A convenient, fast, and cost-effective method to measure the proportion of side population cells in solid tumor cell lines is presented.
Cancer stem cells (CSCs) are an important cause of tumor growth, metastasis, and recurrence. Isolation and identification of CSCs are of great significance for tumor research. Currently, several techniques are used for the identification and purification of CSCs from tumor tissues and tumor cell lines. Separation and analysis of side population (SP) cells are two of the commonly used methods. The methods rely on the ability of CSCs to rapidly expel fluorescent dyes, such as Hoechst 33342. The efflux of the dye is associated with the ATP-binding cassette (ABC) transporters and can be inhibited by ABC transporter inhibitors. Methods for staining cultured tumor cells with Hoechst 33342 and analyzing the proportion of their SP cells by flow cytometry are described. This assay is convenient, fast, and cost-effective. Data generated in this assay can contribute to a better understanding of the effect of genes or other extracellular and intracellular signals on the stemness properties of tumor cells.
Cancer stem cells (CSCs) are subsets of cells with self-renewal ability and multiple differentiation potential, which play a vital role in tumor growth, metastasis, and recurrence1,2. Currently, CSCs have been identified to exist in a variety of malignant tumors, including lung, brain, pancreas, prostate, breast, and liver cancers3,4,5,6,7,8,9. Identification of CSCs in these tumors is mainly based on the presence of surface marker proteins, such as high and/or low expression of CD44, CD24, CD133, and Sca-19,10, but a unique marker that can distinguish CSCs from non-CSCs has not been reported so far. Currently, several techniques are used to identify and purify CSCs in tumor tissue or tumor cell lines. These techniques are designed based on the specific properties of CSCs. Among them, assays and sorting of side population (SP) cells are two of the commonly used methods.
SP cells were originally discovered by Goodell et al.11, when they characterized hematopoietic stem cells in mouse bone marrow cells. When the mouse bone marrow cells were labeled with the fluorescent dye Hoechst 33342, a small group of Hoechst 33342 dimly-stained cells appeared in the two-dimensional dot plot of a flow cytometry assay. Hoechst 33342 is a DNA-binding dye and has at least two binding modes that lead to different spectral characteristics. When viewing fluorescence emission at two wavelengths at the same time, multiple populations can be revealed12. In their assay, the Hoechst 33342 was excited at 350 nm and the fluorescence was measured by using the 450/20 nm band-pass (BP) filter and 675 nm edge filter long-pass (EFLP)11. Compared with whole population of bone marrow cells, this group of cells was enriched with hematopoietic stem cells called SP cells11. SP cells are capable of rapidly expelling Hoechst 33342. The efflux of this dye is related to ATP-binding cassette (ABC) transporters13, which can be inhibited by some agents such as Fumitremorgin C14, Verapamil and Reserpine15,16. After that, different proportions of SP cells were detected in a variety of tissues, organs, tumor tissues, and tumor cell lines17,18,19. These SP cells have many characteristics of stem cells17,19.
This manuscript describes Hoechst 33342 labeling and staining of cultured tumor cells and the analysis of SP cells by flow cytometry. Moreover, optimization of the Hoechst 33342 concentration and the proper blocker selection for a specific tumor cell line using this approach are shown. Finally, the effects of stemness promotion or inhibition signals on the proportion of SP in tumor cells are demonstrated. The experimental examples demonstrate that analysis of SP can be used to explore the effects of various signals, such as gene expression, small inhibitors, activators, cytokines, and chemokines, on tumor stemness. Compared to other methods for isolation and purification of CSCs, such as sorting of CD44+/CD24– population, aldehyde dehydrogenase (ALDH) analysis, and tumor sphere formation assays, this method is easier for manipulation and is cost-effective.
1. Cell preparation
2. Cell staining with Hoechst 33342
3. Analysis by flow cytometry
NOTE: Instructions for use of the flow cytometer software (see Table of Materials) are described in this section and Supplementary Figures 1–10.
4. Data analysis
NOTE: Instructions for the use of the flow cytometry analysis software (see Table of Materials) are described in this section and Supplementary Figures 11–16.
Four experimental SP analyses were performed according to this method. In the first one, we detected the proportion of SP cells in MDA-MB-231, which is a triple negative human breast cancer cell line, under normal conditions. After cell counting, Hoechst 33342 was added into one tube containing 1 x 106 cells to a final concentration of 3 µg/mL. Reserpine and Hoechst 33342 were added to another tube to final concentrations of 40 µM and 3 µg/mL, respectively. PI was added to both tubes. The dot plot of FSC-A (X-axis) versus PI-A (Y-axis) showed three populations: 1) a PI-positive cell population, which represented the dead cells; 2) cell debris; and 3) the main population was PI-negative cells, which were subjected to further analysis (Figure 1A,B). A single-cell population gated from the dot plot of FSC-A (X-axis) versus FSC-W (Y-axis) and the dot plot of SSC-A (X-axis) versus SSC-W (Y-axis) was used to analyze the proportion of SP cells (Figure 1A,B). The SP cells were gated from the dot plot of Hoechst Red-A (X-axis) versus Hoechst Blue-A (Y-axis), and their percentage was about 0.9% in MDA-MB-231 cells (Figure 1A). However, Reserpine greatly decreased the proportion of SP cells (Figure 1B), supporting that the gate-painting for SP is correct.
The second experiment was to determine the suitable staining concentration of Hoechst 33342 in MDA-MB-435 cells. After cell counting, Hoechst 33342 was added to 1 x 106 cells, which were suspended in 1 mL of PBS supplemented with 2% FBS, to different concentration gradients, including 0.5, 1, 1.5, 2, 2.5, 3, 3.5, and 4 µg/mL. As shown in Figure 2A, when the concentration of Hoechst 33342 was too low (i.e., 0.5, 1, 1.5 µg/mL), it was hard to distinguish SP cells from other cell populations, because lots of cells were in a dimly-stained state. When the concentration of Hoechst 33342 was too high (i.e., 2.5, 3, 3.5, 4 µg/mL), the proportion of SP cells decreased until it disappeared. Thus, 2 µg/mL Hoechst 33342 was the best concentration for SP analysis in MDA-MB-435 cells. In addition, to determine the proper blocker for this cell line, Hoechst 33342 and a blocker (Verapamil or Reserpine) were added to 1 x 106 cells, which were suspended in 1 mL of PBS supplemented with 2% FBS, to final concentrations of 2 µg/mL and 40 µM, respectively. As shown in Figure 2B, about 0.4% of cells expelled the dye after Verapamil treatment. However, after Reserpine treatment, the ratio dropped to about 0.1% (Figure 2C). From this experiment, Reserpine was considered a more appropriate blocker for this cell line.
In the third example, A549 cells (human lung adenocarcinoma cells) were pretreated with STAT3 activator-Colivelin20 (100 nM) for 48 hours (h). The STAT3 signaling pathway is important for promoting the stemness features of tumor cells21. As shown in Figure 3, the proportion of SP cells increased upon Colivelin stimulation.
In the last example, T47D cells (human breast cancer cells) were pretreated with 0.1 µM FRA1 inhibitor-SKLB816 (also named 13an) for 48 h22. FRA1 is a reported gatekeeper of epithelial-mesenchymal transition (EMT) and involved in regulation of tumor stemness23. As shown in Figure 4, the proportion of SP cells decreased due to the treatment of FRA1 inhibitor.
Figure 1: Gating strategy of MDA-MB-231 cells in SP analysis. (A) Gating strategy of MDA-MB-231 cells, which were stained with Hoechst 33342 (3 µg/mL) and propidium iodide (PI, 1 μg/mL). (B) Gating strategy of MDA-MB-231 cells, which were treated with Reserpine (40 µM) and stained with Hoechst 33342 (3 µg/mL) and PI (1 μg/mL). Please click here to view a larger version of this figure.
Figure 2: Optimization of Hoechst 33342 concentration and selection of blocker in MDA-MB-435 cells. (A) The SP analysis results of MDA-MB-435 cells, which were stained with Hoechst 33342 (Hoechst) at different concentration gradients (0.5, 1, 1.5, 2, 2.5, 3, 3.5, 4 µg/mL) together with PI (1 μg/mL). (B) The SP analysis result of MDA-MB-435 cells, which were treated with Verapamil (40 µM) and stained with Hoechst 33342 (2 µg/mL) and PI (1 μg/mL). (C) The SP analysis result of MDA-MB-435 cells, which were treated with Reserpine (40 µM) and stained with Hoechst 33342 (2 µg/mL) and PI (1 μg/mL). Please click here to view a larger version of this figure.
Figure 3: The proportion of SP cells was enhanced by STAT3 activator in A549 cells. (A) The SP analysis results of A549 cells pretreated with STAT3 activator-Colivelin (100 nM) or its vehicle control (Ctrl)-H2O for 48 h. A549 cells treated with Reserpine (45 µM) were used as the blocking control. A549 cells were stained with Hoechst 33342 (7 μg/mL) and PI (1 μg/mL). (B) The statistical results of the proportion of SP cells in A549 cells treated with Colivelin (100 nM) and its vehicle control. Data are presented as mean + standard error of the mean (SEM), n = 3 for each group. “*” indicates P < 0.05. Please click here to view a larger version of this figure.
Figure 4: The proportion of SP cells was inhibited by FRA1 inhibitor in T47D cells. (A) The SP analysis results of T47D cells pretreated with SKLB816 (0.1 µM) or its vehicle control (Ctrl)-DMSO for 48 h. T47D cells treated with Reserpine (40 µM) were used as the blocking control. T47D cells were stained with Hoechst 33342 (8 μg/mL) and PI (1 μg/mL). (B) The statistical results of the proportion of SP cells in T47D cells treated with SKLB816 (0.1 µM) and its vehicle control. Data are presented as mean + SEM, n = 3 for each group. “*” indicates P < 0.05. Please click here to view a larger version of this figure.
Supplementary Figure 1: Instructions for flow cytometer software step number 3.1. (A) Click the FOLDER button. (B) Click the Experiment button and then click the New Experiment button. Please click here to download this figure.
Supplementary Figure 2: Instructions for flow cytometer software step number 3.2. (A) Click the OK button. (B) Experiment_001 shows up under FOLDER. Please click here to download this figure.
Supplementary Figure 3: Instructions for flow cytometer software step number 3.3. (A) Click the Experiment_001 button to change the name of Experiment_001 to a specific name (e.g., "20191118-SP"). (B) Click the Enter button. Please click here to download this figure.
Supplementary Figure 4: Instructions for flow cytometer software step number 3.4. (A) Click the New Specimen button. (B) Click the New Tube button. (C) Click the Arrowhead button. Please click here to download this figure.
Supplementary Figure 5: Instructions for flow cytometer software step number 3.5. (A) Click the Parameters button to set up the parameters. Please click here to download this figure.
Supplementary Figure 6: Instructions for flow cytometer software step number 3.6.1.2. (A) Click the Dot Plot button to display the dot plot, click the X-axis and set it to “FSC-A”; click the Y-axis and set it to “PI-A”. (B) Click the Polygon Gate button to create a polygon gate to gate the P1 subset (also known as FSC-A, PI-A subset). Please click here to download this figure.
Supplementary Figure 7: Instructions for flow cytometer software step number 3.6.1.3. (A) Click the Dot Plot button to display the dot plot, click the X-axis and set it to “FSC-A”; click the Y-axis and set it to “FSC-W”. (B) Right-click the dot plot and click the P1 button under Show Populations button. (C) Click the Rectangular Gate button to create a rectangular gate to gate the P2 subset (also known as FSC-A, FSC-W subset). Please click here to download this figure.
Supplementary Figure 8: Instructions for flow cytometer software step number 3.6.1.4. (A) Click the Dot Plot button to display the dot plot, click the X-axis and set it to “SSC-A”; click the Y-axis and set it to “SSC-W”. (B) Right-click the dot plot and click the P2 button under the Show Populations button. (C) Click the Rectangular Gate button to create a rectangular gate to gate the P3 subset (also known as SSC-A, SSC-W subset). Please click here to download this figure.
Supplementary Figure 9: Instructions for flow cytometer software step number 3.6.1.5. (A) Click the Dot Plot button to display the dot plot, click the X-axis and set it to “Hoechst Red-A”; click the Y-axis and set it to “Hoechst Blue-A”. (B) Right-click the dot plot and click the P3 button under the Show Populations button. (C) Click the Polygon Gate button to create a polygon gate to gate the P4 subset (also known as Hoechst Red-A, Hoechst Blue-A subset). (D) Right-click the dot plot, then click the Show Population Hierarchy button to show the population hierarchy. Please click here to download this figure.
Supplementary Figure 10: Instructions for flow cytometer software step number 3.6.1.6. (A) Click the Acquire Data button, then collect 20,000-100,000 events from each sample. Please click here to download this figure.
Supplementary Figure 11: Instructions for flow cytometry analysis software step number 4.1. (A) Open the flow cytometry analysis software and drag one sample file into the software. (B) The sample file is imported. Please click here to download this figure.
Supplementary Figure 12: Instructions for flow cytometry analysis software step number 4.2.1. (A) Double click this sample file. (B) Click the X-axis and set it to “FSC-A”; click the Y-axis and set it to “PI-A”. (C) Click the Create a polygon gate button to create a polygon gate. (D) Click the OK button to obtain the FSC-A, PI-A subset. (E) The FSC-A, PI-A subset is obtained. Please click here to download this figure.
Supplementary Figure 13: Instructions for flow cytometry analysis software step number 4.2.2. (A) Double click the FSC-A, PI-A subset file. (B) Click the X-axis and set it to “FSC-A”; click the Y-axis and set it to “FSC-W”. (C) Click the Create a rectangular gate button to create a rectangular gate. (D) Click the OK button to obtain the FSC-A, FSC-W subset. (E) The FSC-A, FSC-W subset is obtained. Please click here to download this figure.
Supplementary Figure 14: Instructions for flow cytometry analysis software step number 4.2.3. (A) Double click the FSC-A, FSC-W subset file. (B) Click the X-axis and set it to “SSC-A”; click the Y-axis and set it to “SSC-W”. (C) Click the Create a rectangular gate button to create a rectangular gate. (D) Click the OK button to obtain the SSC-A, SSC-W subset. (E) The SSC-A, SSC-W subset is obtained. Please click here to download this figure.
Supplementary Figure 15: Instructions for flow cytometry analysis software step number 4.2.4. (A) Double click the SSC-A, SSC-W subset file. (B) Click the X-axis and set it to “Hoechst Red-A”; click the Y-axis and set it to “Hoechst Blue-A”. (C) Click the Create a polygon gate button to create a polygon gate. (D) Click the OK button to obtain the Hoechst Red-A, Hoechst Blue-A subset. (E) The Hoechst Red-A, Hoechst Blue-A subset is obtained. Please click here to download this figure.
Supplementary Figure 16: Instructions for flow cytometry analysis software step number 4.2.5. (A) Click the Open Layout Editor button to open the layout editor. (B) Drag the SSC-A, SSC-W subset sample file to Layout Editor. (C) Click the Click to save layout window to file button to save the image results. Please click here to download this figure.
There are several key points to keep in mind for the SP assay. The first is the selection of a proper blocker, such as Verapamil or Reserpine, for each cell line, because the "gate" location of the SP cells is determined according to the position at which a large number of SP cells disappear after the addition of the blocker. For the MDA-MB-231 cell line, Reserpine works well. However, for other cell lines, different blockers might work better.
The second is the concentration of Hoechst 33342. The percentage of SP cells increased as the staining concentration of Hoechst 33342 decreased, as the representative data showed. This phenomenon can be explained by the dye uptake kinetics24. Changes in dye concentration affect enrichment of Hoechst 33342 in cells. Therefore, Hoechst 33342 staining concentration is closely related to the results of SP assay. Moreover, uptake and expulsion of Hoechst 33342 vary between cell types. Thus, proper concentrations of Hoechst 33342 need to be explored for different cell lines before the SP analysis.
The third is a good coefficient of variation (CV) of the flow cytometer, which is also critical for the SP analysis25. UV laser power is an important criteria for better CVs12. This protocol uses a commercial flow cytometer (see Table of Materials) to perform the SP assay. In this cuvette-flow-cell instrument, we used the UV laser with a power of 15 mW to get the best CVs. In general, a relatively high UV laser power provides the optimal CVs. For example, 50–100 mW provide the optimal Hoechst signal on jet-in-air instruments12. Some lasers provide lower UV power, which can reduce CVs. In these cases, good laser alignment is critical.
The last is the influence of other factors during the experiment. Cell status, temperature, time of staining, operation of flow cytometry, and other factors may also affect the proportion and quality of the SP assay. For example, changes in cell viability during the preparation of cell suspensions will affect the ratio of SP cells. Therefore, the best experimental conditions need to be explored before performing SP analysis. Considering the above factors, the SP ratio of the same cell line measured by different laboratories may be different. For example, the proportions of MDA-MB-231 cells and A549 cells were reported to be ~0.1%–4.8%26,27,28,29 and ~0.8%–18%30,31,32,33,34, respectively.
Researchers can modify this assay for different applications, such as the study of other tumor cell lines, or primary patient-derived tumor cells. If the protocol does not work well for the cells being tested, the researchers can use MDA-MB-231 cells as the positive control and stain the cells following the given specifications. Because this protocol is very sensitive to the concentration of Hoechst 33342, staining temperature and time etc., the researchers should check all these conditions closely. The percentage of SP in many types of human tumor cell lines is relatively low (~ 0%–37%)19,26,27,28,29,30,31,32,33,34,35,36. If an excessively high or low percentage of SP is observed, it may be due to inappropriate concentrations of Hoechst 33342 or blocker used. If the problem seems to be related to the flow cytometer, technical support should be obtained.
Although the use of SP to analyze and separate CSCs is highly efficient, it still has certain limitations. The first is its high sensitivity to staining conditions12. Many factors, such as the concentration of Hoechst 33342, cell status, temperature, time of staining, operation of flow cytometry, and the blocker selection, can affect the quality of SP analysis. The second is the cytotoxic effect of Hoechst 33342. Hoechst 33342 is a DNA-binding dye, but is toxic to cells when it reaches high concentrations and thus reduces cell activity37.
In summary, SP analysis is one of the most commonly used methods in recent years to identify and purify CSCs in tumor cell lines. Although the method has some limitations, in the absence of specific CSCs surface markers, it is still a method for convenient, rapid, and cost-effective enrichment of CSCs. This method is beneficial for studying the biological functions of CSCs and for the identification of specific surface markers. Moreover, by detecting the effects of various signals on the SP ratio of tumor cells, it can provide clues to the regulatory effect of these signal pathways on CSCs features, and facilitate the discovery of new mechanisms, which can ultimately guide the targeted therapy of tumors.
The authors have nothing to disclose.
This work was funded by the Natural Science Foundation of China 81572599, 81773124, and 81972787; Natural Science Foundation of Tianjin City (China) 19JCYBJC27300; Tianjin People’s Hospital & Nankai University Collaborative Research Grant 2016rmnk005; Fundamental Research Funds for the Central Universities, Nankai University 63191153.
6 well cell culture plate | CORNING | 3516 | 9.5 cm2 (approx.) |
Colivelin | MCE | HY-P1061A | Ser-Ala-Leu-Leu-Arg-Ser-Ile-Pro-Ala-Pro-Ala-Gly-Ala-Ser-Arg-Leu-Leu-Leu-Leu-Thr-Gly-Glu-Ile-Asp-Leu-Pro |
Fetal bovine serum (FBS) | Biological Industries (BIOIND) | 04-001-1ACS | |
Flow cytometer | BD Biosciences | BD LSRFortessa | |
Flow cytometer software | BD Biosciences | FACSDiva | |
Flow cytometry analysis software | BD Biosciences | FlowJo | |
Hoechst33342 | Sigma-Aldrich | B2261 | bisBenzimide H 33342 trihydrochloride |
Polystyrene round bottom test tube | CORNING | 352054 | 12 x 75 mm, 5mL |
Propidium iodide (PI) | Sigma-Aldrich | P4170 | 3,8-Diamino-5-[3-(diethylmethylammonio)propyl]-6-phenylphenanthridinium diiodide |
Reserpine | Sigma-Aldrich | 83580 | (3β, 16β, 17α, 18β, 20α)-11,17-Dimethoxy-18-[(3,4,5-trimethoxybenzoyl)oxy]yohimban-16-carboxylic acid methyl ester |
SKLB816 | Provided by Dr. Shengyong Yang, Sichuan University | ||
Trypsin-EDTA (0.25%), phenol red | Gibco | 25200072 | |
Verapamil hydrochloride | Sigma-Aldrich | V4629 | 5-[N-(3,4-Dimethoxyphenylethyl)methylamino]-2-(3,4-dimethoxyphenyl)-2-isopropylvaleronitrile hydrochloride |