Cytometric cytokine secretion assay: Detection and isolation of cytokine-secreting T cells

2 Cytometric Cytokine Secretion Assay: Detection and Isolation of Cytokine-Secreting T Cells Mario Assenmacher Miltenyi Biotec GmbH, Bergisch-Gladbach,Germany A l e x a n d e r Scheffold and Andreas Radbruch Deutsches RheumaforschungszentrumBerlin, Berlin, Germany

CONTENTS Introduction Basic considerations Protocols

Examples

41,~I,~i,~I,41,~I, I N T R O D U C T I O N The cell-surface affinity matrix technology, otherwise called Secretion Assay or capture assay, represents a new, innovative method for the analysis and enrichment of viable cells according to secreted molecules, such as antibodies or cytokines (Manz et al., 1995). Recent modifications have led to an easy and fast procedure. In the original protocol the cytokine specific affinity matrix was created by cell surface biotinylation, followed by labelling with avidin-conjugated cytokine specific 'catch' antibody. In the actual protocol a cytokine-specific 'catch' antibody is directly attached to the cell surface of leukocytes as a conjugate with a CD45-specific monoclonal antibody. While originally medium of high viscosity, e.g. gelatine or alginate, was used in the secretion period, the actual protocol has been adapted to the use of normal medium. Basically, the secreted product is retained on the cell surface of the secreting cell, making it accessible to powerful technologies for the detection of surface markers. The Cytokine Secretion Assay involves the following steps: (1) A cytokine specific Catch Reagent is attached to METHODS IN MICROBIOLOGY, VOLUME 32 ISBN 0-12-521532-0

Copyright © Elsevier Science Ltd All rights of reproduction in any form reserved

the cell surface of all cells. (2) The cells are then incubated for 30-45 min at 37°C to allow cytokine secretion. The secreted cytokine binds to the cytokine-specific Catch Reagent on the secreting cells and (3) is subsequently labelled with a second cytokine-specific 'detection' antibody, which is usually conjugated to a fluorochrome such as phycoerythrin (PE) for sensitive analysis by flow cytometry. Optionally the captured cytokine is further magnetically labelled with specific antibody conjugated to super-paramagnetic particles for enrichment by MACS (Fig. 1).

Antt-PE ~i~ii~iii~

MicroBeads

~iiiiiiiil ¸

0 ~i!ii!i!ii~

~I~ !i

~i;ii;~;~;~ ~ ~

i;i~;iiiii!iiii~I

~ , i~il!i~/~ ~....

~

antigen-specific resttmulatton

|abel[|ng with Catch Reagent

secretion period

f|uorescent and magnetic tabeil|ng

Figure 1. Principle of the Cytokine Secretion Assay for labelling of cytokinesecreting T cells after antigen-specific stimulation.

Analogous techniques for single-cell analysis but not for isolation of viable cytokine-secreting cells are ELISPOT (see chapter by Hiroi and Kiyono in Section II) or intracellular cytokine staining (ICS) (see chapter by Yssel et al. in Section III), which were developed in the late 1980s and early 1990s, respectively. Comparisons or direct correlations of the different techniques show comparable results with respect to the detected frequencies of cytokine-expressing cells (Oelke et al., 2000b; Ouyang et al., 2000; Pittet et al., 2001). The Cytokine Secretion Assay combines advantageous features and overcomes several limitations of previously used methods. • The Cytokine Secretion Assay allows the isolation of viable cytokine-secreting cells for cell culture, or other downstream experiments (Assenmacher et al., 1998; Brosterhus et al., t 999; Oelke et al., 2000a,b; Ouyang et al., 2000; Farrar et al., 2001; Hu-Li et al., 2001; McNeil et al., 2001; Smits et al., 2001) including adoptive transfer (Becker et al., 2001 ). • It allows the sensitive multiparameter analysis of viable cytokine-secreting cells down to frequencies of 10_6 due to the option of enrichment of cytokinesecreting cells by magnetic cell sorting (MACS) (Brosterhus et al., 1999; Pittet et al., 2001 ).

60

The Cytokine Secretion Assay is especially useful for the detection and isolation of viable antigen-specific T cells after a short restimulation with specific antigen in vitro to induce secretion of cytokines (Brosterhus et al., 1999). T cells secrete cytokines only transiently upon stimulation, therefore normally only very few T cells actually secrete cytokines in peripheral blood or other tissues. However, memory/effector T cells rapidly restart to secrete cytokines after restimulation in vivo or in vitro. With direct flow cytometric analysis the Cytokine Secretion Assay allows the very rapid detection of cytokine-secreting antigen-specific T cells down to frequencies of 0.01-0.1%. Combination with MACS enrichment greatly increases sensitivity of detection and cytokine-secreting antigen-specific T cells can be detected down to frequencies of 10'~ (0.0001%). The Cytokine Secretion Assay combined with MACS enrichment also allows the isolation of viable cytokine-secreting antigen-specific T cells for expansion and functional characterization (Brosterhus et al., 1999; Oelke et al., 2000a, b; McNeil et al., 2001). The Cytokine Secretion Assay can also be used to analyse cytokine secretion by other cells, e.g. IFN-y by NK cells or IL-10 by monocytes. Cytokine-secreting cells can be analysed for co-production of two different cytokines by two-colour Cytokine Secretion Assays, i.e. using a cocktail of two different Cytokine Catch Reagents and Detection Reagents. Cytokine Secretion Assays can directly be combined with peptide/ MHC tetramer staining for functional characterization of antigen-specific T cells (see later). Out of local areas of acute ongoing immune responses sometimes remarkable numbers of cells 'spontaneously' secreting cytokines can be detected and isolated with the Cytokine Secretion Assay, most likely after recent encounter of their relevant target antigens in vivo.

eeeeee

BASIC CONSIDERATIONS The Cytokine Secretion Assay is very useful for the detection and isolation of viable antigen-specific T cells after a short restimulation with specific antigen in vitro to induce secretion of cytokines.

Controls A negative control sample, treated exactly the same as the antigenstimulated sample but without addition of antigen (or with addition of control antigen), should always be included as a measure of spontaneous (probably in-vivo induced) cytokine secretion. (Optional) A positive control may be included in the experiment, e.g. using Staphylococcal Enterotoxin B (Sigma) at 1 ~tg ml ~for 3-16 h. The addition of co-stimulatory agents such as anti-CD28 mAb or anti-CD49d mAb (usually 1 ~g ml ~) may 61

enhance the response to the antigen. If co-stimulatory agents are added to the antigen sample, they also have to be included in the negative control sample. As a high control to verify that all cells are labelled with the Cytokine Catch Reagent and therefore are able to capture cytokine, immediately after labelling of the cells with the Cytokine Catch Reagent a small aliquot of the cells can be incubated in recombinant cytokine (typically at a final concentration of 200-1000 ng ml-9 for 10 min on ice. After washing, the captured cytokine can be detected on all cells by staining with the fluorochrome-conjugated Cytokine Detection Antibody (Manz et al., 1995; Assenmacher et al., 1998; Ouyang et al., 2000).

Kinetics and proposed time schedule Upon stimulation with peptide, antigen-specific T cells can be analysed for IL-2, IL-4, IL-10 and IFN- 7 secretion within approximately 3-6 h. Upon stimulation with protein, kinetics of cytokine expression are slightly slower (due to the time required for processing) and the cells can be analysed within approximately 6-16 h. Human cells are typically collected and prepared for stimulation on the first day. Then antigen (protein) is either added in the late afternoon for overnight (<16 h) stimulation (A) or cells are stored overnight (whole blood at RT; PBMC in culture medium at 37°C, 5-7% CO2) and antigen (peptide) is directly added in the next morning for short (3 h) stimulation (B). Thus the Cytokine Secretion Assay procedure is started in the early (A) or late (B) morning of the second day. Murine cells are typically collected and prepared for peptide stimulation early in the morning of the day. Then peptide is added for short (3 h) stimulation and the Cytokine Secretion Assay procedure is started in the early afternoon of this day. For protein stimulation typically in the afternoon of the first day, murine cells are collected and prepared for stimulation. Then protein is added in the late afternoon for overnight (_<16h) stimulation and the Cytokine Secretion Assay procedure is started in the early morning of the second day. If another type of in vitro stimulation or cell type, e.g. LPS stimulation of monocytes, is used, or if cells stimulated in vivo are directly analysed ex vivo, one can start the cytokine secretion assay with the stimulated cells at Labelling cells with Cytokine Catch Reagent stage of the protocol. Cytokines are usually only transiently secreted upon stimulation. Therefore the optimal time point after stimulation for analysis with the Secretion Assay has to be chosen in the individual experimental set up.

Counterstaining of cytokine-secreting T cells To identify the cells of interest counterstaining with CD4 or CD8 is required. Upon activation of T cells TCR and associated molecules, like 62

CD3, are often downregulated and therefore are not useful markers. Exclusion of dead cells by staining with propidium iodide (PI) or 7-AAD will reduce non-specific background staining. For optimal sensitivity labelling of undesired non-T cells such as monocytes (human) or B cells (mouse) with antibodies conjugated to PerCP is recommended, e.g. using CD14.PerCP (human) or B220(CD45R).PerCP (mouse). These cells can then be excluded together with PI stained dead cells (see Figs 3 and 4). On enrichment of rare antigen-specific T cells exclusion of dead cells is essential, because dead cells can severely disturb analysis of enriched cell fraction.

Critical parameters

¢,

O

Oe-

e-

U

4.1

The most important point within the Cytokine Secretion Assay is to prevent the capture of secreted cytokine by non-secreting cells during the secretion period. Not all of the secreted cytokine is caught by the affinity matrix of the secreting cells, i.e. some cytokine diffuses away from the secreting cell and accumulates in the culture medium. If the concentration of cytokine-secreting cells in the secretion phase is too high (approximately >5 x 104 cytokine-secreting cells ml ~) and thereby the concentration of secreted and accumulated cytokine in the culture medium reaches a certain level (approximately >0.1-1 ng ml 0, all cells start to catch a low amount of cytokine and subsequently are labelled with the Detection Antibody. The resulting fluorescent shift of all cells can be easily recognized in the flow cytometric analysis. This problem (which similarly can occur in ELISPOT assays) can be prevented in two different ways. In the standard protocol the cell density in the secretion phase is adjusted according to the expected frequency of cytokine-secreting cells (among all cells), i.e. for <1-5% to 1-2 x 10'7 cells ml ~and for >1-5% to 1-2 x 10~cells ml ~(for >>20-50% to <10~cells ml '). In addition all washing and incubation steps (except the secretion phase) are done in the cold (ice-cold buffer, incubations on ice, refrigerated centrifuge) to prevent cells from secretion, when they are at high density, e.g. in the pellet or upon labelling with catch or detection antibody. In the rapid protocol the assay is done without intervening washing steps on two samples with different cell densities (A: 10" cells in 200 ~tl and B: 10~ cells in 200 gl). While sample A allows the detection of cytokinesecreting cells at frequencies from approximately 0.01 to 1% (and cells 'shift' at frequencies >>1%), sample B allows the detection of cytokinesecreting cells in the range of 1-10%. The second approach is similar to the serial dilution of cells in an ELISPOT assay. These first two protocols describe the Cytokine Secretion Assay including magnetic enrichment starting from PBMC or equivalent m u r i n e / h u m a n leukocyte-containing populations or starting from whole blood. The third or fourth protocols are very rapid and simple procedures for the direct detection of cytokine-secreting antigen-specific T cells from small numbers of PBMC or equivalent cells or whole blood without magnetic enrichment.

63

QO

eeeee

, PROTOCOLS Reagents

Buffer: Phosphate-buffered saline (PBS) pH 7.2, containing 0.5% (w/v) bovine serum albumin (BSA) and 2 mM EDTA (see recipe). 0.5 M EDTA stock solution: Dissolve 56 g sodium hydroxide (NaOH) in 900 ml distilled water. Add 146.2 g ethylenediamine-tetraacetic acid, adjust pH to 7.5, fill up to 1 1. Prepare buffer with, for example, 4 ml of 0.5 M EDTA stock solution per I 1 of buffer. Culture medium: e.g. RPMI-1640 containing 10% AB or autologous serum for human cells or mouse serum for murine cells.

Additionally for protocols starting with whole blood (1.2 and 2.2): Lysing solution (lx) (do NOT use FACS Lysing solution TM !) 10× stock solution: 41.4 g NH4C1 (1.55 M), 5 g KHCO 3 (100 mM), 1 ml 0.5 M EDTA (1 mM), add 500 ml ddH20, adjust pH to 7.3. prepare 1× Lysing solution fresh from 10× Lysing stock solution. Cytokine Catch Reagent: anti-cytokine monoclonal antibody conjugated to cell surface (CD45) specific monoclonal antibody (Miltenyi Biotec). Cytokine Detection Antibody: anti-cytokine monoclonal antibody conjugated to PE or APC (Miltenyi Biotec). Additional staining reagents, e.g. CD4-FITC or CD8-FITC and CD14PerCP for human cells or CD45R(B220)-PerCP for murine cells (Becton Dickinson). Propidium iodide (PI) or 7-AAD to exclude dead cells from the flow cytometric analysis.

Additionally for protocols with magnetic enrichment: Anti-PE MicroBeads: colloidal super-paramagnetic MicroBeads conjugated to monoclonal mouse anti-PE antibody (Miltenyi Biotec). MACS columns and separators (Miltenyi Biotec). Tissue culture dishes or plates. V-bottom tubes (depending on cell number: 1.5 ml to 15 ml) or deep well plates. Refrigerated centrifuge, 37°C CO2 incubator.

Cytokine Secretion Assay with enrichment and detection (PBMC or equivalent) In vitro stimulation of T cells with specific antigen

1. Prepare human PBMC, murine spleen cells or other leukocytecontaining single cell preparations. (contd.)

64

2. 3.

4.

5.

Do not use any non-human (non-routine) proteins (like BSA or FCS) in buffer or media upon cell preparation, stimulation or freezing to prevent non-specific stimulation. Wash cells b y adding medium, centrifuge at 300g for 10 min, r e m o v e supernatant. Resuspend cells in culture m e d i u m at 1 × 107 cells ml ~and 5 x 10 h cells cm ~ (e.g. 10" cells per 100 gl in a 96-well plate, 10; cells ml ~in a 24-well plate or 10 ~cells per 10 ml in a petri dish). Add antigen immediately, or leave cells overnight at 37°C, 5-7% CO~. High cell density is required for optimal antigenic stimulation. A d d antigen, e.g. peptide (often at 1-10 gg ml ~) or protein (often at 1-100 gg ml '), mix well and incubate at 37°C, 5-7% CO2 for either 3-6 h (peptide) or 6-16 h (protein). Collect cells carefully, rinse d i s h / w e l l with cold buffer.

Labelling cells with Cytokine Catch Reagent 1. Use 10 ~ cells in a 15 ml closable test tube. For larger cell numbers, scale u p all volumes accordingly. For less than 107 cells, use same volume. 2. Wash cells by adding 10 ml of cold buffer, centrifuge at 300 g for 10 rain at 4°C, r e m o v e supernatant completely. Optionally repeat washing step in ice-cold buffer. 3. Resuspend cells in 100 gl Cytokine Catch Reagent diluted in icecold m e d i u m (typically 10-20 ~g ml ') per 10 ~ cells, mix well and incubate for 5 min on ice.

Keep the cells ice-cold to prevent secretion qf cytokines at this step.

Cytokine secretion period 1. A d d w a r m (37°C) m e d i u m to dilute the cells to 10~-10 ~ cells m l ' d e p e n d i n g on the expected frequency of cytokine-secreting cells (among all cells): <1-5%: 1-2 x 10" cells ml ', i.e. add 7 ml per 107 cells >1-5%: 1-2 x 10 ~cells ml ', i.e. add 70 ml per 107 cells (for >>20-50%: <10 ~cells ml 1). 2. Incubate cells in (closed) preparation tube for 45 min at 37°C u n d e r slow continuous agitation/rotation or mix tube every 5 min to avoid sedimentation of the cells.

It is important to prevent close contact of cells to avoid crosscontamination with cytokines.

Labelling cells with Cytokine Detection Antibody 1. Wash cells by adding a m i n i m u m of 1 v o l u m e of ice-cold buffer and place tube on ice. Centrifuge at 300g for 10 min at 4°C. Remove supernatant completely.

(contd.)

65

Keep the cells ice-cold to stop secretion of cytokines after secretion period. 2. Resuspend cells in 100 gl Cytokine Detection Antibody (and additional staining reagents) diluted in ice-cold buffer (typically 1-5 gg ml ~) per 107 cells. Mix well and incubate for 10 min on ice. 3. Wash cells by adding 10 ml of cold buffer, centrifuge at 300g for 10 min at 4°C, remove supernatant completely.

Magnetic labelling and separation (for PE conjugated Cytokine Detection Ab) 1. Resuspend cells in 80 btl of cold buffer per 107 cells, add 20 btl of antiPE MicroBeads, mix well and incubate for 15 min at 8°C. 2. Wash cells by adding 10 ml of cold buffer, centrifuge at 300g for 10 min at 4°C, remove supernatant completely. 3. Resuspend cells in 500 gl of cold buffer per 107 cells; for higher cell numbers use a dilution of 108 cells ml-'. Take an aliquot for FACS analysis and cell count before enrichment. 4. Prepare two MS columns per sample by rinsing with 500 btl of cold buffer. Place the first column into the magnetic field of a suitable MACS Separator (e.g. MiniMACS). 5. Apply the magnetically labelled cells to the column, allow the cells to pass through and wash with 3 x 500 btl of cold buffer. 6. Remove the first column from the separator, place the second column into the separator, and put the first column on top of the second one. Pipette 1 ml of cold buffer on top of the first column, firmly flush out the retained cells using the plunger, directly onto the second one. Two consecutive column runs are required for optimal enrichment. 7. Wash with 3 x 500 btl of cold buffer. 8. Remove the column from the separator, place the column on a suitable collection tube. Pipette 500 btl of cold buffer on top of the column, firmly flush out the retained cells using the plunger. For subsequent cell culture, cells can also be eluted with medium. If part of the cells is analysed by flow cytometry, the medium should not contain phenol red. 9. Proceed to analysis a n d / o r cell culture.

Cytokine Secretion Assay with enrichment and detection (whole blood) 1. Take 5 ml sodium heparinized whole blood in 50 ml conical polypropylene tube. Use sodium heparin as anticoagulant, not EDTA or ACD. Calcium chelating anticoagulants prevent activation. Add antigen immediately, or leave blood overnight at RT. (contd.)

66

2. Add antigen, e.g. peptide (often at 1-10 gg ml ') or protein (often at 1-100 gg ml 1), mix well by vortexing and incubate at 37°C, 5-7% CO2 for either 3-6 h (peptide) or 6-16 h (protein). 3. Add 45 ml of lx Lysing solution. Mix gently and incubate for 10 min at RT. Rotate tube continously or turn tube several times during incubation. 4. Centrifuge cells at 300g for 10 min at RT. Remove supernatant completely. 5. Resuspend cells in 15 ml of cold buffer and transfer into a 15 ml conical propylene tube. 6. Centrifuge cells at 300g for 10 min at 4°C. Remove supernatant completely. 7. Resuspend cells in 200 gl Cytokine Catch Reagent diluted in ice-cold medium (typically 10-20 gg ml ~), mix well and incubate for 5 min on ice. Keep the cells ice-cold to prevent secretion of cytokines at this step. 8. Add 7 ml warm (37°C) culture medium and incubate cells in (closed) preparation tube for 45 min at 37°C under slow continuous agitation/rotation or mix tube every 5 min to avoid sedimentation of the cells. It is important to prevent close contact of cells to avoid cross-contamination with cytokines. 9. Wash cells by adding 7 ml of ice-cold buffer and place tube on ice. Centrifuge at 300g for 10 min at 4°C. Remove supernatant completely. Keep the cells ice-cold to stop secretion of cytokines after secretion period. 10. Resuspend cells in 200 gl Cytokine Detection Antibody (and additional staining reagents) diluted in ice-cold buffer (typically 1-5 gg ml ~). Mix well and incubate for 10 min on ice. 11. Wash cells by adding 10ml of cold buffer, centrifuge at 300 g for 10 min at 4°C, remove supernatant completely. 12. Proceed to magnetic labelling and separation and then to analysis.

Rapid Cytokine Secretion Assay for detection only (PBMC or equivalent) 1. Prepare human PBMC, murine spleen cells or other leukocyte F1 containing single-cell preparations. Do not use any non-human (non-murine) proteins (such as BSA or FCS) in buffer or media upon cell preparation, stimulation or freezing to prevent nonspecific stimulation. 2. Wash cells by adding medium, centrifuge at 300g for 10 min, remove supernatant. (contd.)

67

3. Resuspend cells in culture medium at I x 107 cells ml '. 4. Place 10~'cells in 100 gl culture medium in a >1.5 ml deep well plate or 1.5 ml tube labelled (A). Add antigen immediately, or leave cells overnight at 37°C, 5-7% C02. 5. Add antigen, e.g. peptide or protein (often at 1-10 ~tg ml 1), mix well and incubate at 37°C, 5-7% CO2 for either 3-6 h (peptide) or 6-16 h (protein). 6. Wash cells by adding 1 ml of medium, centrifuge at 300 g for 5 min at RT, remove supernatant. Remove supernatant from plates by simply inverting them once without shaking and let supernatant drain out. 7. Resuspend cells in 200 ~1 medium. Mix well and transfer 20 gl of cell suspension from well/tube (A) into a second well or tube labelled (B) containing 200 gl culture medium (B) to obtain a second sample with a higher dilution of the cells. Tube B is only required for the stimulated sample, but not for the control sample. 8. Add 20 gl Cytokine Catch Reagent (final concentration typically 5-10 gg ml -l) to all wells or tubes (A+B), mix well and incubate cells for 30 min at 37°C under slow continuous agitation/rotation or mix every 5 min to avoid sedimentation of the cells. It is important to prevent close contact of cells to avoid cross-contamination with cytokines. 9. Add 20 gl Cytokine Detection Antibody (and additional staining reagents) (final concentration typically 1-5 gg ml '). Mix well and incubate for 10 min at RT. 10. Wash cells by adding 1 ml of cold buffer, centrifuge at 300g for 5 min at RT, remove supernatant. 11. Resuspend cells in 500 gl of cold buffer and proceed with flow cytometric analysis.

Rapid Cytokine Secretion Assay for detection only (whole blood) 1. Place 300 btl sodium heparinized blood in a >1.5 ml deep well plate or 1.5 ml tube labelled (A). Use sodium heparin as anticoagulant, not EDTA or ACD. Calcium chelating anticoagulants prevent activation. Add antigen immediately, or leave blood overnight at RT. 2. Add antigen, e.g. peptide or protein (often at 1-10 gg ml '), mix well and incubate at 37°C, 5-7% CO2 for either 3-6 h (peptide) or 6-16 h (protein). (contd.)

68

3. Mix well and transfer 30 ~1 of blood from well/tube A into a second well or tube labelled (B) containing 250 ~tl culture medium to obtain a second sample with a higher dilution of the cells. Tube B is only required for the stimulated sample, but not for the control sample. 4. Add 20 ~tl Cytokine Catch Reagent (final conc. typically 5-10 ~g ml ~)to all wells/tubes (A+B), mix well and incubate cells for 30 min at 37°C under slow continuous agitation or mix every 5 min to avoid sedimentation of the cells. It is important to prevent close contact of cells to avoid cross-contamination with cytokines. 5. Add 20 ~1 Cytokine Detection Antibody (and additional staining reagents) (final concentration typically 1-5 ~g ml '). Mix well and incubate for 10 min at RT. 6. Add 1 ml Lysing solution (Ix). Mix gently and incubate for 10 min at RT in the dark. Mix in between. 7. Centrifuge at 300 g for 5 min at RT. Remove supernatant. Remove supernatan t from plates by simply inverting them once without shaking and let supernatant drain out. 8. Wash cells by adding 1 ml of cold buffer, centrifuge at 300 g for 5 min at RT, remove supernatant. 9. Resuspend cells in 500 ~1 of cold buffer and proceed with flow cytometric analysis.

Combination of Cytokine Secretion Assay with peptide/MHC tetramer staining Cytokine Secretion Assays can directly be combined with p e p t i d e / M H C tetramer (or Dimer) staining for functional characterization of antigen-specific T cells (Pittet et al., 2001) (see example later). Upon stimulation with antigen (or some mitogens like anti-CD3) the TCR on the specific T cells can be downregulated very rapidly. This dramatically reduces peptide-MHC tetramer staining. Therefore the peptide-MHC tetramer labelling should be done prior to the stimulation with peptide. After staining with p e p t i d e / M H C tetramer cells are stimulated with the specific peptide for 2-3 h followed by the standard secretion assay procedures. Depending on the type of peptide-MHC tetramer used, the peptideMHC tetramer may (a) or may not (b) stimulate cytokine secretion. In the former case (a) the control sample is stained with peptide-MHC tetramer after the incubation (without peptide or with control peptide) and the cytokine secretion assay. Initially these two types of controls may be compared.

69

Flow cytometric analysis 1. Store samples at 4°C in the dark until analysis. Add PI (0.5 ~g ml") or 7-AAD immediately prior to FACS analysis.

For enumeration of cytokine-secreting antigen-specific T cells (with or without enrichment): 2. Acquire 2 x 10~viable lymphocytes from the fraction before enrichment or from samples without enrichment.

For optimal sensitivity, appropriate and same numbers of viable cells have to be acquired from the antigen-stimulated sample as well as from the control sample. 3. Acquire the complete positive fractions (from the antigenstimulated sample as well as from the negative control sample) after enrichment.

For isolation of cytokine-secreting antigen-specific T cells (e.g. for expansion of isolated cells): 4. Acquire up to 2 x 10~ viable lymphocytes from the fraction before enrichment and acquire an aliquot of the positive fraction to determine the performance of the enrichment.

Enumeration of cytokine-secreting antigen-specific T cells (with or without enrichment) (1) Without or before MACS enrichment frequencies of cytokinesecreting antigen-specific T cells were calculated based on the numbers of cytokine+ (CD8+ or CD4+) T cells and the numbers of (CD8+ or CD4+) T cells in the acquired fractions (see Fig. 2A): % cytokine+ CD4+ T cells among CD4+ T cells = (events in UR quadrant/events in UL + UR quadrants) x 100 = 0.257% (2) After MACS enrichment frequencies of cytokine-secreting antigenspecific T cells were calculated as follows. First, the numbers of cytokine+ T cells in the magnetically enriched fractions were calculated on the basis of total number of cells and the frequency of cytokine+ T cells in the magnetically enriched cell fraction. Frequencies of cytokine+ T cells among total T cells were then calculated on the basis of the numbers of cytokine+ T cells in the magnetically enriched fractions and the numbers of T cells before enrichment of cytokine-secreting cells (see Fig. 2B) (beside dot plot): % cytokine+ CD4+ T cells among CD4+ T cells = (total number of cytokine+ CD4+ T cells after enrichment/ total number of CD4+ T cells before enrichment) x 100 = 0.129%

(contd.)

70

While approach (2) gives highest sensitivity (down to 0.0001%), it is slightly less precise than approach (1), because not all of the cytokine+ T cells present in the sample before enrichment are recovered in the positive fraction after MACS enrichment. Usually the recovery is between 50-70%. The frequencies determined according to approach (2) are about half of the frequencies determined by approach (1). This is also illustrated in the enclosed example, where approach (1) results in 0.251% IFN-¥+ CMV-specific CD4+ cells among CD4+ cells and approach (2) gives 0.129% IFN-7+ CMV-specific CD4+ cells among CD4+ cells. Therefore, if the sensitivity of approach (1) (usually 0.1-0.01%) is sufficient to detect cytokine+ antigen-specific T cells it is preferred. If the sensitivity of approach (1) is not sufficient the extremely sensitive approach (2) should be used. The slight underestimation of the frequencies of cytokine+ antigen-specific T cells does not impair the analysis of antigen-specific T cell immunity, because comparisons between different donors or samples are not influenced and especially at low frequencies (0.01-0.0001%) a factor of 2 is probably biologically irrelevant. Proposed criteria for a significant antigen-specific response are (1) minimal number of 10-20 cytokine+ T cells in the antigen-sfimulated sample and (2) at least 3- to 5-fold higher frequency of cytokine+ T cells among total T cells in the antigen-stimulated sample than in the control sample.

e~,~,4,~,~, E X A M P L E S Detection and isolation of IFN-y-secreting CMV-specific human Th cells The first example shows the detection and isolation of IFN-y-secreting CMV-specific CD4+ T cells using the Cytokine Secretion Assay (Fig. 2). PBMC from a CMV seropositive donor were incubated for 16 h with 5 btg ml ~CMV lysate (A+B) or without antigen (C+D) as described in the protocol above. The IFN-7 Secretion Assay was performed with 10r cells on both samples as described in the protocol B1.1 including enrichment of IFN-7-secreting cells by a MACS separation over two sequential MiniMACS columns. Counterstaining of T cells was performed using CD4-FITC. Monocytes were stained with CD14-PerCP and dead cells were stained with PI. Flow cytometric analysis was performed on the original fractions and the enriched fractions. At least 200 000 events of the original fractions and the complete enriched fractions were acquired. A lymphocyte gate based on forward and side scatter properties (FSC/SSC) was set (like shown in Fig. 3A) and dead cells and monocytes were excluded according to PI and CD14-PerCP staining in a FL-2 versus FL-3 plot (as shown in Fig. 3B). For analysis PE (IFN-7) versus FITC (CD4) staining of gated (viable) lymphocytes is displayed (Fig. 2A-D). 71

O

e"

U

e~ U

after enrichment

before enrichment

B

70. 8%

.

(D

o

~:~.." -

.

•

.

.

.

.

.

•

• ......' .-. ,,:." ?....."

0.006%*

..,. ' . . . . . . .

1292 IFNy÷ CD4 + cells out of 106 CD4 + cells (= 0.129%)

....

D

sO. 3%

_<1 IFNT* CD4*cells out of 106 CD4* cells (_<0.0001%)

> (D "5 o

,. d:~...,,

i •

"

:":;-=,i..: "" IFN-y

IFN- T

* among CD4 ÷ cells

Figure 2. Detection and isolation of IFN-T-secretingCMV-specifichuman Th cells.

Analysis of expression of IFN-T and IL-2 by CMV-specific human CD8+ cells The second example shows the analysis of CMV-specific CD8+ T cells for secretion of IFN- T and IL-2 using a combination of Peptide/MHC Tetramer staining with Cytokine Secretion Assay (Fig. 3). PBMC from a CMV seropositive, HLA-A2+ donor were stained for 1 h at 4°C with a PElabelled CMVpp65495_s03/HLA-A2 Tetramer (D, F and G, I) or incubated without Tetramer (E and H). Cells were then stimulated for an additional 2 h at 37°C with 5 ~tg ml 1CMV peptide pp65495_503 (E, F and H, I) or without peptide (D and G). The IFN-T (D-F) or IL-2 (G-I) Secretion Assay was performed on the samples as described above, but without magnetic labelling/enrichment. Cells were stained with IFN-T Detection Antibody (APC) (D-F) or IL-2 Detection Antibody (APC) (G-I), CD8.FITC, CD14.PerCP and PI. For flow cytometric analysis 200 000 events of each sample were acquired. A lymphocyte gate and an exclusion gate for dead cells and monocytes were set as described earlier (Fig. 3A+B). In addition CD8+ cells were gated based on FL-1 properties (Fig. 3C). For analysis APC (IFN-T) versus PE (Tetramer) staining of gated (viable CD8+) lymphocytes is displayed (D-I). While in this donor the majority (87%) of CMVpp65495_503 specific CD8+ T cells secretes IFN-T upon stimulation (Fig. 3F), only 16% secrete IL-2 (Fig. 3I). 72

A

•

o

., • ~::::,,.:.:.-.~,...

LT.

8 o

i

. . . .

i

. . . .

,

. . . .

i

. . . .

i

. . . .

i

Tetramer Peptide stim.

Tetramer.PE

Tetramer.PE

FSC

+

+

4-

4-

F

E

D

1.. 1 hHuG

0.05%

!

0.07%

I

0.12%

0.00%

I

0.81% •

.

• .,. :.,,

:-

,~,~,,

..,.. .

.

i

0.11% I--

&

.

~6

IFN-y

IFN-y

IFN-y

I

H

G

1.54%

0.00%

0.00%

0.00%

=1 0.97% I

0.18%

{3-

:s (D

.:-?.~-. ". :: 1.7.-.. i

";:'.g"Kf~;a

~r... I -4

,

--,-

0.35%

0.01% ..... i

.......

[

.......

Ik-2

IL-2

IL-2

Figure 3. Analysis of expression of IFN-1 and IL-2 by CMV-specific human CD8+ cells.

Detection and isolation of IFN-7-secreting HEL-specific murine Th cells The last example shows the detection and isolation of IFN-y-secreting HEL-specific murine CD4+ T cells using the Cytokine Secretion Assay (Fig. 4). BALB/c mice, female, 7-10 weeks old, were intraperitoneally (i.p.) immunized with 100 ~tg Henn eggwhite lysozyme (HEL) and 200 ng Pertussis Toxin (PT) in incomplete Freund's adjuvant (IFA) and an additional injection of 200 ng PT in PBS 24 h later. After 2-3 weeks spleen cells (SC) from immunized mice (C-F) or unimmunized control mice (G+H) were incubated in vitro for 16 h with 100 pg ml ~ HEL (C+D and G+H) or without HEL (E+F) in RPMI 1640 medium supplemented with 5% murine (heat-inactivated) serum. IFN-y Secretion Assays were performed with 107 cells on all samples as described earlier including 73

enrichment of cytokine-secreting cells by MACS separation over two sequential MiniMACS columns. Counterstaining of T cells was performed using CD4.FITC. B lymphocytes were stained with CD45R(B220).PerCP and dead cells were stained with PI. Flow cytometric analysis was performed on the original fractions and the enriched fractions. Two hundred thousands events of the original fractions and the complete enriched fractions were acquired. A lymphocyte gate based on forward and side scatter properties (FSC/SSC) was set (Fig. 4A) and dead cells and B cells were excluded according to PI and CD45R(B220).PerCP staining in a FL-2 versus FL-3 plot (Fig. 4B). For analysis PE (IFN-y) versus FITC (CD4) staining of gated (viable) lymphocytes is displayed.

l • .~ :.!, .::

0 (,9 69

~ •

B

.,:~i:'::" ,..: f"

.

;,::j'.: :~.~ :..:-: ,

'~';~i :::'

....

':

•:

,~" -4-,:..!-,.:'..'.: ' ,~...

....

r

cytokine (PE)

FSC

before

117WVO ili Yitl'O HEL

HEL

C

after enrichment

D

0.167%*

D,!:

.¸

67.0%

x...-

: ./....j.:

L HEL

.,:

•" '.i:. "?('~";:':~h ;`:

:,-i

F

0.018%* -

20.1% 67 IFN-T ÷ CD4+ cells out of 106 C D 4 ÷ cells (=0.007%)

. :.......

•

....

•

H

..•.

3.7%

HEL ~-,

:.,...., .-..::

693 IFN-y + CD4* cells out of 106 C D 4 ÷ cells (=0.069%)

.

•

8 IFN-y + C D 4 + cells out of 10 6 C D 4 + cells

(=0.001%)

.i;' .: .. ,. • 9.., .i

* a m o n g CD4* cells IFN-y

IFN-y

Figure 4. Detection and isolation of IFN-Tsecreting HEL-specificmurine Th cells. 74

References Assenmacher, M., L6hning, M., Scheffold, A., Manz, R. A., Schmitz, J. and Radbruch, A. (1998). Sequential production of IL-2, IFN-y and IL-10 by individual staphylococcal enterotoxin B-activated T helper lymphocytes. Eur. J. Immunol. 28, 1534-1543. Becker, C., Pohla, H., Frankenberger, B., Schiiler, T., Assenmacher, M., Schendel, D. J. and Blankenstein, T. (2001). Adoptive tumor therapy with T lymphocytes enriched through an IFN-y capture assay. Nature Medicine 7, 1159-1162. Brosterhus, H., Brings, S., Leyendeckers, H., Manz, R. A., Miltenyi, S., Radbruch, A., Assenmacher, M. and Schmitz, J. (1999). Enrichment and detection of live antigen-specific CD4+ and CD8+ T cells based on cytokine secretion. Eur. J. Immunol. 29, 4053-4059. Farrar, J. D., Ouyang, W., L6hning, M., Assenmacher, M., Radbruch, A., Kanagawa, O. and Murphy, K. M. (2001). An instructive component in T helper cell type 2 (Th2) development mediated by GATA-3. J. Exp. Med. 193, 643-648. Hu-Li, J., Pannetier, C., Guo, L., L6hning, M., Gu, H., Watson, C., Assenmacher, M., Radbruch, A. and Paul, W. E. (2001). Regulation of expression of IL-4 alleles: analysis using a chimeric GFP/IL-4 gene. Immunity 14, 1-11. Manz, R., Assenmacher, M., Pfluger, E., Miltenyi, S. and Radbruch, A. (1995). Analysis and sorting of live cells according to secreted molecules relocated to a cell-surface affinity matrix. Proc. Natl. Acad. Sci. USA 92, 1921-1925. McNeil, A. C., Shupert, W. L., Lyasere, C. A., Hallahan, C. W., Mican, J., Davey, R. T. Jr. and Connors, M. (2001). High-level HIV-1 viremia suppresses viral antigen-specific CD4+ T cell proliferation. Proc. Natl. Acad. Sci. USA 98, 13878-13883. Oelke, M., Moehrle, U., Chen, J. L., Behringer, D., Cerundolo, V., Lindemann, A. and Mackensen, A. (2000a). Generation and purification of CD8+ melan-aspecific cytotoxic T lymphocytes for adoptive transfer in tumor immunotherapy. Clin. Cancer Res. 6, 1997-2005. Oelke, M., Kurokawa, T., Hentrich, I., Behringer, D., Cerundolo, V., Lindemann, A. and Mackensen, A. (2000b). Functional characterization of CD8+ antigenspecific cytotoxic T lymphocytes after enrichment based on cytokine secretion: comparison with the MHC-tetramer technology. Scand. J. hnmunol. 52, 544-549. Ouyang, W., L6hning, M., Gao, Z., Assenmacher, M., Ranganath, S., Radbruch, A. and Murphy, K. M. (2000). Stat6-independent GATA-3 autoactivation directs IL-4-independent Th2 development and commitment, hnmunity 12, 27-37. Pittet, M. J., Zippelius, A., Speiser, D. E., Assenmacher, M., Guillaume, P., Valmori, D., Lienard, D., Lejeune, F., Cerottini, J. C. and Romero, P. (2001). Ex vivo IFN-gamma secretion by circulating CD8 T lymphocytes. Implications of a novel approach for T cell monitoring in infectious and malignant diseases. J. Immunol. 166, 7634-40. Smits, H. H., van Rietschoten, J. G., Hilkens, C. M., Sayilir, R., Stiekema, F., Kapsenberg, M. L. and Wierenga, E. A. (2001). IL-12-induced reversal of human Th2 cells is accompanied by full restoration of IL-12 responsiveness and loss of GATA-3 expression. Eur. J. hnmun. 31, 1056-1065.

75

0

e"

C

U

au