Fast double antibody radioimmunoassay of human granolocyte myeloperoxidase and its application to plasma

181

Journal of Immunological Methods, 137 (1991) 181-191 © 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 002217599100108K

JIM 05856

Fast double antibody radioimmunoassay of human granulocyte myeloperoxidase and its application to plasma J. Pincemail 1, G. D e b y - D u p o n t 2,3 C. Deby 1, A. Thirion 4 G. Torpier 5, M.E. Faymonville 3, P. Damas 3, M. Tomassini 5, M. L a m y 3 and P. F r a n c h i m o n t 2 1 Laboratory of Biochemistry and Radiobiology, University ofLibge, B6, Sart Tilman, 4000 Liege, Belgium, 2 Laboratory of Radioimmunology, University Hospital, B23, Sart Tilman, Liege, Belgium, 3 Department of Anaesthesiology, University Hospital, B33, Sart Tilman, Likge, Belgium, 4 Centre de Transfusion Sanguine, Centre Hospitalier Hutois, 2, rue Trois-Ponts, 5200 Huy, Belgium and 5 Centre d'Immunologie et de Biologie Parasitaire, I N S E R M U.167, CNRS 624, rue Professeur Calmette, 1, PB 245-59019 Lille Cedex, France (Received 25 October 1988, revised received 1 June 1990, accepted 19 November 1990)

The haem enzyme myeloperoxidase (MPO) (EC 1.11.1.7) with a spectral A430/A280ratio > 0.7 and a specific activity of 125 U / m g was purified from isolated human neutrophils. To obtain a radioimmunoassay (RIA) for this enzyme, a specific antiserum against human neutrophil MPO was raised in rabbits and used at an initial dilution of 1/10,000. MPO labelled with t25iodine by a technique of self-labelling in the presence of H202, had a specific activity of 24 mCi/mg. After incubation at room temperature (2 h) and separation by double antibody precipitation in the presence of polyethylene glycol, the sensitivity of the RIA was 21 ng/ml. The RIA showed good precision and accuracy with intra- and interassay coefficients of variation of < 7% for MPO concentrations ranging from 100 to 800 ng/ml, and satisfactory recoveries of known amounts of exogenous MPO in plasma. For the measurement of MPO in blood, the best sampling technique was to collect blood into EDTA. Rapid centrifugation (within 20 min) was necessary for blood collected into heparin. Mean MPO values in normal individuals were 340 ___98 ng/ml in EDTA plasma (n = 152) and 332 + 82 n g / m l in heparinized plasma (n = 34). When MPO was measured 12-16 h after injury in critically ill patients high values (above 1000 ng/ml) were found in 6/15 patients with multiple injuries. In patients with sepsis (n = 22), MPO values were always above 1000 ng/ml. Key words: Myeloperoxidase, human; Neutrophil; Radioirnmunoassay; Trauma; Sepsis

Introduction

Myeloperoxidase (MPO) (EC 1.11.1.7) is a haem enzyme with a molecular weight of approximatively 150,000, located in the azurophilic granules

Correspondence to: J. Pincemail, Laboratory of Biochemistry and Radiobiology, University of Lirge, B6, Sart Tilman, 4000 Lirge, Belgium.

of polymorphonuclear leucocytes (PMNLs). MPO, in the presence of hydrogen peroxide produced by PMNLs during their activation (respiratory burst), catalyses the oxidation of chloride anion to yield hypochlorous acid, a strong oxidant species (Klebanoff, 1988). Klebanoff (1968) demonstrated the bactericidal role of the system H202/MPO / C1- during phagocytosis, and observed that C1can be efficiently replaced by iodide. Other workers have confirmed the antimicrobial, antiparasitic and cytotoxic activity of this system. In

182 fact, MPO appears to be one of the main sources of oxidant species generated in vivo, playing a key role in host defense against pathogenic microorganisms. But MPO is also considered to be harmful in pathologic conditions associated with excessive leucocyte activation (e.g., the adult respiratory distress syndrome) (Badwey and Karnovsky, 1980; Klebanoff, 1988; Simon and Ward, 1988). When leucocytes are stimulated, MPO is released into the extracellular medium (Bentwood and Henson, 1980), and its plasma concentration may therefore be taken as a specific index of leucocyte activation in normal and pathological states. Several techniques have been published for the measurement of MPO either in isolated leucocytes or in blood. Enzymatic assays have been proposed but for serum or plasma these techniques are unsuccessful because of interference from components with pseudo-oxidase or oxidase activity such as haemoglobin, haemoglobin-haptoglobin complexes or ceruloplasmin (Malmquist, 1972; Hansen et al., 1975). Malmquist (1972) developed the first immunological technique for MPO measurement using radioimmunodiffusion but this assay proved to be of low sensitivity. Most of the techniques published subsequently have been immunosorbent assays using purified labelled antiMPO immunoglobulins (Hansen et al., 1975; Olofsson et al., 1977a, b; Neumann et al., 1986; Olsen et al., 1986; Oberg et al., 1987; Venge et al., 1989). All these techniques, except for the ELISA technique of Neumann et al. (1986), are time consuming (more than 24 h of incubation). Two groups of authors have reported the direct labelling of MPO with 125iodine using lactoperoxidase coupled to polyacrylamide (Thorell and Larsson, 1974) or chloramine-T and H 2 0 2 (Olofsson et al., 1977a). However, such labelling was found to be difficult (Hansen et al., 1975) and gave a poorly immunoreactive tracer. This paper describes the development of a rapid double antibody radioimmunoassay for MPO in plasma. Different blood sampling techniques were compared in order to establish the normal plasma concentration of this enzyme. The assay permits raPid diagnosis of P M N L activation in critically ill patients.

Materials and methods

Purification of MPO PMNLs were isolated from 30 lit of blood from healthy donors (Borgeat and Samuelsson, 1979). MPO was purified according to the technique of Bakkenist et al. (1978) with minor modifications. Packed PMNLs were homogenized in 0.34 M sucrose, 0.1 M potassium phosphate (pH 7.3). After centrifugation, the pellets were resuspended in 0.1 M potassium phosphate, 0.1 M NazSO 4 (pH 7.3) containing 1% cetyltrimethylammonium bromide (CTBAB). After centrifugation, the clear green supernatant was saved and the pellets were re-extracted in the same buffer but containing 0.5% CTBAB. After a second centrifugation, the two supernatants were combined and ammonium sulphate added to 80% saturation. The precipitate, dissolved in 0.1 M potassium phosphate (pH 7.3), was applied to an ion exchange column (Sephadex SP50, Pharmacia) and eluted with a continuous gradient of 0.1-0.2 M potassium phosphate (pH 7.3). The eluted peak containing the majority of enzymatically active MPO was further purified by gel filtration on Aca 34 (IBF) eluted with 0.1 M phosphate buffer (pH 7.3) containing 0.01% CTBAB. The purity of the final preparation was checked by the ratio of absorbance at 430 nm versus 280 n m (A430/A280) and by analytical electrophoresis on 9% polyacrylamide gel at pH 4.6 and on gradient polyacrylamide gel (7-15%) in the presence of 0.1% sodium dodecyl sulphate (SDS) and 2-mercaptoethanol (0.1 M Tris-giycine buffer, p H 8.3). Enzyme activity was determined by oxidation of o-dianisidine in the presence of H202, one unit of enzymatic activity being defined as the change in absorbance of 1.0 optical density unit per minute at 460 nm.

Immunization Rabbits were injected with 100 /~g MPO dissolved in 1.0 ml phosphate-buffered saline emulsified with 1.0 ml complete Freund's adjuvant. Animals were injected intradermally at 20 sites on the back (Vaitukaitis et al., 1971). Booster injections were given at monthly intervals with 50 /tg of emulsified MPO. Blood samples were collected 10-12 days after each booster injection by cardiac puncture. 12 days after the last booster

183 injection, the animals were exsanguinated by cardiac puncture.

natant was discarded and the tubes counted in an automatic-gamma-counter (LKB).

Labelling

Validation of the RIA: Specificity, sensitivity, precision, accuracy and reproducibility

MPO was labelled with 1251 by a 'self labelling' method (Deby-Dupont et al., 1987, 1988). 5/~g of MPO in 10/~l of 0.5 M acetate buffer p H 5.5 were added to 1 mCi Naa25I (NEN, specific activity + 1 7 C i / m g ) and 10 #1 H202 (1.9 X 10 -4 M; perhydrol, Merck). After 10 min, the reaction was stopped by dilution with 400 /~1 acetate buffer. The labelled MPO was then isolated from free 125I- and degraded material by gel filtration on Aca 34 (IBF) eluted with 0.05 M phosphate buffer pH 7.4 containing 0.5% bovine serum albumin, 0.05% sodium azide and 0.1% poly-L-lysine (Sigma). The specific activity of the tracer was determined by the method of self displacement by labelled antigen in competition with unlabelled antigen.

RIA procedure All reagents were dissolved in 0.05 M phosphate buffer, p H 7.5, containing 0.5% bovine serum albumin and 0.05% sodium azide. 0.5 ng of 125I-MPO (representing about 20,000 cpm) in 100 /~1 buffer containing 1% normal rabbit serum, 100 /~1 of rabbit anti-MPO serum (initial dilution 1/10,000) and either 100 #1 of buffer, or 100 #1 of increasing amounts of unlabelled MPO standards (25-1,000 ng/ml), or 100 #1 of unknown sample were added to the assay tubes. The tube with buffer alone was used to assess the maximal binding of tracer to antiserum in the absence of unlabelled antigen (B0). Tubes were incubated at room temperature for 2 h. Non-specific binding of tracer was measured by the incubation of a25IMPO in the absence of anti-MPO serum. Separation of antibody bound to antigen from free antigen was achieved using a polyethylene glycol accelerated second antibody precipitation (Wood et al., 1979); 1 ml of incubation buffer containing 0.5% Tween, 6% polyethylene glycol 6000, 0.2% microcristalline cellulose and 0.5% sheep anti-rabbit gammaglobulin serum was added to each tube. After incubation for 15 min at room temperature, antigen-antibody complexes were sedimented by centrifugation (20 minutes at 2,000 g). The super-

The cross-reactivity of the antiserum was tested with plasma, eosinophil and neutrophil proteins at concentrations ranging from 100 to 20,000 n g / m l and also with cellular components of blood. The proteins tested were human serum albumin (Sigma) and haemoglobin (Sigma), leucocyte elastase (Bachem), lactoferrin (Sigma), cathepsin G (Bachem) and peroxidase from eosinophils. The cellular c o m p o n e n t s tested were platelets, lymphocytes, neutrophils and eosinophils. The absence of cross-reaction of the antiserum with eosinophil peroxidase was also tested by immunoelectron microscopy (see below). Sensitivity was defined as the minimum amount of unlabelled MPO which caused a reduction in the percent of tracer bound to antibody which was greater than twice the standard deviation of ten determinations of B0. Precision was estimated from the error profile derived from the whole working range of the assay (Ekins, 1969). The coefficient of variation was calculated for ten determinations of each MPO concentration. Reproducibility was measured by the coefficient of variation calculated for at least five determinations of the same sample within or between assays. Accuracy was estimated by the recovery of known amounts of MPO (100, 200, 400, 600 and 800 ng) added to human plasma and by the parallelism between the dilution curve of an MPO rich plasma and the standard curve.

Immunoelectron microscopy Purified human neutrophils and eosinophils were fixed in 1% glutaraldehyde and embedded in Lowicryl K4M according to the standard procedure of Roth et al. (1981). Thin sections on nickel grids were incubated for 10 min on a drop of Tris-HCl-buffered saline (20 m M Tris-HC1, 0.5 M NaC1) p H 7.4 (TBS) containing 0.5% ( w / v ) ovalbumin, supplemented with 1% heat-inactivated normal goat serum. This was followed by incubation with the anti-human neutrophil

184 MPO antibodies diluted in TBS containing ovalbumin (TBS-OVA) for 2 h at room temperature. The grids were then rinsed with TBS-OVA and incubated on a drop of gold-conjugated goat anti-IgG (1/40) (Janssen Pharmaceutica). Afer incubation for 1 h at room temperature, sections were thoroughly washed with TBS, rinsed with distilled water, and dried. Lowicryl sections were stained with uranyl acetate and lead citrate before examination with a Philips 420 electron microscope.

Normal and pathological values of MPO in blood Blood samples from healthy donors were treated with anticoagulant (either EDTA 1 m g / m l or 12.5 IU h e p a r i n / m l of blood, or 1 ml of a 0.15 M sodium citrate solution for 9 ml of blood) and immediately centrifuged, or were allowed to clot at room temperature (12 h). Plasma or sera were frozen at - 2 0 ° C until assayed. In order to study the effect of a delay between blood sampling and centrifugation, blood was taken from 10 healthy donors either into E D T A or heparin and centrifuged at several times after sampling (immediately (to) , after 10 min (tl) , after 1 h (t2) , 2 h (t3) and 18 h (t4). Values in pathological states were determined in polytrauma and septic patients, in two leukaemic patients and in one patient with periodic agranulocytosis. Blood was taken within the first 6 h after trauma from 15 patients with multiple injuries (three women, 12 men) admitted to the intensive care unit of the university hospital. The blood was anticoagulated with heparin and immediately centrifuged. These patients presented with at least three major injuries (chest, limbs, pelvis) and a mean injury severity score of 45 + 14 (Baker and O'Neil, 1976). 13 out of them were admitted in shock. Blood from 22 patients (ten women, 12 men) with sepsis was taken into E D T A and immediately centrifuged. Seven of these patients were in septic shock at time of blood sampling. The two leukaemic patients were suffering from lymphoid leukaemia and had leucocyte counts of 2600 and 800 cells/mm 3 at the time of blood sampling (into EDTA), with 15% and 1% of granulocytes respectively. In the last patient, a

women with periodic agranulocytosis, five blood samples were taken into E D T A during five different periods of agranulocytosis. All samples were frozen at - 2 0 ° C until assay.

Results

Purification of MPO After ion exchange chromatography (Fig. 1A), enzymatically active MPO was found in the second peak to be eluted (fractions 15-20). This was further purified by gel filtration chromatography where the maximal MPO activity was found in fractions 46 to 50 (Fig. 1B). The ratio of absorbance Aa3o/A280 of the enzyme preparation was >_ 0.7 and its specific activity was 125 U / m g . The analytical polyacrylamide gel electrophoresis at p H 4.6 revealed a single protein band (MW approximately 150,000) with enzymatic activity. The analytical SDS gradient polyacrylamide gel electrophoresis in the presence of mercapto-ethanol revealed three bands of variable staining intensity with molecular weights of 90,000, 60,000 (the most intense band) and 15,000. Immunization After 4 months of immunization, two rabbits produced antisera with similar characteristics of affinity and specificity. These were used in the RIA at an initial dilution of 1/10,000, binding 30-40% of labelled antigen (20,000 cpm) after 2 h of incubation. Under the conditions of the assay, the binding of tracer to the antibodies reached a plateau after 2 h, with a mean non-specific binding of 2.5%. Incubation at 3 7 ° C resulted in a maximal binding of tracer after 1 h, but degradation occurred rapidly. An incubation time of 2 h at room temperature was thus chosen as optimal. Labelling The self labelling of MPO with Na125I in the presence of H 2 0 2 , followed by gel filtration on Aca 34 in phosphate buffer containing 0.1% polyL-lysine, gave a tracer free from degraded MPO. The 125I-MPO obtained by this technique corresponded to the heavy subunit of MPO (MW 60,000), had good immunoreactivity with low non-specific binding, and remained stable and im-

185 IJ / rn I

A280 n m

A280 nm

U/ml

0,4 "1

160

0,30

• 300

0,3 120

0,2

r

0,20 '

• 200

0,2 80 0,1

100

0,10 '

0,1 ' 40

0,0 4~ 0

, 10

0 20

40

30

Fraction

number

0,00

0

30

40

50

6O

Fraction

70

number

Fig. 1. Purification of MPO: chromatographic steps. A: elution profile on a cation exchange column (Sephadex SP50 Pharmacia; 15 x 1.5 cm; potassium phosphate 0.1 M, p H 7.3; NaC1 gradient from 0.1 to 0.2 M). B: elution profile on a gel filtration column (Aca 34 IBF; 50 x 3 cm; 0.1 M phosphate buffer, 0.015[ cetyltrimethyl-ammonium bromide, p H 7.3). Ordinates: [] n, absorbance at 280 n m (left scale); • • , enzymatic activity (right scale).

munoreactive for 40 days after labelling (DebyDupont et al., 1987, 1988). The mean specific activity calculated for five different labelling assays was 24.0 + 1.2 m C i / m g (about 1.5 atoms 125I/molecule of MPO).

Validation of RIA Fig. 2 shows a standard RIA curve, with standard deviations calculated from ten measurements for each unlabelled MPO concentration. The antiserum, used at a dilution of 1/10,000, was specific. Its cross-reaction with human albumin, haemoglobin, leucocyte elastase, cathepsin G and lactoferrin was less than 0.001%. MPO was neither found in platelets (320 × 106 cells) nor in lymphocytes (1.7 x 10 6 cells). In isolated neutrophils, the MPO concentration correlated with

the number of cells (4.5 x 10 -3 to 7.5 x 10 -3 ng of MPO/cell). With pure human eosinophil peroxidase (at concentrations up to 1000 ng/ml), the cross reaction of anti-MPO serum was less than 0.01%. The absence of cross reactivity was confirmed by the electron microscopy studies (Fig. 3). Gold-labelled anti-MPO antibodies were localized in the specific granules of neutrophils whereas they were not present in the specific granules of eosinophils. As a control, gold-labelled antieosinophil peroxidase antibodies were detected in the matrix of the eosinophil granules but were absent from the neutrophil granules. The minimal concentration of MPO detected by the RIA was 21 n g / m l (Fig. 2). The error profile (Fig. 4) indicates excellent reproducibility for unlabelled MPO concentrations ranging be-

186

tween 50 and 400 n g / m l with a coefficient of variation (CV) lower than 2%. For the determination of MPO in plasma, the coefficient of variation (intra- or interassay) never exceeded 7% for M P O values ranging from 100 to 800 n g / m l (Table I). For higher M P O values, it increased above 10%. Amounts of MPO (from 100 to 800 n g / m l ) added to normal plasma were correctly measured with a mean recovery of 96 + 5%. This accuracy of the RIA is confirmed by the parallelism between the standard curve and the dilution curve of M P O rich plasma (Fig. 2).

B/Bo) ~ 100 100 9(3

"~

8C

70

"~'~\

6o

\

50

4O 3C 2C

21ng/ml I

I

I

| 1

I I / /

I

5 0 1 0 0

1~6

1/~

I

1~4

I

. . . . . .

500 1000 MPO

ng/ml 1~1 plasma dllut Ion t~2

• ). Ordinates: Fig. 2. Standard curve of RIA for MPO (e ratio (in percent) between the amount of tracer bound to antibody in the presence (B) and in the absence (B0) of unlabelled MPO. Abscissae: increasing concentrations of unlabelled MPO in n g / m l (logarithmic scale). Limit of sensitivity of the R1A: 21 n g / m l (1"). • . . . . . . • , curve obtained with successive dilutions of a MPO rich plasma.

TABLE I R E P R O D U C I B I L I T Y (INTRA- A N D INTERASSAY) OF PLASMA MPO M E A S U R E M E N T N: number of measurements. SD: standard deviation. CV: coefficient of variation. N

Normal and pathological values Normal values. The mean M P O concentrations measured in blood varied considerably with the anticoagulant used and with the time which elapsed between sampling and centrifugation. Table II presents the mean M P O values measured in the same volunteers in serum or plasma obtained using EDTA, heparin or citrate as anticoagulant and centrifuged immediately after sampling. High values were found in serum and blood anticoagulated with citrate, with a wide range of concentrations. However, when blood was centrifuged immediately after sampling and when erythrocytes as well as leucocytes were discarded before clotting, the M P O values in serum were similar to those in E D T A plasma (300 n g / m l versus 340 n g / m l , and 340 n g / m l versus 360 n g / m l for two healthy volunteers). The mean values of M P O in E D T A or heparinised plasma were comparable (340 and 332 n g / m l respectively). Fig. 5 shows the frequency distribution of M P O in normal individuals.

Mean value (ng/ml)

SD

CV (%)

320 440 520 620 753 880 1,200

19 20 30 40 50 50 190

5.9 4.5 5.7 6.4 6.6 5.7 15.8

TABLE II

Anticoagulant

n

MPO (ng/ml)

SD (ng/ml)

Range

265 367 514 868

13 18 29 100

4.7 4.9 5.7 11.6

EDTA Heparin Citrate None (serum)

152 34 32 37

340 332 1,100 1,200

98 82 400 400

142-450 220-436 282-1,800 330-1,890

Intra-assay 6 8 8 10 10 10 7

EFFECT OF T H E A N T I C O A G U L A N T USED IN BLOOD S A M P L I N G ON MEAN MPO VALUE SD: standard deviation.

lnterassay 6 5 6 5

187

Fig. 3. Immunoelectron microscopic localization of MPO in neutrophils (Bar ~ 1 #m). Gold-labelled antibody is localized in the granules (arrows) within the neutrophils (N). Adjacent eosinophil (E) is the control for the labelling: very few gold particles (arrowheads) are detected over the specific granules. CV

(%)

1

.

10

.

.

.

.

.

.

.

,

100

.

.

.

.

.

.

.

.

i

M]f~

1000

(ng/ml)

Fig. 4. Error profile of RIA for MPO (CV: coefficient of variation).

The effect of delaying centrifugation after blood sampling on the measurement of MPO in plasma is shown in Table III. MPO values in E D T A plasma remained stable whatever the time between sampling and centrifugation. When blood was drawn into heparin, the plasma MPO value remained stable for a delay between sampling and centrifugation not exceeding 20 min, but increased then with time, reaching very high values after 2 h. Pathological values (Fig. 6). In 15 patients with multiple injuries it was observed that in the first hours after trauma, the MPO value in heparinized plasma was above the mean normal value of 332 n g / m l ; MPO values higher than 1000 n g / m l were found in six patients, indicating early activation and degranulation of polymorphonuclear leucocytes. In 22 patients with sepsis the MPO concentrations in EDTA anticoagulated plasma were above 1000 n g / m l with some values higher than

188 MPO ng/ml

Frequency

10000 80

"')*

9000 8000

70 ¸

7000

rio ¸

6000

~o o o

5000

50 ¸

40 ¸

4000

o

3000

~°

2000

oo oo

1000

30'

0

20'

Sepsis n=22

Multiple injuries n=15

Fig. 6. Distribution of MPO values in 22 infected patients and in 15 patients with multiple injuries. *, MPO values above 7000 ng/ml (8800, 25,680, 30,000, 39,000 and 43,480 ng/ml) were measured in five patients. ~ : mean normal value + SD.

10'

0 ~- ~ ~ ~ ~

~ ~ ~

ng/ml

MPO values

Fig. 5. Distribution of MPO values in healthy volunteers. A: plasma ant±coagulated in EDTA (n =152). B: plasma anticoagulated in heparin (n = 34).

10,000 n g / m l , an i n d i c a t i o n of a c t i v a t i o n and d e g r a n u l a t i o n of neutrophils. In the t w o l e u k a e m i c patients, the M P O levels in p l a s m a were u n m e a s u r a b l e , an d in the p a t i e n t with p e r i o d i c agranulocytosis, M P O values were very low (104, 33, 50, 32 a n d 20 n g / m l ) d u r i n g p er i o d s of low g r a n u l o c y t e counts.

TABLE III ROLE OF THE DELAY BETWEEN BLOOD SAMPLING AND CENTRIFUGATION ON PLASMA MPO VALUE (n =10) to, direct centrifugation; tl, centrifugation after 10 min; r2, centrifugation after 1 h; t3, centrifugation after 2 h; t4, centrifugation after 18 h. Anticoagulant

to tI t2 l3

t4

EDTA (1 mg/ml)

Heparin (12.5 IU/ml)

330 + 100 354± 87 340± 100 340_+ 100 355 _+100

332 ± 90 339_+ 92 1,080+_400 > 1,800 > 1,800

Discussion

RIA conditions F o l l o w i n g the t e c h n i q u e of Bak k en i st et al. (1978), we o b t a i n e d a p u r e M P O p r e p a r a t i o n with an a c c e p t a b l e a b s o r b a n c e ratio a n d an electrophoresis p r o f i l e in a g r e e m e n t with p r e v i o u s l y published d a t a (Bakkenist et al., 1978; A n d r e w s an d Krinsky, 1981). In S D S - p o l y a c r y l a m i d e g r a d i e n t gel electrophoresis we s e p a r a t e d the h e a v y a n d light chains of the enzyme, as well as a third c o m p o n e n t with a high m o l e c u l a r weight, identified as a large p r e c u r s o r or a t r an si en t interm e d i a t e of the m a t u r e enzyme, the p r e s e n c e of

189 which was also reported by Olsson (1987), Svensson et al. (1987) and Olsen et al. (1986). This third component was absent in some preparations, which reinforces the hypothesis that it was an intermediate or precursor of MPO. Like the mature enzyme, this precursor was immunoreactive. In the development of the RIA, the major difficulty arose from the labelling of MPO, previously described as difficult for 'unexplained reasons', producing a 12~I-MPO with low specific activity and poor immunoreactivity (Hansen et al., 1975). Thorell et al. (1974) compared three different techniques (chloramine-T, lactoperoxidase and polyacrylamide-coupled lactoperoxidase) and the best specifc activity of a25I-MPO was obtained with coupled lactoperoxidase which nevertheless remained low. Olofsson et al. (1977a) used chloramine-T together with H 2 0 z, but they did not explain the reason for choosing this labelling technique. Taking into account the capacity of MPO to fix halide in the presence of H202 (Klebanoff, 1968), we labelled MPO in the presence of H202 alone and obtained an iodinated molecule with acceptable specific activity (Deby-Dupont et al., 1988). During the purification step by gel filtration chromatography, the use of 0.1% poly-L-lysine in the buffer helped to avoid damage by accelerating the elution of the labelled molecule, so that the purified 125I-MPO could be used for 40 days without loss of immunoreactivity. Our RIA conditions were similar to those used by Hansen et al. (1975), with antiserum at an initial dilution of 1/6400. However, instead of an overnight incubation at 4 ° C we were able to measure MPO after a 2 h incubation at room temperature. No crossreactions of the antiserum were observed with human albumin, haemoglobin, platelet proteins or lymphocyte proteins. In agreement with previously published data of Neumann et al. (1986), elastase, cathepsin G and lactoferrin, three granulocytic proteins, were not detected by the antiserum. Moreover, purified eosinophil peroxidase did not react with anti-MPO serum, as documented by RIA and by electron microscopy. The error profile, the correct recovery of added amounts of MPO in plasma and the intra- and interassay coefficients of variation (comparable with the data of Neumann et al., 1986) indicated good accuracy, precision and reproducibility.

Hansen et al. (1975) and Olofsson et al. (1977a) reported assay sensitivities of 100 n g / m l and 50 n g / m l respectively. We obtained a better sensitivity of 21 n g / m l , sufficient for the measurement of MPO in plasma, but lower than the sensitivity of 0.25 n g / m l obtained by Neumann et al. (1986) with an ELISA technique. Nevertheless, an increase in the dilution of the antiserum (1/80,000) together with an increase of the incubation time (48 h at room temperature) permitted us to reach a sensitivity of 0.5 n g / m l (unpublished data).

Normal and pathological values The mean normal values of plasma MPO published in the literature vary largely with the technique of measurement. Using an ELISA technique, Neumann et al. (1986) found 36 n g / m l , and with a solid-phase immunoassay. Oberg et al. (1987) and Olsson et al. (1979) found 113 and 144 ( + 4 8 ) n g / m l respectively. The values obtained with the double antibody radioimmunoassay were higher, lying between 300 and 500 n g / m l (Hansen et al., 1975). Hansen et al. (1975) and Olofsson et al. (1977a,b) have noted a significant influence of the blood sampling technique on the mean MPO value in the plasma of normal volunteers; for this reason they recommanded the use of E D T A as anticoagulant and rapid centrifugation of blood after sampling. However, they only studied a few blood samples. We also observed that mean normal values of MPO varied over a wide range depending on the anticoagulant used. For plasma antigoagulated with EDTA, we found a value of 340 + 98 n g / m l (mean ___SD) comparable with the normal values (from 300 to 500 n g / m l ) obtained in E D T A anticoagulated plasma by Hansen et al. (1975). This value remained stable, even when blood was centrifuged 18 hours after sampling. In plasma drawn into heparin, the mean MPO value was 332 ng/ml, when the blood was quickly centrifuged after sampling. A delay of more than 20 min between sampling and centrifugation led to increased values of MPO, with very high values (1800 n g / m l ) when the blood was centrifuged 1 h after sampling. In plasma anticoagulated with citrate and in serum, mean normal values increased to 1100 and 1200 n g / m l respectively, despite rapid centrifugation after sampling. Our ob-

190

servations concerning the essential role of the anticoagulant technique used in blood sampling demonstrate the importance of fast and early processing of blood samples in order to avoid in vitro leucocyte activation. In patients with trauma and sepsis, the plasma MPO concentrations were abnormal. Values above 1000 n g / m l were found in six of the 15 patients in the first hours after trauma. In 22 patients with sepsis we found MPO values above 1000 n g / m l , often with very high values, occasionally reaching 10,000 ng/ml. These results indicate early activation and degranulation of neutrophils in trauma and particularly in sepsis, as was previously observed by Jochum et al. (1986) and by Dittmer et al. (1985) in a small series of patients. In two patients with lymphoid leukaemia and in one patient with agranulocytosis, low MPO levels were observed in association with a fall in polymorphonuclear leucocyte counts. However, the number of patients studied is too small for any conclusion to be drawn regarding an association between leucocyte counts and MPO levels. In the literature, such an association has been discussed extensively (Hansen et al., 1975; Olofsson et al., 1977a,b; Olsson et al., 1979).

Acknowledgements The authors wish to thank Mrs. M. De BruynDister, Mrs. J. Bourdon-Neuray, Miss A. Dethier and Mr. R. Vignette for excellent technical assistance, and Mrs. V. Salemme-Casertano for typing the manuscript. They are also indebted Dr. G. Hartstein for careful correction of the manuscript. This work was supported by the Fonds National de la Recherche Scientifique (FRSM) Grant No. 3.4519.86.

References Andrews, P.C. and Krinsky, N.I. (1981) The reductive cleavage of myeloperoxidase in half, producing enzymatically active hemi-myeloperoxidase. J. Biol. Chem. 256, 4211. Badwey, J.A. and Karnovsky, M.L. (1980) Active oxygen species and the function of phagocytic leukocytes. Ann. Rev. Biochem. 49, 695.

Baker, S.P. and O'Neil, B. (1976) The injury severity score: an update. J. Trauma 16, 882. Bakkenist, A.R.J., Wever, R., Vulsma, T., Plat, H. and Van Gelder, B.F. (1978) Isolation procedure and some properties of myeloperoxidase from human leukocytes. Biochim. Biophys. Acta 524, 45. Bentwood, B.J. and Henson, P.M. (1980) The sequential release of granule constituents from human neutrophils. J. Immunol. 124, 855. Borgeat, P. and Samuelsson, B. (1979) Arachidonic acid metabolism in polymorphonuclear leukocytes: Effects of ionophore A23187. Proc. Natl. Acad. Sci. U.S.A. 76, 2148. Deby-Dupont, G., Pincemail, J., Thirion, A. and Deby, C. (1987) A radioimmunoassay for polymorphonuclear leukocytes myeloperoxidase: preliminary results. Arch. Int. Physiol. Biochem. 95, $9. Deby-Dupont, G., Pincemal, J. and Reuter, A.M. (1988) Iodination of polymorphonuclear leucocytes myeloperoxydase. Arch. Int. Physiol. Biochem. 96, B25. Dittmer, H., Jochum, M. and Schmidt-Neuerburg, K.P. (1985) Der PMN-Elastase-Plasmaspiegel, ein biochemischer Parameter der Traumaschwere. Chirurgia 56, 723. Ekins, R.P. (1969) Problems of sensitivity with special reference to optimal conditions. In: M. Margoulis (Ed.), Protein and Polypeptides hormones. Excerpta Medica Foundation ICS 161, Amsterdam, p. 672. Hansen, N.E., Malmquist, J. and Thorell, J. (1975) Plasma myeloperoxidase and lactoferrin measured by radioimmunoassay: relations to neutrophils kinetics. Acta Med. Scand. 198, 437. Jochum, M., Witte, I., Duswald, K.W., Inthorn, D., Welter, H. and Fritz, H. (1986) Pathobiochemistry of sepsis: role of proteinases, proteinase inhibitors and oxidizing agents. Behring Inst. Mitt. 79, 121. Klebanoff, S.J. (1968) Myeloperoxidase-halide-hydrogen peroxide anti-bacterial system. J. Bacteriol. 95, 2131. Klebanoff, S.J. (1988) Phagocytic cells: products of oxygen metabolism. In: J.I. Galfin, I.M. Golstein and R. Snyderman (Eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, New-York, p. 391. Neumann, S., Gunzer, G., Lang, H., Jochum, M. and Fritz, H. (1986) Quantitation of myeloperoxidase from human granulocytes as an inflammation marker by enzyme-linked immunosorbent assay. Fresenius Z. Anal. Chem. 324, 365. Oberg, G., Dahl, R., Ellegaard, J., Sundstrrm, C., Vaeth, M. and Venge, P. (1987) Diagnostic and prognostic significance of serum measurements of lactoferrin, lysozyme and myeloperoxidase in acute myeloid leukaemia (AML): recognition of a new variant, high-lactoferrin AML. J. Haematol. 38, 148. Olofsson, T., Olsson, I., Venge, P. and Elgefors, B. (1977a) Serum myeloperoxidase and lactoferrin in neutropenia. Scand. J. Haematol. 18, 73. Olofsson, T., Olsson, I. and Venge, P. (1977b) Myeloperoxidase and lactoferrin of blood neutrophils and plasma in chronic granulocytic leukaemia. Scand. J. Haematol. 18, 113.

191 Olsen, R.L., Steigen, T.K., Holm, T. and Little, C. (1986) Molecular forms of myeloperoxidase in human plasma. Biochem. J. 237, 559. Olsson, I. (1987) Regulation of myelopoiesis and maturation of phagocyte functions. Eur. J. Clin. Invest. 17, A57. Olsson, I., Olofsson, T., Ohlsson, K. and Gustavsson, A. (1979) Serum and plasma myeloperoxidase, elastase and lactoferrin content in acute myeloid leukaemia. Scand. J. Haematol. 22, 397. Roth, J., Bendayan, M., Carlemalm, E., Villinger, W. and Garavito, M. (1981) Enhancement of structural preservation and immunocyto-chemical staining in low temperature embedded pancreatic tissues. J. Histochem. Cytochem. 29, 663. Simon, R.H. and Ward, P.A. (1988) Adult Respiratory Distress Syndrome. In: J.I. Gallin, I.M. Goldstein and R. Snyder° man (Eds.), Inflammation: Basic Principles and Clinical Correlates. Raven Press, New York, p. 815. Svensson, B.E., Domeij, K., Lindvall, S. and RydeU, G. (1987) Peroxidase and peroxidase-oxidase activities of isolated hu° man myeloperoxidase. Biochem. J. 242, 673.

Thorell, J.l. and Larsson, I. (1974) Lactoperoxidase coupled to polyacrylamide for radioiodination of proteins to high specific activity. Immunochemistry 11,203. Vaitukaitis, J., Robbins, J.B., Nieschlag, E. and Ross, G.T. (1971) A method for producing specific antisera with small doses of immunogen. J. Clin. Endocrinol. 33, 988. Venge, P., Hallgren, R., Stalerdaeim, G. and Olsson, I. (1979) Effects of serum and cations on the selective release of granular proteins from human neutrophils during phagocytosis. Scand. J. Haematol. 22, 317. Wood, W.G., Stella, G., Muller, O.A. and Scriba, P.C. (1979) A rapid and specific method for separation of bound and free antigen in radioimmunoassay systems. J. Clin. Chem. Clin. Biochem. 17, 111. Zgliczynski, J.M., Stelmaseynska, T., Domanski, J. and Ostrowski, W. (1971) Chloramines as indermediates of oxidation reaction of amino acids by myeloperoxidase. Biochim. Biophys. Acta 235, 419.