The Nadia Instrument is designed to encapsulate single cells using barcoded mRNA beads.
Microfluidics droplets are being used to encapsulate cells in revolutionary technology that allows thousands of single cells to be analyzed during the same run. This level of analysis can allow the identification of new cell types or cryptic cell types in the tissue under study, to help uncover the properties of single cells as a way to understand how important processes in biology operate.
Many different approaches are being tried out, including Drop-seq (Macosko E., et al., "Highly Parallel Genome-wide Expression Profiling of Individual Cells Using Nanoliter Droplets." Cell 161:1202) as well as inDrop (Klein, AM., et al. "Droplet Barcoding for Single-Cell Transcriptomics Applied to Embryonic Stem Cells." Cell 161:1187).
The inDrop technique is based upon barcoded beads in lysis buffer, which can undergo deformation. These are encapsulated with a cell including a reverse transcription reaction mix. Once inside the droplet, the cell is lysed, the barcoded bead binds the mRNA and reverse transcription occurs at once within the emulsion.
The Drop-seq is a different method, which encapsulates single cells and non-deformable beads into droplets as shown in Figure 1. The beads capture polyadenylated mRNA through their surface coating with oligonucleotides that have a stretch of (dT)30. Once emulsification occurs and the emulsion breaks, reverse transcription takes place and cDNA is created. The unique DNA barcode on every bead serves to identify the separate cells, as well as the individual mRNA transcriptome from a given cell.
Figure 1 Schematic of the encapsulation of barcoded mRNA capture beads with single cells.
Tens of thousands of single cells are co-encapsulated with uniquely barcoded beads. The cells
are lysed inside the droplets and their mRNA content is captured on the beads. These beads
are subsequently recovered and processed for downstream analysis using NGS technologies.
As more and more commercial instruments become available for automated single cell analysis at high throughput rates, robust and flexible microfluidics platforms for cell encapsulation are becoming a growing reality. Unlike the first frail prototypes, Dolomite Bio’s Nadia Instrument uses rapid and convenient high-throughput technology to analyze single cells at up to 8 cells being encapsulated in parallel.
With its world-leading experience in microfluidics design and manufacture, Dolomite has brought out a fully automated temperature-controlled instrument with homogenization capabilities included in its feature repertoire.
The Nadia Instrument and other similar high-throughput systems must be capable of encapsulating only a single cell into a droplet at a time, with one bead. The typical approach is by diluting the suspensions of cells and beads.
This involves a trade-off between quality and quantity, or in other words, between fewer cells analyzed but with the extraction of high-quality data and a higher throughput but at the expense of low-quality data.
The Nadia Instrument allows over 6000 cells to be analyzed singly, yielding individual transcriptomes, within one 20-minute run. This is the setting usually employed in Drop-seq but the Nadia Instrument can also be run with lower sample quantities, by simply decreasing the run time and the volume of the sample in case the input material is scarce. This article discusses the use of the Nadia Instrument with the Drop-seq protocol to achieve single cell encapsulation with barcoded beads for mRNA capture to sequence single cell RNA.
Materials and Methods
Figure 2 Nadia Instrument
The Nadia Instrument, shown in Figure 2, is a high-throughput droplet microfluidics device designed for single cell transcriptome analysis. Its use in combination with three separate pressure pumps (which have extremely smooth pumping action) leads to the production of droplets with high monodispersity.
It helps the user complete any given application by its clear and detailed instructions at every step. The cooling and stirring of the droplets during the process of encapsulation leads to even suspension of cells and beads, with high quality being maintained. Each run can be specifically designed by the user because parameters like sample volume, sample number, and the speed of stirring by the cell stirrer, can all be changed by the user to meet specific requirements.
How to Make Cell and Bead Preparations and Encapsulation
The Drop-seq on Nadia protocol describes the materials and methods in detail, and this can be obtained from [email protected] on request.
Cell viability testing was carried out to confirm the lack of adverse effects following gentle stirring to keep cells in suspension. This test used 0.4% Trypan Blue Solution (#15250061, ThermoFisher Scientific). The procedure involved stirring MEF cells in a suspension of 6400 cells/µl in a 1:1 solution of PBS-BSA + 0.4% Trypan Blue in the Nadia Instrument for a period of 30 minutes, at 4˚C. They were then compared to other cells which had been kept on ice for the same amount of time, without stirring. The number of unstained or viable cells was found, to indicate cell viability.
Preparation and NGS-analysis of Single Cell DNA Libraries
After encapsulation of cells and beads, the next steps followed the Drop-seq protocol on the Nadia Instrument. That is, six PCR reactions were carried out using aliquots of 2000 beads, corresponding to 600 STAMPs (Single-cell Transcriptome Attached to MicroParticles). A BioAnalyzer was used to determine the quality of DNA once PCR amplification and tagmentation were completed.
Illumina Sequencing and Bioinformatic Pipeline
The DNA libraries formed by tagmentation were sequenced using the NextSeq 500 instrument (Illumina) with a 2x75 bp paired-end run. The read lengths were as follows: 26 bp for read1, 8 bp for the index read, and 116 bp for read 2. Data analysis was performed using a computational workflow from the McCarroll Lab at Harvard, called the dropSeqPipe. The analysis followed Macosko et al (2015).
A set of mixed-species experiments was carried out to measure the quality of performance of the Nadia Instrument and the general effectiveness of the whole workflow. This was based on the Drop-seq on Nadia protocol. Some of the factors that could affect this include:
- The quality of the emulsion (this is measured by the doublet rate since this is influenced by emulsion quality and cell suspension quality – it is defined as the percentage of beads that capture more than one cell)
- The rate of PCR duplication (the ratio of UMIs (Unique Molecular Identifiers) detected to the number of reads per cell)
- The rate of gene capture (the number of genes detected at a specific sequencing depth, which shows how efficient the overall protocol is)
- Reproducibility (how well the results of experiments by a variety of users on different days agree with each other)
Flowing Beads and Cells
The Nadia Instrument has cell stirrers incorporated into the device to produce gentle stirring of the cell and bead suspensions. This helps them remain evenly suspended throughout the run. Even encapsulation is also encouraged, minimizing the chances of doublet formation (either beads or cells).
Gentle rotation of the stirrers, away from the bottom or walls of the chambers, is performed to prevent any cell or bead damage. The measure of emulsion quality is monodispersity, measured as the estimated average size of the droplets and the rate at which beads are encapsulated, as shown in Figure 3.
Figure 3 Droplets produced during a Drop-seq run on the Nadia Instrument.
The quality of the droplets was subjected to analysis over 30 runs carried out by the company or by customers. This showed an average droplet size of 81 µm with a CV (coefficient of variation) of 4 %. To find the rate of bead encapsulation, the number of droplets that contained a bead was divided by the total number of droplets. This was found to be 11%, which is in concordance with the expected theoretical rate of 10%.
Following cell viability testing, as described above, it was found that the two groups of cells used in the separate experiments showed no significant difference, namely, the stirred group and the storage-on-ice group showed 91% and 90% viability respectively.
Preparation of Single Cell cDNA Libraries for NGS-analysis
Once the emulsion was collected from the Nadia Instrument, the beads were treated as per the Drop-seq on Nadia protocol. That is, following PCR amplification (shown in Figure 4A) and tagmentation (shown in Figure 4B), a sample of the cDNA library was analyzed using a BioAnalyzer. This showed an estimate of the size distribution and the quality of the overall sample. The PCR products showed sizes between 400-1000 bp, which agreed with that expected of a mammalian cDNA library, namely, an expected bp range 600-2500 bp.
Figure 4 BioAnalyzer results before (A) and after tagmentation (B).
Performance of the Drop-seq Protocol on the Nadia Instrument
How well did the protocol work? This is measured by looking at the quality of data, using knee- and Barnyard plots. The knee plot supplies the estimated number of beads that have captured cell mRNA (productive beads) in the sample. The beads have a unique cell barcode associated with them, and a corresponding number of NGS reads.
These cell barcodes are sorted in descending order of reads and imaged along with the cumulative fraction of reads. The plot below achieved cDNA library amplification using 12 000 beads. If the dilution is 1 cell in 20 droplets, about 5% of beads should be productive, leading to 600 cellular transcriptomes, in theory. The plot below shows the inflection point is at 537 STAMPs, which is in good agreement with this prediction (according to Poisson statistics, shown in Figure 5).
Figure 5 A knee plot comparing cell barcodes in descending order of reads alongside the cumulative fraction of reads.
A Barnyard plot is used to find the doublet rate in the mixed-species experiment, which shows the STAMPs as separate data points, each of which is linked to a number of transcripts from mouse or human cells, as shown in Figure 6. With the data set used here, the number of STAMPs that was associated with only mouse or only human transcripts was 261 and 258 respectively, while 18 STAMPs associated with mixed species transcripts, corresponding to a 6.6% doublet rate.
Evaluation of PCR Duplication Rate and Gene Capture Efficiency
The measure of PCR duplication and gene capture efficiency is found from a set of four separate Drop-Seq on Nadia runs, which is compared to already published data (Macosko et. al., 2015). The data set comprised two replicates, in which there were 200 STAMPs at a low sequencing depth of approximately 20 000 reads per cell, and two in which there were 600 STAMPs at 50 000 reads per cell and 120 000 reads per cell respectively. Table 1 shows this comparison, with the number of cells by knee plots, the number of NSG reads per cell, the doublet rate and the median number of detected genes and UMIs.
The doublet rates were comparable in all the samples, in general, at an average of 7%. This is in concordance with the published data. The median detected genes and UMIs was affected by the number of cells in the analysis and the sequencing depth per cell.
Figure 6 A Barnyard plot depicting STAMPs as individual data points each associated with a number of human or mouse transcripts
Table 1 Comparison of a set of four independent Drop-seq on Nadia runs to published data
The efficiencies of gene capture were assessed by plotting the cells analyzed for each data set (there were five in all) against the number of genes and the number of reads detected per cell, as shown in Figure 7. This reveals the increase in the number of genes detected per cell, as the sequencing depth goes up. Gene capture was efficient even at low sequencing depth, leading to a more cost-effective NSG analysis, but increasing the sequencing depth will allow less widely expressed gene transcripts to be detected.
Figure 7 The number of detected genes was compared to the read depth across all samples.
If the number of detected UMIs is plotted against the reads per cell, the PCR duplication rate that has occurred during the amplification of cDNA can be estimated, as shown in Figure 8. In this plot, the black line shows the ideal ratio of 1 UMI per read. All the datasets acquired using the Drop-seq on Nadia protocol were in alignment to this line, but the extent of departure was related also to the sequencing depth.
Figure 8 The number of detected UMIs was compared to the read depth per cell across samples. The black line represents the ideal ratio of 1 UMI per 1 read, indicating no PCR duplicates
Reproducibility of Drop-seq Data on the Nadia Instrument
The data shown in this experiment were acquired by different users on different days, and then compared to measure variability between experiments as well as between users when performed on the Nadia Instrument. All data sets were compared in pairs using the criterion “number of UMIs detected per gene across all cells” as shown in Figure 9. An R value between 0.84 and 0.93 was obtained across all data sets, which indicated a good degree of reproducibility between users and replicated experiments.
Figure 9 This plot shows the correlation of four datasets from different users generated on different days on the Nadia instrument. Compared are the numbers of UMIs detected per gene
The emergence of high-throughput methods of analyzing single cells allows scientists to gain insight into how cells express gene profiles, by the thousands, at the level of a single cell, to understand how homogeneous a tissue is, or to detect new cells or mutation clusters. The demand for such methods is, quite understandably, increasing by leaps and bounds.
The Nadia Instrument is an efficient and affordable platform which allows reproducible analysis of encapsulated cells along with barcoded genes which capture mRNA from cells. The instrument produces droplets of high monodispersity, at a 10% cell capture rate. The provision for gentle stirring keeps the beads and cells in suspension and thus allows encapsulation to take place without causing cell damage. This process leads to even bead/cell encapsulation, which is responsible for a low doublet rate of only 6.7%. One run leads to the capture of over 6 000 cells.
With the use of NGS (Illumina), sets containing 200 and 600 cells were subjected to processing and analysis, to assess the level of PCR duplication and the cell capture rate. This was then subjected to comparison analysis using a data set of 1000 STAMPs (Macosko et. al., 2015), which is the reference in this study. The criteria used included the doublet rate, capture rate for UMIs per read, and number of genes detected. The average doublet rate for all the single cells analyzed with the Nadia Instrument was 7%, which agreed well with that in Macosko’s data. The doublet rate depends, however, not only on the quality of droplet formation but the technique of cell preparation and the cell concentration. In their study, Macosko et al found cell doublet rates of 0.36% to 11.3% at cell concentrations of 12.5 cells/µl to 100 cells/µl (Macosko et. al., 2015).
The Nadia Instrument can process cells in concentrations of 300 cells/µl to produce a faster run, while maintaining a low doublet rate, simultaneously conferring the additional benefit of reduced cell death risk and lowering the risk of stress-induced alterations in the transcriptome. This is further reduced by the integrated temperature control feature which keeps the samples at 4˚C during a run.
Again, the run quality and efficiency is assessed by the number of UMIs and detected genes per cell at a given depth of sequencing. These parameters are affected by many factors, such as the rate of binding transcripts, cell lysis, the efficiency of reverse transcription and library amplification, and the depth to which sequencing proceeds for each cell.
The data sets published by the Nadia 600 R1 and the Macosko 1000 STAMP studies have similar sequencing depths, and are therefore ideal for making a comparison of the following factors. It was seen that the use of Nadia achieved a detection rate of about 20,000 UMIs per cell, significantly exceeding the value of 16,000 in the Macosko study. The number of genes detected is more than 5500 per cell with the Nadia protocol, again more than the 4,800 in the other study. Thus it may be claimed that the Nadia Instrument provides comparable or even better results than other platforms available commercially.
This could be attributed to the way in which the Drop-seq protocol carries out reverse transcription outside the droplet, and thus keeps it apart from cell lysis. This means that quite strong lysing agents may be used to achieve almost complete cell lysis for more cell types, while also preventing potential ill effects of a lysis buffer on the efficiency with which the reverse transcription reaction proceeds.
Another intriguing result is that detected gene number was as large as 2000-3000 with low sequencing speeds of 20,000 reads per cell. This has potential for enabling rapid sample characterization or large sample sequencing at cost-effective rates if needed.
The UMIs per oligonucleotide on a bead surface are a tool which is of great utility in distinguishing transcript numbers from PCR duplicates. In the ideal conditions, a single read should be present for each detected UMI in the absence of PCR duplication.
Low duplication of PCR is therefore indicative of data sets close to this condition. The Nadia Instrument generated data sets in good alignment with this ideal line, showing that PCR duplication was extremely minimal, and thus reducing the cost of sequencing. The sequencing depth also affected the ideal ratio of 1:1, causing increasing deviation of the data set as the depth increased.
The reproducibility of data sets generated by the Nadia Instrument was also examined and found to be good, in terms of agreement between results obtained by different users, and from the same experiment on different days. The R-value was compared across data sets and found to show good inter-user and inter-sample reproducibility.
The use of the Drop-seq on Nadia protocol led to good-quality data sets in the form of single cell cDNA libraries. These were analyzed on several parameters to evaluate how well the instrument performed as well as to assess the general effectiveness of the protocol.
This involved measures of emulsion quality, gene capture and PCR duplication rates. It was shown that the protocol allowed excellent gene capture with low PCR duplication, as well as low doublet rates, thus making it a very cost-effective way to analyze single cell transcriptomes.
About Dolomite Bio
Dolomite Bio creates innovative products for high throughput single cell research including the Nadia Instrument and the Nadia Innovate. By encapsulating single cells in microfluidic droplets, our products enable rapid analysis of thousands or millions of individual cells and their biological products.
Key benefits of Dolomite Bio systems:
- Flexible: The systems enable full control of pressures, flow rates and temperatures
- Compatible with a range of applications: Not limited to a specific protocol
- Productised: Complete off-the-shelf, easy to use systems
- High precision with reproducible results: Monodisperse droplets with low doublet rate of cells and beads
Our products are complete systems enable scientists to generate thousands of single cell libraries in just a few minutes.
Our in-house team of biologists and worldwide network of local specialists work with you to provide advice for your application, product demonstrations, installation, training and support.
Sponsored Content Policy: News-Medical.net publishes articles and related content that may be derived from sources where we have existing commercial relationships, provided such content adds value to the core editorial ethos of News-Medical.Net which is to educate and inform site visitors interested in medical research, science, medical devices and treatments.