Measurement of phagocyte chemiluminescence in a microtitre plate format

Journal ofImmunologiealMethods, 112 (1988) 163-168 Elsevier

163

JIM04860

Measurement of phagocyte chemiluminescence in a microtitre plate format Andrea L. Blair 1, Ian A. Cree 1, j. Swanson Beck 1 and Mark J.G. Hastings 2 1 Department of Pathology, University of Dundee, Ninewells Hospital and Medical School, Dundee DDI 9S Y, Scotland, U.K., and 2 Department of Medical Microbiology, Royal Hallamshire Hospital, University of Sheffield Medical School, Sheffield, U.K. (Received 15 February 1988, revised received 17 March 1988, accepted 30 March 1988)

Previously described assays of phagocyte chemiluminescence have required large numbers of cells and have not been able to follow responses from a large number of samples in single experiments. Recently, sensitive luminometers which employ a 96 well microtitre plate format have become available. We describe the application of this equipment to the measurement of phagocyte chemiluminescence using lucigenin to enhance the response and the estimation of the opsonic activity of serum. It was found that as few as 5 X 10 4 cells (polymorphonuclear leukocytes or monocytes) per well and a ratio of 10 : 1 zymosan particles to cells gave good results when opsonised with 10% whole serum. This method allows assays of opsonic activity to be performed in triplicate on large numbers of sera with a relatively small number of phagocytes and should aid the investigation of the role of opsonisation in infectious disease. Key words': Chemiluminescence; Luminometer; Microtiter plate; Monocyte; Polymorphonuclear leukocyte; Zymosan; Phagocytosis

Introduction

Chemiluminescence (CL) provides a simple method of assessing phagocyte function in vitro. The basis of the technique is the detection of oxygen radicals produced by phagocytes during the respiratory burst following phagocytosis of particles or stimulation with various humoral factors (Allen, 1977, Glette et al., 1982; Holt et al., 1984). The CL response can be enhanced by compounds such as luminol or lucigenin. Luminol-dependent CL is linked to the myeloperoxidase-H202

Correspondence to: I.A. Cree, Department of Pathology, Ninewells Hospital and Medical School., Dundee DD1 9SY, Scotland, U.K. Abbreviations: CL, chemiluminescence; HBSS, Hanks' balanced salt solution without phenol red; MNC, mononuclear cell; PMNL, polymorphonuclear leukocyte.

system (Breiheim et al., 1984) whereas lucigenindependent CL is independent of myeloperoxidase and is thought to reflect superoxide production by phagocytes (Williams and Cole, 1981; Allen, 1986; Channon et al., 1987). Luminol- and lucigenin-enhanced CL assays have been used to measure the opsonic activity of serum (Allen, 1977; Robinson et al., 1984) and to investigate phagocyte function (Williams and Cole, 1981; Cree and Beck 1986). When assaying serum opsonic activity by CL it is preferable to perform parallel experiments, if possible in triplicate, with a range of test and control sera. This is difficult to achieve using conventional single channel luminometers, even if these have been designed to handle multiple samples. Furthermore such multiple assays often require large numbers of phagocytes because of the size of the cuvettes used. Recently, a novel luminometer designed to use 96 well microtitre plates has become available (Amerlite,

0022-1759/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

164

Amersham International, Amersham, Bucks, U.K.). This instrument has the potential to measure CL from large numbers of samples in a single run.

Materials and methods

Phagocyte preparation Fresh heparinised whole blood from eight healthy volunteers was separated by density gradient centrifugation using Mono-Poly resolving medium according to the manufacturers instructions (Flow Laboratories, Irvine, U.K.). The resulting polymorphonuclear leukocyte (PMNL) and mononuclear cell (MNC) fractions were washed twice in Hanks' balanced salt solution without phenol red (HBSS, pH 7.5, Gibco Bio-Cult, Paisley, U.K.) before being resuspended in 1-2 ml HBSS. Cell purity and concentration were assessed using a Coulter counter with channelyser and the cells diluted to 1 × 1 0 6 P M N L / m l . The percentage of monocytes present in the MNC fraction was estimated by volume spectroscopy (Potts et al., 1980) and the MNC concentration was adjusted to provide 1 × 10 6 monocytes/ml. Zymosan preparation Zymosan was prepared by the method of Lachmann and Hobart (1978) and diluted to 1 X 107 particles/ml in HBSS before use. Serum and zymosan were pre-incubated at 3 7 ° C for 30 rain to opsonise the zymosan. Final serum concentrations between 1% and 20% were used: in some experiments the effect of complement depletion of the serum samples (at 56 ° C for 30 inin) was also investigated. Chemiluminescence assay CL assays were performed using an Amerlite CL reader (Amersham International). The reader was modified by the manufacturer to give a 20 x increase in sensitivity over the standard machine. Initial experiments using concentrations of 5 x 105, 1 x 106, 5 X 10 6 and 1 X 10 v phagocytes/ml in triplicate established that 1 x 10 6 cells/ml ( P M N L or monocytes) gave a suitable response. This concentration was used in all subsequent experiments. Experiments were then performed in

triplicate to investigate the effect of particle:cell ratio (5:1, 10:1, 20:1 and 4 0 : 1 ) and serum concentration (1%, 5%, 10% and 20%) on the CL response. In each assay, 100 ffl of freshly prepared 10 4 M lucigenin in HBSS was added to individual wells of white polystyrene microtitre plates (Dynatech Laboratories, Billingshurst, Sussex, U.K.) and allowed to equilibrate at 3 7 ° C for 30 min. The cells ( P M N L or MNC) were then added (50 #1) and the plate transferred to a microtitre plate incubator (Amersham International). After a 15 min equilibration period background readings were recorded on the Amerlite analyser, every 5 min for 15 20 rain until a steady value was obtained. The pre-incubated mixture of zymosan and serum was then added (100/~1) and the readings continued at 5 min intervals for a further 60-90 min. Each experiment was performed in triplicate and all results were used in subsequent data analysis. Any outliers that were obviously due to technical error were removed from the data file before statistical analysis. The effect of slight agitation (tapping the plate 3 - 4 times between each reading) was also investigated. Control wells with lucigenin alone and lucigenin with cells were included in each experiment; the difference in volume was made up with HBSS to a total of 250 ffl.

Visual counts To corroborate the results of the CL assay, visual assays of zymosan uptake by the phagocytes were performed (Cree and Beck, 1986). The CL experiments were terminated by the addition of 50 ffl of 4% neutral buffered formaldehyde to each well. The contents of the three wells from each experiment were removed from the plate by pipette and washed twice in Isoton II (Coulter Electronics, Luton, U.K.). Air dried drop preparations were stained with 0.2% methyl green. Each slide was viewed by oil immersion microscopy (Olympus model C H A microscope, Olympus Optical Co., London, U.K.) and the number of zymosan particles adherent to or ingested by 100 cells was counted visually. Results were expressed as the percentage of cells showing uptake (i.e., attachment or ingestion) of ten or more zymosan particles.

165

Analysis of data

Results

All results were transferred from the Amerlite plate reader to an on-line IBM-PC compatible microcomputer (Opus II, Opus Supplies, Redhill, Surrey, U.K.) using a standard commercial communications software package (Mirror, SoftKlone, Florida). The data was then analysed using Supercalc 4 (Computer Associates, San Jose, California).

The peak height of the CL response and the rate of increase of the initial response were related directly to the number of cells in the well (Figs. la and lb). Since CL measurements were made at 5 min intervals, a relatively low cell concentration (1 >( 10 6 c e l l s / m l ) w a s u s e d i n subsequent experi-

6,

(b)

(a)

5, 3, 4, .c

.E z:

3=

2,

40

,•20

60

80

'tOO

t50

0

,~0

40

Time (rain)

60

80

100

120

140

Time (rain)

160

X

Fig. 1. The effect of variation in (a) PMNL number, and (b) monocyte number on CL responses for a standard stimulus of 1 107 zymosan particles/ml (100 #l/well) opsonised with 10% serum. Arrow indicates time of addition of pre-incubated zymosan and serum: each point indicates the mean of three experiments performed in parallel. Key: O, 5 × 105 cells/ml; A, 1 5< 106 cells/ml; II, 5 x 106 cells/ml; O, 1 × 107 cells/ml; zx, no zymosan (1 5< 106 cens/ml); ©, no serum (1 × 106 cells/ml).

0.8,

0.3] (b)

(a)

0.7,

0.6,

"~ 0 . 5 , .E -

0.4'

~

0.3'

¢,

0.2, o.1 o.0

0.0

40

60

80

Time (min)

'tOO

120

140

0

2~, II

40

60

80

lO0

't20

140

Time (min)

Fig. 2. The effect of variation of serum concentration on CL responses for (a) 1 × 106 PMNL/ml (50 #l/well), and (b) 1 × 106 monocytes/ml (50 #l/well) with a standard stimulus of 1 × 107 zymosan particles/ml (100 #l/well). Arrow indicates time of addition of pre-incubated zymosan and serum: each point indicates the mean of three experiments performed in parallel. Key: 0, 1% serum; A, 5% serum; I , 10% serum; O, 20% serum; ©, no zymosan (10% serum); zx, no serum.

166

,2t

0 ,1

(a)

(b)

~. 0

0 3

"o

08

.g g

_c 0.5,

~

0 2.

0.4, O.1 0.2,

0.0

0.0 40

60

80

100

120

140

40

50

T i m e (min)

80

100

120

140

Time (min)

Fig. 3. The effect of variation of particle:cell ratio on CL responses for ( a ) 1 x l 0 6 P M N L / m l (50 ffl/well), and (b) 1 x l 0 6 monocytes/ml (50/~l/well). Arrow indicates time of addition of zymosan (100/~l/well) pre-incubated with 10% serum. Key: 0 , 5 : 1 ratio; -, 10:1 ratio; n, 20 : 1 ratio; o, 40 : 1 ratio; o, no zymosan; z~, no serum (10 : 1 ratio).

ments to permit several measurements to be performed before the peak response was obtained. The ability to perform m a n y experiments in parallel allowed simple estimation of the reproducibility of the results from the C1 assay. The standard deviation (arbitrary units) was normally less than 0.01 and the coefficient of variation was below 10% for positive responses. The effect of gentle plate agitation on the CL response was within the limits of experimental error and no reproducible difference was detected. The CL response of P M N L to opsonised zymosan increased with increasing serum concentration (1%, 5%, 10%, 20%) as shown in Fig. 2a, reflecting the degree of opsonisation of the particles. Fig. 2b shows similar results from experiments using M N C (monocytes), but the time to peak height was longer and the peak height was lower than for PMNL. Complement inactivated sera gave CL responses of a similar level to those of the control without serum (results not shown). The effect of particle : cell ratio was also examined using a constant number of cells (1 x 1 0 6 P M N L or monocytes/ml). The CL response increased as the particle : cell ratio increased (Figs. 3a and 3b). Three descriptive parameters were used to describe the CL response: peak height, m a x i m u m rate of increase in CL, and the area below the curve (Table I). All gave acceptable discrimination

between the results. The area below the curve was found to be useful, since it takes account of the rate of increase in CL and the peak height, but comparison of this parameter between experiments requires standardisation of the duration of the assay, which has to be arbitrarily determined. The results of the visual assays of zymosan uptake by P M N L were compared with the CL results from the same experiment (Table I). There

TABLE I T H E E F F E C T O F V A R I A T I O N IN SERUM CONC E N T R A T I O N ON VISUAL COUNTS OF THE PERC E N T A G E OF PMNL (% CELLS) C O N T A I N I N G TEN OR MORE Z Y M O S A N PARTICLES A N D ON THE CL RESPONSE The results of the CL assay are given for experiments performed in triplicate wells, showing the mean and standard deviation of the peak CL response, the maximum rate of rise in CL (slope), and the area under the curve (total CL). % Serum

% Cells

1 29 5 42 10 46 20 57 No serum 17 No zymosan 0

Peak CL

Slope

Mean

SD

1.8403 2.5350 2.6383 3.2406 0.5883 0.148

0.0119 0.0937 0.1210 0.3156 0.0181 0.0030

Total CL (at 103 min)

0.0987 0.0978 0.1013 0.1028 0.0168 0.0031

104.90 172.18 181.95 214.23 28.55 9.45

167

was good agreement between the CL responses and the visual assay results at the different serum concentrations, showing that differences in opsonic capacity were detectable using this technique.

Discussion The Amerlite microtitre plate reader is capable of measuring CL responses from 96 samples at 2 min intervals. Since the current model of the Amerlite is not temperature-controlled, a microplate incubator at 3 7 ° C was used between readings as the CL response is greater and develops more rapidly at 37 ° C than at 25 ° C (Robinson et al., 1984). We have not observed any diminution of the CL response over the 90 s reading period during which the microtitre plate is removed from the incubator, but if the plate is left at ambient temperature (20 o C) for 5 min there is a decrease in the CL response, although the final peak height is not altered if the plate is returned to the incubator for the remainder of the experiment. By using the microplate incubator several microtitre plates can be set up concurrently, thus increasing the number of sera that can be tested in a single run. If four plates are used, it is possible to compare the opsonic activity of up to 100 serum samples (triplicate wells) at one time, provided that other parameters are kept constant (Allen, 1977). If P M N L at a concentration of 1 x 106 cells/ml are separated from fresh venous blood and used in the assay at the same concentration, only 20 ml of blood would be needed for the four plates. Opsonisation of zymosan has been shown to be largely complement-dependent in other assay systems (Robinson et al., 1984; Vernon et al., 1984; Kemp and Turner, 1986) and the use of complement depletion of serum produced similar effects in these experiments. Our results show the expected effect of serum on zymosan opsonisation and the assay is able to distinguish the effect of different serum concentrations and different particle : cell ratios. Inevitably settling of cells will occur within the wells, but simple agitation of the plate did not produce any detectable difference in the assay results. It is not possible to agitate the plates continuously using the Amerlite, although the use

of a heated tray shaker between readings might be of value. Consequently in these experiments, there was probably phagocytosis of opsonised particles by cells in suspension and by cells settled at the bottom of the wells. The results from other experiments suggest that surface phagocytosis does occur in this assay and this may influence the results in some instances (G. Yeaman, personal communication). The low standard deviations (below 0.01) and a coefficient of variation below 10% indicate the excellent reproducibility of the results. Peak height, slope and area under the curve were all found to be useful parameters for comparing the results of different sera, and the parameter worthy of most attention depends upon the shape of the C1 response. In the zymosan assay, we consider the peak response to give the best discrimination. The Amerlite reader permits a greater number of CL experiments to be performed concurrently using phagocytes, either P M N L or monocytes, from a single donor than was previously possible. The technique is also simple to use and is less time consuming than previous methods. This assay could play an important role in the study of opsonisation and it may also prove useful for the comparison of phagocytic defects between patients.

Acknowledgements We would like to thank Mr. R. Fawkes for preparation of the figures and Mr. G. Yeaman and Dr. M.A. Kerr for their advice. This work was supported by a grant from the Scottish Home and Health Department, to whom we are most grateful.

References Allen, R.C. (1977) Evaluation of serum opsonic capacity by quantitating the initial chemiluminescent response from phagocytosing polymorphonuclear leukocytes. Infect. Immun. 15, 828. Allen, R.C. (1986) Methods in enzymology. Biolumin. Bioluminescence and Chemiluminescence, Part B. Methods Enzymol. 133, 449.

168 Andersen, B.R. and Amirault, H.J. (1979) Important variables in granulocyte chemiluminescence. Proc. Soc. Exp. Biol. Med. 162, 139. Breiheim, G., Stendahl, O. and Dahlgren, C. (1984) Intra- and extracellular events in luminol-dependent chemiluminescence of polymorphonuclear leukocytes. Infect. Immun. 45, 1. Channon, J.Y., Leslie, C.C. and Johnston, Jr., R.B. (1987) Zymosan-stimulated production of phosphatidic acid by macrophages: relationship to release of superoxide anion and inhibition by agents that increase intracellular cyclic AMP J. Leukocyte Biol. 41,450. Cree, I.A. and Beck, J.S. (1986) The influence of killed Mycobacterium leprae and other mycobacteria on opsonised yeast phagocytosis. Clin. Exp. Immunol. 64, 35. Glette, J., Solberg, C.O. and Lehrnann, V. (1982) Factors influencing human polymorphonuclear leucocyte chemiluminescence. Acta Pathol. Microbiol. Immunol. Scand. Sect. C. 90, 91. Holt, M.E., Ryall, M.E.T. and Campbell, A.K. (1984) Albumin inhibits human polymorphonuclear leukocyte luminol-dependent chemiluminescence: evidence for oxygen radical scavenging. Br. J. Exp. Pathol. 65, 231.

Kemp, A.S. and Turner, M.W. (1986) The role of opsonins in vacuolar sealing and the ingestion of zymosan by human neutrophils. Immunology 59, 69. Laehmann, P.J. and Hobart, M.J. (1978) Complement technology. In: D.M. Weir (Ed.), Handbook of Experimental Immunology, Vol. 1, 3rd edn. Blackwell, Oxford, Ch. 5A, p. 9. Potts, R.C., Gibbs, J.H., Robertson, A.J., Brown, R.A. and Beck, J.S. (1980) A simple method for determining the extent of cellular contamination in peripheral blood lymphocyte preparations. J. Immunol. Methods 35, 177. Robinson, P., Wakefield, D., Breitt, S.N, Easter, J.F. and Penny, R. (1984) Chemiluminescent response to pathogenic organisms: normal human polymorphonuclear leukocytes. Infect. Immun. 43, 744. Vernon, J., Kemp, A.S., Van Asperen, P.P., Worsdall, P. and Roy, L.P. (1984) Yeast opsonisation by a visual assay and measurement of neutrophil chemiluminescence. J. Clin. Lab. Immunol. 14, 93. Williams, A.J. and Cole, P.J. (1981) Human bronchoalveolar lavage cells and luminol-dependent chemiluminescence. J. Clin. Pathol. 34, 167.