Questions: Can I perform Cell Hashing in a GEM-X Universal 3’ Gene Expression v4 workflow?
Answer: We have performed minimal testing in-house and consider Cell Hashing with GEM-X Universal 3’ Singleplex a compatible workflow modification. Internally, Cell Hashing has been tested with Gene Expression analysis, targeting higher cell loads. If targeting lower cell numbers (20,000 or lower) with Cell Hashing, additional optimization may be required. Although untested, Cell Hashing is expected to be compatible with Cell Surface Protein (CSP) experiments. Because extensive validation has not been performed, we can only provide limited troubleshooting guidance and Cell Hashing is not considered fully supported.
TotalSeq™-B antibodies can be used for Cell Hashing in GEM-X 3’ assays (see What is the difference between TotalSeq A, B, and C? for details on compatibility). These antibodies can be purchased from BioLegend directly. Prior to your experiment, we strongly recommend that you reach out directly to BioLegend Technical Support (tech@Biolegend.com) for antibody recommendations and guidance on experimental design prior to initiating your experiments. Titrating TotalSeq™ antibodies prior to initiating applications like CITE-seq is critical for optimal results. If the concentration of the hashing antibody is too high, it may overpower the other antibody reads and thus other lower abundant antibodies may be under-represented. Hashing data quality can also be impacted if unbound antibodies are not completely removed during wash steps.
Please note: using Cell Hashing with GEM-X Universal 3’ Multiplexing (OCM) is unsupported and not recommended due to the higher cell multiplet rate per sample than the multiplet rate for GEM-X Singleplex. For more details, see the “Target Cell Number & Chip Loading” section. Listed below are the sample type and species used for in-house testing with TotalSeq™-B hashing antibodies.
General Recommendations for Sample Quality & Compatibility
- Use single cell suspensions with >80% (ideally >90%) viability. If one or more samples in the pool has lower viability, fluorescence activated cell sorting (FACS) after pooling may increase viability and Hashing library quality.
- See the Cell Preparation Guide for more details on sample handling and quality best practices
- Perform pilot experiments to determine if the sample type is suitable for Cell Hashing
Sample Preparation Resources
- 10x Demonstrated Protocol: Cell Surface Protein Labeling for Single Cell RNA Sequencing Protocols
- 10x Technical Note: Quality Control of Cell Surface Protein Labeling using Flow Cytometry
To enable Cell Hashing, the Demonstrated Protocol for Cell Surface Protein Labeling for Single Cell RNA Sequencing Protocols must be modified. The recommended number of cells for this modified protocol is 0.2-2 x 106 cells per sample. The required changes are described below.
Cell Hashing Labeling Protocol
Prepare Antibody Mix Supernatant
-
Add appropriate/manufacturer's recommended amount of antibody-oligonucleotide conjugates to a 1.5-ml microcentrifuge tube
- TotalSeq™ antibodies including hashing antibodies (view Cell Hashing products) should be titrated to determine the optimal labeling concentration prior to use
- Please review the following Technical Note for additional information: Quality Control of Cell Surface Protein Labeling using Flow Cytometry
- The recommended amount of hashing antibody is 0.1-0.25 μg per sample
- Add appropriate volume of the hashing antibody to the Antibody Mix and maintain at 4°C
- Ensure that each sample is stained using an Antibody Mix containing a different hashing antibody, before pooling the samples
- When performing dual staining with Cell Hashing antibodies and TotalSeq™ antibodies, we recommend adding Cell Hashing antibodies into each respective sample’s TotalSeq™ antibody pool. Dual staining has not been tested with GEM-X, but a dual staining Next GEM dataset can be found here.
*Please note: Refer to BioLegend's product performance guarantee for recommended labeling. For example, >500k cells is not covered by BioLegend's product performance guarantee for the Universal Cocktails. Review the BioLegend protocol for their recommended procedure for the reconstitution of lyophilized TotalSeq Universal cocktails.
Labeling Protocol
- Resuspend cell pellet for each sample in 45 μl chilled PBS + 1% BSA
- Add 5 μl Human TruStain FcX to each sample. Gently pipette mix
- Incubate for 10 min at 4°C
- Add the prepared Antibody Mix supernatant (TotalSeq™ cell hashing antibody) to each sample. Each sample should be stained with a different hashing antibody before pooling the samples
- Add chilled PBS + 1% BSA to the cells to bring the total volume to 100 μl. Gently pipette mix 10x (pipette set to 90 μl). For samples containing <70% viable cells, add chilled PBS + 10% FBS.
- Incubate for 30 min at 4°C. If using FACS antibodies, incubate without light exposure.
Recommended Wash Protocol (Protocol 1)
An additional wash step (four total washes) or use of FACs post-pooling may increase Cell Hashing data quality. Wash cells four times, by adding 3.5 ml chilled PBS + 1% BSA to the labeled cells. After each wash, the resuspended cell pellet or cells should be transferred to a new 5-ml tube. Centrifugation speed and time depends upon the sample type. Use Table 2 in the Demonstrated Protocol for guidance. Please see Cell Surface Protein Labeling for Single Cell RNA Sequencing Protocols for additional details.
-
For enrichment of labeled and viable cells by FACS (optional):
- Based on starting concentration and assuming ~50% cell loss, add an appropriate volume chilled PBS + 2% FBS (including a dead cell marker) to obtain a final cell concentration of 5-10 x 106 cells/ml and proceed to FACS (see FACS Guidance in Cell Surface Protein Labeling for Single Cell RNA Sequencing Protocols for additional details)
- After FACS, determine cell concentration and viability
-
If not performing FACS:
- Based on starting concentration and assuming ~50% cell loss, add an appropriate volume chilled PBS + 1% BSA to obtain a concentration of 1,300- 1,600 cells/μl
- Samples should be slowly filtered using 40 µm Flowmi™ Cell Strainer (or appropriate cell strainer; see the Cell Preparation Guide for details)
Sample Multiplexing
- For optimal multiplet detection and optimal signal-to-noise ratios, we recommend pooling samples at 1:1 ratios
- Determine cell concentration and viability of the pooled sample. Proceed immediately to the relevant GEM-X 3’ User Guide with Feature Barcode technology.
- Ideally the viability should be >90% after filtration for optimal capture rate. The presence of a large number of non-viable cells can decrease the efficiency of cell partitioning and recovery within the 10x Genomics chip.
Target Cell Number & Chip Loading
In each GEM-X Universal 3’ library, it is possible to target up to 60,000 cells when performing Cell Hashing with the understanding of the risk.
Exceeding the maximum number of recommended cells per library may increase the undetected multiplet rate and risk for microfluidic failures (i.e., wetting or clogging failures). An important consideration when deciding how many cells to target is the multiplet rate. As more cells are loaded on the 10x chip, it becomes more likely that two or more cells will be partitioned together into the same GEM (cell multiplet).
Below is a Cell Multiplet Table for GEM-X Universal 3’ Singleplex with approximate multiplet rate, cells loaded, cells recovered, and the breakdown of singlets and multiplets.
OCM enables small-scale and low-sample multiplexing using an alternative approach to Cell Hashing. Due to GEM-X 3’ OCM having a higher per sample cell multiplet rate than GEM-X 3’ Singleplex (at the same targeted cell numbers per sample), we would strongly advise against performing Cell Hashing with OCM. Approximate per sample multiplet rates, cells loaded, cell recovered, and the breakdown of singlets and multiplets for OCM are shown below. Overloading leads to higher cell multiplets per sample and higher wasted data per emulsion.
Once a target cell number has been decided, please use the GEM-X 3’ Singleplex Cell Loading Table below when preparing the Sample Master Mix for GEM-X 3’ Chip loading. Select suspension concentrations have been excluded from the table and gray cells indicate volumes that would exceed the allowable cell suspension volume in each reaction.
cDNA Amplification & Library Construction Modifications
cDNA Amplification
- For cDNA Amplification, cycling conditions are recommended to be updated based on the target cell number
-
The recommended starting point for cycle number optimization is shown below
- The optimal cycle number is a trade-off between generating sufficient final mass for libraries & minimizing PCR amplification artifacts
Library Construction
- For GEM-X Universal 3’ Gene Expression v4 Library Construction, cycling conditions are unchanged
Cell Hashing Library Construction
- When targeting >20,000 cells, we recommend reducing the number of SI-PCR cycles from 10 total cycles to 8 total cycles
*Further optimization of cycle number may be needed based on target protein expression levels and number of antibodies used for labeling. The final library concentration (nM) is lowered by 3.5-4 fold with every 2 cycle reduction in SI-PCR cycles.
Sequencing
We recommend a sequencing depth of a minimum 5,000 read pairs/cell for each Hashing library for obtaining sufficient coverage. If sequencing GEM-X 3’ Hashing libraries with Gene Expression libraries, they should be sequenced in a 28 x 10 x 10 x 90 bp configuration. If sequencing GEM-X 3’ Hashing libraries alone, Read 2 can be shortened so that the libraries can be sequenced in a 28 x 10 x 10 x 25 bp configuration.
Data Analysis
We recommend using Cell Ranger versions 9.0+, for the analysis of multiplexed cells and samples using antibody hashtags. Please see here for more guidance on how to run the pipeline.
Datasets
- Analysis Guide: Demultiplexing and Analyzing 5’ Immune Profiling Libraries Pooled with Hashtags
- Published Dataset: 60k Human PBMCs Stained with 8 TotalSeq™-B Human Hashtags (Donor 1-4, 4 samples)
See this 10x Genomics software page for more information on sample multiplexing.
Products: Single Cell Gene Expression