Ten color flow cytometry discriminates thymocyte subpopulations

Figure 2 shows the 10-color staining and gating scheme we have developed to comprehensively discriminate thymocyte subpopulations across development using a single stain. This scheme leaves the PE channel open for assessment of the relative expression of an additional antigen in each of the subpopulations. The ratios of the major populations of DN, DP, SP4, and SP8 cells and relative staining patterns were similar to those previously published [20–22]. These populations were further parsed into the indicated subsets, which again showed expected staining patterns previously published with less comprehensive staining methods. Importantly, this staining scheme allowed for the discrimination of subsets from the earliest DN1 populations throughout development to the most mature SP populations, without the need for cell sorting or multiple stains (Figure 1). Table 1 shows the average percentages of each of the 21 subpopulations from 10 individual mice. Of note, the use of CD25 expression relative to CD28 expression not only discriminated between pre-β-selected and post-β-selected cells [10], but also showed a DN CD44 negative, CD28 positive, CD25 intermediate population that we recently defined as DN3c [13]. This staining scheme also illuminates DN4 subsets which we define as DN4a (CD25^-CD28^int), DN4b (CD25^-CD28^hi), and DN4c (CD25^-CD28^-) [13]. The DN4b cells express high levels of TCR and thus appear to be a small subset of mature thymocytes that fail to express CD4 and CD8 (data not shown). Our previous OP9-DL1 fate mapping assays showed that DN3c cells give rise to DN4a cells which represent preDP cells in the DN compartment[13]. The advantage of this polychromatic scheme is the ability to enumerate and characterize all these populations simultaneously without the need for pre-sorting or separate stains. We also validated the ten-color stain by comparing it with a simpler six-color stain that used different combinations of antibodies. As shown in Additional File 1 Figure S1, the distributions of the major thymocyte subsets were very similar between the two staining schemes.

Assessment of CD124 and CD126 expression during T cell development

To show the utility of the 10-color staining scheme, we assessed surface expression of two potentially important cytokine receptors across thymocyte development (Figure 3). The ratios of the median fluorescent intensities for CD124 (IL-4Rα) and CD126 (IL-6Rα) relative to the corresponding isotype controls, and adjusted for FSC differences, are shown in Figure 3A (and summarized in Additional File 2 Table S1). These stains revealed two distinct patterns of expression. CD124 expression was very high on DN1a and DN1b early T cell precursors (ETPs), whereas CD126 showed very low expression on the ETPs. As the cells progressed from ETPs to the DN2 populations, they retained expression of CD124 but expression of this receptor waned as the thymocytes approached the β-selection checkpoint (DN3a stage). As we have previously published, TCR^- and TCR^lo DP cells showed little expression of CD126 but high expression of CD124 [23]. The most mature populations of SP4 and SP8 cells expressed both cytokine receptors. Of particular note, CD126 was highly expressed only on the most mature SP4 and SP8 populations (CD24^lo cells) and two very small DN1 subsets (DN1d, DN1e) that are reportedly not efficient precursors to mature T cells [8].

IL-6 signaling via pSTAT1 correlates with expression of IL-6Rα in the CD24^loSP4 and CD24^loSP8 populations

An IL-6 response assay was performed to determine whether the limited IL-6Rα (CD126) expression pattern shown in Figure 3A correlated with a similar pattern of responsiveness to IL-6 (Figure 3B). The CD24^hiSP4, CD24^loSP4, CD24^hiSP8, and CD24^loSP8 subsets were sort purified as described in the Materials and Methods. Cells were assessed for response to IL-6 as previously described [17, 23]. Cells were treated for 20 minutes in the presence or absence of 25 ng/ml of IL-6 and responsiveness assessed by intracellular staining, using mAbs specific to the phosphorylated form of STAT1 (pSTAT1). As predicted from the pattern of surface expression of the IL-6Rα, IL-6 responsiveness was most evident in the mature CD24^loSP4 and CD24^loSP8 populations.

Flow Cytometer Setup and Data Acquisition

Flow cytometric data were acquired using a BD LSR II Flow Cytometer (BD Biosciences, San Jose, CA). This instrument is an air-cooled 4-laser benchtop flow cytometer with the ability to detect up to 12 individual fluorescent signals. The lasers and filters used were supplied with the instrument from BD Biosciences. The lasers used in this study were the following: solid state, 50 mW, 488 nm (blue); solid state, 50 mW, 405 nm (violet); solid state, 150 mW, 532 nm (green); helium-neon, 70 mW, 633 nm (red). The photomultiplier (PMT) setup is depicted in Additional File 3 Figure S2 and used a trigon configuration for blue, violet, and red laser paths and an octagon configuration for the green detector arrays. Dichroic long pass mirrors were used on the inner rings, and band pass filters used on outer rings. A 2-Blue, 2-Violet, 5-Green, 3-Red fluorochrome excitation configuration was used in this study (Additional File 3 Figure S2). Instrument performance was validated using BD Cytometer Setup and Tracking (CS&T) beads (part #: 910723, Lot ID: 75445, BD Biosciences, San Jose, CA). This validation was performed routinely and prior to each of the large phenotyping experiments. Triggering was set on the forward scatter (FSC) signal. FACS DIVA 6.0 software (BD Biosciences) was used to acquire digital data.

Fluorochrome Compensations and PMT Settings

Although PMT settings and compensations should be adjusted for each instrument and stain setup, we include here the settings used for our instrument which were the following: the FSC setting included both area and width, with a threshold of 10,000, and a PMT Voltage of 315. The SSC setting included area with a PMT voltage of 300. All other PMT settings (log scale) were as follows: Pacific Blue = 380, APC = 545, APC-Cy7 = 568, FITC = 539, PerCP = 642, PE = 577, PE-Texas Red = 501, PE-Cy5.5 = 479, PE-Cy7 = 551, and Pacific Orange = 468. The PMT voltage settings were manually set using unstained cells as a reference. Each histogram was adjusted to allow a full view of the peak at approximately the same intensity for each fluorochrome.

Unstained and single fluorochrome-mAb stained thymocytes were used to establish baseline instrument application settings and compensation settings for each fluorochrome measured. To determine and validate compensation settings for fluorescence on small populations, cells of interest were sort enriched prior to use. Preliminary experiments were designed to first assess the staining patterns of small panels of mAbs and then the staining patterns observed in these smaller panels were compared with those previously observed in simple two-, three-, or four-color experiments. After comparable results were obtained, an additional interest marker was added, and separate "fluorescence minus one" (FMO) controls were prepared to ensure that the staining patterns and all subset populations of the original markers were not changed by the addition of the new marker. FMO control histograms are shown in Additional File 4 Figure S3. Using this sequence, the 10-color staining panel was eventually successfully validated.

Cell Sorting

For the pSTAT assays shown in Figure 3B, the cell subsets indicated in the figure legend were sort-purified using a two-laser Moflo cell sorter (Beckman-Coulter, Fullerton, CA) and Summit v4.3 software. For the validation of compensation settings for the small DN thymocyte subsets, cells were stained with a single fluorochrome and sort-enriched using the MoFlo cell sorter prior to use on the LSR II. These cells were stained and positively enriched using one of the following mAb or mAb combinations: cKit-APC, CD25-PerCP, CD44-APC-Cy7, or Lin-biotin/TCR γδ-biotin followed by streptavidin-PE-Texas Red.

Data Analysis

Data were analyzed using FACS DIVA 6.0 software (BD Biosciences). For all experiments, cell doublets and clusters were gated from the analysis using doublet discrimination (FSC width versus FSC area). Cell debris and small particles were excluded by gating out events with low forward scatter. The gating strategy is shown in Figure 2. CD4 and CD8 double positive and single positive populations that were negative for the non-T lineage markers (Lin) and TCR-γδ were further discriminated as follows: DP (CD4⁺CD8⁺), SP4 (TCR⁺CD4⁺CD8^-), and SP8 (TCR⁺CD4^-CD8⁺). DP TCR negative, DP TCR intermediate, and DP TCR high were separated by TCR-β chain (H57-597) expression. CD4^-CD8^- (DN) populations were further defined as follows: DN1 (Lin^-CD25^-CD44⁺), DN2 (Lin^-CD25⁺CD44⁺), DN3 (Lin^-CD25⁺CD44^-) and DN4 (Lin^-CD25^-CD44^-). Five subsets of TCR negative DN1 thymocytes (DN1a-DN1e), as defined by c-Kit and CD24 expression were discriminated as previously demonstrated [8]. DN3 and DN4 were also further defined based on CD25 and CD28 expression. DN3 cells were divided into CD28 low/CD25 high (DN3a), CD28 intermediate/CD25 high (DN3b) and CD28 intermediate/CD25 intermediate (DN3c). DN4 cells were divided into CD28 intermediate (DN4a), CD28 high (DN4b), and CD28 low (DN4c) cells as described in the results section.

In Figure 3, median fluorescent intensity (MFI) ratios were calculated based on the FSC-adjusted MFI values of the indicated specific cytokine receptor stains divided by the FSC-adjusted MFI values from the corresponding isotype control stains. FSC adjustment (PE MFI/FSC MFI) was done in order to account for cell size differences in the different subsets. This also allowed for normalization across the multiple experiments. MFI ratio averages and standard deviations were calculated from a minimum of five separate experiments for each cytokine receptor. To do overlay plots, FCS data were exported, and analyzed by FlowJo (Tree Star, Inc, Ashland, OR). Statistical analysis was generated using GraphPad Prism v. 4.00 (La Jolla, CA).

Mice

Mice used in this study were C57BL/6 females ranging from 6 to 8 weeks of age that were bred and housed at U.S.D.A. approved animal facilities located at the University of Oklahoma in Tulsa and the University of Tulsa. Animal care and all animal experiments were done in accordance with procedures outlined in the Guide for the Care and Use of Laboratory Animals (National Research Council). Protocols were reviewed and approved by the Institutional Animal Care and Use Committees of the University of Oklahoma Health Sciences Center and the University of Tulsa.

Tissue harvest and cell staining

Thymuses were harvested and placed into complete tumor media (CTM) as previously described [28]. Thymuses were pressed through 70-μm nylon screens to generate single thymocyte suspensions [17]. Cells were treated with RBC lysis buffer (Sigma-Aldrich, St. Louis, MO), washed into CTM and counted using a hemocytometer. Cells were incubated with mAb against mouse CD16/CD32 (Fc Block, BD Biosciences) to block potential Fc-mediated binding and then stained at a density of 1 × 10⁸ cell/ml with primary mAbs as indicated in Additional File 5 Table S2 for 45 minutes at 4°C in the dark. After two washes, the cells were further stained with Streptavidin PE-Texas Red (1:1600) for 30 minutes at 4°C in the dark. The monoclonal antibodies used in the 10-color stain experiments were individually tested. Optimal antibody titers (Additional File 5 Table S2) were determined for each monoclonal antibody and/or their secondary stain (Streptavidin PE-Texas Red) by staining thymocytes with mAb titers ranging from 1:50 to 1:1600, depending on the fluorochrome. For Annexin-V (Biolegend) staining, cells were washed twice after surface marker staining, then resuspended in Annexin-V binding buffer (Biolegend containing Annexin-V-FITC (1:100), and incubated for 15 minutes at room temperature in the dark.

STAT phosphorylation assays

Assessment of IL-6-induced STAT-1 phosphorylation began with surface staining of total thymocytes with anti-CD4, CD8, and CD24 mAbs (Additional File 5 Table S2). These cells were sort-purified into the following populations: CD24^loSP4, CD24^hiSP4, CD24^loSP8, and CD24^hiSP8. Sorted cells were then incubated in the presence or absence of 25 ng/ml IL-6 (R&D Systems, Minneapolis, MN) for 20 minutes at 37°C. Intracellular stains were performed using Invitrogen/Caltag Fix & Perm reagents with an additional methanol step as previously described[23]. STAT-1 phosphorylation status was assessed with PE-conjugated anti-pSTAT1 (pY701) (BD Biosciences) using a PE-conjugated mouse IgG2a (BD Biosciences) as an isotype control.