The common assays for detecting DNA synthesis in proliferation cells are [3H] thymidine and BrdU incorporation (1). The traditional method of [3H]thymidine incorporated into DNA is detected by autoradiography, which is cumbersome and time-consuming. BrdU immunostaining method is a more convenient approach with simplified procedures and high efficiency. However, the major impediment of BrdU immunostaining is the complementary bases pairing in double-stranded DNA, which blocks the binding of the anti-BrdU antibody to BrdU. Hence, strong denaturing conditions such as digestion with DNase or concentrated hydrochloric acid have to be applied for the exposure of BrdU epitopes. So the intensity of BrdU staining highly depends on the conditions (2). Therefore, a simple and rapid nucleic acid labeling technique with considerable sensitivity and stability is still needed for the studies of cell kinetics, DNA synthesis, and cellular proliferation in vitro and in vivo. In particular, techniques which are environmentally friendly and do not destroy the ultrastructure of sample are highly desirable. The recently developed EdU incorporation (3) overcomes the disadvantages mentioned above, and can thus be used to label the replicated DNA in the context of well preserved cellular and chromatin ultrastructure. As a thymidine analog, EdU is used by replacing the methyl group at position 5 of deoxyuridine with an alkyne group. It is reported that EdU is readily incorporated into cellular DNA during the S phase of the cell cycle as shown in NIH 3T3 cells (3), neuron (4, 5), Xenopus egg extracts (3), human T-cell leukemia Jurkat cell line (6), human breast cancer cell lines SK-BR-3 and BT474 (7), murine mesenchymal stem cells (8), mouse epithelial cells (9) as well as the spleen cells and the lymph node cells of mouse (10, 11). The terminal alkyne group of EdU can react with fluorescent azides in a Cu(I)-catalyzed [3+2] cycloaddition (“click” chemistry). This reaction enables detection of EdU incorporation into cells by fluorescent microscopy or fluorescence activated cell sorter. Recently, EdU incorporation combined with fluorescent microscopy is widely used in vitro and in vivo. However, few groups have used flow cytometry analysis of EdU incorporation and cell surface antigens in evaluating the proliferations of mouse T lymphocyte subsets (10). In their study, multiple cell surface antibodies simultaneously labeling T cells was not mentioned and the factors affecting the entire assay were not fully investigated. Moreover, the manufacturer's suggested protocols (Cell-Light™ EdU Cell Proliferation Detection from RiBoBio, Guangzhou, China) often result in a significantly weakened fluorescence intensity of cell surface antibodies such as CD3e-PerCP-Cy5.5, CD4-FITC, CD8a-PE. Hence, we investigated the effect of reagents and applications related to polychromatic flow cytometry in this assay.
Finally, we propose a feasible protocol of flow cytometry assay to detect the proliferations of T cell subsets by EdU incorporation and the labeling of cell surface antigens.
MATERIALS AND METHODS
Spleen cells were isolated from C57BL/6J mice, filtered through 40 μm nylon cell strainers, and lymphocytes were separated by EZ-Sep™ mouse lymphocyte separation medium (Dakewe, Beijing, China). Red blood cells were lysed with ammonium chloride. The remaining cells were maintained in RPMI 1640 with 10% FCS (Gibco, Grand Island, NY). The study was approved by the Ethics Committee of the National Center for Clinical Laboratories.
Lymphocytes (1 × 106/ml) were stimulated with phytohemagglutinin (PHA) (10 μg/ml, Sigma, Shanghai, China) for 72 h in 24-well plates. Different final concentrations (0, 0.5, 1, 5, 10, 50, 100, 150, 200, 250 μM) of EdU (RiBoBio, Guangzhou, China) were added to the plates 12 h before harvesting the cells.
Additional experiments were performed with a 50 μM final concentration to incubate 30 min, 1 h, 2 h, 3 h, 4 h, 8 h, 12 h, 16 h, 20 h, and 24 h before harvesting the cells.
Immunofluorescence Staining of Cell Surface Antigens
Cells were collected and centrifuged at 300g for 5 minutes, and supernatant was removed. Red fluorescent reactive dye (Invitrogen, Carlsbad, CA) was added, incubated for 30 minutes at room temperature (RT), and washed with 1 ml PBS (Gibco, San Diego, CA) with 1% FCS. After that, cell surface antibodies (anti-mouse CD3e-PerCP-Cy5.5, CD4-FITC, CD8a-FITC, CD8a-PE or their isotype controls from BD Pharmingen, San Diego, CA) were added, incubated for 15 minutes at RT, and washed with 1 ml PBS with 1% FCS.
Fixation and Permeabilization
For fixation, 100 μl 4% paraformaldehyde (Solarbio, Beijing, China) was added into the cells, incubated for 15 minutes at RT, and washed once with 1 ml PBS with 1% FCS. Next, cells were resuspended in 100 μl of PBS with 0.05% saponin (Sigma, Shanghai, China) or Tris buffer saline (10 mM Tris–HCl, 100mM NaCl, pH 7.4) with 0.1% Triton X-100, incubated the cells for 15 minutes and washed again with 1 ml PBS with 1% FCS.
As a comparison study, 100 μl fixation/permeabilization solution (BD pharmingen, San Diego, CA) was used instead of 4% paraformaldehyde and permeabilization detergent reagents in another experiments.
In the Click reaction, the staining solution was prepared with 100 mM Tris (pH 8.5), 0.5 mM CuSO4, 10 μM fluorescent azide (Apollo® 643 azide from RiBoBio, Guangzhou, China), and 50 mM ascorbic acid. Amounts of 5, 25, 100, 200, 500 μl staining solution was added into cells (about 1 × 106 cells), incubated for 30 minutes at RT, and then rinsed once or twice with 1 ml PBS with 0.05% saponin.
The stimulated cells labeled with Red fluorescent reactive dye, anti-mouse CD3, CD4, CD8, Apollo® 643 azide were stored in 300 μl PBS with 1% FCS and 0.09% sodium azide at 4°C and were analyzed at 1, 3, 7, 14, 21 d, respectively. Other stimulated cells were stored in FCS with 10% DMSO at −80°C or in liquid nitrogen, and were labeled only before flow cytometry analysis at the same timepoint as mentioned above.
Flow cytometry was performed with a FACSCalibur Flow Cytometer, which is equipped with two lasers: 488 nm and 635 nm. FITC, PE, PerCP-Cy5.5, and red fluorescent reactive dye are excitated by the 488 nm laser and the emitted light is collected through 530/30, 585/42, 670 LP, and 585/42 band pass filters, respectively. Apollo® 643 azide is excitated by the 635 nm laser and the emitted light is collected through a 661/16 band pass filter. The data were analyzed by the BD CellQuest analysis software (BD Bioscience, San Diego, CA). Lymphocytes were gated based on SSC/FSC characteristics. Analysis of the total events was performed after exclusion of dead cells. CD3+ cells were gated within the viable lymphocytes gate. EdU+ cells were determined based on the negative control: cells cultured without EdU were stained in parallel by the panel of antibodies and Apollo® 643 azide. Furthermore, all measurements were performed under the same compensation conditions. The additions of kinds of fluorescence were decided by the aim of the experiments.
Incorporation of [3H]thymidine was performed as previously described (12). Around 2 × 105 cells were cultured with 10 μg/ml PHA for 72 h followed by a pulse of 12 h with the addition of 0.5 μCi of [3H]thymidine (Perkin Elmer, Waltham, MA). [3H]thymidine incorporation was determined by the LS6500 scintillation counter (Beckman coulter, Miami, FL). Results were expressed as stimulation index (SI), the ratio of the mean counts per minute (cpm) following stimulation divided by the mean cpm without stimulation.
Results were analyzed by GraphPad Prism 5 software (GraphPad, La Jolla, CA) and statistical analyses were performed using unpaired two-tailed Student's t test and correlation analysis. Results were considered statistically significant when P < 0.05 (*) and P < 0.01 (**).
The Effects of EdU Concentration and EdU Incubation Time
In Invitrogen's protocol (Click-iT® EdU Flow Cytometry Assay Kits) and RiBoBio's protocol, the recommended EdU concentration is 10 μM. But according to previous report (10), the optimal EdU concentration is 50 μM. Therefore, we determined whether the EdU+ cells were related to EdU concentration. Different final concentrations of EdU were added to the wells 12 h before harvesting the cells. As shown in Figure 1A, we noted that the number of EdU+ cells significantly increased in EdU concentration from 0.5 μM to 50 μM, as the peak value was seen when the concentration of EdU was 50 μM, but it decreased significantly as the EdU concentration increased from 100 μM to 250 μM. Furthermore, the number of EdU+ cells in EdU concentration to 10 μM was significantly higher than 1 μM (P < 0.01), but not different from 50 μM (P > 0.05).
Besides, in two manufacturers' protocols, the incubation time is not given clearly. Therefore, the effect of EdU incubation time on the number of EdU+ cells for T lymphocytes was also estimated in this study. A 50 μM final concentration of EdU was added to the wells before harvesting the cells. The number of EdU+ cells detected by flow cytometry increased significantly following increase of EdU incubation time from 30 min to 12 h (Fig. 1B), but there was no increase in the number of EdU+ cells in the non-stimulated groups (data not shown). A further increase in EdU incubation time from 16 h to 24 h did not change significantly the number of EdU+ cells.
Fixation and Permeabilization
Formaldehyde-based fixatives preserve tissue and subcellular architecture to a large extent and have been the preferred compounds for decades in standard pathology procedures. In RiBoBio's protocol, fixation is ignored. However, in Invitrogen's kits, fixation is performed using 4% paraformaldehyde. Furthermore, in their protocols, Triton X-100 (RiBoBio) and saponin (Invitrogen) are recommended as permeabilization detergent reagents, respectively. As we know, Triton X-100 is widely used in EdU incorporation with microscopy (10, 13). Hence, we considered whether fixation was necessary for the entire assay and which permeabilization detergents reagents were better for flow cytometry analysis. Compared with the fixed and permeabilized cells (Figs. 2B, 2E, and 2H), the number of the unfixed but permeabilized lymphocytes decreased significantly (P < 0.05), the sizes of the cells shrunk, and the fluorescence intensity of CD3+ cells was much weaker (174.75 ± 8.52 vs. 20.21 ± 1.89, P =0.0022) (Figs. 2C, 2F, and 2I), although the same number of cells were used before different treatments.
Next, the effect of saponin and Triton X-100 on the staining of cell surface antibodies was determined. Compared with saponin treatment (Fig. 3B in gate R1) and control (Fig. 3A in gate R1), the size and granularity of lymphocytes treated with Triton X-100 (Fig. 3C in gate R1) were much smaller, which led to undistinguishable differences between the lymphocytes and cell debris. Also, the signal-to-noise ratios (SNRs) for CD3e-PerCP-Cy5.5 were worse with Triton X-100 treatment (3.79 ± 0.23 vs. 58.66 ± 0.93, P < 0.01) (Figs. 3F and 3E). However, the SNRs for CD3e-PerCP-Cy5.5 with neither saponin nor Triton X-100 treatment were similar to control (105.56 ± 0.89, P < 0.01, Fig. 3D).
In APC BrdU Flow Kit from BD pharmingen™, fixation and permeabilization are performed simultaneously. But in Click-iT® EdU Flow Cytometry Assay Kits from Invitrogen, fixation and permeabilization are performed separately. Therefore, the best way used in EdU flow cytometry assay was verified here. Judged from their size and granularity, the cells treated by the two-step method (Fig. 3H) were similar to control group (Fig. 3G), but were different from the ones treated by the one-step treatment (Fig. 3I). Furthermore, the SNRs for CD3e-PerCP-Cy5.5 in these three groups were not identical: control (64.96 ± 1.03, Fig. 3J) >two step (11.04 ± 0.46, Fig. 3K) >one step (4.30 ± 0.41, Fig. 3L) (P < 0.01).
Impact of Click Reaction on the Staining of Cell Surface Antigens and EdU+ Cells
Cell surface antigens are commonly stained before the click reaction. But the click-iT® EdU Flow Cytometry Assay Kits from Invitrogen, PE, PE-tandem, or Qdot® antibody conjugates are required to be used after performing the Click reaction. Therefore, cell surface antibodies were added before or after Click reaction in order to determine whether Click reaction had an influence on the use of PE antibody conjugates (Fig. 4). The SNRs of CD8a-PE and CD4-FITC before Click reaction (Figs. 4B and 4E) did not change significantly compared to that after Click reaction (CD8a-PE: 63.46 ± 1.17 vs. 61.48 ± 1.92, P = 0.15; CD4-FITC: 27.22 ± 1.07 vs. 25.82 ± 0.84, P = 0.0556, Figs. 4C and 4F), but both were worse than control (Figs. 4A and 4D, P<0.05). For CD3e-PerCP-Cy5.5, the change of SNRs was also similar with the results above. Thus, Click reaction solution may have significant effects on the dissociation of three cell surface antibodies and antigens.
In two manufacturers' protocols, the recommended volume of Click reaction solution is 500 μl, but according to Yu's results, 250 μl does not affect EdU detection (14). In view of the facts mentioned above, the proper volume of Click reaction solution was determined. Figure 4G shows the association between EdU+ cells and the volumes of Click reaction solution. When 100 μl Click reaction solution was added, the highest number of EdU+ cells was achieved, but this was not different from the other groups except 5 μl group.
The Effect of Extent of Wash After Click Reaction and Correlate Analysis Between EdU and [3H]thymidine Incorporation
In the method of detecting EdU incorporation by fluorescence microscopy, the cells were washed several times with TBS with 0.1% Triton X-100 after Click reaction (10). However, for Click-iT® EdU Flow Cytometry Assay Kits, the cells were washed only once with 3 ml saponin-based permeabilization and wash reagent. We had shown the permeabilization reagent saponin is better than Triton X-100. Therefore, the proper extent of wash with PBS with 0.05% saponin after click reaction was determined.
The extent of wash has no significant influence on the SNRs of cell surface antibodies (CD3e-PerCP-Cy5.5, CD4-FITC, CD8a-PE) (data not shown). For the cells washed once, the Apollo® 643 azide intensity of the EdU- population in Fig. 5C and D was significantly stronger than the cells washed twice in Figs. 5E and 5F (CD4-FITC: 19.49 ± 3.04 vs. 7.17 ± 2.86, P = 0.0022; CD8a-PE: 25.82 ± 3.95 vs. 7.76 ± 1.96, P = 0.0022), respectively. However, the proportion of EdU+ cells in CD4+ T cells between one wash and two washes did not change (10.25 ± 1.58 vs. 10.59 ± 1.47, P = 0.6245). Therefore, sufficient wash only reduced the nonspecific fluorescence intensity of Apollo® 643 azide (Fig. 5).
Yu et al (10) reported that EdU incorporation was associated with [3H]thymidine incorporation, which was considered as the “gold standard” for measuring T-cell proliferations. Here, we confirmed the association between EdU and [3H]thymidine incorporation into PHA stimulated murine T lymphocytes. As demonstrated in Figure 5G, in the range of 10% to 135%, the SIs calculated from EdU incorporation were strongly correlated with [3H]thymidine incorporation (r = 0.90).
Influence of Storage Time on Surface and Intracellular Staining at Different Temperatures
To evaluate whether time of storage at 4°C, −80°C and in liquid nitrogen could affect the number of EdU+ cells, we tested murine spleen lymphocytes samples stored at 4°C, −80°C and in liquid nitrogen from 1 days to 21 days (Fig. 5H). No major differences in the percentages of EdU+ cells among CD4+ T cells were observed for storage periods up to 21 days for spleen lymphocytes (P ≥ 0.10). Compared to 4°C, the percentages of EdU+ cells among CD4+ T cells did not change markedly when samples were frozen in parallel at −80°C and in liquid nitrogen (P ≥ 0.26). After PHA stimulation, CD8+ T cells decreased significantly, therefore, only the proportions of CD3+ cells, CD4+ T cells among lymphocytes were analyzed (Fig. 5I). The results showed that they did not change when samples were stored at 4°C, −80°C, and in liquid nitrogen. This indicates that either storage condition above can be used for the analysis of T lymphocytes proliferation.
Although flow cytometry analysis of the proliferations of T lymphocytes subsets, which was performed through labeling of cell surface antigens combined with EdU incorporation, has been used to measure T-cell proliferation in vivo (8, 11), it has not been well established to detect T-cell proliferation in vitro. Therefore, we tried to optimize the conditions for this assay in vitro and to establish a feasible and accurate protocol for the detection of T-cell proliferation.
A few general conclusions could be drawn from this study. First, the number of EdU+ cells was associated with EdU concentration, EdU incubation time and the volume of Click reaction solution. We found that 10–50 μM EdU efficiently labeled T cells. However, the increasing EdU concentrations from 150 to 250 μM caused significant decrease of the labeled cells, which further proved the suppression of EdU on cell proliferation (10, 14, 15). Also, the optimal EdU incubation time was determined, which was from 8 to 12 h. As to the volume of Click reaction solution, 100 μl, 1/5 of recommended volume in two manufacturers' protocols, was well suited for the staining of initiating cells to 1 × 106. Considering that fixative may induce change in antibody binding capacity because of potential alteration on epitope exposure and affinity (16), Click reaction solution was recommended to be added before fixation. Besides, after Click reaction, sufficient wash was helpful for the reduction of the nonspecific fluorescence intensity of Apollo® 643 azide, but had no effect on the proportion of EdU+ cells.
Second, fixation was implicated to play a pivotal role in maintaining the cell morphology and the fluorescence intensity of cell surface antibodies, which varied with cell types and surface markers (17); therefore, a proper fixation for lymphocytes was important. Our results revealed that this step was better to be performed before permeabilization, and the suitable incubation time was 15 minutes for lymphocytes.
Third, for permeabilization the detergent reagent saponin was milder than Triton X-100 which could lyse cell membrane more thoroughly; this result was similar to the reports in primary adult epidermal keratinocytes (9). Furthermore, saponin not only lysed the cell membrane properly, but also maintained the fluorescence intensity of cell surface antibodies, which was consistent with the previous observations (9, 13, 18).
Fourth, strong correlation between EdU and [3H]thymidine incorporation indicated EdU incorporation was as accurate as [3H]thymidine incorporation to detect PHA stimulated murine T lymphocytes proliferation; this was similar to previous report on anti-CD3 stimulated murine T lymphocytes (10).
Finally, EdU labeled cells could be stored at 4°C, −80°C or in liquid nitrogen for storage periods up to 21 days, with the stable proportions of EdU+ cells in CD4+ T cells and surface staining in lymphocytes.
In summary, we described here an optimized technology of EdU incorporation and the labeling of cell surface antigens combined with flow cytometry analysis (Table 1), which could be used to detect the proliferations of T lymphocyte subsets accurately. Different factors contribute to this method include EdU concentration, EdU incubation time, fixation, permeabilization, Click reaction, wash reagents, extent of wash, which may have significant effects on final results; therefore, these factors should be scrutinized in immunofluorescence experiments. This method does not require strong denaturing conditions, anti-BrdU antibodies, and radioactive substance as required in the aforementioned methods. However, there are still some limitations in this method compared with BrdU incorporation and [3H]thymidine incorporation. For example, the cost of EdU incorporation is higher than that of [3H]thymidine incorporation (10). The present study not only can aid in the optimization of flow cytometry staining panels to quantify T-cell proliferation in mouse, but also can serve as a particular protocol of T-cell proliferation detection by the labeling of cell surface antigens combined with EdU incorporation.
Table 1. The protocol for flow cytometry analysis of the proliferations of T-lymphocyte subsets by the labeling of cell surface antigens combined with EdU incorporation
Isolation and culture of lymphocytes
Immunofluorescence staining of cell surface antigens