Higher Throughput, More Flexible Single-Cell Multiomics Analysis

BD Biosciences describes a platform that can isolate, barcode, and analyze single cells at high throughput without sacrificing sample integrity


By Aruna Ayer, PhD, and Cynthia Sakofsky, PhD

Single-cell multiomics provides a comprehensive view of the cellular heterogeneity and the complex interplay between multiple layers of cellular omics, namely genome, epigenome, transcriptome, and proteome. Researchers are increasingly applying single-cell multiomics to unveil cellular complexity across many fields, including cancer research, drug discovery, infectious disease research, and more.

With more affordable next-generation sequencing options, single-cell multiomics will increasingly be used as a common approach to profiling cells and tissues. Understandably, the number of single-cell technologies has also rapidly grown in recent years and includes droplet- and microwell-based platforms, as well as microfluidics-free and instrument-free single-cell workflows.

Even with the myriad options for single-cell assays that these technologies provide, researchers are often challenged with technical issues that can impact biological outcomes, such as batch effects or experimental run-to-run variability, sample loss, low cell capture rate, high cell multiplet rate, and high background noise, to name a few. Hence, there is a need for a single-cell platform that can overcome these challenges and provide an accurate and reproducible representation of cellular information in diverse samples, while also providing higher and flexible throughput that can provide cost savings and boost experimental efficiency.

A system for single-cell multiomics analysis

The high-throughput BD Rhapsody™ HT Xpress System leverages a microwell-based single-cell partitioning technology with minimal benchtop equipment to perform single-cell analysis. With no fluidic pumps or microfluidics, there is no clogging of channels with precious samples. The product features a flexible eight-lane cartridge design with the ability to run up to eight times the number of lanes as the on-market BD Rhapsody Express System with similar workflow time and performance.

With the eight-lane cartridge, one can process up to half a million cells per cartridge. Additionally, in combination with the new BD Flex Single-Cell Multiplexing Kit (SMK), a user can run up to 24 samples per lane or up to 192 samples per cartridge. Single-cell data obtained from each lane is reproducible and concordant between lanes and cartridges, with no lane-to-lane contamination within a cartridge.

Figure 1. A t-SNE plot generated for a sample that was loaded at 5,000 cells per lane versus 65,000 cells per lane shows no batch effect between lanes at varying cell input (top left). Differential gene expression correlations between 5,000 and 65,000 cell loads on different lanes show high concordance (top right). A bar graph shows the percentages of expected sample tags identified in different lanes—results that demonstrate absence of lane-to-lane contamination (bottom).

A demonstration of system capabilities

Samples stained with different sample tags from an SMK showed no unexpected sample tag data in neighboring lanes (Figure 1). The HT Xpress also features partial use of the cartridge, allowing users to run the unused lanes of the same cartridge at a different time for up to four months.

The HT Xpress offers cell retention of samples with minimal cell loss. Cell capture rates are typically >80%, even with cells of varying sizes and fragility, including neutrophils, T cells, and natural killer cells, as well as nuclei. The system enables the introduction of cells into a cartridge and allows them to settle into individual microwells by gravity. This process results in minimal cell manipulation and assures cell and mRNA integrity.

The stochastic pattern of cells falling into microwells follows a Poisson distribution, which can be used to theoretically estimate the number of wells that contain more than one cell, that is, a multiplet. When a multiplet occurs, the transcriptomes and/or proteomes of two or more cells are captured on a single barcoded bead simultaneously, rendering the data obtained from these cells unusable since the individual cell information cannot be deconvoluted.

The HT Xpress includes a scanner component, that can visualize cells and beads in the wells, providing an empirical estimate of multiplet rates. There is high concordance between both theoretical and scanner estimates of multiplets, even at high cell inputs. In fact, with cell inputs close to 60,000 per lane, scanner multiplet rates have been shown to be reproducibly <10% (Figure 2).

The scanner used in the HT Xpress requires just a field upgrade to the Rhapsody scanner. The scanner enables not only empirical multiplet rate estimates, but also visual quality control (QC) of cell viability and step-by-step QC metrics of the cartridge workflow, including cell and bead capture, wash steps, and bead retention rates.

Figure 2. Varying cell loads (100 to 65,000 cells) in the different lanes of the eight-lane cartridge demonstrate a predictable trend in cell capture and cell multiplet rates.

Implications for scientific discovery

Such a visual in-process QC after single-cell partitioning allows users to make more informed decisions about their experiments prior to library preparation and sequencing, which can save a user a substantial amount of money. Additionally, the biomolecules captured on the beads can be stored for up to four months after the cDNA conversion step, allowing users more flexibility to subsample beads for initial shallow sequencing to evaluate library quality and performance metrics before deciding to potentially proceed with further sequencing of the entire experiment.

The HT Xpress fits within an end-to-end single-cell multiomics solution that is supported by BD Rhapsody single-cell multiomics assay kits and the BD Rhapsody Analysis Bioinformatic Pipeline tool. The pipeline tool generates detailed output files that can be used for comprehensive secondary analyses.

The latest pipeline version also automatically generates a sharable HTML file that highlights QC and summary metrics, with an additional interactive t-distributed stochastic neighbor embedding (t-SNE) plot displaying features such as single bioproduct (gene or protein) expression data and immune cell-type calling information (Figure 3).

Figure 3. Example of the interactive portion of an HTML file showing a single bioproduct expression graph displaying a t-SNE on the left and a histogram on the right for individual bioproducts (top), and an immune-cell-type experimental graph showing a t-SNE plot with each cell annotated based on a cell-type prediction algorithm (bottom).

Aruna Ayer, PhD, is a senior director, and Cynthia Sakofsky, PhD, serves as a staff scientist at BD Biosciences. To learn more about the BD Rhapsody single-cell multiomics solutions, visit bdbiosciences.com.