This protocol outlines the procedure for rapidly dissociating human and mouse tumor samples for single-cell RNA sequencing.
Human tumor samples hold a plethora of information about their microenvironment and immune repertoire. Effective dissociation of human tissue samples into viable cell suspensions is a required input for the single-cell RNA sequencing (scRNAseq) pipeline. Unlike bulk RNA sequencing approaches, scRNAseq enables us to infer the transcriptional heterogeneity in tumor specimens at the single-cell level. Incorporating this approach in recent years has led to many discoveries, such as identifying immune and tumor cellular states and programs associated with clinical responses to immunotherapies and other types of treatments. Moreover, single-cell technologies applied to dissociated tissues can be used to identify accessible chromatin regions T and B cell receptor repertoire, and the expression of proteins, using DNA barcoded antibodies (CITEseq).
The viability and quality of the dissociated sample are critical variables when using these technologies, as these can dramatically affect the cross-contamination of single cells with ambient RNA, the quality of the data, and interpretation. Moreover, long dissociation protocols can lead to the elimination of sensitive cell populations and the upregulation of a stress response gene signature. To overcome these limitations, we devised a rapid universal dissociation protocol, which has been validated on multiple types of human and murine tumors. The process begins with mechanical and enzymatic dissociation, followed by filtration, red blood lysis, and live dead enrichment, suitable for samples with a low input of cells (e.g., needle core biopsies). This protocol ensures a clean and viable single-cell suspension paramount to the successful generation of Gel Bead-In Emulsions (GEMs), barcoding, and sequencing.
Although many advancements in cancer research have led to the development and FDA approval of agents targeting inhibitory receptors expressed on immune and tumor cells, known as checkpoint blockade inhibitors, therapy resistance and identifying mechanisms that drive patient response remain challenging1. The complex challenge of characterizing tumor heterogeneity on its molecular diversity and the intricate interplay between tumor cells and the immune microenvironment necessitates new approaches to dissect this complex ecosystem at the single-cell resolution. Understanding the molecular intricacies within tumors and their microenvironments is pivotal for advancing therapeutic strategies and deciphering the complex biology underlying cancer progression2. Single-cell RNA sequencing (scRNA-seq) has become increasingly popular because it provides high-resolution analysis of individual cells within complex tumor samples3,4.
Tumor heterogeneity, encompassing genetic, epigenetic, and phenotypic diversities, poses a challenge in uncovering the intricacies of cancer biology5. The conventional methods of bulk RNA sequencing tend to obscure the unique expression profiles and phenotypes of heterogeneous cell populations present within tumors, as these methods average signals across entire cell populations4. In contrast, scRNA-seq allows the dissection of individual cell types, revealing diverse gene expression, repression patterns, and cellular states otherwise overlooked4,6. The premise of scRNA-seq lies in its capacity to decode individual cells' genetic and molecular signatures, holding immense promise in discovering new cancer therapeutics7.
By deciphering the gene expression profiles of tumor cells and the surrounding stromal and immune cells, researchers can identify novel therapeutic targets and develop precision genetic medicine strategies tailored to individual patients6. Furthermore, dissecting the immune repertoire within the tumor microenvironment provides crucial insights into the interplay between tumor cells and immune effectors, paving the way for immunotherapeutic interventions3. Despite advances in using scRNAseq on clinical samples, several challenges during the processing steps can harm the cell's RNA integrity, viability, and quality of the data generated. Moreover, processing tissue with low cell viability can increase ambient RNA levels in the droplets generated during the encapsulation of individual cells, resulting in cross-contamination and incorrect annotation of cell types, while long dissociation protocols performed at 37 °C can lead to the upregulation of a stress response gene signature in multiple cell types8,9,10.
The stepwise procedure (Figure 1A) outlined in this protocol commences with rapid mechanical and enzymatic dissociation of tumors, followed by filtration and the removal of dead cells, depending on the viability of each tumor sample. At present, this procedure has been verified in melanoma11, colorectal12, head and neck squamous cell carcinoma13, pancreatic and lung (unpublished data) tumor samples. These techniques ensure the generation of high-quality viable cell suspensions crucial for downstream analyses14. Notably, removing red blood cells, devoid of synthesized RNA, becomes imperative to purify the sample15. Subsequent evaluation of cell counts and the elimination of dead cells serve as a prerequisite for successful 10x Genomics Gel bead-in Emulsion (GEM) generation, barcoding, and increase the recovery of transcripts and sequenced cells without jeopardizing the exclusion of sensitive populations (e.g., neutrophils and epithelial cells)16. These steps are fundamental in unlocking the potential of single-cell resolution analysis within tumor samples9,10, uncovering novel gene expression profiles and immune signatures necessary for driving innovative therapeutic interventions, and identifying new tumor vulnerabilities.
This study complied with all institutional guidelines regarding human tissue sampling. Informed consent was received from patients, and identifiable sample data was anonymized. Samples are collected in the operation room, placed in a solution of RPMI, saline, or PBS, and confirmed cancerous via the pathology department before use in research. All steps, except when indicated, should be carried out at 4 °C or on ice. Work inside a biosafety hood when processing human tissue. See the Table of Materials for details on all materials, reagents, and instruments used in this protocol.
1. Obtaining the sample
2. Mechanical and enzymatic digestion
3. Sample filtering
4. Red blood cell lysis
5. Cell counting
6. Dead cell removal – Kit 1
NOTE: Work inside a biosafety hood when using the dead cell removal reagents, as these can easily get contaminated.
7. Dead cell removal – Kit 2
NOTE: Work inside a biosafety hood when using the kit reagents.
A melanoma biopsy suspended in RPMI was obtained and immediately placed on ice. The sample was transferred into a 1.5 mL microcentrifuge tube containing 420 µL of RPMI and digestion enzymes, minced into small pieces using scissors, and subjected to subsequent enzymatic digestion for 15 min in a 37 ˚C vertically positioned thermal mixer (Figure 1). Following digestion, the sample was filtered through a 50 µm sterile disposable filter, and the filter was washed with fresh RPMI to increase cell yield (Figure 1). The remaining parts of the tissue left after the washes were discarded, as these are unwanted biological materials (e.g., adipose tissue).
Based on the color of the resulting cell pellet, 15 mL of ACK lysis buffer was needed to remove the red blood cells (Figure 2). Following resuspension in 3 mL of RPMI media, the total cell count was 1 × 106/mL, with viability that was <80% (Figure 3A). This indicated that one cycle of live cell enrichment using Kit 1 (protocol section 6) was needed before proceeding to 10x Genomics loading (Figure 4). If this sample had over 1 million cells and had a viability of less than 10%, the Dead Cell removal – Kit 2 would have been utilized instead (Figure 5). The sample was washed with isolation media, and 10 µL of Annexin V and biotin selection cocktail were added, followed by 20 µL of the magnetic beads and 60 µL of isolation media, as described in protocol section 6 (Figure 4). After this procedure, the sample was resuspended in 500 µL of media and recounted. The live cell count was 4.9 × 105/mL with viability above 90% (Figure 3B), and the sample proceeded immediately to GEM generation according to the manufacturer's instructions.
Figure 1: Tumor sample processing overview. (A) Illustration of the step-by-step process of the tissue dissociation protocol. (B) The digestion of a sample using a vertically positioned thermal mixer. Please click here to view a larger version of this figure.
Figure 2: Pre-ACK and Post-ACK treatment of a mildly pigmented Melanoma sample. (A) Image shows a melanoma sample that was filtered and pelleted after the digestion step before red blood cell (RBC) lysis using ACK. (B) Image shows the same melanoma sample after RBC removal using ACK buffer. Please click here to view a larger version of this figure.
Figure 3: View of cells on hemocytometer under trypan blue zoomed in at 200x magnification. (A) Image showing the appearance of blue cells before live cell enrichment, representing dead cells, with viability lower than 80%. (B) Image showing the sample after live cell enrichment using the Kit 1 protocol (protocol step 6), showing increased viability above 90%. Please click here to view a larger version of this figure.
Figure 4: Overview of the live-dead sample cell isolation using Kit 1. This step is performed when under one million or over one million cells with viability >10% are detected after cell count (protocol step 5). Please click here to view a larger version of this figure.
Figure 5: Overview of the live-dead cell isolation using the dead cell removal microbeads in Kit 2. This step is performed when there are over one million cells and a viability of <10%. Please click here to view a larger version of this figure.
Figure 6: Validation of single-cell analysis of an unmanipulated sample after one cycle of live cell enrichment using Kit 1's modified protocol. Upper part shows a UMAP analysis demonstrating consistency in cell groups and types prior to and after live cell enrichment. Lower part summarizes the sequencing outputs generated by Cell Ranger. Please click here to view a larger version of this figure.
This protocol describes human or murine tumor samples dissociation into a single-cell suspension for scRNA sequencing using the 10x Genomics microfluidic system. In processing tissues that often come from rare cancers or are part of ongoing clinical trials or long experiments for murine tumors, the sample must be optimally and carefully preserved between all steps. All steps should be carried out rapidly on ice or at 4 ˚C to keep the native cellular RNA profiles and prevent degradation, as working at room temperature for extended periods will affect the transcriptomic quality17,18.
Several steps should be taken to ensure the loss of cells throughout the procedure is minimal, especially when processing small needle core biopsies17. Ensuring that all fragments of the tumor sample are included in the digestion will lead to the most successful result, as more cells extracted from the sample will improve future 10x Genomics encapsulation and sequencing and the total number of cells and transcripts recovered for analysis. If the sample becomes lodged in the tube it arrives in, media can be added, the tube inverted, and transferred to the dish for processing. If the initial sample is fragmented or the solution is cloudy, indicating the presence of cells in suspension, ensure that all the sample is collected for enzymatic digestion by centrifuging the sample with the solution it arrived in and resuspending it in media volume for digestion (protocol section 1).
When aspirating, ensure that the pellet is not disturbed throughout the procedure. Leave extra volume when aspirating, which can be removed with a P200 pipette. If the pellet is disturbed, the volume can be dispensed back into the tube and re-centrifuged. Post ACK Lysis or post live dead removal, the sample may be in volumes as low as 25 µL. Thus, it is critical not to lose any supernatant containing the cells. Proper mechanical digestion is essential to ensure the cells of interest are exposed to enzymatic digestion. Longer enzymatic digestion using the appropriate enzymes and thermal mixer might be needed for larger sample sizes9,19,20. While specific enzyme names are proprietary, the manufacturer notes enzyme A as a nuclease and enzymes H, D, and R as nonspecific proteases. Passing as much of the tumor sample through the filter is vital to ensuring the highest cell yield possible. Successful filtering will result in the filter not having any or only a tiny visible sample caught in it. Some pieces of the sample might not be able to be filtered through if it has high amounts of adipose or connective tissue. If there are visible pieces of the sample on the top of the filter, continue using the back of the syringe plunger to grind and push all the sample through using the twisting motion while also washing the top of the filter with more media.
The success of the red blood cell lysis step is determined by viewing the sample underneath the microscope once the sample is loaded on the hemocytometer for counting and viability15,21. If enough ACK is added to the sample to clear out red blood cells, there would be no red blood cells or visible redness in the pellet. If red blood cells are visible in the sample, additional ACK should be added, and the sample should be re-centrifuged at 450 × g, 4 ˚C for 5 min. Once done, the supernatant should be aspirated, and the pellet of cells can be resuspended in RPMI media for visualization and counting.
To ensure an optimal 10x Genomics run, the cell count should be between 700 and 1,200 cells/µL (7 × 105– 1.2 × 106/mL). A high concentration will increase the doublet rate (two cells in one droplet) during droplet formation. If the count is too high, an appropriate amount of media can be added to the sample and reloaded on a hemocytometer for recounting. If the concentration is too low, the sample can be re-centrifuged at 450 × g, 4 ˚C for 5 min, the supernatant aspirated and resuspended in a lower volume of media before recounting. The sample is ready to be loaded for 10x Genomics if the live cell viability is above 80% (Figure 3).
If the sample's viability is below 80%, a live-dead procedure is needed before 10x Genomics loading16. If one round of live-dead removal has not brought the viability percentage above 80%, another round should be performed if there are over 1 × 105 cells total. Higher viability will prevent clogs in the subsequent 10x Genomics protocol, increase cell recovery, and reduce the presence of ambient RNA. Kit 1's protocol has been optimized and miniaturized for rapidly dissociating smaller tissue samples with lower cell counts. The reduced volume of reagents and shortened incubation times increase cell yield and prevent a stress response gene signature (FOS, JUN, and JUNB) on various cells9,22. Alternative methods to enrich live cells include staining and sorting. However, this would result in reduced cell yield, prolonged staining incubation, which could affect cellular states (e.g., aCD3e antibodies will activate T cells), and stress induced by the machine's pressure on the cells during sorting9,22. The example shown in Figure 6 illustrates a UMAP generated after 10X Genomics Gene Expression (GEX) sequencing and the conversion of the generated cell barcodes into FASTQ files that were then concatenated and aligned via CellRanger. Subsequent sequencing results and live-dead removal comparisons were plotted, demonstrating cell populations in a sample before and after live-dead removal and improved sequencing quality. The lung sample shown in Figure 6 indicated an issue with the fraction of reads, as it was below 70%. This warning was generated due to high ambient RNA in a population of lysed or dead cells. However, after the removal of dead cells, this error was no longer present and resulted in a higher percentage of fraction reads in cells23.
Single-cell RNA sequencing is a costly and precarious process that has dramatically advanced the field of cancer immunology and cancer research. It allows for further interrogation of genes, cellular states, and programs expressed and repressed in specific cell types, divulging more about the behavior and heterogeneity of cancers. Suboptimal or low-viability tissue is more likely to generate low-quality data that can create artifacts that take away from the reality of tumor biology, emphasizing the critical importance of proper tumor dissociation and sample preparation.
The authors have nothing to disclose.
This study was supported by the Adelson Medical Research Foundation (AMRF), the Melanoma Research Alliance (MRA), and U54CA224068. Figure 1, Figure 4, and Figure 5 were created with BioRender.com.
1x PBS | Corning | 21-040-CV | |
10 mL serological pipette | Corning | 357551 | |
1000 μL low-retention pipette tips | Rainin | 30389213 | |
1000 μL low-retention wide pipette tips | Rainin | 30389218 | |
1000 μL pipette tips | Rainin | 30389212 | |
10x Magnetic Separator | 10x Genomics | 120250 | |
15 mL conical tubes | Corning | 430052 | |
2 mL aspirating pipette | Corning | 357558 | |
20 μL low-retention pipette tips | Rainin | 30389226 | |
20 μL pipette tips | Rainin | 30389225 | |
200 μL low-retention pipette tips | Rainin | 30389240 | |
200 μL pipette tips | Rainin | 30389239 | |
50 mL Polypropylene Conical Tube | Falcon | 352098 | |
60 x 15 mm Tissue Culture Dish | Falcon | 353004 | |
ACK Lysing Buffer | Gibco | A10492-1 | |
BD 1 mL syringe | Becton, Dickinson and Company | 3090659 | |
Bright-Line Hemocytometer | Hausser Scientific | 551660 | |
Calcium Chloride Solution | Sigma Aldrich | 2115-100ML | |
Cell Ranger | 10x Genomics | N/A | https://www.10xgenomics.com/support/software/cell-ranger/latest |
Cell counting slides for TC20 cell counter | Bio-Rad | 1450015 | |
CellTrics 30μm sterile disposable filters | Sysmex | 04-004-2326 | |
CellTrics 50μm sterile disposable filters | Sysmex | 04-004-2327 | |
Dead Cell removal Kit | Miltenyi Biotec | 130-090-101 | Store at -20 °C; kit 2 |
Dissecting Forceps, Fine Tip | VWR | 82027-386 | |
DNA LoBind Microcentrifuge tubes, 1.5 mL | Eppendorf | 22431021 | |
EasySepTM Dead Cell Removal (Annexin V) Kit |
STEMCELL Technologies | 17899 | Store at 4 °C; Kit 1 |
Eppendorf centrifuge 5425R | Eppendorf | 5406000640 | |
Eppendorf centrifuge 5910R | Eppendorf | 2231000657 | |
Eppendorf Easypet 3 Electronic Pipette controller | Eppendorf | EP4430000018 | |
Eppendorf Thermomixer F1.5 Model 5384 | Eppendorf | EP5384000012 | |
FBS | Gibco | 26140-079 | Store at 4 °C, use in a Biological hood |
German Stainless Scissors | Fine Science Tools | 14568-12 | |
Leica Dmi1 Microscope | Leica Microsystems | 454793 | |
LS Column | Miltenyi Biotec | 130-042-401 | |
MACS Multistand | Miltenyi Biotec | 130-042-303 | |
Pipet-Lite LTS Pipette L-1000XLS+ | Rainin | 17014382 | |
Pipet-Lite LTS Pipette L-200XLS+ | Rainin | 17014391 | |
Pipet-Lite LTS Pipette L-20XLS+ | Rainin | 17014392 | |
Quadro MACS Magnet | Miltenyi Biotec | 130-091-051 | |
RPMI Medium 1640 (1x) | Gibco | 21870-076 | Store at 4 °C |
TempAssure 0.2 mL PCR 8-Tube strips | USA Scientific | 1402-4700 | |
Trypan blue stain 0.4% | Thermo Fisher Scientific | T10282 | |
Tumor Dissociation Kit, Human | Miltenyi Biotec | 130-095-929 | Store at -20 °C, prepare enzymes according to kit instructions: Reconstitute lyophilized Enzyme H vial in 3 mL of RPMI 1640 Reconstitute lyophilized Enzyme R vial in 2.7 mL of RPMI 1640 Reconstitute lyophilized Enzyme A vial in 1 mL of Buffer A supplied with the kit. |
Tumor Dissociation Kit, Mouse | Miltenyi Biotec | 130-096-730 | Store at -20 °C, prepare enzymes according to kit instructions: Reconstitute lyophilized Enzyme D vial in 3 mL of RPMI 1640 Reconstitute lyophilized Enzyme R vial in 2.7 mL of RPMI 1640 Reconstitute lyophilized Enzyme A vial in 1 mL of Buffer A supplied with the kit. |
UltraPure Distilled Water | Invitrogen | 10977-015 | Store at 4 °C |