The current protocol presents a fast, efficient, and gentle method for isolating single cells suitable for single-cell RNA-seq analysis from a continuously growing mouse incisor, mouse molar, and human teeth.
Mouse and human teeth represent challenging organs for quick and efficient cell isolation for single-cell transcriptomic or other applications. The dental pulp tissue, rich in the extracellular matrix, requires a long and tedious dissociation process that is typically beyond the reasonable time for single-cell transcriptomics. For avoiding artificial changes in gene expression, the time elapsed from euthanizing an animal until the analysis of single cells needs to be minimized. This work presents a fast protocol enabling to obtain single-cell suspension from mouse and human teeth in an excellent quality suitable for scRNA-seq (single-cell RNA-sequencing). This protocol is based on accelerated tissue isolation steps, enzymatic digestion, and subsequent preparation of final single-cell suspension. This enables fast and gentle processing of tissues and allows using more animal or human samples for obtaining cell suspensions with high viability and minimal transcriptional changes. It is anticipated that this protocol might guide researchers interested in performing the scRNA-seq not only on the mouse or human teeth but also on other extracellular matrix-rich tissues, including cartilage, dense connective tissue, and dermis.
Single-cell RNA sequencing is a powerful tool for deciphering in vivo cell population structure, hierarchy, interactions, and homeostasis1,2. However, its results strongly depend on the first step of this advanced analysis – the preparation of a single-cell suspension of perfect quality out of the complex, well-organized tissue. This encompasses keeping cells alive and preventing unwanted, artificial changes in gene expression profiles of the cells3,4. Such changes might lead to the inaccurate characterization of population structure and misinterpretation of the collected data.
Specific protocols for the isolation out of a wide range of tissues have been developed5,6,7,8. They usually employ mechanical dissociation in combination with further incubation with various proteolytic enzymes. These typically include trypsin, collagenases, dispases, papain6,7,8,9, or commercially available enzyme mixtures such as Accutase, Tryple, etc.5. The most critical part affecting the transcriptome quality is enzymatic digestion. It was shown that prolonged incubation with enzymes at 37 °C influences the gene expression and causes the upregulation of many stress-related genes10,11,12,13. The other critical parameter of the isolation process is its overall length, as it has been shown that cell transcriptomes change after the tissue ischemia14. This protocol presents an efficient protocol for gentle isolation of single cells from mouse and human teeth, faster than other, previously utilized protocols for isolation of cells from complex tissues5,6,9,11,13,15,16.
This protocol presents how to quickly dissect soft tissue from the hard tooth and prepare a single-cell suspension suitable for scRNA-seq. This method employs only one centrifugation step and minimizes the effect of unwanted transcriptional changes by reducing the tissue handling and digestion time and keeping the tissue and cells at 4 °C most of the time. The procedure showcases the isolation of cells from mouse incisors, molar, and human wisdom teeth as an example, but principally should work for other teeth in various organisms. The complete protocol is schematically visualized in Figure 1. This protocol has been recently used to generate a dental cell type atlas obtained from mouse and human teeth1.
All animal experiments were performed according to the International and local regulations and approved by the Ministry of Education, youth and sports, Czech Republic (MSMT-8360/2019-2; MSMT-9231/2020-2; MSMT-272/2020-3). This protocol was tested with both male and female wildtype C57BL/6 and CD-1 mice and with genetically modified Sox10::iCreERT2 mice17 (combined with various reporter systems) on a C57BL/6 background. Experiments with human samples were performed with the approval of the Committees for Ethics of the Medical Faculty, Masaryk University Brno & St. Anne´s Faculty Hospital in Brno, Czech republic.
1. Experimental set-up and preparation of solutions
2. Preparation of experimental animal/s and human tooth
3. Tissue dissection
4. Preparation of single-cell suspension
5. Fluorescence-activated cell sorting (FACS)
Exemplary isolation of single cells was performed from two mandibular incisors from one 6-week-old C57BL/6 mouse male. Following this protocol, a single-cell suspension was prepared, and subsequently, single-cell sequencing was performed. The prepared single-cell suspension was analyzed and sorted using FACS (Figure 2). Firstly, the FSC-A (forward scatter, area) and SSC-A (side scatter, area) plotting was applied, and an appropriate gating strategy was used to select a population with expected size and granularity to filter our cell debris and cell doublets or aggregates (Figure 2A). This selected population (P1), counting 38% of all events, was further used, and FSC-A and FSC-H (forward scatter, height) parameters were applied to remove the remaining cell doublets (Figure 2B). The population without cell doublets (P2) counting 95% of P1 can be subsequently used for scRNA-seq. Alternatively, additional gating can be used to select the population of interest (e.g., expression of fluorescent proteins or live/dead staining). To check the number of live/dead cells in the final suspension, the PI (propidium iodide) staining was performed (Figure 2C). The P3 population containing PI– (living) cells was 98.4% out of the parent P2 population and 35.5% out of the total events. The total number of filtered out, dead (PI+) cells was 1887.
The final number of cells obtained without viability staining suitable for RNA-seq (P2) counted 118,199 cells from two mouse incisor pulps. This means the number of almost 60,000 living cells from one mandibular incisor.
To clarify the number of immune cells in the final single-cell suspension, two approaches were used. Firstly, the CD45 antibody staining and subsequent FACS analysis were used. As a complementary method, the total number of immune cells (CD45+) in scRNA-seq data was analysed. FACS analysis showed 14.44% of CD45+ cells (13.20% alive and 1.24% dead) (Figure 3A). Analysis of scRNA-seq data showed 10.90% of CD45-expressing cells (Figure 3B). The decrease of CD45+ cells in scRNA-seq data can be caused by additional thresholding during scRNA-seq analysis.
These representative data on the example of mouse incisor show that the given protocol in combination with strict gating strategy is efficient in obtaining a high number of cells out of a single mouse tooth without the necessity of additional use of viability staining. The ratio of immune (CD45+) cells was minor (13.2%). Moreover, it was previously shown that the immune cells are essential in maintaining tooth homeostasis, so removing them from scRNA-seq analysis during the FACS would be counterproductive in some applications.
Figure 1: Schematic representation of the protocol. Different steps, including temperature conditions and expected time, are represented to prepare single-cell suspension from mouse and human teeth. Please click here to view a larger version of this figure.
Figure 2: Example of the gating strategy. FSC-A and SSC-A gating was used to produce the P1 gate, reflecting the cell population with expected size and granularity and filtering out the cell and extracellular matrix debris and most large events (A). Subsequently, the P1 population was plotted in FSC-H and FCS-A plot, which filtered out cell doublets (B). This P2 population was then analyzed for the presence of dead cells by propidium iodide (C). The number of events/cells per gate are represented in (D). (FSC-A – forward scatter, area; SSC-A – side scatter, area; FSC-H – forward scatter, height; PI – propidium iodide). Please click here to view a larger version of this figure.
Figure 3: Quantification of the immune cells. Quantification of the immune cells was performed by FACS analysis of cells stained with anti-CD45 antibody and Live/Dead analysis using propidium iodide staining (A). Further quantification of immune cells was performed during scRNA-seq analysis (B). (CD45-APC – anti-CD45 allophycocyanin conjugated antibody; PI – propidium iodide; t-SNE – t-distributed stochastic neighbor embedding). Please click here to view a larger version of this figure.
Supplementary Figure 1: Overview of the mandible dissection process. Dashed lines illustrate suggested cuts. TMJ – temporomandibular joint, m. masseter – musculus masseter. Please click here to download this File.
Supplementary Table 1: The compositions of the solutions used in the study. Please click here to download this Table.
Studying teeth and bones on the cellular or molecular level is generally challenging since cells forming these tissues are surrounded by different kinds of hard matrices19. One of the main goals for performing single-cell RNA-seq on dental tissue is the need to obtain cells of interest fast and without any artificial changes in their transcriptomes. To accomplish this, a highly efficient protocol suitable for isolating cells from mouse and human tooth pulps was developed, which allows for quick generation of single-cell suspensions for all transcriptomic applications. This was ensured by fast tissue isolation, minimizing the steps of tissue and cell manipulations, and streamlining the mechanical and enzymatic digestion.
The most critical steps of this protocol are fast tissue processing and adequate single-cell suspension preparation8,9. A manual approach is used to obtain dental pulps without utilizing a dental drill or other heat-generating devices. Overheating may cause an artificial expression of heat shock proteins and other genes, ultimately leading to the analyzed gene expression patterns being unrepresentative of the original tissue20. Manual tissue harvesting may be a challenging step that will likely need some training beforehand. The pulp is then cut into small pieces and enzymatically digested at 37 °C. Except for the 15-20 min of enzymatic digestion, the whole protocol is performed at 4 °C. The tissue processing and especially the enzymatic digestion were minimized to the shortest possible time since more prolonged incubation at 37 °C can cause changes in gene expression patterns10. Mechanical removing of the dentin is recommended before enzymatic digestion. Dentin and the pulp-attached predentin contain a large amount of collagen, and its excessive presence might decrease the effectiveness of the digesting solution. After being removed from the body (or death of organisms), it was shown that cells start to modify their gene expression patterns quickly12. Therefore, cell isolation and processing should be carried out as fast as possible. The current protocol reduces the processing time to 35-45 min from isolating the tissue (euthanizing animal) to preparing single-cell suspension.
One alternative modification of this technique is cell preservation for later use. This is achieved by methanol fixation. Methanol-fixed cell suspension can be stored for up to 1 month at -80 °C, as described in the protocol21. However, whenever possible, perform scRNA-seq directly, since it was shown that the single-cell data from methanol-fixed single-cell suspensions might suffer from increased expression of stress-related genes and contamination with ambient RNA22. This step might need additional modification according to the manufacturer's protocols.
Before the first application of this protocol, performing several validation steps are recommended to test the technique. From our experience, we suggest testing the aforementioned critical steps of the protocol. Additionally, we suggest testing the effectiveness of the collagenase P solution and testing the handling of the tissue dissociation step. Specifically, around the first 5 min after the initiation of collagenase P incubation, the pieces of tissue should aggregate together. This is a common situation. Aggregates are disintegrated every 3-4 min using a 1 mL pipette, and with increasing time, they should become smaller until barely visible.
Furthermore, it is recommended to perform cell counting in a cell counting chamber before centrifugation and before and after filtering to detect possible cell losses due to suboptimal supernatant removal. If the final single-cell suspension needs to be purified, FACS can be used. Cell sorting enables not only to remove debris or dead cells but importantly enables to enrich final suspension with fluorescently labeled cells13,19. To avoid shear stress or clogging of the cell sorter, a wide nozzle (85 µm or 100 µm) is used. This will further improve the viability of the sorted cells.
This technique was designed and tested on both mouse and human teeth. The major limiting factor is the small number of cells in the reduced dental pulps of the teeth of older mice (molars) and humans. Suppose a larger number of cells need to be obtained or cells from the teeth of older patients are to be acquired. One possible solution is to process a higher number of teeth and merge them into a single batch, subsequently processed as one sample.
Living cells of human dental pulp were firstly isolated more than twenty years ago using an enzyme mixture of collagenase I and dispase23. Since then, isolations of dental pulp cells became widely utilized, and several techniques have been used5,6,7,8. The critical significance of the method presented here is the adaption of all isolation steps to make the isolation fast and gentle to ensure the high quality of the final cell suspension for scRNA-seq. Higher cell yield can be obtained by more prolonged incubation with enzymes. This protocol provides an efficient solution for quickly obtaining single cells from mouse and human teeth of suitable quality for single-cell RNA-sequencing. This technique is expected to be widely used for other tissues or organisms with just slight technical modifications.
The authors have nothing to disclose.
J.K. was supported by the Grant Agency of Masaryk University (MUNI/H/1615/2018) and by funds from the Faculty of Medicine MU to junior researcher. J.L. was supported by the Grant Agency of Masaryk University, (MUNI/IGA/1532/2020) and is a Brno Ph.D. Talent Scholarship Holder – Funded by the Brno City Municipality. T.B. was supported by the Austrian Science Fund (Lise Meitner grant: M2688-B28). We thank to Lydie Izakovicova Holla and Veronika Kovar Matejova for their help with the obtaining of human teeth. Finally, we thank Radek Fedr and Karel Soucek for their kind assistance with FACS sorting.
APC anti-mouse CD45 Antibody | BioLegend | 103112 | |
Bovine Serum Albumin Fraction V | Roche | 10735078001 | |
CellTrics 50 µm, sterile | Sysmex | 04-004-2327 | |
Collagenase P (COLLP-PRO) | Roche | 11213857001 | |
CUTFIX Scalpel Blades, Fig. 10 | AESCULAP | 16600495 | |
CUTFIX Scalpel Blades, Fig. 11 | AESCULAP | 16600509 | |
Fetal Bovine Serum (South America), Ultra low Endotoxin | Biosera | FB-1101/500 | |
Hank's balanced salt solution (HBSS) without Ca2+ and Mg2+ | SIGMA | H6648 | |
Industrial low lint wipes, MAX60 | Dirteeze | MAX60B176 | |
Industrial strong tweezers, style 660, bent serrated pointed tips, 150mm | Value-Tec | 50-014366 | |
Methyl alcohol A.G. | Penta | 21210-11000 | |
Propidium Iodide Solution | BioLegend | 421301 | |
Scalpel handle | CM Instrumente | AG-013-10 | |
Serological pipettes 10mL, individually wrapped, 200 pcs. | CAPP | SP-10-C | |
Stereo microscope | Leica | EZ4 | |
Surgical scissors (9cm) | CM Instrumente | AI-430-09Y | |
Tissue culture dish Ø 100 mm | TPP | 93100 |