samples. Despite having revolutionized modern biology, scRNA-seq remains technically challenging and often cost-prohibitive. Current solutions on the market enable efficient analysis of a fraction of cells when a large initial pool of tens of thousands of cells is available. However, a streamlined protocol for handling scRNA-seq library preparation of rare cell populations is still missing.

Here, we develop a novel plate-based assay for scRNA-seq that is affordable and highly sensitive. This workflow enables the library preparation from scarce samples of rare cell populations. In this new workflow (Fig. 1), we used the DispenCell™ Single-Cell Dispenser and its dedicated sensing tips according to standard protocol (SEED Biosciences, Switzerland), in combination with a dedicated and optimized single-cell library preparation system (MERCURIUS™ High-sensitivity BRB-seq kit, Alithea Genomics, Switzerland).

Figure 1. Novel workflow for single-cell RNA sequencing

Single-cell dispensing

First, we tested the capacity of DispenCell to dispense single-cells in the dedicated lysis buffer prior to sequencing library preparation. A lymphoblastoid cell line was thawed from a frozen vial, passed twice and then harvested in a single-cell suspension. Cells were filtered using a 10 μm cell strainer (Miltenyi Biotec, Germany), counted and diluted in the RNAse-free methylcellulose-based dispensing medium (DispenMe, SEED Biosciences) to a final concentration of 2.104 cells/mL.

Then, we used the DispenCell to load 200 cells from the cell suspension in the sensing tip. We set the DispenCell to distribute the single-cells individually into wells of a 384- well plate pre-loaded with 4.2 μL of lysis buffer (HS BRBseq kit, Alithea Genomics). Dispensing of the first 96 wells with the device took \~7 minutes. Then, it was stopped as we had enough samples for downstream analysis.

Single-cell quality control

Immediately after dispensing, we used DispenSoft, a software developed by SEED Biosciences to perform a single-cell quality control post-processing. Each well of the plate is represented in a color-coded matrix (Fig 2.a).

The software marks the wells in green when they contain a single-cell while the red wells have zero or more than one cell and should be discarded. To discriminate between debris and cells, a threshold was set by gating on the size-based histogram (Fig 2.b). Only the peaks which are higher than the threshold are counted as cells.

Here, we obtained 77 wells with a single-cell (80% plate-filling rate). Using DispenSoft, the raw impedance data can also be used for documenting the process. Each impedance profile can be examined manually. A single sharp peak is the signature of a single-cell (Fig 2.c) while multiple peaks result from multiple cells. This demonstrates that DispenCell is capable of dispensing single-cells in a lysis buffer in a simple, rapid and fully traceable process.

Figure 2. Single-cell quality control. a) Map of the plate; b) size-based histogram; c) Example of a single-cell signature.

cDNA amplification and sequencing

The sequencing libraries were produced following the High-sensitivity BRB-seq protocol. After single-cell dispensing in the lysis buffer containing the oligo-dT primer, the plate was sealed, quickly spun down in a centrifuge and heated-up at 72°C for 3 minutes to allow primer annealing. RT enzyme mix was added to the each well and the reverse-transcription reaction was performed following the protocol. The double-stranded cDNA was first amplified with 20 cycles of PCR and purified using AMPure beads. Then, 1 ng of double-stranded cDNA was tagmented and a second PCR amplification using Illumina Nextera compatible indexes was performed (12 cycles). The final libraries were purified using AMPure beads and run on a Fragment Analyzer to assess the fragments size distribution (QC, see Figure 3). All profiles demonstrated similar fragment size distribution in the expected range and a single smooth peak around 600 bp, showing good library quality and good reproducibility (Fig. 4). The libraries were sequenced using the NextSeq instrument at 0.5 Mio reads per library. The sequencing reads were aligned to the human genome (GRCh38.102) using STAR (2.7.9a).

Figure 3. Fragment size distribution of a single-cell library

Upon sequencing, all the single-cell libraries allowed for the detection of a high number of genes (around 10,000 genes in total), with a good proportion of poorly expressed genes (around ⅓ of the total number of detected genes, shown in green on Fig. 4). Those results confirm high data quality and high reproducibility across the wells with single-cell samples.

Figure 4. Number of detected genes for each single-cell library sequenced.

Conclusion

We showed that using this novel workflow high quality sequencing data can be generated. Furthermore, we demonstrated for the first time that DispenCell is compatible with downstream RNA sequencing. Finally, we showed that the MERCURIUS™ High-sensitivity BRB-seq kits from Alithea Genomics together with the DispenCell are best suited for end-to-end simple and affordable library preparation and sequencing.

Download PDF