- Email: [email protected]
Determination of intracellular cytokines by flow-cytometry following whole-blood culture
Journal of Immunological Methods 209 Ž1997. 67–74
Determination of intracellular cytokines by flow-cytometry following whole-blood culture W.A Carrock Sewell ) , Margaret E. North, A. David B. Webster, John Farrant MRC Immunodeficiency Research Group, Department of Clinical Immunology, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF, UK Received 7 April 1997; revised 31 July 1997; accepted 26 August 1997
Abstract Various methods have been reported for measuring intracellular cytokines in peripheral blood mononuclear cells isolated by density-gradient centrifugation. In this report, we describe a whole-blood method for the determination of intracellular cytokines ŽIFN-g , TNF-a and IL-2. that uses small-volume Ž500 m l. blood samples. Directly conjugated anti-cytokine antibodies and commercial cell membrane fixation and permeabilisation reagents were used. Blood was cultured in a 1:3 dilution with a combination of PMA and ionomycin to reveal the cytokine synthetic potential of each cell, together with monensin to increase the sensitivity by retaining cytokines within the cell to detectable levels. The optimum concentrations of PMA Ž10 ngrml Ž16.2 nmolrl.., ionomycin Ž2 m molrl. and monensin Ž3 m molrl. were determined. Kinetic studies showed maximal cytokine expression after 2 h of culture for TNF-a and IFN-g and 4 h for IL-2. Assessment of TNF-a and IFN-g production within the CD4 and CD8 lymphocytes from 10 normal volunteers showed that considerably more CD8q than CD4 q cells produced IFN-g . This technique could be used by routine immunology laboratories and will be of use in studies to determine whether cytokine assays are of value in the investigation of immune disorders. q 1997 Elsevier Science B.V. Keywords: Intracellular cytokines; Flow cytometry
1. Introduction
Abbreviations: CVID, common variable immunodeficiency; ECD, energy coupled dye; FITC, fluorescein isothiocyanate; FL1,2,3,4, fluorescence channel 1, 2, 3, 4; IFN-g , interferongamma; IL-2, interleukin-2; PBSrArA, phosphate-buffered saline with azide and albumin; PerCP, peridinin chlorophyll protein conjugate; PE, phycoerythrin; PECy5, phycoerythrin-cyanin 5; PMA, phorbol myristate acetate; RPMI1640, Roswell Park Memorial Institute medium 1640; TNF-a , tumour necrosis factor-alpha ) Corresponding author. Tel.: q44-171-7940500 ext. 4932; fax: q44-171-8302224.
The ability to assess cytokine profiles in different immune cells is fundamental to understanding both normal immunoregulation and its dysfunction in disease. Traditionally, cytokines in supernatants from lymphocyte cultures are measured by ELISA or bioassays ŽThorpe et al., 1992.. This provides a useful but incomplete picture, since the relative contribution of individual cell subsets to cytokine production is not determined. With advances in the techniques of cell fixation and permeabilisation it has became possible to identify individual cytokine-
0022-1759r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 2 2 - 1 7 5 9 Ž 9 7 . 0 0 1 5 0 - 6
68
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
producing cells by indirect immunofluorescence microscopy; initially by two-layer staining ŽSander et al., 1991. and more recently with directly conjugated anti-cytokine antibodies. Such methods are useful but laborious and have only a limited ability to distinguish the lymphocyte’s phenotype as well as cytokine production. However, the advent of intracellular cytokine assessment by flow cytometry, staining permeabilised cells with fluorochrome-conjugated anti-cytokine antibodies, permits large numbers of cells of known phenotype to be examined ŽPicker et al., 1995; Prussin and Metcalfe, 1995; Schauer et al., 1996.. Various techniques to measure intracellular cytokines have been reported for peripheral blood mononuclear cells isolated by density gradient centrifugation ŽPrussin and Metcalfe, 1995; North et al., 1996.. These need comparatively large volumes of blood Ž10–20 ml., which is disadvantageous in paediatric studies. Also, the process of cell separation may affect cell activation status ŽMaino et al., 1995. and perturb cytokine production Žsee the whole-blood paper by Maino et al., 1996.. To assess potential cytokine production, various culture conditions Žinvolving changes in culture time, type and dosage of stimulants. have been proposed. Since cytokine production by normal resting cells is minimal, most published techniques use a strong supra-physiological stimulus in culture to demonstrate the potential of each cell to synthesise cytokines. Many of these methods use culture for 10–12 h with stimulants such as phorbol myristate acetate ŽPMA. that activate cells via the protein kinase C pathway; this down-regulates surface CD4 on T-cells ŽPetersen et al., 1992., making identification of CD4q cells and subsequent assessment of cytokines within the Thelper cell subset very difficult. Here we report the development of a robust, rapid, whole-blood flow-cytometric assay for the
determination of IL-2, IFN-g and TNF-a in CD4q and CD8q lymphocytes that requires small volumes of blood, 2 Žor 4. h of culture and which can be performed using equipment available in most clinical immunology laboratories. As an example of its use, comparisons were made of the propensity of CD4q and CD8q lymphocytes from normal subjects to make TNF-a and IFN-g .
2. Materials and methods 2.1. Subjects Normal subjects were healthy laboratory and office personnel Ž5 male, 5 female; mean age 39.7 years.. Informed consent was obtained from all subjects and local ethical committee approval was granted for the study. 2.2. Cell culture and staining Blood Ž1 ml. was collected into lithium heparin and immediately processed; preliminary experiments showed irreproducible changes in cytokine production, such as increased TNF-a , if analysis was not started within approximately 4 h of sampling Ždata not shown.. To increase the quantity of cytokine available for detection within each cell, monensin was added to all cultures to block the export of newly synthesised cytokine from the Golgi apparatus ŽJung et al., 1993.. Blood Ž250 m l. was added to culture medium ŽRMPI1640, Gibco BRL, 500 m l. containing monensin with PMA and ionomycin Ž‘stimulated cultures’. or without Ž‘unstimulated cultures’.. Cultures were carried out in ventilation-capped 5 ml polystyrene round-bottomed plastic tubes ŽGreiner Labortechnik, Gloucestershire..
Fig. 1. Four-colour analysis of IFN-g production by CD4q and CD8q cells. List-mode data were collected through an acquisition gate, to exclude the majority of residual red cells, debris, monocytes and granulocytes. Lymphocytes ŽR1., CD4q ŽR2. and CD8q ŽR3. cells were defined by regions on the corresponding dot-plots. Histograms of the cytokine fluorescence signal were gated using algebraic summation of the regions CD4q AND lymphocytes for CD4q cells and similarly for CD8q cells. Regions for IFN-gq CD4q ŽR4. and IFN-gq CD8q cells ŽR5. were defined on the cytokine histograms using limits defined using the IFN-gy populations in the unstimulated cells. Panel Ža. shows unstimulated cells Žcultured with monensin only.. Panel Žb., from cells stimulated with PMA and ionomycin; the percentage of IFN-gq CD4q in R4 is 8.9% and IFN-gq CD8q lymphocytes in R5 is 26.8%.
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
69
70
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
Both stimulated and unstimulated cultures were incubated at 378C in a humid, 5% CO 2 atmosphere. After incubation, following the manufacturer’s instructions, red cells were lysed using a platform-appropriate lysing solution, namely Optilyse C Ž2 ml for 10 min, Coulter. for use with Coulter Epics-XLMCL four-colour flow cytometers. Other platforms require different lysing solutions, for example Optilyse B ŽCoulter. if an Ortho or Becton-Dickinson flow cytometer is used. After lysing red cells, the remaining cells were washed with phosphate buffered saline Ž2 ml. containing sodium azide Ž0.02%, Sigma. and bovine serum albumin Ž 0.2%, Sigma . ŽPBSrArA. at 300 g for 5 min and the cell pellet resuspended. The cells were then fixed by the addition of Cytoperm Reagent A Ž250 m l, Serotec. for 15 min, following which the cells were washed again in PBSrArA and then permeabilised by the addition of Cytoperm Reagent B Ž250 m l, Serotec.. The exact composition of these commercial solutions is not publicly available, but they are quality-controlled and according to the manufacturers, Reagent A is formaldehyde based and Reagent B is detergent based. Aliquots Ž50 m l. of cells in Reagent B were then added to tubes containing directly-conjugated monoclonal antibodies against various combinations of cytokines and cell surface markers. Saturating concentrations of monoclonal antibodies were used, previously determined in other experiments for the anticytokine antibodies Ždata not shown. or according to the manufacturer’s recommendations for the surface markers. For the anti-cytokine antibodies Žconcentration 100 m grml., 2 m l were used. All samples were stained in the dark at room temperature for 30 min. Samples were then washed once in PBSrArA and stored in 0.5% paraformaldehyde in phosphate buffered saline Ž500 m l. at 48C until acquisition within 24 h. 2.3. Monoclonal antibodies The following monoclonal antibodies were used: anti-human CD4rPE, CD8rPerCP, CD8rFITC, CD69rFITC ŽBecton Dickinson., CD4rECD, CD8rECD ŽCoulter., CD3rPECy5 ŽImmunotech., anti-human TNF-arFITC Žclone B-D9., IFN-
grFITC Žclone B-B1., IL-2rFITC Žclone B-G5. ŽSerotec.. 2.4. Flow cytometry and data analysis A Coulter Epics-XL-MCL four-colour flow cytometer was used for these experiments. Listmode data from 10 000 events was acquired without colour compensation; colour compensation was subsequently applied to the listmode data using the Colour Compensation Toolbox within the WinList Version 3.0 flow-cytometry analysis software ŽVerity, Topsham, ME USA.. The compensation standards used were lymphocytes stained with strongly-positive single-colour monoclonal antibodies, eg. CD3 or CD8, in separate tubes, for each of the four fluorochromes: FITC, PE, ECD and PErCy5. WinList Version 3.0 ŽVerity. was used to analyse the listmode data. For each analysis, dot-plot graphs of forward-scatter versus side-scatter were drawn, and a lymphocyte region R1 was defined. Other regions were defined as illustrated in Fig. 1, namely R2 for CD4 q cells, and R3 for CD8 q cells. These regions encompassed the ‘bright’ cells for each specificity. Separate histograms of cytokine fluorescence were then constructed for CD4 q and CD8 q T-cells. Cytokine-positive CD4 q or CD8 q T-cells were defined by setting regions with the lower-limits for cytokine positivity determined from the unstimulated cultures. Data were expressed as the percentage of CD4 q or CD8 q cells within the lymphocyte gate positive for cytokine. The early activation surface antigen CD69 was also measured on these fixed and permeabilised cells to determine the activation status of the CD4 q and CD8 q cells. In the time-course experiments, surface CD4 became down-regulated with prolonged simulation with PMA, making positive identification of CD4 q cells difficult. To determine the degree of CD4 down-regulation, four-colour flow cytometry was used, measuring CD3, CD4 and CD8 with three separate colours, and cytokine fluorescence in the fourth. The mean CD4 channel fluorescence was measured on CD3q8y cells cultured for 0, 2, 4 and 6 h with PMA and ionomycin. 2.5. Determination of optimal stimulant dose Blood from 5 normal volunteers was cultured for 2 h with ionomycin at 2 m molrl, monensin at a
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
Fig. 2. Determination of optimum stimulant concentrations. Mean"SEM production of CD69, TNF-a , IL-2 and IFN-g by CD4q and CD8q cells is shown Ž ns 5.. In Ža. ionomycin Ž2 m molrl. and monensin Ž3 m molrl. concentrations were held constant and cells were incubated for 2 h with PMA at 5, 10 and 15 ngrml. In panel Žb. the PMA concentration was set at 10 ngrml Ž16.2 nmolrl. and ionomycin concentrations of 1, 2, 5 and 10 m molrl were used. Concentrations of PMA of 10 ngrml Ž16.2 nmolrl., with ionomycin at 2 m molrl gave the maximal cytokine responses for the majority of cytokines tested.
71
72
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
concentration of 3 m molrl and various final concentrations of PMA Ž5, 10 and 15 ngrml.. Intracellular TNF-a and IFN-g production and surface CD69 expression were assessed in CD4q and CD8q cells. The experiment was then repeated using the optimum PMA concentration of 10 ngrml Ž16.2 nmolrl. with various concentrations of ionomycin Ž1, 2, 5 and 10 m molrl.. 2.6. Determination of optimal culture time Blood from 3 normal volunteers was cultured at 378C with monensin Ž3 m molrl. with or without PMA and ionomycin at the optimum concentrations determined above, for various time intervals from 0 to 6 h. Cytokine ŽIL-2, TNF-a and IFN-g . production was again determined in CD4q and CD8q T-cells. 2.7. Comparison of TNF-a and IFN-g production between CD4 q and CD8 q cells Since IL-2 was not optimal after 2 h of culture, only TNF-a and IFN-g , as examples of Th1 cytokines, were measured in CD4q and CD8q cells for this experimental series. Blood from 10 normal volunteers was analysed using the optimum conditions for time and stimulant concentrations determined above, namely the blood was diluted 1r3 giving a final concentration of PMA, 10 ngrml Ž16.2 nmolrl.; ionomycin, 2 m molrl; monensin, 3 m molrl and incubated for 2 h.
3. Results 3.1. Determination of the optimum stimulant dose As shown in Fig. 2Ža. the optimum PMA concentration Ži.e. the concentration producing maximal cytokine expression. was found to be 10 ngrml Ž16.2 nmolrl., when the ionomycin concentration was held constant at 2 m molrl. When this concentration of PMA was used with various concentrations of ionomycin, illustrated in Fig. 2Žb., a 2 m molrl concentration of ionomycin produced the maximal cytokine production. In subsequent experiments, these concentrations were used, together with monensin at a concentration of 3 m molrl.
Fig. 3. Kinetics of cytokine production under optimal conditions. Whole blood was cultured with PMA Ž10 ngrml Ž16.2 nmolrl.., ionomycin Ž2 m molrl. and monensin Ž3 m molrl., for the times shown within Ža. CD4q and Žb. CD8q cells. Stimulation for 2 h gave maximal IFN-g and TNF-a production; 4 h of stimulation resulted in the optimal IL-2 production. Error bars show the SEM.
3.2. Determination of the optimum culture time When blood was cultured with PMA 10 ngrml Ž16.2 nmolrl. and ionomycin 2 m molrl, with monensin 3 m molrl, as determined above, the optimum stimulation time for maximum TNF-a , IL-2 and IFN-g production was assessed. The results are shown in Fig. 3. For both CD4q and CD8q cells, a 2 h stimulus gave clear discrimination of the CD4q
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
73
down-regulation of surface CD4 expression with increasing culture time is illustrated in Fig. 4 and shows a small degree of CD4 down-regulation by 2 h, giving acceptable CD4q cell discrimination, more down-regulation at 4 h but still sufficient to distinguish CD4q from CD4y cells, but poor discrimination after 6 h of stimulation.
Fig. 4. Down-regulation of surface CD4 expression with increasing culture time. The mean"SEM mean CD4 channel fluorescence for three normal donors is shown. Acceptable discrimination of CD4q cells is shown compared to CD8q cells both in unstimulated CD4q cells, and cells stimulated for 2 h. Discrimination is possible, but not ideal, at 4 h, and not possible at 6 h.
3.3. Comparison of TNF-a and IFN-g production between CD4 q and CD8 q cells The variation in cytokine production following 2 h of stimulus for ten normal donors is shown in Fig. 5. There was a considerably greater proportion of CD8 q cells than CD4 q cells producing IFN-g .
cells, maximal IFN-g and TNF-a production and a 4 h stimulus yielded peak IL-2 production. The 4. Discussion
Fig. 5. Th1-like cytokine production in normal donors following 2 h incubation with PMA and ionomycin. The percentage of CD4q Ža. and CD8q Žb. cells producing CD69, TNF-a and IFN-g is illustrated. Each symbol represents a different donor. IFN-g production was greater in CD8q than CD4q cells.
We have developed a rapid, small volume, whole-blood technique for the determination of intracellular cytokines, using commercially available reagents. The advantages of a whole blood method include the small volume of blood required and the avoidance of cell separation procedures which may affect cell populations, cytokine production and activation status. The choice of PMA and ionomycin as stimulants, independent of accessory cells, was demonstrated in earlier work ŽSander et al., 1991., where these reagents were found to be superior to phytohaemagglutinin ŽNorth et al., 1996.. The stimulus is deliberately supra-physiological in an attempt to activate as many cells as possible capable of producing a given cytokine. Antigen-specific stimuli were not used in this study, as the antigen-specific T-cell frequency in normals, as well as in disease states Žfor example common variable immunodeficiency. is very low ŽKondratenko et al., 1997.. Furthermore, it is not yet clear whether antigen stimulation through the T-cell receptor, in combination with co-stimuli, is capable in vitro of inducing adequate cytokine expression measurable by current flow-cytometric technologies. The kinetic study demonstrates that short culture times Že.g. 2 h. for whole blood samples are satisfac-
74
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
tory for IFN-g and TNF-a ; longer times resulted in the detection of lower levels of these two cytokines. The optimal conditions used are a compromise between maximising cytokine production and minimising down-regulation of surface CD4 which prevents efficient discrimination of CD4 q cells. However, for IL-2, 2 h of stimulation was insufficient so 4 h was chosen, accepting the increased difficulties in discriminating CD4 q cells. There are notable differences in the kinetics of cytokine production in this whole-blood system when compared to other studies using PBMC, illustrating the physiological importance of the whole-blood culture system used here. For example, cytokine production in whole-blood systems has been reported to be greater than purified PBMC preparations ŽDe Groote et al., 1992.. Interestingly, the short 2 h optimal whole-blood culture period determined here does not result in significant loss of surface CD4, maintaining the ability to characterise CD4q cells unequivocally. This means that three-colour flow cytometry is suitable for determining intracellular TNF-a and IFN-g production, without recourse to four-colour methods, such as, for example, those following the CD4 population by defining it as CD3-positive CD8-negative. The detection of CD4 in our previous paper ŽNorth et al., 1996. with 12 h of stimulation was less than satisfactory due to this down regulation of the CD4 antigen on prolonged phorbol stimulation. In the kinetics experiment, with more than 4 h of culture with PMA and ionomycin, the CD4q cells were not easily distinguishable from the CD4y population. Stimulation protocols described in previously published whole-blood techniques to measure intracellular cytokines Žeg. Elson et al., 1995. would not, therefore, be able to delineate CD4q populations clearly because of the down-regulation of CD4 that occurs with prolonged stimulation via the PKC pathway. Modifications of the present assay will permit the measurement of cytokines within further subsets of the cells studied, for example CD45ROqCD4q cells. Furthermore, as directly conjugated reagents are produced against additional cytokines and chemokines, with different fluorochromes, future developments will include the ability to assess multiple inducible biologically relevant proteins within single cells. This will permit a wider use of the assessment
of T h 1 and T h 2-like profiles within different diseases, with the potential for monitoring therapy.
References De Groote, D., Zangerle, P.F., Gevaert, Y., Fassotte, M.F., Beguin, Y., Noizat-Pirenne, F., Pirenne, J., Gathy, R., Lopez, M., Dehart, I. et al., 1992. Direct stimulation of cytokines ŽIL-1 beta, TNF-alpha, IL-6, IL-2, IFN-gamma and GM-CSF. in whole blood. I. Comparison with isolated PBMC stimulation. Cytokine 4, 239–248. Elson, L.H., Nutman, T.B., Metcalfe, D.D., Prussin, C., 1995. Flow cytometric analysis for cytokine production identifies T-helper 1, T-helper 2 and T-helper 0 cells within the human CD4qCD27-lymphocyte subpopulation. J. Immunol. 154, 4294–4301. Jung, T., Schauer, U., Heusser, C., Neumann, C., Rieger, C., 1993. Detection of intracellular cytokines by flow cytometry. J. Immunol. Methods 159, 197–207. Kondratenko, I., Amlot, P.L., Webster, A.D.B., Farrant, J., 1997. Lack of specific antibody response in common variable immunodeficiency ŽCVID. associated with failure in production of antigen specific memory T-cells. Clin. Exp. Immunol. 108, 9–13. Maino, V.C., Sunbi, M.A., Ruitenberg, J.J., 1995. Rapid flow cytometric method for measuring lymphocyte subset activation. Cytometry 20, 127–133. Maino, V.C., Ruitenberg, J.J., Suni, M.A., 1996. Flow cytometric method for analysis of cytokine expression in clinical samples. Clin. Immunol. Newslett. 16, 95–98. North, M.E., Ivory, K., Funauchi, M., Webster, A.D.B., Lane, A.C., Farrant, J., 1996. Intracellular cytokine production by human CD4q and CD8q T-cells from normal and immunodeficient donors using directly conjugated anti-cytokine antibodies and three-colour flow cytometry. Clin. Exp. Immunol. 105, 517–522. Petersen, C.M., Christensen, E.I., Andresen, B.S., Moller, B.K., 1992. Internalization, lysosomal degradation and new synthesis of surface membrane CD4 in phorbol ester-activated Tlymphocytes and U-937 cells. Exp. Cell Res. 201, 160–173. Picker, L.J., Singh, M.K., Zdraveski, Z., Treer, J.R., Waldrop, S.L., Bergstresser, P.R., Maino, V.C., 1995. Direct demonstration of cytokine synthesis heterogeneity among human memoryreffector T-cells by flow cytometry. Blood 86, 1408–1419. Prussin, C., Metcalfe, D.D., 1995. Detection of intracytoplasmic cytokine using flow cytometry and directly conjugated anti-cytokine antibodies. J. Immunol. Methods 188, 117–128. Sander, B., Andersson, J., Andersson, U., 1991. Assessment of cytokines by immunofluorescence and the paraformaldehyde–saponin procedure. Immunol. Rev. 119, 65–93. Schauer, U., Jung, T., Krug, N., Frew, A., 1996. Measurement of intracellular cytokines. Immunol. Today 17, 305–306. Thorpe, R., Wadhwa, M., Bird, C.R., Mire-Sluis, A.R., 1992. Detection and measurement of cytokines. Blood Rev. 6, 133– 148.
Determination of intracellular cytokines by flow-cytometry following whole-blood culture W.A Carrock Sewell ) , Margaret E. North, A. David B. Webster, John Farrant MRC Immunodeficiency Research Group, Department of Clinical Immunology, Royal Free Hospital School of Medicine, Rowland Hill Street, London NW3 2PF, UK Received 7 April 1997; revised 31 July 1997; accepted 26 August 1997
Abstract Various methods have been reported for measuring intracellular cytokines in peripheral blood mononuclear cells isolated by density-gradient centrifugation. In this report, we describe a whole-blood method for the determination of intracellular cytokines ŽIFN-g , TNF-a and IL-2. that uses small-volume Ž500 m l. blood samples. Directly conjugated anti-cytokine antibodies and commercial cell membrane fixation and permeabilisation reagents were used. Blood was cultured in a 1:3 dilution with a combination of PMA and ionomycin to reveal the cytokine synthetic potential of each cell, together with monensin to increase the sensitivity by retaining cytokines within the cell to detectable levels. The optimum concentrations of PMA Ž10 ngrml Ž16.2 nmolrl.., ionomycin Ž2 m molrl. and monensin Ž3 m molrl. were determined. Kinetic studies showed maximal cytokine expression after 2 h of culture for TNF-a and IFN-g and 4 h for IL-2. Assessment of TNF-a and IFN-g production within the CD4 and CD8 lymphocytes from 10 normal volunteers showed that considerably more CD8q than CD4 q cells produced IFN-g . This technique could be used by routine immunology laboratories and will be of use in studies to determine whether cytokine assays are of value in the investigation of immune disorders. q 1997 Elsevier Science B.V. Keywords: Intracellular cytokines; Flow cytometry
1. Introduction
Abbreviations: CVID, common variable immunodeficiency; ECD, energy coupled dye; FITC, fluorescein isothiocyanate; FL1,2,3,4, fluorescence channel 1, 2, 3, 4; IFN-g , interferongamma; IL-2, interleukin-2; PBSrArA, phosphate-buffered saline with azide and albumin; PerCP, peridinin chlorophyll protein conjugate; PE, phycoerythrin; PECy5, phycoerythrin-cyanin 5; PMA, phorbol myristate acetate; RPMI1640, Roswell Park Memorial Institute medium 1640; TNF-a , tumour necrosis factor-alpha ) Corresponding author. Tel.: q44-171-7940500 ext. 4932; fax: q44-171-8302224.
The ability to assess cytokine profiles in different immune cells is fundamental to understanding both normal immunoregulation and its dysfunction in disease. Traditionally, cytokines in supernatants from lymphocyte cultures are measured by ELISA or bioassays ŽThorpe et al., 1992.. This provides a useful but incomplete picture, since the relative contribution of individual cell subsets to cytokine production is not determined. With advances in the techniques of cell fixation and permeabilisation it has became possible to identify individual cytokine-
0022-1759r97r$17.00 q 1997 Elsevier Science B.V. All rights reserved. PII S 0 0 2 2 - 1 7 5 9 Ž 9 7 . 0 0 1 5 0 - 6
68
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
producing cells by indirect immunofluorescence microscopy; initially by two-layer staining ŽSander et al., 1991. and more recently with directly conjugated anti-cytokine antibodies. Such methods are useful but laborious and have only a limited ability to distinguish the lymphocyte’s phenotype as well as cytokine production. However, the advent of intracellular cytokine assessment by flow cytometry, staining permeabilised cells with fluorochrome-conjugated anti-cytokine antibodies, permits large numbers of cells of known phenotype to be examined ŽPicker et al., 1995; Prussin and Metcalfe, 1995; Schauer et al., 1996.. Various techniques to measure intracellular cytokines have been reported for peripheral blood mononuclear cells isolated by density gradient centrifugation ŽPrussin and Metcalfe, 1995; North et al., 1996.. These need comparatively large volumes of blood Ž10–20 ml., which is disadvantageous in paediatric studies. Also, the process of cell separation may affect cell activation status ŽMaino et al., 1995. and perturb cytokine production Žsee the whole-blood paper by Maino et al., 1996.. To assess potential cytokine production, various culture conditions Žinvolving changes in culture time, type and dosage of stimulants. have been proposed. Since cytokine production by normal resting cells is minimal, most published techniques use a strong supra-physiological stimulus in culture to demonstrate the potential of each cell to synthesise cytokines. Many of these methods use culture for 10–12 h with stimulants such as phorbol myristate acetate ŽPMA. that activate cells via the protein kinase C pathway; this down-regulates surface CD4 on T-cells ŽPetersen et al., 1992., making identification of CD4q cells and subsequent assessment of cytokines within the Thelper cell subset very difficult. Here we report the development of a robust, rapid, whole-blood flow-cytometric assay for the
determination of IL-2, IFN-g and TNF-a in CD4q and CD8q lymphocytes that requires small volumes of blood, 2 Žor 4. h of culture and which can be performed using equipment available in most clinical immunology laboratories. As an example of its use, comparisons were made of the propensity of CD4q and CD8q lymphocytes from normal subjects to make TNF-a and IFN-g .
2. Materials and methods 2.1. Subjects Normal subjects were healthy laboratory and office personnel Ž5 male, 5 female; mean age 39.7 years.. Informed consent was obtained from all subjects and local ethical committee approval was granted for the study. 2.2. Cell culture and staining Blood Ž1 ml. was collected into lithium heparin and immediately processed; preliminary experiments showed irreproducible changes in cytokine production, such as increased TNF-a , if analysis was not started within approximately 4 h of sampling Ždata not shown.. To increase the quantity of cytokine available for detection within each cell, monensin was added to all cultures to block the export of newly synthesised cytokine from the Golgi apparatus ŽJung et al., 1993.. Blood Ž250 m l. was added to culture medium ŽRMPI1640, Gibco BRL, 500 m l. containing monensin with PMA and ionomycin Ž‘stimulated cultures’. or without Ž‘unstimulated cultures’.. Cultures were carried out in ventilation-capped 5 ml polystyrene round-bottomed plastic tubes ŽGreiner Labortechnik, Gloucestershire..
Fig. 1. Four-colour analysis of IFN-g production by CD4q and CD8q cells. List-mode data were collected through an acquisition gate, to exclude the majority of residual red cells, debris, monocytes and granulocytes. Lymphocytes ŽR1., CD4q ŽR2. and CD8q ŽR3. cells were defined by regions on the corresponding dot-plots. Histograms of the cytokine fluorescence signal were gated using algebraic summation of the regions CD4q AND lymphocytes for CD4q cells and similarly for CD8q cells. Regions for IFN-gq CD4q ŽR4. and IFN-gq CD8q cells ŽR5. were defined on the cytokine histograms using limits defined using the IFN-gy populations in the unstimulated cells. Panel Ža. shows unstimulated cells Žcultured with monensin only.. Panel Žb., from cells stimulated with PMA and ionomycin; the percentage of IFN-gq CD4q in R4 is 8.9% and IFN-gq CD8q lymphocytes in R5 is 26.8%.
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
69
70
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
Both stimulated and unstimulated cultures were incubated at 378C in a humid, 5% CO 2 atmosphere. After incubation, following the manufacturer’s instructions, red cells were lysed using a platform-appropriate lysing solution, namely Optilyse C Ž2 ml for 10 min, Coulter. for use with Coulter Epics-XLMCL four-colour flow cytometers. Other platforms require different lysing solutions, for example Optilyse B ŽCoulter. if an Ortho or Becton-Dickinson flow cytometer is used. After lysing red cells, the remaining cells were washed with phosphate buffered saline Ž2 ml. containing sodium azide Ž0.02%, Sigma. and bovine serum albumin Ž 0.2%, Sigma . ŽPBSrArA. at 300 g for 5 min and the cell pellet resuspended. The cells were then fixed by the addition of Cytoperm Reagent A Ž250 m l, Serotec. for 15 min, following which the cells were washed again in PBSrArA and then permeabilised by the addition of Cytoperm Reagent B Ž250 m l, Serotec.. The exact composition of these commercial solutions is not publicly available, but they are quality-controlled and according to the manufacturers, Reagent A is formaldehyde based and Reagent B is detergent based. Aliquots Ž50 m l. of cells in Reagent B were then added to tubes containing directly-conjugated monoclonal antibodies against various combinations of cytokines and cell surface markers. Saturating concentrations of monoclonal antibodies were used, previously determined in other experiments for the anticytokine antibodies Ždata not shown. or according to the manufacturer’s recommendations for the surface markers. For the anti-cytokine antibodies Žconcentration 100 m grml., 2 m l were used. All samples were stained in the dark at room temperature for 30 min. Samples were then washed once in PBSrArA and stored in 0.5% paraformaldehyde in phosphate buffered saline Ž500 m l. at 48C until acquisition within 24 h. 2.3. Monoclonal antibodies The following monoclonal antibodies were used: anti-human CD4rPE, CD8rPerCP, CD8rFITC, CD69rFITC ŽBecton Dickinson., CD4rECD, CD8rECD ŽCoulter., CD3rPECy5 ŽImmunotech., anti-human TNF-arFITC Žclone B-D9., IFN-
grFITC Žclone B-B1., IL-2rFITC Žclone B-G5. ŽSerotec.. 2.4. Flow cytometry and data analysis A Coulter Epics-XL-MCL four-colour flow cytometer was used for these experiments. Listmode data from 10 000 events was acquired without colour compensation; colour compensation was subsequently applied to the listmode data using the Colour Compensation Toolbox within the WinList Version 3.0 flow-cytometry analysis software ŽVerity, Topsham, ME USA.. The compensation standards used were lymphocytes stained with strongly-positive single-colour monoclonal antibodies, eg. CD3 or CD8, in separate tubes, for each of the four fluorochromes: FITC, PE, ECD and PErCy5. WinList Version 3.0 ŽVerity. was used to analyse the listmode data. For each analysis, dot-plot graphs of forward-scatter versus side-scatter were drawn, and a lymphocyte region R1 was defined. Other regions were defined as illustrated in Fig. 1, namely R2 for CD4 q cells, and R3 for CD8 q cells. These regions encompassed the ‘bright’ cells for each specificity. Separate histograms of cytokine fluorescence were then constructed for CD4 q and CD8 q T-cells. Cytokine-positive CD4 q or CD8 q T-cells were defined by setting regions with the lower-limits for cytokine positivity determined from the unstimulated cultures. Data were expressed as the percentage of CD4 q or CD8 q cells within the lymphocyte gate positive for cytokine. The early activation surface antigen CD69 was also measured on these fixed and permeabilised cells to determine the activation status of the CD4 q and CD8 q cells. In the time-course experiments, surface CD4 became down-regulated with prolonged simulation with PMA, making positive identification of CD4 q cells difficult. To determine the degree of CD4 down-regulation, four-colour flow cytometry was used, measuring CD3, CD4 and CD8 with three separate colours, and cytokine fluorescence in the fourth. The mean CD4 channel fluorescence was measured on CD3q8y cells cultured for 0, 2, 4 and 6 h with PMA and ionomycin. 2.5. Determination of optimal stimulant dose Blood from 5 normal volunteers was cultured for 2 h with ionomycin at 2 m molrl, monensin at a
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
Fig. 2. Determination of optimum stimulant concentrations. Mean"SEM production of CD69, TNF-a , IL-2 and IFN-g by CD4q and CD8q cells is shown Ž ns 5.. In Ža. ionomycin Ž2 m molrl. and monensin Ž3 m molrl. concentrations were held constant and cells were incubated for 2 h with PMA at 5, 10 and 15 ngrml. In panel Žb. the PMA concentration was set at 10 ngrml Ž16.2 nmolrl. and ionomycin concentrations of 1, 2, 5 and 10 m molrl were used. Concentrations of PMA of 10 ngrml Ž16.2 nmolrl., with ionomycin at 2 m molrl gave the maximal cytokine responses for the majority of cytokines tested.
71
72
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
concentration of 3 m molrl and various final concentrations of PMA Ž5, 10 and 15 ngrml.. Intracellular TNF-a and IFN-g production and surface CD69 expression were assessed in CD4q and CD8q cells. The experiment was then repeated using the optimum PMA concentration of 10 ngrml Ž16.2 nmolrl. with various concentrations of ionomycin Ž1, 2, 5 and 10 m molrl.. 2.6. Determination of optimal culture time Blood from 3 normal volunteers was cultured at 378C with monensin Ž3 m molrl. with or without PMA and ionomycin at the optimum concentrations determined above, for various time intervals from 0 to 6 h. Cytokine ŽIL-2, TNF-a and IFN-g . production was again determined in CD4q and CD8q T-cells. 2.7. Comparison of TNF-a and IFN-g production between CD4 q and CD8 q cells Since IL-2 was not optimal after 2 h of culture, only TNF-a and IFN-g , as examples of Th1 cytokines, were measured in CD4q and CD8q cells for this experimental series. Blood from 10 normal volunteers was analysed using the optimum conditions for time and stimulant concentrations determined above, namely the blood was diluted 1r3 giving a final concentration of PMA, 10 ngrml Ž16.2 nmolrl.; ionomycin, 2 m molrl; monensin, 3 m molrl and incubated for 2 h.
3. Results 3.1. Determination of the optimum stimulant dose As shown in Fig. 2Ža. the optimum PMA concentration Ži.e. the concentration producing maximal cytokine expression. was found to be 10 ngrml Ž16.2 nmolrl., when the ionomycin concentration was held constant at 2 m molrl. When this concentration of PMA was used with various concentrations of ionomycin, illustrated in Fig. 2Žb., a 2 m molrl concentration of ionomycin produced the maximal cytokine production. In subsequent experiments, these concentrations were used, together with monensin at a concentration of 3 m molrl.
Fig. 3. Kinetics of cytokine production under optimal conditions. Whole blood was cultured with PMA Ž10 ngrml Ž16.2 nmolrl.., ionomycin Ž2 m molrl. and monensin Ž3 m molrl., for the times shown within Ža. CD4q and Žb. CD8q cells. Stimulation for 2 h gave maximal IFN-g and TNF-a production; 4 h of stimulation resulted in the optimal IL-2 production. Error bars show the SEM.
3.2. Determination of the optimum culture time When blood was cultured with PMA 10 ngrml Ž16.2 nmolrl. and ionomycin 2 m molrl, with monensin 3 m molrl, as determined above, the optimum stimulation time for maximum TNF-a , IL-2 and IFN-g production was assessed. The results are shown in Fig. 3. For both CD4q and CD8q cells, a 2 h stimulus gave clear discrimination of the CD4q
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
73
down-regulation of surface CD4 expression with increasing culture time is illustrated in Fig. 4 and shows a small degree of CD4 down-regulation by 2 h, giving acceptable CD4q cell discrimination, more down-regulation at 4 h but still sufficient to distinguish CD4q from CD4y cells, but poor discrimination after 6 h of stimulation.
Fig. 4. Down-regulation of surface CD4 expression with increasing culture time. The mean"SEM mean CD4 channel fluorescence for three normal donors is shown. Acceptable discrimination of CD4q cells is shown compared to CD8q cells both in unstimulated CD4q cells, and cells stimulated for 2 h. Discrimination is possible, but not ideal, at 4 h, and not possible at 6 h.
3.3. Comparison of TNF-a and IFN-g production between CD4 q and CD8 q cells The variation in cytokine production following 2 h of stimulus for ten normal donors is shown in Fig. 5. There was a considerably greater proportion of CD8 q cells than CD4 q cells producing IFN-g .
cells, maximal IFN-g and TNF-a production and a 4 h stimulus yielded peak IL-2 production. The 4. Discussion
Fig. 5. Th1-like cytokine production in normal donors following 2 h incubation with PMA and ionomycin. The percentage of CD4q Ža. and CD8q Žb. cells producing CD69, TNF-a and IFN-g is illustrated. Each symbol represents a different donor. IFN-g production was greater in CD8q than CD4q cells.
We have developed a rapid, small volume, whole-blood technique for the determination of intracellular cytokines, using commercially available reagents. The advantages of a whole blood method include the small volume of blood required and the avoidance of cell separation procedures which may affect cell populations, cytokine production and activation status. The choice of PMA and ionomycin as stimulants, independent of accessory cells, was demonstrated in earlier work ŽSander et al., 1991., where these reagents were found to be superior to phytohaemagglutinin ŽNorth et al., 1996.. The stimulus is deliberately supra-physiological in an attempt to activate as many cells as possible capable of producing a given cytokine. Antigen-specific stimuli were not used in this study, as the antigen-specific T-cell frequency in normals, as well as in disease states Žfor example common variable immunodeficiency. is very low ŽKondratenko et al., 1997.. Furthermore, it is not yet clear whether antigen stimulation through the T-cell receptor, in combination with co-stimuli, is capable in vitro of inducing adequate cytokine expression measurable by current flow-cytometric technologies. The kinetic study demonstrates that short culture times Že.g. 2 h. for whole blood samples are satisfac-
74
W. Carrock Sewell et al.r Journal of Immunological Methods 209 (1997) 67–74
tory for IFN-g and TNF-a ; longer times resulted in the detection of lower levels of these two cytokines. The optimal conditions used are a compromise between maximising cytokine production and minimising down-regulation of surface CD4 which prevents efficient discrimination of CD4 q cells. However, for IL-2, 2 h of stimulation was insufficient so 4 h was chosen, accepting the increased difficulties in discriminating CD4 q cells. There are notable differences in the kinetics of cytokine production in this whole-blood system when compared to other studies using PBMC, illustrating the physiological importance of the whole-blood culture system used here. For example, cytokine production in whole-blood systems has been reported to be greater than purified PBMC preparations ŽDe Groote et al., 1992.. Interestingly, the short 2 h optimal whole-blood culture period determined here does not result in significant loss of surface CD4, maintaining the ability to characterise CD4q cells unequivocally. This means that three-colour flow cytometry is suitable for determining intracellular TNF-a and IFN-g production, without recourse to four-colour methods, such as, for example, those following the CD4 population by defining it as CD3-positive CD8-negative. The detection of CD4 in our previous paper ŽNorth et al., 1996. with 12 h of stimulation was less than satisfactory due to this down regulation of the CD4 antigen on prolonged phorbol stimulation. In the kinetics experiment, with more than 4 h of culture with PMA and ionomycin, the CD4q cells were not easily distinguishable from the CD4y population. Stimulation protocols described in previously published whole-blood techniques to measure intracellular cytokines Žeg. Elson et al., 1995. would not, therefore, be able to delineate CD4q populations clearly because of the down-regulation of CD4 that occurs with prolonged stimulation via the PKC pathway. Modifications of the present assay will permit the measurement of cytokines within further subsets of the cells studied, for example CD45ROqCD4q cells. Furthermore, as directly conjugated reagents are produced against additional cytokines and chemokines, with different fluorochromes, future developments will include the ability to assess multiple inducible biologically relevant proteins within single cells. This will permit a wider use of the assessment
of T h 1 and T h 2-like profiles within different diseases, with the potential for monitoring therapy.
References De Groote, D., Zangerle, P.F., Gevaert, Y., Fassotte, M.F., Beguin, Y., Noizat-Pirenne, F., Pirenne, J., Gathy, R., Lopez, M., Dehart, I. et al., 1992. Direct stimulation of cytokines ŽIL-1 beta, TNF-alpha, IL-6, IL-2, IFN-gamma and GM-CSF. in whole blood. I. Comparison with isolated PBMC stimulation. Cytokine 4, 239–248. Elson, L.H., Nutman, T.B., Metcalfe, D.D., Prussin, C., 1995. Flow cytometric analysis for cytokine production identifies T-helper 1, T-helper 2 and T-helper 0 cells within the human CD4qCD27-lymphocyte subpopulation. J. Immunol. 154, 4294–4301. Jung, T., Schauer, U., Heusser, C., Neumann, C., Rieger, C., 1993. Detection of intracellular cytokines by flow cytometry. J. Immunol. Methods 159, 197–207. Kondratenko, I., Amlot, P.L., Webster, A.D.B., Farrant, J., 1997. Lack of specific antibody response in common variable immunodeficiency ŽCVID. associated with failure in production of antigen specific memory T-cells. Clin. Exp. Immunol. 108, 9–13. Maino, V.C., Sunbi, M.A., Ruitenberg, J.J., 1995. Rapid flow cytometric method for measuring lymphocyte subset activation. Cytometry 20, 127–133. Maino, V.C., Ruitenberg, J.J., Suni, M.A., 1996. Flow cytometric method for analysis of cytokine expression in clinical samples. Clin. Immunol. Newslett. 16, 95–98. North, M.E., Ivory, K., Funauchi, M., Webster, A.D.B., Lane, A.C., Farrant, J., 1996. Intracellular cytokine production by human CD4q and CD8q T-cells from normal and immunodeficient donors using directly conjugated anti-cytokine antibodies and three-colour flow cytometry. Clin. Exp. Immunol. 105, 517–522. Petersen, C.M., Christensen, E.I., Andresen, B.S., Moller, B.K., 1992. Internalization, lysosomal degradation and new synthesis of surface membrane CD4 in phorbol ester-activated Tlymphocytes and U-937 cells. Exp. Cell Res. 201, 160–173. Picker, L.J., Singh, M.K., Zdraveski, Z., Treer, J.R., Waldrop, S.L., Bergstresser, P.R., Maino, V.C., 1995. Direct demonstration of cytokine synthesis heterogeneity among human memoryreffector T-cells by flow cytometry. Blood 86, 1408–1419. Prussin, C., Metcalfe, D.D., 1995. Detection of intracytoplasmic cytokine using flow cytometry and directly conjugated anti-cytokine antibodies. J. Immunol. Methods 188, 117–128. Sander, B., Andersson, J., Andersson, U., 1991. Assessment of cytokines by immunofluorescence and the paraformaldehyde–saponin procedure. Immunol. Rev. 119, 65–93. Schauer, U., Jung, T., Krug, N., Frew, A., 1996. Measurement of intracellular cytokines. Immunol. Today 17, 305–306. Thorpe, R., Wadhwa, M., Bird, C.R., Mire-Sluis, A.R., 1992. Detection and measurement of cytokines. Blood Rev. 6, 133– 148.