Flow cytometric analyses disclose intercellular communications in FasL-stimulated T cells: Results and trouble shooting


Life depends on the ability of cells to communicate with one another. Much of this crosstalk occurs at cell–cell contacts and is regulated by complex structural interface: neurological and immunological synapses that transmit cell–cell signals through the extracellular space, relying on mechanisms of ligand-receptor signalling across tight cell–cell junctions. In addition to these well-known examples, other cellular structures involved in cell–cell communication have been identified (1). A frequent result of cell–cell communication is collective population behavior that can potentially lead to complex phenotypes not observed in separate individual cells, for example, in development or tumor formation (2, 3). Immune responses against pathogens or any foreign antigens require fine immune regulation, where cellular communications are mediated by either soluble or cell surface molecules. Generally, absorption, exosome production and uptake, internalization, and membrane nanotube formation are the probable mechanisms through which the membrane fragments and cytoplasmic content are transferred from one cell to another (4). The identification of new structures involved in cell-to-cell communication has led to the development of new analytical and imaging tools, which have allowed us to enhance our understanding of the ubiquitous phenomenon of cell-to-cell communication. Such tools include quantitative cell microscopy, now mainly represented by two distinct techniques: flow cytometry (FCM) and image cytometry (IC) (5). FCM allows analysis of cell suspensions, cell-by-cell quantification of optical signals, rapid analysis (several hundred cells per second), and cell sorting, whereas IC allows precise localization and quantification of optical signals emitted by each point of the observed field and image analysis (morphology and morphometry).

FasL Stimulation of T Cells

The Fas/CD95 surface receptor mediates rapid death of various cell types, including autoreactive T cells, which can trigger autoimmunity. In this study we investigate novel aspects of Fas signaling that define a “social” dimension to receptor-induced apoptosis. At least in some cell types, FasL can be stored in granules and then rapidly released at the cell surface in response to external stimuli. Alternatively, FasL can be cleaved from the membrane by matrix metalloproteases and act as a soluble cytokine (6). In its membrane-bound form, FasL acts as a ligand for Fas and can trigger Fas-induced death.

In our experiments, CD4+ T cells were purified from PBMC by negative selection using the MACS system. CD4+ T cells and Jurkat cells were both treated with FasL at a final concentration of 0.5 μg/ml for a maximum of 2 h. Our previous findings (7) showed that Fas stimulation led to an intriguing (and previously unsuspected) asymmetry of death propagation according to cell conjugation and rapidly induced extensive membrane nanotube formation between neighboring T cells. The experiments were performed using flow cytometric (FC) and morphologic (fluorescence and time lapse) analyses. For the measurement of intercellular exchange by FCM, we used the total mean fluorescence method subtracted for background with the separate staining of T cells by means of red PKH-26 or carbocyanine membrane dye DiI and various green probes, including the cytosolic stain 5-(and -6)-carboxyfluorescein diacetate succinimidyl ester (CFSE). Exchanges were investigated between the same cell types, in PHA and IL-2 stimulated CD4+ T cells and Jurkat cells. Double labeling experiments using DiI (or PKH 26) and CFSE were performed to exclude the possibility that the uptake of cytosolic and/or membrane fragments was merely an artefact caused by cells sticking to one another.

Results and Troubleshooting

As shown in Figure 1A, co-incubation of variably fluorescent, activated, FasL-treated T cells resulted in a marked increase in fluorescence intensity in the FL-1 channel, reflecting the uptake of the CFSE stained portion (FasL-promoted ratio of exchange; RE: 9.7 and 11.3 after 30 and 60 min, respectively vs. 2.5 and 2.8 for untreated RE, P < 0.03).

Figure 1.

A: Fluorescence bivariate analysis, CFSE versus PKH26 of CD4+ lymphocytes untreated (a) and FasL treated (b). The events shown are those selected by gate R1, drawn on putative single cells, 30 min. after FasL addition. A strong shift for FL1 fluorescence of PKH26+ cells is visible. B: Contour plot FSC versus SSC representing scatter characteristics of IL-2 stimulated CD4+ cells. Untreated cells are shown in a) whereas b) shows FasL treated cells. Gates 1 and 2 are drawn around putative single cells and cellular aggregates, respectively. FasL-treatment leads to a strong increase in cell conjugation as demonstrated by this representative experiment: 12% cell conjugates found in (a) increasing to 32% in (b). C: Fluorescence bivariate analysis, CFSE versus PKH26 of CD4+ lymphocytes untreated (a) and FasL treated (b). The events shown are those selected by gate R2, drawn around cellular aggregates and conjugates. As shown by the events distribution, the percentage of double fluorescent cells remains stable, whereas the percentage of aggregates and conjugates in the scatter area increases sharply. Furthermore, grey arrows highlight the significant FL1/fluorescence shift of red-coloured events. D: Histograms reporting shift for FL1 (CFSE) fluorescence of PKH26+ cells selected by gate R2, drawn, as cited in the text, around cellular aggregates and conjugates, as visible in the scatter area. The empty histogram represents CFSE fluorescence of CD4+ untreated cells, whereas the filled histogram represents CFSE fluorescence of FasL-treated cells. The following values are reported: FasL-promoted RE: 20.3 and 24.5 after 30 and 60 min versus 9 and 10.3, respectively for untreated RE, P < 0.003).

Conversely, there was only a slight increase in the fluorescence intensity in the FL-2 channel (particularly for PKH26) of T cells showing that the observed signal was indeed the consequence of a specific and different membrane/cytosolic transfer and not caused by cell adherence. Such data were obtained excluding debris, dead cells, and also aggregates, following the guidelines set out in recent literature (8). Nevertheless, these virtual aggregates, as they appear in the FC dot plot (Fig. 1B), may contain events of interest for the evaluation of this biological phenomenon.

In fact, after FasL administration, we found a rapid increase in cell conjugates found in the scatter-area of aggregates in both Jurkat and stimulated CD4+ cells. These events, revealed by fluorescence expression, highlight a double positive subpopulation (Fig. 1C) but also a reliable fluorescence shift, particularly for CFSE (Figs. 1C and 1D).

The scatter-area enclosing aggregates (Fig. 1B, R2) may highlight multiple cellular contacts and, taking this region into consideration, may result in an overestimation of intercellular communication. On the other hand, excluding it from the analysis may lead to an underestimation of the phenomenon.

We adopted the expedient of only evaluating the doublets as double fluorescent events (Fig. 1C) excluding them only (Fig. 1D) in order not to suppress the entire scatter-area of aggregates. Nevertheless, we must keep in mind that, although excluded, such double fluorescent cells may represent developing physiological and functional cell contacts and not just casual aggregates or artifactual “double positive” cells.

Moreover, experiments on unstimulated equivalent cells and those performed by diluting the cell suspensions triggered by FasL confirm FasL proficiency as a cell conjugate promoter, also providing the greatest values for CFSE (> DiI> PKH dyes) uptake, if compared to unstimulated cells.

Membrane Nanotubes and Microvesicles

Intercellular transfer, revealed by CFSE uptake, may be independent (formation of tunneling nanotubes, TNTs) or dependent [exosomes and microvesicles (MVs)] on diffusion through the extracellular medium. For the analyses of TNT structures (Fig. 2A) and intercellular transfer, co-cultures of green-stained cells and red-stained cells were mixed in a 1:1 ratio. Mixed cells were seeded on cover slips coated with polylysine in the presence of Fas stimulation, fixed in 2% glutaraldehyde in PBS for 30 min, washed in PBS, and then processed for microscopic imaging. For FACS analyses, donor and acceptor populations were analyzed at 488 nm excitation wavelength. We also investigated the effects of various concentrations of cytoskeleton polymerization inhibitors (i.e., cythocalasin B and latrunculin A, data not shown) on TNT formation and stability.

Figure 2.

A: Fas signalling increases cell conjugation and the formation of connecting membrane nanotubes. Representative images combining PKH67 and DiI stainings. (Aa) Unstimulated CD4+ T cells; (Ab) FasL-stimulated CD4+ T cells after 60 min. The blebbing cell (*) is conjugated with a cell showing no evident sign of membrane blebs. We used a Leica TCS SP5 II confocal microscope with oil-immersed objectives. B: Histograms related to CD4+ T recipient cells, mixed with ultracentrifuged pellets DiI labelled from unstimulated (a) and 1h FasL-stimulated cells (b), after 30 min from their addition. ○ Polysciences beads (1 μm Ø); • Dako Cytocount beads (5,2 μm Ø); * CD4+ cells; equation image MV area (slope). C: Contour plots describing CD4+ T recipient cells plated 30 min with DiI+ MVs from unstimulated (a) and FasL stimulated (b) cells. A well-defined cluster of DiI+ cells appears in contour plot (b), whereas in (a) only a few events are visible. Absolute numbers are given by means of Dako Cytocount beads. D: Western blot quantification of CD4 + T cell exosomes in response to FasL treatment. Whole cellular lysate proteins and exosomes from CD4 + T cells were separated on SDS-PAGE and electroblotted to nitrocellulose membrane. Blots were probed with antibodies against Alix, a well-established marker for exosomes. Samples were loaded as follows: (1) cellular pellet from FasL treated cells; (2) cellular pellet from untreated cells; (3) exosomal pellet from FasL-treated cells; (4) exosomal pellet from untreated cells. Molecular mass markers are shown on the left. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

MVs are submicron structures shed from the cell membrane in a final step of the budding process. After being released into the microenvironment, they are free to move and carry signaling molecules to distant cells; hence, they represent a communication system (9–11).

We performed differential centrifugation coupled with membrane filtration in order to eliminate large contaminating extracellular vesicles, though small contaminating vesicles may still have been present in the preparation (12).

To visualize the MVs by FCM, the axis for forward scatter versus side scatter plot required log-scale. The beads of different sizes (1 μm Polysciences Invitrogen, Carlsbad, CA and 5.2 μm DakoCytoCount beads) were used as a kind of internal standard for size and to obtain an absolute count for both MVs and cellular MV uptake. In fact, we coincubated ultracentrifugated pellets (about 2 μg and 0.5 μg for FasL-stimulated and unstimulated cells, respectively, obtained from the same amount of cells—4 × 106) with recipient cells (0.2 ×106 cells). Preliminary FC experiments were carried out together with Western blotting analyses for Alix detection. Our FC preliminary approach seems to have the ability to detect MVs released by 4 × 106 cells just 1 h from their seeding with and without FasL addition (Fig. 2B). Furthermore, analyses of unstimulated/recipient cells co-incubated with MV pellets (from both unstimulated and stimulated cells) not only shows a major DiI uptake for MV pellets from stimulated cells (Fig. 2C), but also a mitochondria impairment (by Nonyl Acridine Orange staining, data not shown), suggesting a numerical and functional distinction between MVs from FasL-stimulated and unstimulated cells. These data were confirmed by Western blot analysis. Using Alix as a specific marker for exosome detection, we found an increase in exosomes in response to FasL-treatment compared to the control condition (Fig. 2D).


In conclusion, combining different dye-loading techniques with advanced fluorescence-activated FCM facilitates the identification of different cell populations and subcellular components (MVs) revealing FasL as a promoter of cell conjugates, microvesicle release, and TNT formation (as attested by CFSE uptake) in both CD4+ and Jurkat cells (data not shown). Hence, FCM appears to be a powerful and fundamental tool for the analysis of cell-to-cell communication, able to quantify the degree of the phenomenon, and, at least in part, to characterize it (MVs vs. TNTs; CFSE vs. DiI).