Gene expression profiling is a powerful method that allows one to relate the expression of specific genes to a particular cellular phenotype. For example, when comparing cancer tissue to healthy tissues within an organ.

However, gene expression patterns often vary even among cells within the same tissue sample. Traditional methods of gene expression analysis on bulk samples do not allow researchers to distinguish the contributing variation from individual  cells.

Gene expression analyses performed on single cells can offer a powerful method to resolve sample heterogeneity and reveal hidden biology in multicellular tissue samples. This approach allows in-depth analysis of complex biological phenomena such as cancer, development, and microbial ecology.

Described below is a complete workflow of single cell gene expression analysis offered by BD.


Figure 1. A complete workflow. A. Single cell suspension is prepared and stained with antibodies to surface marker, B,C. Cells are analyzed and sorted using the BD FACS sorter based on desired markers into BD Precise 96-well plates. D,E. The BD Precise assay protocol is followed to generate a sequencing library. F. After sequencing, data deconvolution is performed with an automated algorithm using a Seven Bridges Genomics pipeline. G. mRNA transcript counts were analyzed to dissect the biological relevance.

To learn more about individual stages of the process and solutions offered by the company, visit the BD webpage.

Isolation of Single Cells

The first step in a workflow of single cell gene expression analysis is isolation of single cells from a sample. Single cells or small numbers of cells can be isolated and prepared by limiting dilutions, laser capture micro-dissection, microfluidics technologies, or fluorescence-activated cell sorting (FACSTM).

FACS has emerged as a leading platform for this purpose because it is fast, enables high-throughput processing of a cell mixture, and ensures the delivery of single cells or a small number of cells into a chosen receptacle at a very high purity level.

Prior to sorting the cells, samples can be stained with fluorescent antibodies specific to cell surface markers or intracellular marker proteins, or nucleic acid dyes. This makes it is possible to sort live cells, as well as different cell types, based on the expression levels of target cell markers.

Figure 2 shows how cell sorting is performed in a BD FACS sorting instrument.


Figure 2. Cell sorting process. This diagram shows how the cells passing through the fluid stream are interrogated and selected based on fluorescence signal or light scatter. The fluid breaks into drops carrying the selected cells, which are electrically charged and collected.

A FACS instrument incorporates the fluidic, electronic, and optical systems of a flow cytometer.

When a sample is introduced into a FACS sorter, the fluidic system funnels the cells into a single stream so that they pass one at a time through a laser beam or a series of laser beams.

When each cell passes through the laser beam(s), it scatters light and may emit fluorescence light based on cell surface or intracellular fluorescent staining. Light scatter and fluorescence signals are collected and detected, and a decision is made whether the cell is of interest and should be sorted.

Flow cytometry cell sorters can collect the cells into collection tubes, trays, and slides, as well as 96-well or 384-well tissue culture and PCR plates.

BD makes a variety of instruments for sorting single cells to fit the researcher’s experimental need, the lab’s budget, and the operator’s flow sorting experience and preference.

To learn more about FACS sorters offered by BD, visit the BD Biosciences website.

Generation of gene expression sequencing libraries

To obtain sequencing information from low input samples, such as single cells, amplification of the sequencing library is required. At this stage, it is important to control for the introduction of artefacts, such as amplification bias.

Molecular barcoding is an integral part of BDTM Precise assays - ultra-sensitive RNA sequencing assays designed to examine large numbers of low input mRNA samples in a high-throughput manner. To improve workflow and mitigate amplification bias, these assays append three layers of DNA barcoding to each mRNA molecule.

  1. A molecule-specific barcode (Molecular Index) is used to identify the molecule of origin of each sequenced read in an amplified sequencing library. This provides a more accurate representation of transcript abundance in the original sample prior to amplification. 
  2. Well-specific barcodes are used to link individual molecules from the same single cell. They also allow for pooling of individual cells to simplify the library preparation protocol. 
  3. Plate-specific barcodes are used to allow content from multiple plates to be run in a single sequencing run.

Custom BDTM Precise Targeted assays can be designed to interrogate approximately 80-240 genes simultaneously. Pre-designed panels are also available, covering specific biological areas, such as breast cancer, regulatory T cells, prostate cancer and other cell types.


Figure 3. BD Precise targeted assay workflow. Each well of the BD Precise targeted plate is preloaded with lysis buffer, dNTPs, and an oligo dT primer containing sample and molecular indices along with a universal priming site. When cells are sorted into the wells on the plate, the cells are lysed and mRNA contents are released. During cDNA synthesis, individual transcripts are tagged with a unique molecular index to provide a more accurate representation of transcript abundance in the original sample prior to amplification. Following cDNA synthesis, the sample indices allow all of the wells to be pooled together into a single tube to simplify sample handling. Two rounds of PCR amplification are done to amplify the specific genes of interest and to add sequencing adaptor and plate-specific indices. The amplified library is run on an Illumina NGS instrument.

In a BDTM Precise whole transcriptome analysis (WTA) assay, any transcript with a poly A tail will be interrogated. This assay best suited for gene expression discovery studies.


Figure 4. BD Precise WTA assay workflow. Cell lysis and cDNA synthesis are identical to the BD Precise targeted assay. Following cDNA synthesis, an optimized second-strand reaction is performed on the pooled samples, followed by adapter ligation to the 3' end of the first-strand cDNA which, along with the RT primer’s universal sequence, is used as a primer-binding site for whole transcriptome amplification. Library preparation is performed by random priming from the 3' ends of the amplified transcriptome, followed by size selection and PCR amplification, which adds Illumina flow cell binding sequences and optional plate indices.

BD has developed a series of BDTM Precise targeted assays, and a WTA assay to fit the specific needs of researchers.

To learn more about BDTM Precise kits, visit the BD website.

Data Analysis

The BDTM Precise assay sequencing data analysis pipeline is provided via Seven Bridges Genomics, a secure online cloud environment.

The pipeline takes FASTQ files from the sequencing run and collapses molecular indices using a proprietary algorithm. The number of molecules per gene per cell is reported in a Microsoft® Excel® (or tabular) file, which can then be imported into statistical software packages such as R for secondary data analysis.


Figure 5. Data processing steps for BD Precise assays.

Single cell sorting experiment

The following experiment demonstrates the data, which can be obtained in BD single cell analysis workflow.

Peripheral blood mononuclear cells (PBMCs) were isolated from human whole blood and stained with an antibody cocktail containing the cell surface markers CD16 PE, CD56 PE, CD3 FITC, CD45 PerCP-CyTM5.5, CD4 PerCP-CyTM7, CD8 APC-Cy7, and CD19 APC (BD MultitestTM 6-color TBNK Reagent, Cat. No. 644611).

Pan T cells were identified by CD45+CD3+ staining; B cells were identified by CD45+CD3-CD19+; and NK cells were identified by CD45+CD3-CD16/CD56+.

Using BD FACS sorter, thirty cells from each of the T, B, and NK populations were sorted into each BDTM Precise plate.

For the single T, B, and NK cells sorted from this experiment, mRNA expression was measured by the BDTM Precise targeted assay using a panel targeting 100+ genes expressed by immune cells in human peripheral blood (Figure 6). Genes included those that encode surface proteins, cytokines, and transcription factors.


Figure 6. Number of reads and transcripts obtained from single cells. A. Read counts from each cell in a 96-well plate (log scale). Individual cells were sorted into a BD Precise plate, featuring plate indexing that allows multiple plates (up to 4,000 cells) to be pooled into a single MiSeq run if desired. B. The average transcript count across 96 single cells from the same plate shown in Panel A after the reads have been collapsed into absolute molecule counts. Data illustrates absolute mRNA quantitation in single cells across a large measurement range. This information can be obtained from primary data analysis. NTC: no template control.

Once the transcript counts are obtained from the automated data analysis pipeline, the sequencing data can be further analyzed by various methods for biological relevance.

The following figure links mRNA expression measured by the BDTM Precise assay with protein expression measured by antibody staining.


Figure 7. Linking protein expression and mRNA expression using index sorting and BD Precise assay. Each cell in the TBNK sorting experiment was plotted on a 2-D space according to principal component analysis of the detected mRNA gene expression profiles (PC1 and PC2 are shown). Three populations were observed based on gene expression profiles, representing T, B, and NK cells. Top: each cell is colored by detected transcript level from the BD Precise assay. Markers shown: TRAC (TCR alpha constant region), GNLY (granulysin, secreted protein), IRF8 (transcription factor, intracellular). Bottom: each cell is colored by surface protein level as measured by FACS: CD3, CD16/56, and CD19, respectively. The cell type specific surface protein expression agrees largely with the cell types identified by gene expression. The index sorting feature enables the linking of protein expression with mRNA expression.

The combined workflow of BD FACS sorting combined with BDTM Precise assays provides investigators with a complete solution for high-throughput single-cell gene expression sequencing analysis.

To learn more about the offer, visit the BD website.

Publication: A Complete Workflow from Single Cell Isolation to mRNA Sequencing Analysis. Wang, A et al. (White paper by BD). Click here to view.