The immune system consists of various subsets of immune cells such as T-cells, B-cells, macrophages, NK cells, NKT cells, dendritic cells, and so on. These immune cells are connected to each other through networks of cytokines or chemokines as well as through direct cell–cell interaction (1). The immune system functions by detecting and distinguishing various subsets of the immune cells using various cell surface markers. Flow cytometry (FCM) and fluorescence microscopy are very useful and powerful methods for analyzing several surface markers simultaneously (2–4). Recent improvements in fluorescence reagents and equipment enable us to analyze about 20 different markers at one time (5, 6). In analysis with FCM or fluorescence microscopy, fluorophores should be activated with a laser or light in special equipment and their fluorescence should be distinguished from other fluorescence by optical filters (2, 5, 7). Consequently, the number of markers that can be analyzed with these methods is limited by the number of fluorophores that can be used in these methods.
Photobleaching is a dynamic process in which fluorochrome molecules irreversibly undergo photoinduced chemical destruction upon exposure to excitation light and thus, lose their ability to fluoresce. Multiple photochemical reaction pathways are involved in the photobleaching of fluorescein (8, 9). Photobleaching is a factor that limits fluorescence detectability. To overcome the problem, attempts have been made to maximize detection sensitivity by developing devices such as CCD cameras as well as high-numerical aperture objective lenses (10). Alternatively, less photolabile fluorophores have been developed (11). In contrast, photobleaching has been the basis of a fluorescence measurement technique, fluorescence recovery after photobleaching, which is utilized for measuring the lateral diffusion of fluorescent probes or endogenous fluorescent compounds in biological membranes (12).
Recently, we have established a cell-microarray system that enables high-throughput analysis of cellular responses of more than 10,000 lymphocytes at the single-cell level simultaneously (13). In the cell-microarray system, the chip has 45,000–234,000 microwells that capture single lymphocytes. Because the cell position on the chip is fixed, we can repeatedly analyze the fluorescence signals of the same cell on a single-cell basis. In this study, we present a novel method for detecting and analyzing multiple cell markers using a single fluorophore by applying the cell microarray system and photobleaching. We named this method a single fluorophore multiplex analysis on a chip (SFMAC). Using this method, we can analyze multiple numbers of biological markers with a single fluorophore from a small sample of blood. In addition, we can retrieve objective cells from the chip and analyze them at the protein level or DNA level. The SFMAC system might expand the horizons of fluorescence-based cell analyses.
MATERIALS AND METHODS
Biotinylated mouse CD4-Ab, streptavidin (Sav)-phycoerythrin (PE), Sav-Cy3, Sav-FITC, and Sav-PE-Cy5 were all purchased from BD Pharmingen (San Diego, CA). To analyze cell surface markers on human lymphocytes, PE conjugates of mAbs to CD3, CD19, or CD25 were purchased from Beckman-Coulter (Hialeah, FL), and PE-CD4 or CD8 mAbs from Immunotech (Marseille, France). For staining intracytoplasmic proteins, PE-IFNγ-mAb and PE-IL-4-mAb were purchased from Immunotech, and PE-human-FoxP3 mAb (clone PCH101) from eBioscience (San Diego, CA).
To prepare human lymphocytes, we collected heparinized peripheral blood from healthy individuals after obtaining written informed consent. The collected blood samples (100 μl) were hemolyzed by adding 3 ml of hemolysis reagent (Nichirei, Tokyo, Japan) and used for preparing a cell microarray. Where indicated, peripheral blood lymphocytes were prepared by centrifugation on a Ficoll-Hypaque gradient. To prepare mouse lymphocytes, we collected spleen cells from 6- to 12-week-old C57Bl/6 female mice and used them as a lymphocyte fraction after hemolysis with NH4Cl buffer, as described recently (13). The mice experiments were performed in accordance with protocols approved by the animal experiment committees at the University of Toyama.
Cell Array and Photobleaching
Microwell array chips had a regular array of 45,000 microchambers 10 μm in diameter and 15-μm deep, which could capture only single lymphocytes as previously described (14). For depositing a single cell in each microwell, cells were first stained with the biotinylated mAb and fluorophore-conjugated Sav or with PE-conjugated mAb. The stained cells were added onto the microwell array chips. After the chip was laid still for 5 min, cells outside of wells were gently washed away with PBS by removing the cell suspension buffer gently from the chip and readding PBS gently onto the chip to cover it over. We repeated the washing several times. Because the size and shape of wells are optimized to lymphocytes so that they can stay in wells during gentle washing, and the larger cells such as granulocytes or monocytes are washed away (Supporting Fig. 1, this material is available online). We then scanned the chip with a cell-scanner (CRBIOIIe-FITC, Hitachi Software Engineering, Tokyo, Japan) and analyzed the fluorescence of each cell with software (TIC-Analysis, Hitachi Software Engineering). After analysis, we irradiated the cells stained with PE, Cy3, or PE-Cy5 with green light (wavelength, 520–550 nm) or the cells stained with FITC with blue light (wavelength, 470–490 nm) for photobleaching the fluorophores under a fluorescence microscope (BX51WI, Olympus, Tokyo, Japan). After irradiation, photobleaching was confirmed by observing the cells under the fluorescence microscope or the cell-scanner. During the photobleaching process, we could chase the fluorescence signals of positive cells because cell position was fixed on the chip during the experiment.
A Single Fluorophore Multiplex Analysis on a Chip
For multiplex analysis, we removed the PBS on the chip and added another PE-conjugated mAb solution to cover the well area. Five minutes after the incubation, we removed the Ab solution gently from the chip and readded PBS gently onto the chip to cover it over. We repeated this washing step at least two more times to remove Ab solution. Thereafter, the photobleaching and staining steps were repeated as earlier. As previously described, the cell-scanner and computer recorded the address and fluorescence intensity of cells at each scan (14, 15). We could freely combine the information of analyzed cells before and after staining or photobleaching. So, we could correlate the results from different mAbs. During washing and staining processes, most of the cells stay in the wells (Supporting Fig. 2b).
Detection of Intracytoplasmic Cytokines or Intracellular Proteins
To detect intracytoplasimic cytokines, the blood cells were stained with PE-CD3 mAb, hemolyzed, and arrayed on the microwell array chips. The arrayed cells were analyzed with the cell-scanner, and the cells on the chip were stained with PE-CD4 mAb and PE-CD8 mAb, as earlier, and analyzed with the cell-scanner. The cells were then stimulated with 40 ng/ml phorbol 12-myristate 13-acetate (PMA, Sigma, St Louis, MO) and 1 μg/ml Ca2+ ionophore (Ionomycin, Sigma) in the presence of 20 μg/ml Brefeldin A in RPMI1640 containing 10% fetal bovine serum for 4 h. After the simulation, the cells were washed with PBS three times, incubated with fixation/permeabilization buffer (eBioscience) for 30 min, and incubated with permeabilization buffer (eBioscience) for 10 min. Intracytoplasmic cytokines were detected using PE-mAbs against IFNγ or IL-4. For detecting an intracellular protein, FoxP3, cells were fixed and permeabilized as above and stained with PE-mAb against human-FoxP3 for 15 min. Cells were then washed, and the fluorescence intensity of each cell was measured and analyzed with the cell-scanner system. FoxP3 positive cells were determined when compared with control cells (CD3+CD4+CD25− or CD3+CD4−).
Sensitivity of Various Fluorophores to Photobleaching
We hypothesized that we could repeatedly stain and analyze cells on a microwell array chip with multiple marker Abs using a single fluorophore if the fluorophore is photobleached. To verify the hypothesis, we first examined the effect of blue- or green-light irradiation on the photobleaching of various fluorescent probes that were used for FCM or fluorescence microscopy. We first incubated mouse spleen cells with biotinylated-CD4 mAb. We then divided them into four groups. Each was stained with the saturated amount of either PE-, FITC-, Cy3-, or PE-Cy5-conjugated Sav. The stained cells were arrayed on a microwell array chip, and the fluorescence of each cell was monitored by a cell-scanner. As shown in Figure 1A, the fluorescence intensity of PE, Cy3, or PE-Cy5 was high compared with that of FITC when the fluorescence intensity of positive cells was compared with that of negative cells. When we irradiated PE-, Cy3-, or PE-Cy5-stained cells with green light or FITC-stained cells with blue light, PE and PE-Cy5 were photobleached in 30–60 s, whereas FITC and Cy3 were rather stable. Because the photostability of PE-Cy5 is dependent on the PE-moiety, we used PE-conjugated Abs in the following experiments. We then examined whether the fluorescence of PE recovered from the photobleached state. To that end, we first stained mouse splenocytes with biotinylated CD4 mAb and PE-conjugated Sav, and then performed photobleaching experiment as described in Figure 1A for total 60 s. Thereafter, we kept the chip in the dark and observed it with a cell-scanner at indicated time points in Figure 1B. The result showed that after PE was photobleached with green light, the fluorescence of PE did not recover at all. Next, human lymphocytes were stained with PE-conjugated CD4-mAb, arrayed on a microwell array chip, and observed under a fluorescence microscope. As shown in Figure 1C, CD4+ cells were recognized in a regular array. As we continuously observed the cells under the fluorescence microscope, the fluorescence of PE was quickly attenuated within 15 s and almost completely faded out within 30 s. When the border of the green light was observed after photobleaching, we could clearly recognize the bleached and unbleached areas (Fig. 1D). Because the PE did not regain the ability to fluoresce after it was photobleached, the results showed that PE might be a suitable fluorophore for multiplex analysis with a single fluorophore using the photobleaching phenomenon.
Applicability of Photobleaching Procedure to Multiplex Marker Analysis
To assess the applicability of the photobleaching process to multiplex marker analysis using a single fluorophore, we first stained human peripheral blood samples with PE-CD3-mAb, and then hemolyzed the blood cells. We applied the stained and hemolyzed blood cells to a microwell array chip and analyzed their fluorescence with the cell-scanner. The major cell population in white blood cells is neutrophils. Because their diameter is 12–15 μm and that of microwells is 10 μm, most of neutrophils could not enter into microwells and were removed from the chip during the washing processes. Consequently, analyzed cells that remained on the chip were mostly lymphocytes (>90%). We confirmed this by harvesting cells on the chip and analyzing them with a flow cytometer (Supporting Fig. 1). CD3+ T-lymphocytes were detected as shown in Figures 2A and 2F. Photobleaching of PE under a fluorescence microscope reduced the fluorescence intensity of PE-CD3+ cells to the same level as that of unstained cells (Figs. 2B and 2G). By repeating the staining and bleaching procedures, we could detect PE-CD4 (Figs. 2C and 2H) and PE-CD19 (Figs. 2E and 2J) positive cells in the same sample. Because the positions of the cells on the chip are fixed, we could combine the data of each cell in Figures 2A and 2C and prepare a dot plot as in Figures 2K or 2L, which show the CD3+CD4+ cells and CD3+CD4− cells or CD3+CD19− and CD3−CD19+ cells. These results corresponded to those analyzed by FCM with multicolor analysis (Table 1). In contrast to FCM, some background noise, such as CD4dim, CD19bright, or CD3+CD19+ population, might be seen in SFMAC, probably because of the binding of PE-antibody to Fcγ-receptor on remained monocytes or intrinsic bias of chip analysis. The result demonstrates that we can analyze several surface markers on lymphocytes from a small volume of peripheral blood by a single fluorophore using a photobleaching procedure and microwell array chips.
Table 1. Comparison of the data between cell-microarray system (SFMAC) and flow cytometry (FCM)
The number indicates the percentage among lymphocytes. For cytokine or FoxP3-positive cells, the number indicates the percentage among the indicated population. The average ± SD of three independent experiments in a representative case is shown. Statistical significance is evaluated with Mann-Whitney U-test.
66.9 ± 5.5
67.1 ± 5.2
31.2 ± 3.4
32.7 ± 3.5
28.2 ± 3.8
28.4 ± 2.8
12.3 ± 4.8
10.3 ± 0.9
CD4+IFNγ+ (in CD4)
2.1 ± 0.4
6.8 ± 1.0
CD8+IFNγ+ (in CD8)
33.9 ± 5.4
44.6 ± 6.5
CD4+IL-4+ (in CD4)
0.3 ± 0.1
2.6 ± 0.7
CD8+IL-4+ (in CD8)
0.1 ± 0.1
0.1 ± 0.1
CD4+CD25+FoxP3+ (in CD4)
3.6 ± 0.5
4.1 ± 0.2
Multiplex Analysis of Cytokine-Secreting Cells with a Single Fluorophore
Next, we analyzed cytokine secretion as well as cell surface markers on T-lymphocytes with the SFMAC. First, we stained whole blood cells in 100 μl of peripheral blood with PE-conjugated CD3-mAb, hemolyzed, and arrayed them on a chip, and then analyzed them with the scanner. We sequentially stained and analyzed cells with PE-conjugated CD4-mAb and CD8-mAb using the SFMAC method (Fig. 3A). We detected 51% CD4+ T-lymphocytes and 46% CD8+ T-lymphocytes among the CD3+ T-lymphocytes. The percentage of each population was comparable with that of the FCM analysis (Table 1, Supporting Fig. 3). We then applied the SFMAC for the detection of cytokine-producing cells. We stimulated the stained cells with PMA and Ca2+ ionophore in the presence of Brefeldin A on the chip. After 4 h, we stained the cells using the cell-permeabilization assay. As shown in Figure 3A, we were able to detect IFNγ- or IL-4-secreting cells in the CD4+ or CD8+ T-lymphocytes. When the percentages of cytokine-producing cells were compared with those analyzed by FCM, the percentage of IFNγ-producing CD8+ cells analyzed with SFMAC was almost comparable to those analyzed with FCM, although those of IFNγ- or IL-4-producing CD4+ T-cells analyzed with SFMAC were lower. (Table 1 and Supporting Fig. 3). The results show that we can analyze the ratio of CD4/CD8 subpopulations in lymphocytes prepared from 100 μl of peripheral blood.
Multiplex Analysis of Intracellular Protein Expression in Human Lymphocytes with a Single Fluorophore
Finally, we examined whether we could use the SFMAC method to analyze cells that expressed intracellular proteins of interest. For this purpose, we stained human hemolyzed blood cells with PE-conjugated CD3-, CD4-, and CD25-Ab on a chip (Fig. 3B). After the cells were stimulated with PMA and Ca2+ ionophore as earlier, we permeabilized the cells on the chip and stained them with PE-conjugated IFNγ-mAb, followed by PE-conjugated FoxP3-Ab. As shown in Figure 3B, we were able to detect FoxP3, a key transcriptional factor in regulatory T-cells (16), in addition to IFNγ. As shown in Figure 3B, CD4+CD25high cells, which also expressed FoxP3, did not produce IFNγ as previously reported (17). Taken together, these results show that the cell microarray system enables us to detect a variety of markers including intracellular proteins as well as cell surface markers using a single fluorophore.
In this study, we have proposed a procedure for analyzing cells with multiple markers using a single fluorescent probe, PE, and a cell-microarray system (SFMAC system). Using this system, we were able to detect various surface markers, secreted cytokines, and intracellular proteins in lymphocytes that were prepared from a limited small volume (100 μL) of blood samples. We previously demonstrated that the cell-microarray system enables us to analyze individual lymphocytes repeatedly (13–15). Here, we have demonstrated that the photobleaching technique enables us to use multiple markers that were labeled with PE. In addition, the SFMAC enables us to not only analyze cells but also to retrieve and utilize them in a live state.
PE is used for FCM but not for conventional fluorescence microscopy because it is susceptible to photobleaching. In the cell microarray system, we monitor the fluorescence signals of cells with a laser scanner that gains fluorescence signals from a single cell in about 50 μs. The short scanning time enables the detection of fluorescence signals of PE without concern for the effects of photobleaching. Similarly, laser scanning confocal microscopy can be used to analyze PE conjugates without affecting the photostability of PE, indicating that we might be able to analyze tissue sections with multiple markers using only PE conjugates by a laser scanning confocal microscope. Recently, slide-based optical live cell array technology was made available by Molecular Cytomics (Boston, MA) (18). They supply a range of microwells with diameters from 15 to 250 μm. Although the size and shape of their microwells are different from those of our microwell array chip, we might be able to utilize their chip for the SFMAC analysis.
In the detection of cytokines, the numbers of cytokine-producing cells (especially IFNγ- or IL-4-producing CD4+ T-cells) detected by SFMAC were lower when compared with those by FCM. The difference might be due to the difference in the stimulation condition. In SFMAC analysis, the cells were arrayed into each microwell and stimulated in the separated state, whereas cells were stimulated with 96-well plate (u-bottom) for FCM analysis, where cells could contact with each other. More suitable condition for stimulating cells on the chip should be established in the future. The number of FoxP3+ cells was corresponding between two methods, which indicate that SFMAC analysis could be used for analyzing intracellular markers.
In summary, we have developed a novel method of analyzing multiplex cell markers using a single photobleachable fluorophore with a cell-microarray system. The method can be used to analyze not only cells but also tissue sections if the positions of the samples are fixed on a slide. Our method may simplify the reagents that are required for analyzing multiple markers and the equipment used for analysis by employing a single fluorophore. In the future, the development of new fluorophores that emit strong and uniform fluorescence and that enable complete and immediate photobleaching might contribute to a more feasible system for multiplex analysis.
This work was performed under the auspices of the Toyama Medical Bio-Cluster Project.