Question: Can I process neutrophils (or other granulocytes) using 10x Single Cell applications?
Answer: Single-cell analysis of neutrophils and granulocytes generally remains a significant challenge. This is because neutrophils (and other granulocytes) have relatively low RNA content and relatively high levels of RNases and other inhibitory compounds, resulting in fewer transcripts detected in GEMs, and less usable sequencing reads. Furthermore, neutrophils are particularly sensitive to degradation after collection and cannot be isolated via density gradient centrifugation (as is the case for other nucleated blood cells, often referred to as PBMCs).
A few points to consider when optimizing the 10x Genomics Single-Cell Assays for neutrophils:
1) Immediate processing is critical. Ideally, neutrophils should be processed immediately after collection, and any delay longer than two hours is likely to result in a failed run. Isolation of neutrophils from frozen samples may also present significant challenges, and fresh samples are always preferred.
2) Supplementing RNase inhibitor into the wash and resuspension buffer may also help preserve your RNA. We have tested the RNase inhibitor (Sigma Aldrich PN-3335399001).
3) Neutrophils may not react well to long incubation periods on ice. They prefer room temperature during the sample prep.
4) Increasing the PCR cycles by two extra cycles during cDNA amplification in a non-enriched sample may help bring the neutrophil-specific transcripts above the background.
5) When isolating neutrophils from whole blood, Red Blood Cells should be removed from the sample using an RBC lysis buffer such as ACK (ammonium chloride potassium). This will reduce the sequencing required by enriching the sample for nucleated cells, including neutrophils. Instructions for isolating leukocytes (i.e., PBMCs and granulocytes) from whole blood samples can be found in the Isolation of Leukocytes, Bone Marrow, and Peripheral Blood Mononuclear Cells for Single Cell RNA Sequencing Demonstrated Protocol.
6) Sorting neutrophils before sequencing may be desirable to improve the number of neutrophils assayed and reduce total sequencing costs. Sorted cells should be deposited directly into 1x PBS with 0.04% BSA + RNAse inhibitor (i.e., the recommended input buffer for 10x single-cell assays). The sorted single-cell solution should be used as the 10x assay input. Forward and side-scatter should be used to exclude cell doublets.
7) Cell Ranger software performs an unsupervised analysis of the sample. If you have a sample that contains two very distinct sub-populations, one with lots of RNA/cell and another with very little RNA/cell, the latter may be incorrectly labeled as background. Thus, CellRanger may filter out the neutrophils that do not have many mRNAs. We recommend using the --force-cells parameter and setting its value to the target cell recovery.
As neutrophils exhibit the largest amount of retained introns compared to other cell types in peripheral blood samples (Wong et al., 2013, Ulrich & Guigo, 2020), the "Intron Mode" parameter in CellRanger(--include-introns) can be used to recover reads assigned to introns. By activating this parameter, intronic reads will be considered in the analysis, increasing the overall UMIs per barcode and thus, increasing neutrophil discovery.
We also have a tutorial on preserving and annotating the neutrophils in your data.
We have successfully identified granulocytes in whole blood samples processed using the leukocyte isolation instructions outlined in Option 1 of the Isolation of Leukocytes, Bone Marrow, and Peripheral Blood Mononuclear Cells for Single Cell RNA Sequencing Demonstrated Protocol.
The data can be accessed here :
3'GEX datasets and 5'GEX datasets
A few neutrophil-relevant 10x Genomics papers are also listed below:
- https://www.biorxiv.org/content/10.1101/792200v1.full.pdf
- https://www.sciencedirect.com/science/article/pii/S1097276519301376
- https://science.sciencemag.org/content/360/6390/758
- https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5969381/
- https://link.springer.com/content/pdf/10.1007/s13238-020-00752-4.pdf
With 3' v4 and 5' v3 assays, data generated internally was performed by following a subset of best practices listed in this article (considerations 1, 2, and 5 listed above). Additional cycles for cDNA Amplification or adjustments to the Cell Ranger pipeline were not required.
Products: Single Cell Gene Expression, Single Cell Immune Profiling, Fixed RNA Profiling Gene Expression