Flow Cytometry Protocol for Tracking Immune Cell

Materials of Flow Cytometry Protocol for Tracking Immune Cell

Figure 1. Cell Isolation and Cell Culture. (Bleichrodt Robert-Jan,et al.; 2019)

1. Complete medium (CM). RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum (FBS), 25 mM HEPES, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 2 mM fresh glutamine, and 50 mg/mL gentamicin sulfate and 5 × 10⁻⁵ M b-mercaptoethanol.
2. Phosphate-buffered saline (PBS). Prepare 10× stock with 1.37 M NaCl, 27 mM KCl, 100 mM Na₂HPO₄, and 18 mM KH₂PO₄. Adjust to pH 7.4 with HCl if necessary. Sterilize by 0.2-mm filtration and store at room temperature. Prepare working solution by diluting one part with nine parts of tissue culture grade water.
3. 10% Acid citrate dextrose in PBS.
4. Formaldehyde, 10%, methanol free, ultrapure. Dilute to 2% in PBS (pH 7.4) and store at 4–8°C.
5. Hanks’ balanced salt solution (HBSS) without phenol red, magnesium- and calcium- free. Store at room temperature until opened, then at 4–8°C.
6. Histopaque^®-1077. Store at 4–8°C and use at room temperature.
7. IL-2. Dilute stock (2.2 × 10⁶ IU/ mL) in sterile HBSS to 1 × 10⁵ IU/mL and store at −80°C. Do not refreeze after thawing; store at 4–8°C and discard thawed product after 7 days.
8. Phytohemagglutinin-HA (PHA). Prepare stock of 10 mg/mL in CM from powder, sterilize by 0.2-mm filtration, and store frozen at −80°C in single-use aliquots. To induce polyclonal T-cell proliferation, incubate human peripheral blood mononuclear cells (hPBMC) at a final PHA concentration of 5 mg/mL.
9. TRIMA filters. WBC-retaining filters were obtained from the pheresis facility at Roswell Park Cancer Institute, after informed consent from donors, and used to isolate hPBMC in quantity for some of the studies described here.
10. Versene, 0.48 mM EDTA·4Na in PBS. Adjust to pH 7.4 with HCl if necessary.
11. K562 cell line. Also available for purchase, #CCL-243.

Antibodies

1. Anti-CD3 (clone OKT3) and anti-CD28. Azidefree, unconjugated preparations, each at a concentration of 1.0 mg/mL.
2. Fluorochrome-conjugated monoclonal antibodies (mAbs). CD4 phycoerythrin cyanine 7, CD25 allophycocyanin, and CD127 phycoerythrin; CD45 Pacific Blue.

Flow Cytometry Reagents

1. 7-Aminoactinomycin D (7-AAD). Reconstitute powdered solid to 1 mg/mL in PBS and store at −20°C. Make a weekly working stock by diluting thawed 1-mg/mL stock to 100 mg/mL in PBS and store at 4–8°C. Add 4 mL of working stock to each 100 mL of cells (4 mg/mL final) and allow the cells to stand on ice for 30 min prior to data acquisition.
2. FCM buffer. PBS (pH 7.2) supplemented with 1% BSA, 0.1% sodium azide, and 40 mg/mL tetrasodium ethylenediaminetetraacetic acid.
3. Human IgG block. Reconstitute human IgG Cohn fraction II and III globulins to 12 mg/mL in RPMI 1640 supplemented with 25 mM HEPES, 20 mg/mL gentamicin sulfate, and 2 mg/mL BSA. Store frozen at −20°C until use. Once thawed, store at 4–8°C for no longer than 1 month.
4. LIVE/DEAD^®Fixable Violet. Reconstitute with DMSO according to the manufacturer’s instructions. Store frozen at −20°C for no longer than 6 months. Thaw a fresh aliquot daily and dilute 1:50 in PBS. Add 5 mL per test and incubate for 30 min in a buffer free of exogenous protein before washing and fixing in 2% formaldehyde for assessment of viability by flow cytometry.
5. Spherotech AccuCount Ultra-Rainbow Fluorescent Particles. These are used for single platform enumeration of absolute cell numbers by flow cytometry.

Cell Tracking Dyes

1. CellVue^®Claret, PKH26, and PKH67 fluorescent cell linker kits. The kits contain 1 mM of stock solutions in ethanol and cell-labeling diluent for general cell membrane labeling. Store tightly capped at room temperature to avoid evaporation of ethanol and associated increases in dye concentration. If any dye solids are visible, sonicate the dye stocks to resolubilize before use and verify that dye absorbance remains within the range specified in the Certificate of Analysis available for each kit.
2. 2. 5-(and-6)-carboxyfluorescein diacetate, succinimidyl ester. As the nonfluorescent CFDA precursor diffuses across the plasma membrane into the cytoplasm, its acetate substituents are cleaved by nonspecific esterases, forming the fluorescent amino-reactive product, carboxyfluorescein succinimidyl ester.

Flow Cytometer and Other Equipment

1. For routine data acquisition, any flow cytometer capable of acquiring forward and side scatter, PacBlue, FITC, PE, PerCP/PECy5, PECy7, and APC would be appropriate. The data were acquired using an LSRII flow cytometer fitted with a 25-mW 407-nm violet diode laser, a 20-mW 488-nm blue optically pumped semiconductor laser, and a 20-mW 635-nm HeNe laser. From the 407-nm laser, PacBlue or LIVE/DEAD Fixable Violet was detected using a 450/50-nm bandpass filter; from the 488-nm laser, CFSE or PKH67, PKH26, 7-AAD, and CD4 PECy7 fluorescence were detected using 530/30-, 575/26-, 685/35-, and 780/60-nm bandpass filters, respectively. CellVue Claret fluorescence was excited using the 633-nm line and collected using a 660/20-nm bandpass filter.
2. For sorting experiments, any flow cytometer capable of collecting peak area, width, and height and sorting based on forward and side scatter, PE, PECy7, and APC would be appropriate. The data were acquired on a FACSAria II flow cytometer fitted with a 100-mW 355-nm solid state UV laser, a 100-mW 405-nm violet diode laser, a 50-mW 488-nm blue optically pumped semiconductor laser, and a 40-mW 639-nm red diode laser. From the 488-nm laser, CD127 PE and CD4 PECy7 fluorescence were detected using 575/26-nm and 780/60-nm bandpass filters, respectively. CD25 APC was excited using the 639-nm line and collected using a 660/20-nm bandpass filter.
3. A tube rotator was used for monocyte depletion of hPBMC and preparation of accessory cells.

Notes of Flow Cytometry Protocol for Tracking Immune Cell

1. Lymphocytes and monocytes are typically isolated from anticoagulated blood using standard Ficoll-Hypaque density centrifugation techniques prior to labeling, but cryopreserved PBMCs, adherent cell lines (harvested using trypsinization), and non-adherent lines are also suitable for staining. Cells may be labeled while adherent by flooding the culture dish or flask with dye solution. However, this typically gives considerably more heterogeneous intensity distributions, especially for membrane dyes, and makes their interpretation in dye dilution proliferation assays more complex. Labeling of single cell suspensions is, therefore, generally preferred.
2. Labeled cells are typically placed back into culture for in vitro assays or injected into animal models for in vivo functional studies. Standard sterile technique should, therefore, be followed throughout the labeling protocols.
3. Amount of dye required for bright but nontoxic staining will in general increase as total number and/or size of cells to be stained increases. However, exact concentrations resulting in over-labeling and loss of function will vary depending on cell type and class of tracking dye used. Therefore, appropriateness of final cell concentration and final tracking dye concentration used for labeling should always be verified by comparing viability and functionality of labeled versus unlabeled cells. Similarly, both cell and dye concentrations used for labeling should be reported in any publication.
4. Total number of cells to be stained will depend on the number of replicates and controls required by the experimental protocol. Staining intensities are most easily reproduced when staining is done in volumes ranging from 0.5 to several milliliters. Once an approximate cell concentration has been established based on these factors, a preliminary dye titration experiment is recommended to determine (or verify) the optimal concentration of the tracking dye.
5. Exogenous protein reduces labeling efficiency for both protein and membrane dyes and is, therefore, normally removed by washing the cells with a protein-free buffer such as PBS or HBSS prior to staining. However, when labeling must be done at relatively low cell concentrations due to limited number of cells or other experimental concerns, addition of exogenous protein may aid in avoiding over-labeling and resultant loss of cell viability or functionality.
6. Selection of proliferation tracking dye(s) for a given study is typically based on spectral compatibility with other fluorochromes to be used in the study, and the ability to achieve acceptable starting intensity without adverse effect on function. CFSE, PKH26, and PKH67 can be excited using a 488-nm laser line (absorption maxima: 492, 551, and 490 nm, respectively) and have emission maxima of 516, 565, and 504 nm, respectively. CellVue Claret (absorption maximum: 654 nm) can be excited using red diode or HeNe laser lines at 633–635 nm or a 647-nm Kr laser line and has an emission maximum of 677 nm.
7. CFSE is highly lipophilic but poorly soluble in ethanol or other polar organic solvents. If CFSE is purchased as a bulk powder, it should be accurately weighed out and made up as a 5-mM stock solution (MW 557.47 g/mol) in freshly opened anhydrous DMSO. Although the entire contents of dye vial can be dissolved in a calculated volume of DMSO, the final dye concentration should be confirmed spectrophotometrically (e.g., by absorption at 490 nm) and adjusted as needed for consistency, since exact weights contained may otherwise vary sufficiently from vial to vial to require re-titration of new versus old dye stocks in order to avoid toxicity.
8. Aliquots of 5 mM CFSE dye stock in DMSO can be stored in a dessicator at −20°C for several months but repeated freezing and thawing of a given aliquot should be avoided since DMSO is hygroscopic and takes up moisture from the air. The presence of water leads to reduced labeling efficiency due to hydrolysis of both the diacetate ester moieties required for entry into cells and the succinimidyl ester moieties required for covalent reaction with amino groups under physiologic conditions.
9. If the necessary weighing or spectrophotometric instrumentation is not available in the laboratory, by adding anhydrous DMSO to commercially available singleuse vials containing pre-weighed amounts of CFSE. However, it should be noted that cost per milligram of dye is greater for such vials than for dye purchased in bulk powder form.
10. When staining cells at concentrations<1 × 10⁷/mL, inclusion of exogenous protein (e.g., 5% v/v FBS or 1% v/v serum albumin) in the resuspension buffer is suggested to avoid over-labeling and loss of cell function (see Notes 5 and 14). Resuspension in a serum-free culture medium will also reduce labeling efficiency and potential for over-labeling, due to the presence of free amino acids that compete for reaction with CFSE. Alternatively, if addition of exogenous protein must be avoided due to other experimental considerations. The time between initial dilution and initiation of cell labeling should be minimized since hydrolysis begins immediately upon dilution of the DMSO stock into aqueous solution and proceeds very rapidly.
11. Platelets present in variable amounts act as “hidden” sources of added protein or membrane that can affect labeling efficiency even when hPBMC and dye concentrations are carefully reproduced. Addition of a final low-speed wash step (5 min at 300 × g) minimizes platelet contamination of hPBMC and improves consistency of staining with both protein- and membrane-labeling dyes.
12. Ensure that the 5 mM CFSE stock in DMSO is completely thawed prior to preparation of the working stock, but minimize the length of time that the DMSO stock is exposed to ambient conditions to limit the uptake of moisture. The CFSE working stock solution should be clear and colorless. If there is any sign of yellowing, it should not be used, since this indicates conversion to carboxyfluorescein, the charged fluorescent hydrolysis product which will not enter cells.
13. This concentration was chosen such that following a 24-h stabilization period, the fluorescence intensity of non-dividing lymphocytes should fall within the third and fourth decade of a four-decade log amplifier when unstained cells are placed in the first decade (see also Notes 20–22).
14. For an hPBMC concentration of 1 × 10⁷/mL, staining at a final concentration of 0.5–1.0 mM CFSE is recommended to avoid over-labeling. CFSE labels proteins indiscriminately and if the function of critical residues is modified by labeling, it can interfere with signal transduction pathways, proliferation, and other cell functions even when cell viability remains acceptable. More extensive labeling increases the likelihood of altered cell function(s) and the extent of labeling is a function of dye concentration, cell concentration, labeling time, and labeling conditions (temperature, mixing, etc.). Concentrations given here for both cells and dye should, therefore, be taken only as a starting point and verified in each user’s experimental system.
15. Obtaining reproducible starting intensities from study to study requires accurately reproducing both dye and cell concentrations. Therefore, cell counting using a Coulter Counter or other automated cell counter rather than manual counting using a hemocytometer is recommended, since results of replicate hemocytometer counts often vary by as much as 15–20%.
16. Since uptake of CFSE into cells and reaction with free amino groups occur very rapidly, it is critical to disperse the dye solution quickly and evenly throughout the cell suspension immediately after addition.
17. Once formed by hydrolysis, carboxyfluorescein is sensitive to photobleaching. Therefore, covering with aluminum foil or placing in a dark location is recommended to protect tubes or wells containing CFSE-labeled cells from exposure to high intensity light or prolonged exposure to room light.
18. Inclusion of protein in the stop solution is essential, since it reacts with and inactivates free CFSE. Free amino acids in culture medium further aid in the inactivation. Alternatively, PBS or HBSS containing 1–2% serum albumin may be used as a stop solution.
19. For starting cell numbers of 10⁷ or more, recoveries of at least 85% and viabilities of at least 90% should be obtained for freshly drawn hPBMC. However, recoveries typically decrease at lower cell numbers and may also be lower for preparations in which the cells are older or have been subjected to other stresses (e.g., pheresis, elutriation, or cryopreservation and thawing). Staining intensity and CV will vary for different cell types, but a bright symmetrical fluorescence intensity profile coupled with poor recovery and/or viability usually indicates substantial overlabeling and the need to increase cell concentration, decrease dye concentration, or both. Conversely, heterogeneous and/ or dim staining (<2 log separation from unstained control) coupled with good recovery and viability suggests underlabeling and the need to decrease cell concentration, increase dye concentration, or both.
20. The great majority of cell-associated CFSE is lost within the first 24–48 h as cells export and/or degrade unreacted dye and labeled proteins/peptides that are short-lived, damaged by the covalent labeling process, destined for secretion, etc. This typically results in (a) up to a 1-log decrease in mean fluorescence intensity and (b) an approximately 5–10% decrease in CV, since T0 intensities reflect primarily variations in cell size, whereas T24–48 intensities reflect primarily variations in cellular content of stable long-lived proteins.
21. Some labeling protocols recommend a washout period in which freshly stained cells are incubated in CM at 37°C for 30–60 min to hasten fluorescence stabilization. Our experience has been that a relatively short washout period of this type is not sufficient to prevent transfer of exported dye to unlabeled cells present in a co-culture, suggesting that labeled cells continue to export labeled proteins beyond the washout period. Therefore, at least a 24-h stabilization period in culture is recommended before labeled cells are used for in vitro cytotoxicity or proliferation assays. For in vivo cytotoxicity or proliferation assays, if labeled cells are to be reinfused within less than 24 h post-staining, it is critical to include control animals in which minimal killing or proliferation is expected in order to verify that ex vivo labeling conditions resulted in adequate resolution between stably labeled cells and unlabeled cells.
22. T₀ CFSE-labeled samples are NOT appropriate for use as compensation controls because they are so much brighter than T24 samples that they typically cannot be run on the same intensity scale when unstained cells are placed within the first decade. For proliferation assays based on CFSE dye dilution, it is, therefore, critical to select a staining concentration that gives adequate separation from unstained cells at 24 h without unacceptable fluorescence overlap into spectral channels used to detect other reagents. It is also essential to ensure that sufficient cells are prepared to provide unstimulated T24 samples for use as compensation controls. For an example of typical controls to set up for a proliferation assay. Inability to achieve adequate color compensation for a CFSE-labeled but unstimulated (i.e., non-proliferating) T24 sample indicates the need to reduce CFSE concentration, increase cell concentration, or both during the staining step, while recognizing that this may reduce the number of daughter cell generations that can be resolved from unstained cells.
23. Fixation of CFSE-labeled cells in EtOH or methanolfree formaldehyde leads to further loss of cell-asso-ciated CFSE (30–50% decrease in fluorescence intensity), most likely due to leakage of small but stably labeled peptides or proteins out of the cells as membrane permeabilization occurs. Fortunately, the decrease from fresh to fixed cells does not appear to affect the shape of dye dilution profiles, which can therefore still be used to deduce cell proliferation history so long as the weaker fluorescence of fixed cells does not compromise the ability to resolve the desired number of generations from unstained cells.
24. For dye dilution proliferation assays using hPBMC (or PBMC from other species), it may be necessary to use an independent method such as 3H-thymidine incorporation to verify that cell function is unaltered by labeling with tracking dye at the chosen concentration.
25. Like most membrane-intercalating dyes, PKH26, PKH67, and CellVue Claret contain not only aromatic chromophores but also lipophilic alkyl tails, all of which readily adsorb to the walls of polystyrene tubes or plates. This can substantially reduce labeling efficiency, particularly when working at relatively low dye concentrations (2 mM or less). Use of polypropylene tubes is strongly recommended to minimize adsorptive dye loss and maximize reproducibility of labeling.
26. The PKH and CellVue dyes are incorporated into membranes based on strong hydrophobic forces that drive partitioning from the aqueous phase, in which the dyes are highly insoluble, into cell membranes, where they remain stably intercalated due to noncovalent interactions with the lipid bilayer. This means the final staining intensity is strongly affected by both the amount of dye and the amount of membrane present in the staining step. Dye concentrations required for bright but nonperturbing staining, therefore, vary with cell type/size as well as cell concentration, making it important to publish and accurately reproduce both dye and cell concentrations in order to obtain reproducible results.
27. Membrane-intercalating dyes are even more lipophilic than CFSE due to long chain alkyl moieties that mediate stable insertion into the membrane bilayer. Although the dyes are modestly soluble in polar organic solvents such as ethanol, the alkyl chains tend to self-associate in aqueous solutions, reducing membrane-labeling efficiency. The presence of salts substantially increases the extent of micelle and aggregate formation and decreases general membrane-labeling efficiency, although phagocytic cells can become differentially labeled by taking up dye aggregates. Use of conical rather than roundbottom tubes is highly recommended, since this facilitates more complete removal of salt-containing media or buffers prior to cell labeling.
28. The technique used to remove supernatant prior to resuspension in Diluent C or during wash steps can, in our experience, substantially impact both quality of staining and cell recovery when labeling with membrane dyes. Tube inversion and blotting typically leaves ~100 mL of supernatant. Subsequent aspiration to reduce the amount of fluid risks loss of cells at the top of the pellet that have been loosened as the fluid drains back down the side of the tube. The best method that we have observed for maximizing fluid removal while minimizing cell loss is to fit a sterile disposable 200-mL pipette tip at the end of a vacuum aspirator, which reduces the aperture size and makes it easier to position the point of suction accurately relative to the cell pellet.
29. The fact that the PKH and CellVue dyes label cells by partitioning into membranes has two other significant consequences beyond the negative effect of salts on staining efficiency. First, staining occurs even more rapidly than for CFSE, with the partitioning of dye into membranes being essentially complete within 15–30 s after admixing 2× cells and 2× dye solutions. Second, to obtain bright, homogeneous labeling, it is critically important that admixing of cells and dye be done so that all cells are exposed to the same concentration of dye at the same time. This is much more easily achieved by admixing similar volumes of dye and cells than by trying to disperse a small volume of ethanolic dye into a much larger volume of cell suspension.
30. Diluent C is an aqueous isotonic, iso-osmotic, salt-free staining vehicle that contains neither organic solvents nor physiologic salts. Although Diluent C is designed to maximize dye solubility and minimize cell toxicity for short periods (5–15 min), the longer cells and dye stand in Diluent C, the more likely that (a) staining efficiency will decrease due to dye aggregation and (b) decreased cell viability or function may result from lack of physiologic salts. When multiple samples are to be labeled, it is highly recommended that processing through the first wash be completed for each sample before the next sample is stained. However, remaining steps may then be carried out in parallel for all samples.
31. In theory, results should be the same whether 2× cells are admixed with 2× dye or vice versa. However, our experience has been that adding cells to dye makes it easier to reliably obtain homogeneous staining. Adding dye to cells runs the risk that any cells on the wall of the tube may remain unstained if they are above the level of admixed solution or stain at lower intensity if they do not mix with bulk staining solution until after trituration is completed.
32. Some literature protocols suggest the use of CM, which contains 10% FBS, as the stop reagent. However, our experience has been that use of neat FBS results in more efficient removal of unbound dye (due to its higher protein concentration) and also reduced likelihood of forming dye aggregates large enough to sediment with cells during subsequent wash steps (due to its lower ionic strength). If CM is used, a larger volume (5× rather than 1×) is recommended to ensure sufficient protein is present to adsorb all unbound dye.
33. Even when polypropylene tubes are used (see Note 25), some dye adsorption to tube walls may occur. Therefore, washing efficiency is improved if cells are transferred to a fresh polypropylene tube after aspiration of the stop solution and resuspension of the cell pellet for the first wash. This is particularly important if CM is used as the stop reagent (see Note 32), since carryover of dye particles may result in inadvertent labeling of other cell types present in the culture.
34. In contrast to CFSE, the fluorescence intensity of PKH26, PKH67, or CellVue Claret labeled cells does not decrease significantly in the absence of cell proliferation, and the intensity is also stable to fixation with neutral methanol-free formaldehyde (e.g., 1–2% paraformaldehyde dissolved in PBS). A small aliquot of cells fixed at T0 can, therefore, be used as a compensation control to evaluate the overlap of membrane dye signal into spectral regions used to detect other reagents. For an example of typical controls to set up for a lymphocyte proliferation assay. As with CFSE, the inability to achieve adequate color compensation indicates the need to reduce dye concentration, increase cell concentration, or both, during the staining step.
35. Paradoxically, our experience and that of others suggest that while over-staining with membrane dyes may not reduce viability or functionality, it can sometimes result in increased T24 fluorescence intensity compared with T0 intensity, either in the absence or presence of cell proliferation. The most likely explanation is stacking and self-quenching of dye in the plasma membrane at T0 that is relieved as the dye redistributes into intracellular membranes via normal membrane trafficking mechanisms. Dye dilution appears to proceed linearly with cell division once stacking/quenching is relieved, presumably because total number of molecules per cell does not change, and membrane dyes are typically pH insensitive in physiologic ranges. However, for assay systems where proliferation monitoring is to be initiated immediately, cell and/or dye concentration(s) in the staining step should be chosen to ensure that dye dilution proceeds linearly from T0. Finally, it should be noted that although CFSE over-labeling may also result in stacking/quenching, any modest increase from T0 to T24 would be completely undetectable in the face of the substantial division-independent dye loss seen during the same period.
36. Regardless of which class of tracking dyes is used, labeling prior to LAK induction with IL-2 is preferable to post-induction labeling. This is particularly true if CFSE or other protein dyes are used, since staining immediately before assay initiation runs the risk that early dye efflux could result in unintended transfer of label to target cells. It is also recommended when labeling with membrane dyes, since it gives the cells time to recover from thermal or other stresses incurred during the staining procedure.
37. Use of target cells obtained from high-density cultures should be avoided, due to the presence of significant numbers of apoptotic and/or dead cells. Such cells will readily stain with tracking dyes but will give highly variable staining with 7-AAD or other dyes used to determine viability by dye exclusion, making it much more difficult to establish the intensity limit above which a target cell is to be considered nonviable.
38. Cell and dye concentrations were selected to give staining intensities that (a) were on scale in the last decade of a 4-log intensity scale when unstained cells were placed in the first decade, and (b) could readily be compensated in adjacent spectral channels. It should be noted that a slight peak asymmetry is of somewhat less concern when labeling targets or effectors for a cytotoxicity assay than when labeling hPBMC for a proliferation assay. It is however, important to ensure that 100% of cells are labeled with the chosen tracking dye so that a clear distinction can be made between cells that are labeled, and cells that are unlabeled or those that are labeled with a different dye.
39. Assay and staining/instrument setup controls, as well as additional controls required at the time of initial assay setup or useful for troubleshooting.
40. For longer term assays, it may be desirable to set up the plate with a border of CM-filled wells around the periphery in order to minimize evaporation-associated variability in cell and cytokine concentrations in the assay wells.
41. CD45 was used here as a gating parameter because both LAK and K562 are CD45+. The PacBlue conjugate was selected here for spectral compatibility with the other probes used in the assay. Because the enumeration beads also fluoresce in the PacBlue channel (450/50), they also will be included in the “CD45” gate. A low SSC threshold was used to reduce debris but required substantial care to ensure that all events corresponding to both dead and live targets and effectors were included above the threshold. An alternative strategy would be to set a threshold on CD45 PacBlue that included all bead and cellular events (live and dead), avoiding the necessity for a side scatter threshold.
42. Choice of optimized labeling conditions for tracking dye(s) should have already established that stained cells will be on scale, without events accumulating in the highest channel. This condition should be established when compensation is set to 0% and voltages are adjusted to place unstained control cells in the first decade, but sufficiently above the left axis such that events do not accumulate in the lowest channel. In addition, it is critical to recognize that cells labeled with tracking dyes can be extremely bright, requiring substantially decreased detector voltage settings compared with those used for the detection of immunofluorescence. If the tracking dye signal in a secondary channel (set to a higher detector voltage) is greater than its intensity in the primary channel, it will be impossible to compensate properly for tracking dye overlap in the secondary channel. In this case, staining with concentrations of tracking dye(s) may be required.
43. CellVue Claret staining does not affect the cytolytic potential of LAK cells. Somewhat surprisingly for a 4-h assay, the use of 7-AAD to quantify dead target cells (Method 1) appears to be less sensitive than the use of enumeration beads to count remaining viable target cells (Method 2). To test the hypothesis that a too-stringent CD45 region or scatter region might have excluded some dead K562 targets from the PKH67 + 7-AAD + gate in Method 1, we evaluated the effect of extending region R1 to include events with lower CD45 intensity, or expanding region R2 to incorporate events with lower forward scatter. These manipulations (especially in combination) gave decreased rather than increased values for percent cytotoxicity. This was true using either Method 1 or 2 and did not reduce the differential between the two. In order to enumerate dead cells with a viability dye, Method 1 requires that cells remain intact enough to contain a nucleus. It has been demonstrated in longer term assays that killed cells may fragment, enucleate, and thus become uncountable by this method. Even in a relatively short-term assay, Method 2 – which relies on enumeration of live cells remaining in the sample – is inherently a more accurate way to quantify cytotoxicity. An alternate hypothesis considered was that Method 1 might be more sensitive to the quadstat region used to determine % dead targets than Method 2, due to a consistent decrease in PKH67 intensity observed for dead versus live K562 targets, which was most pronounced at low E:T ratios. For example, gMFI = 299 for PKH67 + 7-AAD+ events, gMFI = 498 for PKH67 + 7-AAD− events, and approximately 8% of PKH67 + 7-AAD+ events fall in the lower right quadrant (i.e., are PKH67dim). Lowering the quadstat boundary to include ³98% of PKH67dim 7-AAD+, however, actually resulted in decreased values for percent cytotoxicity, not increased values, due to increased levels of noise in the live K562 region. Further analysis of the reduced PKH67 intensity observed for 7-AAD+ targets suggested that it was at least partially due to membrane transfer from K562 targets to LAK effectors, as evidenced by increases in both PKH67 and CD45 intensity observed for live CellVue Claretpositive LAK cells at low E:T ratios.
44. Number of non-viable effectors present in the assay can be calculated similarly, using a CellVue Claret versus 7-AAD histogram and then dividing the number of dead effectors by the total number of effectors. This may be helpful in troubleshooting if target cell killing is lower than expected and/or in longer term assays in which some effector cell death is expected.
45. During the analysis of suppression assays, fluorescent monoclonal antibodies (mAb) such as anti-CD4 can be used to discriminate Teff and Treg from accessory cells, which are required to maximize Teff proliferation if soluble anti-CD3 is used for stimulation rather than plate-bound anti-CD3 (data not shown).
46. Anti-coagulated peripheral blood or apheresis packs used for stem or mononuclear cell collections are suitable sources of both lymphocytes and accessory cells. WBC-retaining TRIMA filters were used as a starting point for the studies described here because they represent a ready source of large numbers of hPBMC (average recovery: 1.2 ± 0.4 × 109 cells; n = 39), an important consideration when using multiparameter flow cytometry and sorting to isolate low-frequency subpopulations such as Treg cells.
47. Be sure to rotate tubes on their horizontal axis to avoid foaming, sloshing from side to side, or end-over-end mixing, which will inhibit monocyte aggregation.
48. Typical lymphocyte recoveries from a TRIMA filter are 7.1 ± 3.1 × 10⁸ cells. Typical composition is 71.8 ± 7.1% T cells, 3.3 ± 3.6% monocytes, and 24.7 ± 6.4% B/NK cells as determined by CD45 FITC/CD14 PE/CD3 PerCP staining and flow cytometric analysis (n = 39).
49. Though not further used in the studies described here, the cell pellet obtained after cold aggregation and 1 × g fractionation is an excellent source of relatively pure monocytes. These may be harvested with a 10-mL pipette, being careful not to disrupt the aggregates. To remove platelets, overlay the aggregates onto 5 mL of FBS and let them to sediment at 1 × g on ice for 15 min. Recover pellet, and wash twice with 15 mL of Versene to remove platelets and once with 15 mL of CM. Typical recoveries are 2.1 ± 1.2 × 108 monocytes with 74 ± 10% purity (n = 39).
50. All procedures should be carried out on ice with cold reagents.
51. All mAbs should be used at concentrations previously determined to be saturating. In general, we have found that the concentration of mAb determined to be saturating for 100 mL of whole blood will also be saturating when staining up to 3–5 × 10⁶ lymphocytes in a volume of 100 mL.
52. After sorting, we routinely obtain 1–4 × 10⁶ Treg and 1–2 × 10⁷ Teff cells from a starting preparation of 1.5 × 10⁸ monocytedepleted lymphocytes. Accessory cells were simultaneously sorted with the Treg and Teff cells until 1.5 × 107 cells had been recovered.
53. Ideally, given sufficient cell numbers, samples should be prepared in triplicate.
54. Assay and staining/ instrument setup controls, as well as additional controls required at the time of initial assay setup or useful for troubleshooting (if sufficient cells are available to accommodate their inclusion).
55. Single-color compensation controls are best prepared from cells treated in the same fashion as the test samples. This is particularly important for CFSE, which exhibits rapid nonproliferation- related intensity losses during the first 24 h and more slowly thereafter. If insufficient numbers of CellVue Claret-stained Treg cells are available for both test and control samples, it is possible, although not ideal, to stain other lymphocytes under identical cell concentration and dye conditions for use as Treg surrogates in establishing appropriate detector voltages and compensation settings.
56. Since many wells in the 96-well plate will remain empty, it is recommended that wells surrounding test and control cells for the assay be filled with CM to minimize evaporation in the assay wells.
57. Although 7-AAD can be used as an alternative to LIVE/ DEAD Fixable Violet reagent, color compensation for CellVue Claret is simpler with the latter since there is greater spectral separation than with 7-AAD. In our experience, labeling cells with LIVE/DEAD Fixable Violet reagent can successfully be performed in the presence of low levels of exogenous protein, although this will reduce the separation between viable and nonviable populations. Here we chose to incubate simultaneously with human IgG, anti-CD4 PECy7, and LIVE/DEAD reagent in order to minimize cell losses associated with sequential staining and additional wash steps.
58. Choice of optimized labeling conditions for both Treg (here CellVue Claret stained) and Teff (here CFSE stained) should have already established that stained cells will be on scale, without events accumulating in the highest channel. This condition should be established when compensation is set to 0% and voltages are adjusted to place unstained control cells in the first decade, but sufficiently above the left axis so that events do not accumulate in the lowest channel.
59. Total number of cells to be collected depends upon the actual population of interest. In a cell proliferation experiment, the population of interest may be distributed over a broad range of intensities representing up to seven or eight generations. Therefore, to model and calculate the number of cells in each generation accurately, a large number of cells may need to be collected. It is most appropriate to set a stop region on the specific population of interest and collect sufficient cells for analysis within this region. In this example, if the population of interest is Treg cells, it is recommended that a stop region be set on viable Treg cells and a minimum of 5,000 events be collected within this region. When studying rare cells, it may be necessary to simply run the sample tube nearly dry in order to collect the maximum possible number of events.