Clonal expansion is a key feature of antigen-specific T cell response. However, the cell cycle of antigen-responding T cells has been poorly investigated, partly because of technical limitations. We describe a flow cytometric method to analyze clonally expanding antigen-specific CD8 T cells in spleen and lymph nodes of vaccinated mice.
The cell cycle of antigen-specific T cells in vivo has been examined by using a few methods, all of which possess some limitations. Bromodeoxyuridine (BrdU) marks cells that are in or recently completed S-phase, and carboxyfluorescein succinimidyl ester (CFSE) detects daughter cells after division. However, these dyes do not allow identification of the cell cycle phase at the time of analysis. An alternative approach is to exploit Ki67, a marker that is highly expressed by cells in all phases of the cell cycle except the quiescent phase G0. Unfortunately, Ki67 does not allow further differentiation as it does not separate cells in S-phase that are committed to mitosis from those in G1 that can remain in this phase, proceed into cycling, or move into G0.
Here, we describe a flow cytometric method for capturing a "snapshot" of T cells in different cell cycle phases in mouse secondary lymphoid organs. The method combines Ki67 and DNA staining with major histocompatibility complex (MHC)-peptide-multimer staining and an innovative gating strategy, allowing us to successfully differentiate between antigen-specific CD8 T cells in G0, in G1 and in S-G2/M phases of the cell cycle in the spleen and draining lymph nodes of mice after vaccination with viral vectors carrying the model antigen gag of human immunodeficiency virus (HIV)-1.
Critical steps of the method were the choice of the DNA dye and the gating strategy to increase the assay sensitivity and to include highly activated/proliferating antigen-specific T cells that would have been missed by current criteria of analysis. The DNA dye, Hoechst 33342, enabled us to obtain a high-quality discrimination of the G0/G1 and G2/M DNA peaks, while preserving membrane and intracellular staining. The method has great potential to increase knowledge about T cell response in vivo and to improve immuno-monitoring analysis.
Naïve T cells undergo clonal expansion and differentiation upon antigen-priming. Differentiated T cells display effector functions that are essential for antigen clearance and for the maintenance of antigen-specific memory, which is key for long-lasting protection. During the first steps of the primary response, naïve T cell interaction with antigen-presenting cells (APCs) within specialized niches in lymphoid organs is critical to induce the huge T cell proliferation that characterizes the clonal expansion phase1,2,3. T cell-APC interaction is finely regulated by concentration and persistence of antigen, co-stimulatory signals, and soluble factors (cytokines and chemokines) that influence the quantity and quality of the T cell clonal progeny4,5,6,7.
Despite intensive studies of T cell clonal expansion, it is still not known whether antigen-primed T cells complete their entire cell cycle at the site of antigen recognition, or whether they migrate to other organs during cell cycle progression. This lack of knowledge is due to the properties of available tools for cell cycle analysis. These include monoclonal antibodies (mAbs) specific for the nuclear marker, Ki67, and cell dyes that either identify cells that have undergone the S-phase of the cell cycle (e.g., Bromodeoxyuridine (BrdU)) or discriminate among daughter cells and their ancestors (e.g., carboxyfluorescein succinimidyl ester (CFSE)).
However, cell-labeling dyes, such as CFSE and BrdU, do not allow the determination as to whether cells found in a particular organ proliferated locally or rather migrated to this site after division8,9. Moreover, the intranuclear protein, Ki67, is only able to distinguish cells in G0 (Ki67-negative cells) from those in any other cell cycle phase (Ki67-positive cells). Thus, Ki67 analysis does not distinguish cells in active proliferation (i.e., in S, G2, or M) from those in G1, which may either quickly progress to division or stay for long periods in G1 or revert to quiescence10,11.
Here, we describe a new flow cytometric method for cell cycle analysis of antigen-specific CD8 T cells12 from the spleen and lymph nodes (LNs) of vaccinated mice (Figure 1). The method exploits a combination of Ki67 and DNA staining that was previously used to analyze the cell cycle of mouse bone marrow (BM) hematopoietic cells13,14. Here, we successfully applied Ki67 plus DNA staining, together with the recently published innovative gating strategy12, to the analysis of CD8 T cell clonal expansion. We were able to clearly discriminate between antigen-specific CD8 T cells in G0, in G1, and in S-G2/M phases in the spleen and draining LNs of vaccinated mice.
Mice were housed at Plaisant Animal Facility, and the work was performed under Italian Ministry of Health authorization number 1065/2015-PR. The protocol followed the animal care guidelines according to national and international laws and policies (UE Directive 2010/63/UE; Italian Legislative Decree 26/2014).
1. Preparation of medium and staining solution
2. Mouse treatment
3. Isolation of draining LN, spleen, and BM cells
4. Staining of spleen, LN, and BM cells
5. Fixation/permeabilization
6. Intracellular staining
7. Preparation of compensation bead samples
8. Instrument and compensation setup and experimental sample acquisition at the flow cytometer
NOTE: Refer to flow cytometer settings (Table 2) for the cytometer configuration.
9. Data Analysis
The cell cycle phases of cells from spleen, LNs, and BM of Balb/c mice were analyzed using the fluorescent DNA dye, Hoechst, and an anti-Ki67 mAb, according to the protocol summarized in Figure 1. This staining allowed the differentiation of cells in the following phases of cell cycle: G0 (Ki67neg, with 2N of DNA defined as DNAlow), G1 (Ki67pos, DNAlow), and S-G2/M (Ki67pos, with a DNA content comprised between 2N and 4N, or equal to 4N of DNA defined as DNAintermediate/high).
We first performed cell cycle analysis of BM cells to reproduce previously published results13,14 and then analyzed the cells of interest, i.e., CD8 T cells. Figure 2 shows a typical example of cell cycle analysis of BM cells (Figure 2A). The protocol yielded a low coefficient of variation (CV) of G0/G1 and G2/M DNA peaks, indicating the excellent quality of the DNA staining (Figure 2B, showing an example with CV < 2.5; CV was always < 5 in all the experiments).
We then applied the same protocol to antigen-specific CD8 T cells from vaccinated mice. BALB/c mice were vaccinated against the antigen gag of HIV-1 by using Chad3-gag for priming and MVA-gag for boosting, both engineered to carry HIV-1 gag. At day (d) 3 post-boost, we analyzed the frequency of gag-specific CD8 T cells from the spleen and draining LNs. We took advantage of the newly defined gating strategy for T cells in the early phase of immune response, which in contrast to the conventional strategy, is appropriate for detecting highly activated antigen-responding CD8 T cells12. We executed the novel strategy in five subsequent steps. In step 1, we excluded doublets or aggregates by DNA-A/ -W gate, and in step 2, we identified live cells by dead cell marker exclusion. In step 3, we identified the population of interest using a non-conventional "relaxed" FSC-A/ SSC-A gate (Figure 3A) instead of the canonical narrow lymphocyte gate12. After gating on CD3+CD8+ cells (step 4 of Figure 3A), we identified gag-specific CD8 T cells by using two different MHC multimers, i.e., Pent-gag and Tetr-gag (step 5 of Figure 3A). We used two multimers instead of one to improve the sensitivity of gag-specific CD8 T cell detection in vaccinated mice, without increasing the staining background in untreated mice (Figure 3B and C, step 5). Thus, we successfully distinguished untreated mice (0.00% and 0.00% antigen-specific CD8 T cells in LNs and spleen, respectively) from vaccinated mice (0.46% and 0.29% antigen-specific CD8 T cells in LNs and spleen, respectively, Figure 3B and C).
Notably, the protocol allowed us to have an extremely low background in the antigen-specific CD8 T cell gate of LNs and spleen of untreated mice (usually 0.00% and at maximum 0.02%). The comparison of gag-specific and not gag-specific FSC-A / SSC-A plots showed that the gag-specific cells had high SSC-A and FSC-A (Figure 3D), confirming the need to use a "relaxed" FSC-A/ SSC-A gate to capture these cells. We then evaluated the percentages of gag-specific CD8 T cells in different cell cycle phases (Figure 4A). We found that gag-specific CD8 T cells in the spleen and even more in the draining LNs contained a high proportion of cells in S-G2/M phases at day 3 post-boost (18.60% and 33.52%, respectively).
Furthermore, we found that gag-specific CD8 T cells in S-G2/M phases had high FSC-A and SSC-A, when overlaid onto the total CD8 T cells from the same organ (Figure 4B). CD62L expression by gag-specific CD8 T cells was low, as expected for activated T cells, except for a few cells in G0 in the LNs (Figure 4C). Altogether, these results confirmed that the "relaxed" gate (step 3 of Figure 3A, B, and C) was required to include all of the proliferating antigen-specific CD8 T cells12. The protocol was extremely valuable for a "snapshot" evaluation of cell cycle phases of antigen-specific CD8 T cells at the time of analysis and of CD62L expression by cells in different cell cycle phases.
Figure 1: Scheme of the protocol for cell cycle analysis of antigen-specific CD8 T cells. Please click here to view a larger version of this figure.
Figure 2: Cell cycle analysis of BM cells. BM cells from untreated Balb/c mice were stained and analyzed by flow cytometry. (A) Example of gating strategy. We gated on single cells in the DNA-A/-W plot (left) and subsequently on live cells by dead cell dye exclusion (middle). Then, a "relaxed" FSC-A/SSC-A gate was used for all BM cells (right). (B) Example of cell cycle analysis of BM cells (left). We used a combination of Ki67 and DNA staining to identify cells in the following phases of cell cycle: G0 (bottom left quadrant, Ki67neg-DNAlow cells), G1 (top left quadrant, Ki67pos-DNAlow), S-G2/M (top right quadrant, Ki67pos-DNAintermediate/high). Fluorescence Minus One (FMO) control of Ki67 mAb (middle) and DNA histogram (right) are shown. In the DNA histogram plot, the left and right gates correspond to the G0/G1 and the G2/M DNA peak, respectively, and the numbers represent the coefficients of variation (CV) of each peak. In all the other plots, the numbers represent cell percentages in the indicated gates. The figure shows 1 representative experiment out of 5. In each experiment, we analyzed pooled BM cells from 3 mice. Please click here to view a larger version of this figure.
Figure 3: Analysis of antigen-specific CD8 T cells from LNs and spleen. Balb/c mice were primed intramuscularly (i.m.) with Chad3-gag and boosted i.m. with MVA-gag. At day 3 post-boost, draining LN and spleen cells from vaccinated and untreated control mice were stained and analyzed by flow cytometry. (A) Scheme of the gating strategy in five steps to identify single cells (Step 1); live cells (Step 2); lymphocytes (Step 3); CD8 T cells (Step 4); and gag-specific cells (Step 5). (B–C) Example of plots: analysis of cells from (B) LNs and (C) spleen of untreated (top) and vaccinated (bottom) mice. We identified single cells on the DNA-A/ -W plot in Step 1. Then, in Step 2, we selected live cells by dead cell dye exclusion. In Step 3, we used a non-canonical "relaxed" gate for lymphocytes. In Step 4, we identified CD8 T cells by their double expression of CD3 and CD8. We then identified gag-specific cells and not gag-specific in Step 5, based on their capacity to bind fluorochrome-labelled H-2kd-gag-Pentamer (Pent-gag) and H-2kd-gag-Tetramer (Tetr-gag), or not, respectively. (D) FSC-A/SSC-A profiles of gag-specific (blue) and not gag-specific (grey) cells after gating as described above. Numbers represent cell percentages in the indicated gates. The figure shows 1 representative experiment out of 5. In each experiment, we analyzed pooled spleen and pooled LN cells from 3 vaccinated mice and 3 untreated mice. Please click here to view a larger version of this figure.
Figure 4: Cell cycle analysis of antigen-specific CD8 T cells. Mice were vaccinated as in Figure 3 and cell cycle analysis of gag-specific cells was performed at day 3 post-boost, after gating in 5 steps as in Figure 3. (A) Example of cell cycle analysis of gag-specific CD8 T cells from LNs (top) and spleen (bottom) of vaccinated mice. Cell cycle phases were identified as in Figure 2B. The panels represent cells in G0, in G1, and in S-G2/M (left) and Fluorescence Minus One (FMO) control of Ki67 mAb (right). Numbers represent cell percentages in the indicated gates. (B) FSC-A/SSC-A dot plots showing gag-specific CD8 T cells in S-G2/M phases (in red) overlaid onto total CD3+CD8+ T cells (in grey) from LNs (top) and spleen (bottom) of vaccinated mice. (C) Offset histograms showing CD62L expression by gag-specific CD8 T cells in G0 (green), in G1 (blue), and in S-G2/M (red) from LNs (top) and spleen (bottom) of vaccinated mice. The y-axes indicate normalized number of events. The figure shows 1 representative example out of 5 independent experiments with a total of 15 mice. Please click here to view a larger version of this figure.
Supplementary Material: Flow cytometer settings. Please click here to download this file.
Although T cell clonal expansion has been intensively studied, some aspects remain unknown, mostly because the tools available to investigate it are few and have their own drawbacks. From this perspective, we set up a highly sensitive flow cytometric method to analyze cell cycle of antigen-specific CD8 T cells at early times after vaccination in a mouse model. The protocol is based on a combination of Ki67 and DNA staining, which was previously used to analyze the cell cycle of BM hematopoietic cells in mice13,14. To adapt the protocol to antigen-specific CD8 T cells, we had to consider a few critical issues, including the choice of the DNA dye, the appropriate conditions to obtain comparable DNA staining across different samples, and the gating strategy for data analysis.
Many dyes are available for DNA staining, including propidium Iodide and 7-aminoactinomycin D; we chose Hoechst because it was compatible with membrane staining and the mild fixation / permeabilization protocol required for Ki67 staining. At the same time, staining with Hoechst allowed us to obtain a DNA histogram of excellent quality, i.e., the G0/G1 and G2/M DNA peaks had a much lower coefficient of variation (CV) than DNA peaks usually obtained with other DNA dyes, e.g., DRAQ519. Indeed, Hoechst can stain DNA even in live cells20.
Some strategies were used to avoid the fluctuation in Hoechst intensity in different samples of the same experiment. Hoechst staining was performed just before sample acquisition at the flow cytometer to minimize the decline of dye intensity during time. For those interested in reproducing the protocol in big experiments with numerous samples, we recommend performing Hoechst staining on a few samples at a time. One other drawback is that Hoechst intensity can be heavily influenced by cell number during incubation with the dye. For this reason, we strongly recommend always using the same number of cells and the same volume per sample for DNA staining. If a high number of cells is required for acquisition at the flow cytometer, we recommend preparing two or more identical samples and then merging them just before the Hoechst staining step.
A key point of the protocol is the gating strategy for data analysis. We recently published a novel strategy for T cell analysis at early times of the immune response, which allowed us to increase the sensitivity of detection of antigen-specific T cells12. We applied this strategy to the data shown here as follows. First, we excluded cell aggregates in the DNA-A/W plot. Second, after gating out dead cells, we used a fairly large lymphocyte gate in the FSC/SSC plot ("relaxed gate"). By this strategy, we were able to include highly activated antigen-specific CD8 T cells in S-G2/M that are usually missed by current gating strategies, as these cells have high FSC-A and SSC-A. In summary, the data analysis represents a critical part of the method, which is essential to obtain a sensitive detection of activated /proliferating antigen-specific T cells.
The method prevents the possibility of missing critical T cell data at early phases of immune response and opens new perspectives for T cell immuno-monitoring. A future improvement might be to include staining for phospho-histone 3 that would allow differentiation between G2 and M21. A current limitation is that cells have to be fixed and permeabilized to stain for the nuclear marker, Ki67. Thus, cells cannot be used for other types of analysis such as sorting and subsequent functional analysis. Moreover, DNA dyes, including Hoechst, usually interfere with the genomic DNA analysis and are not suitable for this type of evaluation. Identification of membrane markers that correlate with different cell cycle phases and that can be stained on live cells could overcome this limitation. In conclusion, the method has great potential for the evaluation of activated/proliferating T cells in several contexts such as vaccination, infection, immuno-mediated diseases, and immuno-therapy.
The authors have nothing to disclose.
This work was supported by Reithera, by MIUR project 2017K55HLC_006, and by 5 × 1000 grant from Associazione Italiana Ricerca sul Cancro (AIRC). The following tetramer was obtained through the NIH Tetramer Facility: APC-conjugated H-2K (d) HIV gag 197-205 AMQMLKETI.
1-200 μL universal fit bulk packed pipet tips | Corning | CLS4866-1000EA | |
2.4G2 anti-FcγR mAb | BD | 553141 | 10 μg/ml final concentration |
5 ml syringe plunger | BD Emerald | 307733 | |
15 ml conical tubes | MercK Millipore | SBHA025SB | |
60 mm TC-treated Cell Culture Dish | Falcon | 353002 | |
70 μm cell strainer | Falcon | 352097 | |
96-well Clear Round Bottom TC-treated Culture Microplate | Falcon | 353077 | |
Anti-Rat/Hamster Ig,k/Negative Control Compensation Particles | BD- Bioscience | 552845 | |
Beta-mercaptoethanol | Sigma | M3148 | |
Bovine Serum Albumin | Sigma | A07030 | |
BUV805 Rat Anti-Mouse CD8a | BD- Bioscience | 564920 | 4 μg/ml final concentration |
Dulbecco's Phosphate Buffer Saline w/o Calcium w/o Magnesium | Euroclone | ECB4004L | |
Eppendorf Safe-Lock Tubes, 1.5 mL | Eppendorf | 30120159 | |
Ethanol | Sigma | 34852-1L-M | |
Ethylenediaminetetraacetic Acid Disodium Salt solution (EDTA) | Sigma | E7889 | |
Fetal Bovine Serum | Corning | 35-079-CV | |
Filcon, Sterile, Syringe-Type 70 μm | Falcon | 352350 | |
Fixable Viability Dye eFluor 780 | eBioscience | 65-0865-14 | 1:1000 final concentration |
Foxp3 / Transcription Factor Staining Buffer Set | eBioscience | 00-5523-00 | This Set contains fixation/permeabilization concentrate and diluent, and permeabilization buffer 10x |
H-2k(d) AMQMLKETI allophycocyanin (APC)-labelled tetramer | provided by NIH Tetramer Core Facility | 6 μg/ml final concentration | |
H-2k(d) AMQMLKETI phycoerythrine (PE) labelled pentamer | Proimmune | F176-2A-E – 176 | 10 μL / sample |
Hoechst 33342, Trihydrochloride, Trihydrate – 10 mg/mL Solution in Water | ThermoFisher | H3570 | |
Ki-67 Monoclonal Antibody (SolA15), FITC | eBioscience | 11-5698-82 | 5 μg/ml final concentration |
L-Glutamine 100X (200 mM) | Euroclone | ECB3000D | |
Millex-HA Filters 0,45 µm | BD | 340606 | |
Penicillin/Streptomycin 100X | Euroclone | ECB3001D | |
PE/Cyanine7 anti-mouse CD62L Antibody | Biolegend | 104418 | 0.2 μg/ml final concentration |
PerCP-Cy™5.5 Hamster Anti-Mouse CD3e | BD- Bioscience | 551163 | 4.4 μg/ml final concentration |
Red Blood Cell Lysis Buffer | Sigma | R7757 | |
Round-Bottom Polystyrene Tubes, 5 mL | Falcon | 352058 | |
RPMI 1640 Medium without L-Glutamine with Phenol Red | Euroclone | ECB9006L | |
Software package for analyzing flow cytometry data | FlowJo | v.10 | |
Software for acquisition of samples at flowcytometer | BD FACSDiva | v 6.2 | |
Trypan Blue Solution | Euroclone | ECM0990D |