0 1990 Wiley-Liss, Inc.
Cytometry 11:907-912 (1990)
Flow Cytometric Evaluation of Nitro Blue Tetrazolium (NBT) Reduction in Human Polymorphonuclear Leukocytes Andrea Fattorossi,' Roberto Nisini, Soccorsa Le Moli, Guido De Petrillo, and Raffaele D'Amelio Gruppo Igiene ed Immunologia Reparto Medicina, D.A.S.R.S. Aeroporto Pratica di Mare, (A.F., R.N., S.L.M., R.D.A.) and Fondazione Andrea Cesalpino-I Clinica Medica Universita degli Studi La Sapienza, (G.D.P.)Roma, Italy
Oxidative metabolic burst of activated human polymorphonuclear leukocytes (PMN) is most commonly investigated in clinical practice by evaluating nitroblue tetrazolium (NBT) reduction at the single cell level. Reduced NBT precipitates where the redox reaction has taken place and can be visualized as PMN-associated dark blue granules of formazan in light microscopy. Although widely used and not technically demanding, this method remains subjective and labor intensive, especially when large numbers of samples need to be investigated. We developed a new flow cytometry technique in which PMN membrane was rendered fluorescent by a short incubation with fluorescein-conjugated Concanavalin A. PMN were then incubated with NBT and increasing doses of a suitable stimulus, such as phorbol myristate acetate (PMA). Formazan has a distinct peak of absorption at 520 nm that represents the peak of emission of fluorescein. As a consequence, formazan quenches the PMNassociated fluorescence. Data show that
Polymorphonuclear leukocytes (PMN) are of paramount importance for maintaining the immunologic integrity of the individual. This requires PMN to perform several functions, such as ingest and kill foreign organisms (18).Killing is largely performed by a series of redox reactions that take place after activation of a membrane-bound pyridine nucleotide oxidase (10-15). This process is commonly referred to as the respiratory burst. In vitro triggering of the respiratory activity can be obtained by certain soluble agents such a s various phorbol diesters, including phorbol myristate acetate (PMA) (1). This latter is currently used in clinical practice to investigate the respiratory burst of PMN in normal and pathological conditions. Nitroblue tetrazolium
a dose-dependent reduction of fluorescence can be obtained using graded amounts of PMA in normal PMN cultures. PMN-associated fluorescence remains unchanged in control patients with chronic granulomatous (CGD) disease, a disorder characterized by a selective impairment of PMN oxidative metabolism. Electronic cell size increases upon PMA incubation in normal PMN, irrespective of the presence of NBT. Conversely, forward light scatter intensity decreases in the presence, but not in the absence, of NBT indicating that the phenomenon is due to the capacity of formazan to absorb/ scatter the incident light. The present method for easily detecting NBT reducing activity at single cell level by flow cytometry makes use of commonly available, inexpensive reagents and standard instrumentation. It could become a useful test for clinical purposes. Key terms: Polymorphonuclear granulocytes (PMN), oxidative burst, flow cytometry
(NBT) reduction has served as a useful marker to visualize the redox capabilities of PMN upon incubation with PMA (18). NBT is a dye that, after oxidation, precipitates at the site of reaction as dark-blue granules of formazan. The most widely used methods inolve the assessment of NBT reduction microscopically (NBT slide test) or spectrophotometrically, after extraction of
'Address reprint requests to Reparto Medicina, D.A.S.R.S., Gruppo Igiene ed Immunologia Aeroporto Pratica di Mare, 00040 POMEZIA, Rome, Italy.
908
FATTOROSSI E T AL
formazan with suitable solvents. The microscopy-based technique is tedious and time consuming and only provides semiquantitative data, whereas the spectrophotometry-based assay, although reproducible and quantitative, only measures the mean activity of a bulk population of cells and requires large numbers of purified PMN. One of the most valuable aspects of flow cytometry is the ability to make exact measurements on individual cells in time of minutes. This allows for the identification of homogeneous subsets in a heterogeneous population or that of heterogeneities in apparently homogeneous subsets. The aim of the present study was to develop a flow cytometry assay for the measurement of NBT reduction by human PMN using commercial reagents and instruments. The method is based on the observation that membrane-bound oxidase system reduces NBT to formazan a t the outer cytoplasmic membrane level (1) where a fluorescein-conjugated probe has been previously bound. Formazan is a lipophilic molecule with a n absorption curve of incident light that peaks at 520 nm (17). As this value represents the peak of emission spectrum of fluorescein and both substances are in intimate contact on the cell membrane, a quenching of fluorescein emission light, that represents a n estimate of formazan production, can be measured by flow cytometry.
Fluorescent Labeling Fluorescein isothiocyanate (F1TC)-conjugated Concanavalin A (F-Con A) (Sigma Chemical Co., St. Louis, MO) was chosen as a cell surface probe. F-Con A was preferred over other commercially available fluoresceinated probes because i t was found to bind readily to many proteins from human PMN membrane (14). As previously reported (13), Con A does not influence appreciably PMN activities in the conditions of time and temperature used here. Labeling was carried out by incubating cell suspension (either purified PMN o r whole blood after erythrocyte lysis) with F-Con A (Sigma Chemical Co., St. Louis, MO) (25 pg/ml final concentration) for 10 min in a n ice bath. Cells were then washed once to remove unbound F-Con A and resuspended a t 1O”iml in RPMI 1640. Virtually all PMN were stained and assessed by flow cytometry and epifluorescence microscopy equipped with a standard FITC filter and transmitted light set (Leitz, Dialux 20). NBT Reduction Assay NBT reduction was performed with only minor modifications to the standard procedure (4).F-Con A labeled PMN were incubated for 10 min a t 37°C in RPMI 1640 containing NBT and graded amounts of PMA (from 3 to 100 ngiml). At the end of incubation, preparations were centrifuged a t 100 g at 4°C and finally resuspended in 0.3 ml of ice-cold PBS, routinely used for running samples in FACS Analyzer (Becton Dickinson) (5). Controls included unlabeled PMN incubated with the same amount of NBT and PMA, and F-Con A labeled PMN incubated with PMA but in the absence of NBT. In microscopy, the NBT reducing capacity of PMN was expressed as the percentage of cells with membrane-associated formazan granules according to the experimental protocol routinely used for the conventional NBT slide test (4). Cell suspensions were kept in a n ice bath until used. Quantitative NBT reduction assay was performed in parallel as previously detailed (4). Briefly, PMN were incubated at 37°C for 20 min either alone or in the presence of graded amounts of PMA. Samples were then centrifuged and the pellet resuspended with pyridine to extract the reduced NBT. Absorbance was determined spectrophotometrically at 520 nm and taken as a measure of NBT reduction capacity.
MATERIALS AND METHODS Collection of Human PMN For this study, we selected eight normal subjects as well as two patients with chronic granulomatous disease (CGD), a n inherited disorder characterized by the nonfunctioning of PMN oxidative metabolism. In the present CGD patients, routine metabolic studies, including 02-production, chemiluminescence, and NBT slide test, had revealed a total defect of PMN respiratory activity (4). In a preliminary series of experiments, PMN were studied after isolation on density gradient, as previously described ( 5 ) . Briefly, venous blood (5 ml) was mixed with a commercially purchased gelatin solution (Hemagel, Istituto Behring, Italy) and allowed to sediment. The leukocyte-rich supernatant was then layered on Lymphoprep (Nyegaard, Oslo, Norway), centrifuged, and PMN collected a t the bottom of the test tube. In other experiments, centrifugation through Lymphoprep was omitted. Erythrocytes were removed using a lysis step with 0.155 M NH,C1 at 4°C. This allowed the cells to be examined with minimal in vitro modification and required a much smaller Flow Cytometry amount of blood (0.4 ml). PMN were washed with cold RPMI 1640 (Flow Labs, UK) and resuspended a t 106/ml Flow cytometry was performed with a FACS Anain the same medium (viability > 95% by trypan blue or lyzer equipped with a mercury lamp and filtered emisethidium bromide exclusion). Conical polypropilene sion at 488 nm. Cell size and FITC fluorescence due to tubes were used in all experiments to minimize cell F-Con A were measured using the standard Becton adherence and possible unwanted selection of PMN Dickinson equipmentifilter sets. Analog ratio of cell subsets (5). PMN could be easily identified in flow cy- size to fluorescence was also recorded. In some expertometry by gating on appropriate volume and/or scat- iments, flow cytometry was performed using a FACS ter signals. 420 (Becton Dickinson) equipped with a n argon ion la-
NBT REDUCTION TEST IN FLOW CYTOMETRY
909
FIG.1. Dual parameter contour plots of electronic volume (y axis, log scale) vs. fluorescence (x axis, log scale) of normal PMN incubated with NBT and PMA (50 ngiml). Data refer to a typical experiment and were obtained with FACS Analyzer. A Unlabeled and unstimulated
PMN. B: F-Con A labeled unstimulated PMN: all PMN are fluorescent. C: Upon incubation with PMA a reduction in the number of fluorescent PMN and an increase in size can be observed.
ser operated a t 200 mW power (excitation 488 nm) and suitable Becton Dickinson filter sets for FITC. For studies with this cytometer, forward scatter was substituted for electronic volume as a measure of cell size. Fluorescent Monodispersed Carboxylated Microspheres (polysciences Inc., Warrington, PA) served as a means of standardizing fluorescence and scatter signals on a daily basis. For all experiments, cell size and scatter thresholds were established to include only PMN in the analysis. For each experimental condition, 10,000 events were counted. Log amplification was used for all fluorescence analyses and for electronic volume measurements. Light scatter signals were collected with linear amplification. Data were acquired in list mode and processed with a Consort 30 Data Management System (Becton Dickinson). Results are displayed as single- and dual-parameter correlated histograms.
says for assessing NBT reduction. The present flow cytometry technique gives results comparable to the other routinely used assays in discerning patients with impaired oxidative metabolism from healthy subjects. When PMN preparations were examined by light microscopy, the majority of formazan granules appeared located on the outer side of the cell membrane, although a variable amount of granules were also visible inside the cytoplasm. Reducing PMA concentration to suboptimal doses induced PMN to become "blueish" instead of dark blue with a wide range of staining intensity. This made the assessment of the degree of positivity extremely difficult and subjective, and even dependent on the operator's experience and on the quality of optical system used.
Modification of Cell Size FACS Analyzer measurements indicated a n increase in cell size upon incubation with PMA (Figs. 1 and 2b). RESULTS Although the magnitude of the effect was not the same Fluorescence of PMA-Stimulated F-Con a s fluorescence, it was readily quantifiable a s the perA-Labeled PMN centage of cells that had a n electronic volume greater As illustrated in Figures 1 and 2a, the intensity of than a level preselected by using unstimulated PMN. fluorescence was reduced by PMA in a dose-dependent Comparable results were obtained with normal, F-Con manner and a n increasing number of cells fell below a A labeled or unlabeled PMN incubated with or without fluorescence threshold established by using unlabeled NBT in the presence of PMA (not shown). The possicells. In preliminary experiments, PMN were incu- bility that cell aggregation induced by F-Con A andlor bated with FITC-conjugated bovine serum albumin to PMA was responsible for the phenomenon is unlikely. establish fluorescence background level. However, this PMN were used at a density (106/ml) t h a t has been procedure did not consistently modify the autofluores- reported to minimize cell-cell adhesion (10) and pacence of unlabeled PMN and was therefore not used for tients with CGD did not respond with a n increase in subsequent studies. If NBT was omitted from the incu- cell size to PMA stimulation (Fig. 2b). In some cases, however, when a n occasional doublet bation mixture, no change in fluorescence could be obtained (not shown). F-Con A labeled PMN from CGD (or other aggregate) population was observed, appropatients did not modify their fluorescence, irrespective priate electronic gates were established before analysis. When FACS 420 was used and forward scatter subof PMA concentration (Fig. 2a). Table 1 compares results obtained by using three as- stituted for electronic cell volume as a measure of cell
FATTOKOSSI ET AT,
A
7"
I
0
10'
1o2
1o3
10'
1o2
1o 3
70 1
35
0 0
*0°
1
100
-
o! 0
C
1
1
10'
1o2
1o3
P M A [ng/ml] FIG.2. C,rzph,- 7howing the graded modifications occurring in two normal . + and two CGD 43- . 0 PMN preparations assayed in parallel a s in Materials and Methods. Data refer to the whole blood procedure and were obtained with FACS Analyzer. A: Reduction of fluorescence emission. B: Increase of electronic volume. C: Increase of analog ratio of electronic volume to fluorescence.
size, a decreased signal was noticed upon incubation with PMA and NBT (Fig. 3c). If this latter was omitted from the culture mixture, the forward scatter increased (Fig. 3b). As i t can be seen in Figure 2c, the analog ratio of electronic volume and fluorescence emission
may be measured and could be useful for better visualizing the observed changes.
DISCUSSION In routine clinical practice, the oxidative activity of PMN is commonly investigated by NBT reduction slide test assay. PMN-reducing NBT become dark and are scored in light microscopy. In the present paper we show that a n equivalent NBT reduction test can be performed in flow cytometry. Our data show that FCon A labeled PMN can be induced to lose their fluorescence if incubated in the presence of NBT and using PMA as stimulus. The magnitude of the phenomenon depends on the amount of PMA added and is not observed in PMN from CGD patients that lack oxidative function. These observations indicate that the reduction of fluorescence intensity observed in normal PMN is a function of the amount of reduced NBT and thus of the respiratory activity. It has been reported that NBT reduction mainly takes place in the outer side of the cytoplasmic membrane where membrane-bound redox enzymes have been described (1,10,15). Formazan granules would act like a barrier filter since the absorption spectrum of formazan and emission spectrum of FITC show a high degree of overlap (17). The possibility exists, however, that formazan granules also absorb or scatter exciting light, thus contributing to the overall reduction of measured fluorescence emission. Formazan granules are also located inside the cell where they migrate after a phagocytic process of membrane-bound granules has taken place (1).It is likely that these internal formazan granules do not significantly interfere with fluorescence emission of membrane-bound F-Con A; however, they might be of relevance in reducing excitation light. Although fluorescence intensity decreased following PMA concentration, its complete disappearance could not be obtained. In our experience, 50 nglml represent the highest effective concentration of PMA: at this dosage about 50-60% of F-Con A labeled normal PMN lost their fluorescence and could be classified as having fluorescence background only. In optical microscopy, this same concentration of PMA currently induces a higher number of PMN to be classified as having reduced NBT. The reason for this apparent discrepancy is that there is always some uncertainty in establishing a n NBT positivity at optical microscopy level, as there is a wide range in the amount of formazan granules associated with a given cell. Even using optimal PMA concentrations, there are cells with one or two formazan granules that are routinely scored positive, as are cells completely loaded with granules. In our system, the measurement of fluorescence quenching is obtained when a consistent amount of formazan is produced. This makes it possible to grade the extent of cell activation. Available data indicate that circulating PMN are not a n homogeneous population and even normal PMN re-
911
NBT REDUCTION TEST IN FLOW CYTOMETRY
TABLE 1. N B T Reduction Assays in PMA-Stimulated Human PMN“
Subjects
NBT slide test (%Ib
Quantitative NBT test Flow cytometry (OD)“
(%Id
0.04 0.05 0.40 0.10
2 0
I* I
CGD patients 3
1 2
4
Normal subjects (n
=
8)
88 k 3
*
55 2 2.1
“Optimal concentration for each assay: 1 pgiml in NBT slide test and quantitative assay; 50 ng/ml in flow cytometric assay.
bPercentof PMN with detectable formazan granules. “Optical density. dPercent of reduction of fluorescent PMN.
portedly respond in a n heterogeneous manner to a variety of stimuli (6,8,12).The wide range of NBT reduction we observed here is likely to reflect this fact. There are conflicting reports on the quality of the oxidative burst in PMN responding to a n appropriate stimulus. “All-or-none” (7) as well as graded response models ( 2 ) have been described using different flow cytometry based techniques. The present data are in line with the latter hypothesis and also indicate that our test can detect those phagocytes that are not capable of being consistently activated, even in the presence of a n optimal concentration of a suitable stimulus. It is possible, however, that the variability of F-Con A labeling will induce a few NBT reducing PMN to be mistakenly scored negative. Although this fact does not prevent pathological conditions to be detected, e.g., CGD patients, further studies are required to increase the efficiency of fluorescence quenching and thus improve the resolution of the present system. In this regard i t is worth mentioning that, in a n early series of experiments, FITC-conjugated heat-aggregated human immunoglobulins were used instead of F-Con A with similar results. This suggests that the quality of the fluorescent ligand is not relevant for our purposes. The electronic volume of PMN augmented upon incubation with PMA and irrespective of the presence of NBT. The lack of influence of NBT on the phenomenon suggests that the increase in cell volume is a direct consequence of PMA stimulation rather than of formazan granules precipitation. In apparent contrast with these data, measurements made with FACS 420 showed that a reduction of the PMN size occurred upon incubation with NBT and PMA. Conversely, PMN size seemed to increase when NBT was omitted from the incubation medium. As with other laser-powered flow cytometers, FACS 420 measures cell size as a function of forward light scatter. Formazan granules have been reported to absorb andior scatter the laser light (3). It is also well established that light scattering is dependent on a number of variables, in addition to the actual cell size. Cell shape, surface changes and relative refractive indices of the intracellular structures can all contrib-
0
100
200
FORWARD SCATTER FIG.3. Forward scatter distribution of normal PMN upon incubation with NBT and PMA (50 ngiml). Linear scale. Data refer to a typical experiment, representative of three, and were obtained with FACS 420. A: Unstimulated PMN. B: Forward scatter signal of PMAstimulated PMN increases if NBT is omitted. C : Forward scatter signal of PMN decreases upon incubation with PMA in the presence of NBT.
Ute to the measured forward scattering (16). With this background information, and with the data obtained with electronic volume measurements, we conclude that a n actual swelling occurs in PMA stimulated normal PMN, and that formazan granules are responsible for lowering the forward scatter signal. Since light
912
FATTOROSSI ET AL.
scattering has a marked angle dependence, the collection angle of the cytometer may be of major importance. Our machine was used at a 1"to 15" collection angle. Similar results were obtained using a "FACScan" (Becton Dickinson) cytometer that has a fixed 1" to 10" angle. It will be of interest to determine whether other instruments with different light collection angles behave in a similar manner. In some experiments, the decrease in forward scatter signal rendered pre-established electronic gates for PMN in whole blood preparations unsuitable. Thus, care is required when using the present technique if the available instrumentation measures cell size only as a function of narrow angle scattered light. In all experiments performed with FACS Analyzer, the decrease of fluorescence was associated with a n increase of electronic size. This observation prompted us to include the analog ratio of volume to fluorescence as a n additional measurement of PMN activation. Data show that this derived parameter is indeed useful to improve the distinction between resting and stimulated PMN a t the different PMA concentrations. Unfortunately, only few commercially available flow cytometers have the capacity of computing the analog ratio of one measured parameter to another. Several papers have described the flow cytometry measurement of the oxidative metabolic response of human PMN using dichlorofluorescein diacetate (DCFH) (2,9,10). DCFH oxidation cannot be regarded as a n equivalent to the conventional NBT reduction test in clinical use for a number of years. Oxidation of DCFH is mediated by intracellular H,O, (7), whereas NBT reduction mainly depends on membrane-bound oxidase system and O2 production (15).To our knowledge, a study comparing both methods has not yet been performed. Flow cytometric evaluation of intracellular NADPH concentration has a measure of PMN stimulation has been reported (6). This technique requires ultraviolet excitation and appears out of the average possibility of most clinical immunology-oriented flow cytometry facilities, where light source is commonly tuned to 488 nm. In conclusion, the described method allows the most frequently used test for measuring respiratory burst of PMN, the NBT reduction slide test, to be performed by flow cytometry using commercially available reagents and instruments. This technique circumvents the problem of counting only a limited number of cells, which is typical of the conventional microscopy-based assay, it provides objective results, and it is less tedious to perform for the operator. The present method is not technically demanding, i t is not time consuming, and only requires a small amount of blood. It could be a valuable alternative for routine clinical application.
REFERENCES 1. Baehner RL: Subcellular distribution of nitrobluetetrazolium reductase (NBT.R) in human polymorphonuclear leukocytes (PMN). J Lab Clin Med 86:785-792, 1975. 2. Bass DA, Parce JW, Dechatelet LR, Szejda P, Seeds MC, Thomas M: Flow Cytometric studies of oxidative product formation by neutrophils: A graded response to membrane stimulation. J Immunol 130:1910-1917, 1983. 3. Blair OC, Carbone R, Sartorelli AC: Differentiation of HL-60 promyelocytic leukemia cells monitored by flow cytometric measurement of nitro blue tetrazolium (NBT) reduction. Cytometry 6: 54-61, 1985. 4. D'Amelio R, Bellavite P, Bianco P, De Sole P, Le Moli S, Lippa S, Seminara S, Vercelli B, Rossi F, Rocchi G, Aiuti F: Chronic granulomatous disease in two sisters. J Clin Immunol 4:220-227, 1984. 5. Fattorossi A, Nisini R, Pizzolo JG, D'Amelio R: New, simple flow cytometry technique to discriminate between internalized and membrane-bound particles in phagocytosis. Cytometry 10:320325, 1989. 6. Fritsche R, De Weck AL: Chemiluminescence microscopy reveals functional heterogeneity in single neutrophils undergoing oxygen burst. Eur J Immunol 18:817-820, 1988. 7. Hafeman DG, McConnell HM, Gray JW, Dean PN: Neutrophil activation monitored by flow cytometry: stimulation by phorbol diester is an all-or-none event. Science 215:673-675, 1982. 8. Harvath L, Leonard EJ: Two neutrophil populations in human blood with different chemotactic activities: separation and chemoattractant binding. Infect Immun 36:443-449, 1982. 9. Hasui M, Hirabayashi Y, Kobayashi Y: Simultaneous measurements by flow cytometry of phagocytosis and hydrogen peroxide production of neutrophils in whole blood. J Immunol Methods 117:53-58, 1989. 10. Hirabayashi Y, Taniuchi S, Kobayashi Y: A quantitative assay of oxidative metabolism by neutrophils in whole blood using flow cytometry. J Immunol Methods 82:253-259, 1985. 11. Hoffstein ST,Friedman RS, Weissman G: Degranulation, membrane addition, and shape change during chemotactic factor-induced aggregation of human neutrophils. J Cell Biol95:234-341, 1982. 12. Klempner MS, Gallin JI: Separation and functional characterization of human neutrophil subpopulations. Blood 55:659-664, 1978. 13. McPhail LC, Henson PM, Johnston RB: Respiratory burst in human neutrophils. Evidence for multiple mechanisms of activation. J Clin Invest 67:710-716, 1980. 14 Schmalstieg FC, Rudloff HE, Anderson DC: Binding of the adhesive protein complex (LFA-l/Mac-l/pl50,95)to concanavalin A. J Leukocyte Biol 39:193-203, 1986. 15 Schopf RE, Mattar J, Meyenburg W, Scheiner 0, Hammann KP, Lemmel EM: Measurement of the respiratory burst in human monocytes and polymorphonuclear leukocytes by nitro blue tetrazolium reduction and chemiluminescence. J Immunol Methods 67:109-117, 1984. 16 Shapiro HM: Parameters and probes. In: Practical Flow Cytometry. Alan R Liss, New York, 1985, pp 84-158. 17 Stellmach J: Fluorescent redox dyes. 1. Production of fluorescent formazan by unstimulated and phorbol ester- or digitonin-stimulated Ehrlich ascites tumor cells. Histochemistry 80:137-143, 1984. 18 Van Der Valk P, Herman CJ: Biology of disease. Leukocyte functions. Lab Invest 57:127-137, 1987.