Journal of Immunological Methods, 137 (1991) 55-63 1991 Elsevier Science Publishers B.V. 0022-1759 /91 /$03.50 A DONIS 0022175991000938
55
e
JIM 05836
Oxidation of homovanillic acid as a selective assay for eosinophil peroxidase in eosinophil peroxidase-myeloperoxidase mixtures and its use in the detection of human eosinophil peroxidase deficiency Renzo Menegazzi 1, Giuliano Zabucchi 1, Patrizia Zuccato 1, Rita Cramer and Pierluigi Patriarca 1 I
Istituto di Patologia Generale, Uniuersita di Trieste, Trieste, Italy, and Trieste, Italy
l
1,
Clara Piccinini
2
Laboratorio di Ricerche Cliniche, Ospedale Maggiore,
(Received 20 Jul y 1990. revised received 8 October 1990, accepted 29 October 1990)
Biochemical assays for peroxidase activity do not usuall y dist inguish between different peroxidases. The guaiacol assay, for example, which is one of the most commonly used assays for peroxidase activity, is sensitive to both eosinophil peroxidase (EPa) and the peroxidase of neutrophils, i.e., myeloperoxidase (MPa) , thus preventing distinction of the two peroxidases in mixed neutrophil-eosinophil populations. In this paper we describe a simple and sensitive method for selective assays of EPa in EPa-MPO mixtures or mixed populations of neutrophils and eosinophils. The method is based on the peroxidase-mediated oxidation of homovanillic acid (BVA) under appropriate assay conditions in which EPO is still very active in catalyzing the reaction whilst MPa-mediated BV A oxidation is almost undetectable. Optimal assay conditions were as follows: pH 10.5, 10 J.tM hydrogen peroxide, 0.8 mM HVA and an incubation time of 120 min at 37 0 C. Under these conditions the assay permits EPO activities as low as 0.025 guaiacol U /mI to be measured even in the presence of 0.175 guaiacol U j ml of MPO. In mixed neutrophil-eosinophil cell suspensions the test permits the detection of as few as 5 X 10 3 eosinophils even in the presence of about 700 X 10 3 neutrophils (eosinophils : neutrophils ratio 1 : 140) with no appreciable interference by the latter cells. The method described here has been applied to studies of human EPa deficiency and proved to be successful in the identification of individuals with partial EPO deficiency, which is not feasible with non quantitative methods (for example, cytochemistry) or unse1ective biochemical assay of peroxidase activity. Key words: Eosinophil peroxidase; Myeloperox idase; Eosinophil peroxidase deficiency
Introduction Myeloperoxidase (MPO) and eosinophil peroxidase (EPa ) are haem-proteins contained in Correspondence 10: R. Menegazzi , Istituto di Patologia Generale, Universita di Trieste. via A. Fleming, 22, 34127 Trieste, Italy.
large amounts in the azurophilic granules of neutrophils (Bainton et al., 1968) and the specific granules of the eosinophils (Bai nton et al., 1970), respectively. Although MPO and EPO differ in several respects, for example molecular weight (Bakkenist et al., 1978; Carlson et aI., 1985), isoelectric point (Agner, 1941; Carlson et al., 1985), spectroscopic features (Bolscher et al., 1984; Bak-
56
kenist et al., 1978), etc., it is not practical to rely on such differences to distinguish between the two peroxidases when both are present in a mixture. A simpler and more widely applicable approach would be the assay of the peroxidase activity of MPO-EPO mixture by methods selectively sensitive for one or other enzyme. Specific human MPO assays have been described and have been successfully applied to the detection of heterozygosity in primary MPO deficiency (Dri et al., 1982 ; Cramer et aI., 1984). The selective assay of human EPO in neutrophil-eosinophil mixtures has also been achieved by measuring alanine decarboxylation in the presence of bromide, but the general applicability of this radioisotopic method is limited (Cramer et aI., 1981). The use of bromide or iodide in the assay of peroxidaze-catalyzed oxidation of tetramethylbenzidine has been proposed by Bozeman et aI. (1990) as a means of distinguishing EPO from MPO, since the activity of the former enzyme is greatly enhanced in the presence of either halide. However, the assay is apparently not selective for EPO , since, under the same experimental conditions, the activity of MPO also increases. Other authors have used a substrate which is more easily oxidized by EPO than MPO to selectively assay EPO in mixed populations of mouse neutrophils and eosinophils (Strath et aI., 1985). This method, however, might not be applicable to man because human neutrophils have a much greater MPO activity than mouse neutrophils (Kleb anoff et al., 1978). Finally, oxidation of iodide has been proposed as a means of distinguishing EPO from MPO in the presence of the cationic detergent cetyltrimethylammonium bromide (CTAB) (Bos et aI., 1981). Clouding of the reaction mixture, probably due to formation of a CTAB-iodide complex, is a drawback of this assay. In the present paper we propose a simple fluorimetric assay that permits quantification of EPO activity without interference from MPO in both mixed neutrophil-eosinophil suspensions and in mixtures of purified MPO and EPO. The assay is based on the EPO-mediated oxidation of homovanillic acid (H VA). We also demonstrate that the assay may also be used as a simple screening test for the detection of human EPO deficiency, in particular partial deficiency.
Materials and methods Human peroxidases Myeloperoxidase was purified according to the method described by Zabucchi et aI. (1990). Briefly, human polymorphonuclear leukocytes were lysed by sonication and the peroxidase extracted by means of 0.1 M sodium acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1% cetyltrimethylammonium bromide (CTAB , Eastman Kodak, Rochester, NY, U.S .A.). Purification of the peroxidase was achieved by ion exchange chromatography on Sulphopropyl Sephadex C-50 (Pharmacia, Uppsala, Sweden) followed by gel filtration chromatography on Ultrogel AcA34 (LKB, Bromma, Sweden). The purity of the MPO was assayed by determining the ratio of absorbance at 428 nm - the Soret band for the heme group of the native MPO (Bakkenist et aI., 1978) to that at 280 nm, which estimates protein concentration. This provides the so-called reinzhalt (RZ) ratio. The RZ value of MPO prepared as described above was > 0.8, which is in agreement with the previously reported RZ value for pure MPO (Bakkenist et aI., 1978). Eosinophil peroxidase was purified following the method described by Menegazzi et aI. (1986). The enzyme was extracted from pure granule-rich eosinophil fragments , prepared according to Zabucchi et al. (1985), by means of 0.1 M sodium acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1% CTAB. The crude extract was then applied to an Ultrogel AcA44 column (LKB) from which EPO was eluted with 0.025 M Sodium acetate buffer, pH 4.7, containing 0.02% CTAB. The purity of the peroxidase was assessed by its RZ value (ratio between absorbance at 413 nm and 280 nm) (Carlson et al., 1985), which was found to be > 0.9. Before being used in the HV A oxidation assay, both MPO and EPO were extensively dialyzed against 0.05 M Sodium phosphate buffer, pH 7.4. Granulocyte isolation Peripheral blood from normal subjects « 0.4 X 10 6 eosinophilsjml) was collected in ACD solution (Don Baxter Lab. , Trieste, Italy). After the addition of EDTA (1 roM final concentration), the red cells were removed by dextran sedimentation (1 ml 4.5% dextran in saline was added to 5 ml of
57
blood). Polymorphonuclear leukocytes (i.e., a mixed neutrophil-eosinophil population) were separated from mononuclear cells by centrifuging the post-dextran white cell-rich plasma for 20 min at 1000 X g on isotonic Percoll (density = 1.077 g/rnl) (Pharmacia). A 90 s hypotonic treatment was used to remove residual erythrocytes from the granulocyte-rich pellet. The cells were then centrifuged, washed once in phosphate-buffered saline (PBS), suspended in the same medium and counted electronically (Coulter' Counter ZB I, Luton, U.K.). The percentage of neutrophils and eosinophils in the final cell suspension was determined by differential counts carried out on W right-Giemsa-stained smears. Neutrophil and eosinophil isolation The granulocyte-containing pellet obtained as described above was washed once in PBS containing 13 mM sodium citrate and 0.5% bovine serum albumin (BSA, Miles Laboratories, Goodwood, South Africa). Granulocytes were then suspended in isotonic Percoll containing 13 mM sodium citrate and 0.5% BSA. The density of the Percoll suspension was 1.0853 ± 0.0002 g/rnl, as measured at 20 0 C by a DMA 45 density meter (A. Paar, Graz, Austria), and the Percoll osmotic value was 290 ± 2 mosM. The cell suspension was layered onto Percoll with a density higher than 1.1 g/rnl and then centrifuged at 1000 X g for 20 min at 20 0 C. Neutrophils, whose normal density is lower than 1.085 gyrnl (Gartner et al., 1980), were collected from the top of the gradient, washed once in PBS and suspended in PBS. Differential counts showed that the final cell suspension usually contained only a few contaminating eosinophils (less than 1%). However, only 100% neutrophil populations were selected as starting material for the preparation of the neutrophil extract. The cell ring formed at the interface between Percoll 1.085 and 1.1 was collected and treated with a hypotonic solution at 4 0 C to remove the contaminating red cells. As judged by differential counts carried out on Wright-Giemsa stained smears, the resulting eosinophil suspension usually contained 85-100% eosinophils, the remaining cells being neutrophils with dense pycnotic nuclei. Unless otherwise stated, only eosinophil suspensions containing more than 95%
eosinophils were used to prepare eosinophil extracts.
Preparation of neutrophil and eosinophil extract Both neutrophil and eosinophil populations were lysed by sonication (4 cycles of 10 s each at 0 0 C) with an ultrasonic disintegrator (Lab sonic 2000, Braun, F.R.G.) operating at maximum power with a needle tip. The suspensions were then centrifuged for 10 min at 560 X g to pellet nuclei and unbroken cells. The supernatants were centrifuged for 15 min at 12,000 X g and the granulecontaining pellets were then suspended in 0.1 M Na-acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1 % CTAB. The mixtures were incubated for 6-12 h in melting ice and then centrifuged for 15 min at 12,000 X g. Before being used in the HVA assay, the supernatants, which contained most of the solubilized peroxidase, were extensively dialyzed against 0.05 M sodium phosphate buffer, pH 7.4.
Peroxidase activity (1) Guaiacol oxidation assay. The method described by Cramer et al. (1984) was followed. Peroxidase activity was expressed as guaiacol units (GU), i.e. p.mol of guaiacol oxidized/min. (2) Homooanillic acid oxidation assay. The method described by Guilbault et al. (1966) was used with some modifications. Briefly, various amounts of either purified peroxidase, granulocyte extract or granulocyte population were diluted, unless otherwise stated, in 0.1 M glycine-NaOH buffer, pH 10.5. The mixtures were then incubated in polypropylene tubes at 37 0 C for the required period of time in the presence of HVA (Sigma Chemicals Co., St. Louis, MO, U.S.A.) and hydrogen peroxide (Merck, Darmstadt, F.R.G.). At the end of the incubation, the mixtures were transferred to a 1.0 em cuvette and the fluorescence was measured with a 650 IDS Fluorescence Spectrophotometer (Perkin Elmer Corp., Norwalk, CT, U.S.A.) equipped with a thermoelectric bath set at 37 0 C. The operating conditions of the spectrophotofluorimeter were as follows: Aexcitation = 315 nm, slit 2, Aemi"ion = 425 nm, slit 4. Fluorescence values were expressed as fluorescence units (FU).
58
Results Evidence for MPO- and EPO-mediated HVA oxidation The ability of MPO and EPO to catalyze HVA oxidation was first tested using (a) neutrophil and eosinophil extracts and (b) purified MPO and EPO. Under the experimental conditions used that is a peroxidase activity of 0.1 GU /ml, 100 p.M H 20 2 , 0.8 mM HVA in 0.1 M phosphate buffer pH 7.4 - regardless of the source of the peroxidase activity, both enzymes catalyzed the reaction, EPO being more active than MPO (10.1 FU/l20 min versus 3.9 FU/120 min) and, in turn, less potent than a plant-derived peroxidase (horseradish peroxidase) tested for comparison (l0.1 FU/120 min versus 24.5 FU/120 min). Characteristics of EPO-mediated HVA oxidation The effect of pH on MPO- and EPO-mediated HVA oxidation was studied. Fig. 1A shows that MPO-mediated HVA oxidation was maximal at pH 8, rapidly declined with increasing pH and became undetectable at pH 10.5. EPO behaved differently in that it exhibited a plateau of activity from pH 8 to pH 9 and was still quite active at pH 10.5. These results suggest that the assay of HV A oxidation at pH 10.5 could be exploited to distinguish EPO from MPO in samples containing both peroxidases. Fig. 1B shows that EPO-mediated HVA oxidation was maximal at H 2 0 2 concentrations ranging from 10 to 25 u M and that, at higher concentrations of the peroxide, the reaction was inhibited in a dose-dependent manner. Fig. Ie shows that EPO-mediated HVA oxidation was strongly dependent on HVA concentration up to 0.8 mM, while at higher concentrations the activity of the reaction increased only slightly. Time course experiments, illustrated in Fig. 1D , show that EPO-mediated HVA oxidation progressively increased with time until at least 120 min of incubation. Results not shown here indicated that the reaction was independent of temperature over the range 20-37 0 C. On the basis of these results the following optimal assay conditions were selected: 10 p.M H 20 2 , 0.8 mM HVA and 120 min incubation at 37 0 C. In all of these experiments MPO-mediated HVA oxidation was either very low (no more than 2-3% of that mediated by
EPO) or undetectable. Table I shows that HVA oxidation in the presence of EPO was strongly inhibited by both sodium azide, a known inhibitor of peroxidases, and by catalase, an enzyme that degrades hydrogen peroxide. These results suggest that HVA oxidation is a true EPO-mediated reaction. Effect of EPO and MPO concentration on HVA oxidation Various amounts of either a neutrophil extract or an eosinophil extract, containing equal peroxidase activity as measured by guaiacol oxidation, were tested in the HVA assay. Fig. 2 shows that MPO activity was undetectable at peroxidase concentrations of 0.2 GU/ ml or less. MPO-mediated HVA oxidation increased with increasing peroxidase concentration, but still remained very low compared with the HVA oxidation activity exhibited by equivalent amounts (in terms of guaiacol units) of EPO. Similar results were obtained using highly purified MPO and EPO , although the activities of both enzymes appeared somewhat lower than those exhibited by the neutrophil and eosinophil extracts. Further experiments investigated the possible interference of MPO with the EPO-mediated BVA oxidation when the two peroxidases were simultaneously present in the assay (Table II). Over the range of activities studied (0.1-0.2 GU / rnl), MPO-mediated HVA oxidation was undetectable and, moreover, MPO did not interfere with the reaction catalyzed by EPO . HV A oxidation measured in mixtures containing both MPO and EPO was comparable to that observed in the presence of EPO only, indicating that MPO does not affect the ability of EPO to oxidize HVA. It is concluded that the HV A oxidation activity of an EPO-MPO mixture, whose total peroxidase activity has been adjusted to 0.2 GU / rnl, is accounted for exclusively by EPO. HVA ox idation by mixed neutrophil-eosinophil cell suspensions The guaiacol oxidation activity of several granulocyte populations, containing variable numbers of neutrophils and eosinophils, was adjusted to 0.2 GU / rnl and the BVA oxidation activity tested. As shown in Fig. 3, no HV A oxidation
59
FU FU
50
20
40 15 30 10 20
5
10
0
0
5
6
7
8
9
10
2.5
11
5
10
pH
A
H 20
B
FU
25
50
100
200
2 (}-tMJ
FU
30
30
20
10
o 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
c
HVA (mMJ
20
D
40
60
80
100
120
minutes
Fig. 1. A: EPO- and MPO-mediated HVA oxidation as a function of pH. The peroxidase activity of both an eosinophil (e) and a neutrophil extract (.) was diluted in the incubation mixture to 0.2 au jml. The buffer solutions used were the following: for pH 5 and 6, sodium acetate buffer 0.1 M; for pH 7.4, sodium phosphate buffer 0.1 M; for pH 8, Tris-HCI buffer 0.1 M; for pH 9, 10 and 10.5. glycine-NaOH buffer 0.1 M. The assay mixtures. containing 100 I-lM H 202 and 0.8 mM HVA, were incubated for 120 min at 37 a C. B: EPO- and MPO-mediated HVA oxidation as a function of H 202 concentration. The peroxidase activity of both an eosinophil (white columns) and a neutrophil (black columns) extract was diluted to 0.2 GU /ml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 0.8 mM HV A. The assay mixtures were incubated for 120 min at 37 a C. C: EPO- and MPO-mediated HVA oxidation as a function of HVA concentration. The peroxidase activity of both an eosinophil (e) and a neutrophil (.) extract was diluted to 0.2 GUjml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 10 JLM H 202 . The assay mixtures were incubated for 120 min at 37 a C. D: time course of EPO- and MPO-mediated HVA oxidation at 37 a C. The peroxidase activity of both an eosinophil (e) and a neutrophil (...) extract was diluted to 0.2 GUjml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 10 I-lM H 202 •
60 TABLE I
TA BLE II
INHIBITORS OF EPO-MEDIATED HVA OXIDATION
EF F ECT OF MPO ON EPO·MED IATED HVA OXIDATION
The assay was performed in 0.1 M glycine-NaOH buffer, pH 10.5. The mixtures were incubated for 120 min at 37 0 C in the presence of 10 IlM H 20 2 and 0.8 mM HV A. Flu ores cence was detected as described in the materials and methods section. Values are the means of two dupli cate expe riments. Comp ounds added
FU
EPO b EPO + NaN 3 EPO + catalase EPO + boiled catalase
8.00 1.50 1.15 7.50
% inhi bi tion
a
81.2 85.6 6.2
FU = fluorescence units. EPO (eosinophil extract) = 0.1 GU /ml; N aN) = 2 mM; catalase = 4000 U / ml. a
b
FU 25
20
/
15 I
I
•
""
"
" .... .... "
""
.... "
,,"
The assay was performed in 0.1 M glycine-NaOH buffer, pH 10.5. The assay mixtures were incuba ted for 120 min at 37 0 C in the prese nc e of 10 }lM H 20 2 and 0.8 mM HVA . Fluorescence was de tected as de scribed in the mate rials and meth ods sec tion. Valu es are the mean of two representative dup lica te experi ments. Peroxidas e activity (G U / ml) in the assay
FU •
EPO b EPO EPO EPO MPO c M PO M PO MP O EP O EPO EPO EPO EPO
1.20 3.47 7.15 15.40 0.10 0.05 0.00 0.00 0.07 1.15 3.35 7.20 14.80
a b C
0.025 0.050 0.100 0.200 0.200 0.175 0.150 0.100 0.000 + MPO 0.025 + MP O 0.050 + MPO 0.100 +MPO 0.200 + MPO
0.200 0.175 0.150 0.100 0.000
FU = fluorescence un its. EPO = eosinophil extract. MPO = neutrophil extract. FU 10
•
8
I
10
,
•
I
/
6
I
I
5
•
4
2 0.1
•
•
I
0 .2
0.3
0 .4
0 .5
GU/ml
I
•
•
0 10
Fig. 2. EPO - and MPO -med iated H VA oxid at ion as a fun ction of peroxidase concentration . Purified prepa rations of EPO and MPO were obtained as describ ed in the materials and methods section. Eosinophil (e--e) and neutrophil (. - - . ) extracts, and purified EPO (e - -- - - - e ) and MPO (.------.) were diluted to different concentra tions in 0.1 M glycine-NaOH buffer, pH 10.5, con taining 10 IlM H P 2 and 0.8 mM HVA. The mixtures were incuba ted for 120 min at 37 0 C.
•
••• •
20 EOSINOPHILS (x10
40
50
3)
Fig. 3. HVA oxidation activit ies of granulocyte populations containing various amounts of eosinophils. Granulocyte suspensions (containing 1.5 to 12 % cos ino phils) were sonicated and their peroxid ase activity was dilu ted to 0.2 Gl.Jy'ml in 0.1 M glycine-NaO H bu ffer, pH 10.5, co n taining 10}lM H 20 2, 0.8 mM HVA and 0.05 % T riton X-100.
61 TABLE III EPO ACTIVITY IN GRANULOCYTE POPULATIONS ISOLATED FROM NORMAL AND EPO-DEFICIENT SUBJECTS Aliquots of each granulocyte population, corresponding to a final peroxidase activity of 0.2 GU/mI, were diluted in 0.1 M gtycine-NaOH buffer, pH 10.5, containing 10 iLM H 2 0 2 , 0.8 mM HVA and 0.05% Triton X-l00. The assay mixtures were incubated for 120 min at 37 0 C. The percentage of eosinophils in the granulocyte populations was determined by differential counts carried out on Wright-Giemsa stained smears. Fluorescence was determined as described in the materials and methods section.
% eosinophils in cell suspensions Normal subjects
L.c.
F.V. Z.F. Z.M. (mother) Z.Fa. (brother) Z.E. (sister) a b
3.8±0.9 4.8 3.9 8.3 3.2 5.2 3.0
b
FU "/10 6 eosinophils 169.1 ± 15.8 b
55
e
JIM 05836
Oxidation of homovanillic acid as a selective assay for eosinophil peroxidase in eosinophil peroxidase-myeloperoxidase mixtures and its use in the detection of human eosinophil peroxidase deficiency Renzo Menegazzi 1, Giuliano Zabucchi 1, Patrizia Zuccato 1, Rita Cramer and Pierluigi Patriarca 1 I
Istituto di Patologia Generale, Uniuersita di Trieste, Trieste, Italy, and Trieste, Italy
l
1,
Clara Piccinini
2
Laboratorio di Ricerche Cliniche, Ospedale Maggiore,
(Received 20 Jul y 1990. revised received 8 October 1990, accepted 29 October 1990)
Biochemical assays for peroxidase activity do not usuall y dist inguish between different peroxidases. The guaiacol assay, for example, which is one of the most commonly used assays for peroxidase activity, is sensitive to both eosinophil peroxidase (EPa) and the peroxidase of neutrophils, i.e., myeloperoxidase (MPa) , thus preventing distinction of the two peroxidases in mixed neutrophil-eosinophil populations. In this paper we describe a simple and sensitive method for selective assays of EPa in EPa-MPO mixtures or mixed populations of neutrophils and eosinophils. The method is based on the peroxidase-mediated oxidation of homovanillic acid (BVA) under appropriate assay conditions in which EPO is still very active in catalyzing the reaction whilst MPa-mediated BV A oxidation is almost undetectable. Optimal assay conditions were as follows: pH 10.5, 10 J.tM hydrogen peroxide, 0.8 mM HVA and an incubation time of 120 min at 37 0 C. Under these conditions the assay permits EPO activities as low as 0.025 guaiacol U /mI to be measured even in the presence of 0.175 guaiacol U j ml of MPO. In mixed neutrophil-eosinophil cell suspensions the test permits the detection of as few as 5 X 10 3 eosinophils even in the presence of about 700 X 10 3 neutrophils (eosinophils : neutrophils ratio 1 : 140) with no appreciable interference by the latter cells. The method described here has been applied to studies of human EPa deficiency and proved to be successful in the identification of individuals with partial EPO deficiency, which is not feasible with non quantitative methods (for example, cytochemistry) or unse1ective biochemical assay of peroxidase activity. Key words: Eosinophil peroxidase; Myeloperox idase; Eosinophil peroxidase deficiency
Introduction Myeloperoxidase (MPO) and eosinophil peroxidase (EPa ) are haem-proteins contained in Correspondence 10: R. Menegazzi , Istituto di Patologia Generale, Universita di Trieste. via A. Fleming, 22, 34127 Trieste, Italy.
large amounts in the azurophilic granules of neutrophils (Bainton et al., 1968) and the specific granules of the eosinophils (Bai nton et al., 1970), respectively. Although MPO and EPO differ in several respects, for example molecular weight (Bakkenist et al., 1978; Carlson et aI., 1985), isoelectric point (Agner, 1941; Carlson et al., 1985), spectroscopic features (Bolscher et al., 1984; Bak-
56
kenist et al., 1978), etc., it is not practical to rely on such differences to distinguish between the two peroxidases when both are present in a mixture. A simpler and more widely applicable approach would be the assay of the peroxidase activity of MPO-EPO mixture by methods selectively sensitive for one or other enzyme. Specific human MPO assays have been described and have been successfully applied to the detection of heterozygosity in primary MPO deficiency (Dri et al., 1982 ; Cramer et aI., 1984). The selective assay of human EPO in neutrophil-eosinophil mixtures has also been achieved by measuring alanine decarboxylation in the presence of bromide, but the general applicability of this radioisotopic method is limited (Cramer et aI., 1981). The use of bromide or iodide in the assay of peroxidaze-catalyzed oxidation of tetramethylbenzidine has been proposed by Bozeman et aI. (1990) as a means of distinguishing EPO from MPO, since the activity of the former enzyme is greatly enhanced in the presence of either halide. However, the assay is apparently not selective for EPO , since, under the same experimental conditions, the activity of MPO also increases. Other authors have used a substrate which is more easily oxidized by EPO than MPO to selectively assay EPO in mixed populations of mouse neutrophils and eosinophils (Strath et aI., 1985). This method, however, might not be applicable to man because human neutrophils have a much greater MPO activity than mouse neutrophils (Kleb anoff et al., 1978). Finally, oxidation of iodide has been proposed as a means of distinguishing EPO from MPO in the presence of the cationic detergent cetyltrimethylammonium bromide (CTAB) (Bos et aI., 1981). Clouding of the reaction mixture, probably due to formation of a CTAB-iodide complex, is a drawback of this assay. In the present paper we propose a simple fluorimetric assay that permits quantification of EPO activity without interference from MPO in both mixed neutrophil-eosinophil suspensions and in mixtures of purified MPO and EPO. The assay is based on the EPO-mediated oxidation of homovanillic acid (H VA). We also demonstrate that the assay may also be used as a simple screening test for the detection of human EPO deficiency, in particular partial deficiency.
Materials and methods Human peroxidases Myeloperoxidase was purified according to the method described by Zabucchi et aI. (1990). Briefly, human polymorphonuclear leukocytes were lysed by sonication and the peroxidase extracted by means of 0.1 M sodium acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1% cetyltrimethylammonium bromide (CTAB , Eastman Kodak, Rochester, NY, U.S .A.). Purification of the peroxidase was achieved by ion exchange chromatography on Sulphopropyl Sephadex C-50 (Pharmacia, Uppsala, Sweden) followed by gel filtration chromatography on Ultrogel AcA34 (LKB, Bromma, Sweden). The purity of the MPO was assayed by determining the ratio of absorbance at 428 nm - the Soret band for the heme group of the native MPO (Bakkenist et aI., 1978) to that at 280 nm, which estimates protein concentration. This provides the so-called reinzhalt (RZ) ratio. The RZ value of MPO prepared as described above was > 0.8, which is in agreement with the previously reported RZ value for pure MPO (Bakkenist et aI., 1978). Eosinophil peroxidase was purified following the method described by Menegazzi et aI. (1986). The enzyme was extracted from pure granule-rich eosinophil fragments , prepared according to Zabucchi et al. (1985), by means of 0.1 M sodium acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1% CTAB. The crude extract was then applied to an Ultrogel AcA44 column (LKB) from which EPO was eluted with 0.025 M Sodium acetate buffer, pH 4.7, containing 0.02% CTAB. The purity of the peroxidase was assessed by its RZ value (ratio between absorbance at 413 nm and 280 nm) (Carlson et al., 1985), which was found to be > 0.9. Before being used in the HV A oxidation assay, both MPO and EPO were extensively dialyzed against 0.05 M Sodium phosphate buffer, pH 7.4. Granulocyte isolation Peripheral blood from normal subjects « 0.4 X 10 6 eosinophilsjml) was collected in ACD solution (Don Baxter Lab. , Trieste, Italy). After the addition of EDTA (1 roM final concentration), the red cells were removed by dextran sedimentation (1 ml 4.5% dextran in saline was added to 5 ml of
57
blood). Polymorphonuclear leukocytes (i.e., a mixed neutrophil-eosinophil population) were separated from mononuclear cells by centrifuging the post-dextran white cell-rich plasma for 20 min at 1000 X g on isotonic Percoll (density = 1.077 g/rnl) (Pharmacia). A 90 s hypotonic treatment was used to remove residual erythrocytes from the granulocyte-rich pellet. The cells were then centrifuged, washed once in phosphate-buffered saline (PBS), suspended in the same medium and counted electronically (Coulter' Counter ZB I, Luton, U.K.). The percentage of neutrophils and eosinophils in the final cell suspension was determined by differential counts carried out on W right-Giemsa-stained smears. Neutrophil and eosinophil isolation The granulocyte-containing pellet obtained as described above was washed once in PBS containing 13 mM sodium citrate and 0.5% bovine serum albumin (BSA, Miles Laboratories, Goodwood, South Africa). Granulocytes were then suspended in isotonic Percoll containing 13 mM sodium citrate and 0.5% BSA. The density of the Percoll suspension was 1.0853 ± 0.0002 g/rnl, as measured at 20 0 C by a DMA 45 density meter (A. Paar, Graz, Austria), and the Percoll osmotic value was 290 ± 2 mosM. The cell suspension was layered onto Percoll with a density higher than 1.1 g/rnl and then centrifuged at 1000 X g for 20 min at 20 0 C. Neutrophils, whose normal density is lower than 1.085 gyrnl (Gartner et al., 1980), were collected from the top of the gradient, washed once in PBS and suspended in PBS. Differential counts showed that the final cell suspension usually contained only a few contaminating eosinophils (less than 1%). However, only 100% neutrophil populations were selected as starting material for the preparation of the neutrophil extract. The cell ring formed at the interface between Percoll 1.085 and 1.1 was collected and treated with a hypotonic solution at 4 0 C to remove the contaminating red cells. As judged by differential counts carried out on Wright-Giemsa stained smears, the resulting eosinophil suspension usually contained 85-100% eosinophils, the remaining cells being neutrophils with dense pycnotic nuclei. Unless otherwise stated, only eosinophil suspensions containing more than 95%
eosinophils were used to prepare eosinophil extracts.
Preparation of neutrophil and eosinophil extract Both neutrophil and eosinophil populations were lysed by sonication (4 cycles of 10 s each at 0 0 C) with an ultrasonic disintegrator (Lab sonic 2000, Braun, F.R.G.) operating at maximum power with a needle tip. The suspensions were then centrifuged for 10 min at 560 X g to pellet nuclei and unbroken cells. The supernatants were centrifuged for 15 min at 12,000 X g and the granulecontaining pellets were then suspended in 0.1 M Na-acetate buffer, pH 4.7, containing 0.1 M sodium sulfate and 0.1 % CTAB. The mixtures were incubated for 6-12 h in melting ice and then centrifuged for 15 min at 12,000 X g. Before being used in the HVA assay, the supernatants, which contained most of the solubilized peroxidase, were extensively dialyzed against 0.05 M sodium phosphate buffer, pH 7.4.
Peroxidase activity (1) Guaiacol oxidation assay. The method described by Cramer et al. (1984) was followed. Peroxidase activity was expressed as guaiacol units (GU), i.e. p.mol of guaiacol oxidized/min. (2) Homooanillic acid oxidation assay. The method described by Guilbault et al. (1966) was used with some modifications. Briefly, various amounts of either purified peroxidase, granulocyte extract or granulocyte population were diluted, unless otherwise stated, in 0.1 M glycine-NaOH buffer, pH 10.5. The mixtures were then incubated in polypropylene tubes at 37 0 C for the required period of time in the presence of HVA (Sigma Chemicals Co., St. Louis, MO, U.S.A.) and hydrogen peroxide (Merck, Darmstadt, F.R.G.). At the end of the incubation, the mixtures were transferred to a 1.0 em cuvette and the fluorescence was measured with a 650 IDS Fluorescence Spectrophotometer (Perkin Elmer Corp., Norwalk, CT, U.S.A.) equipped with a thermoelectric bath set at 37 0 C. The operating conditions of the spectrophotofluorimeter were as follows: Aexcitation = 315 nm, slit 2, Aemi"ion = 425 nm, slit 4. Fluorescence values were expressed as fluorescence units (FU).
58
Results Evidence for MPO- and EPO-mediated HVA oxidation The ability of MPO and EPO to catalyze HVA oxidation was first tested using (a) neutrophil and eosinophil extracts and (b) purified MPO and EPO. Under the experimental conditions used that is a peroxidase activity of 0.1 GU /ml, 100 p.M H 20 2 , 0.8 mM HVA in 0.1 M phosphate buffer pH 7.4 - regardless of the source of the peroxidase activity, both enzymes catalyzed the reaction, EPO being more active than MPO (10.1 FU/l20 min versus 3.9 FU/120 min) and, in turn, less potent than a plant-derived peroxidase (horseradish peroxidase) tested for comparison (l0.1 FU/120 min versus 24.5 FU/120 min). Characteristics of EPO-mediated HVA oxidation The effect of pH on MPO- and EPO-mediated HVA oxidation was studied. Fig. 1A shows that MPO-mediated HVA oxidation was maximal at pH 8, rapidly declined with increasing pH and became undetectable at pH 10.5. EPO behaved differently in that it exhibited a plateau of activity from pH 8 to pH 9 and was still quite active at pH 10.5. These results suggest that the assay of HV A oxidation at pH 10.5 could be exploited to distinguish EPO from MPO in samples containing both peroxidases. Fig. 1B shows that EPO-mediated HVA oxidation was maximal at H 2 0 2 concentrations ranging from 10 to 25 u M and that, at higher concentrations of the peroxide, the reaction was inhibited in a dose-dependent manner. Fig. Ie shows that EPO-mediated HVA oxidation was strongly dependent on HVA concentration up to 0.8 mM, while at higher concentrations the activity of the reaction increased only slightly. Time course experiments, illustrated in Fig. 1D , show that EPO-mediated HVA oxidation progressively increased with time until at least 120 min of incubation. Results not shown here indicated that the reaction was independent of temperature over the range 20-37 0 C. On the basis of these results the following optimal assay conditions were selected: 10 p.M H 20 2 , 0.8 mM HVA and 120 min incubation at 37 0 C. In all of these experiments MPO-mediated HVA oxidation was either very low (no more than 2-3% of that mediated by
EPO) or undetectable. Table I shows that HVA oxidation in the presence of EPO was strongly inhibited by both sodium azide, a known inhibitor of peroxidases, and by catalase, an enzyme that degrades hydrogen peroxide. These results suggest that HVA oxidation is a true EPO-mediated reaction. Effect of EPO and MPO concentration on HVA oxidation Various amounts of either a neutrophil extract or an eosinophil extract, containing equal peroxidase activity as measured by guaiacol oxidation, were tested in the HVA assay. Fig. 2 shows that MPO activity was undetectable at peroxidase concentrations of 0.2 GU/ ml or less. MPO-mediated HVA oxidation increased with increasing peroxidase concentration, but still remained very low compared with the HVA oxidation activity exhibited by equivalent amounts (in terms of guaiacol units) of EPO. Similar results were obtained using highly purified MPO and EPO , although the activities of both enzymes appeared somewhat lower than those exhibited by the neutrophil and eosinophil extracts. Further experiments investigated the possible interference of MPO with the EPO-mediated BVA oxidation when the two peroxidases were simultaneously present in the assay (Table II). Over the range of activities studied (0.1-0.2 GU / rnl), MPO-mediated HVA oxidation was undetectable and, moreover, MPO did not interfere with the reaction catalyzed by EPO . HV A oxidation measured in mixtures containing both MPO and EPO was comparable to that observed in the presence of EPO only, indicating that MPO does not affect the ability of EPO to oxidize HVA. It is concluded that the HV A oxidation activity of an EPO-MPO mixture, whose total peroxidase activity has been adjusted to 0.2 GU / rnl, is accounted for exclusively by EPO. HVA ox idation by mixed neutrophil-eosinophil cell suspensions The guaiacol oxidation activity of several granulocyte populations, containing variable numbers of neutrophils and eosinophils, was adjusted to 0.2 GU / rnl and the BVA oxidation activity tested. As shown in Fig. 3, no HV A oxidation
59
FU FU
50
20
40 15 30 10 20
5
10
0
0
5
6
7
8
9
10
2.5
11
5
10
pH
A
H 20
B
FU
25
50
100
200
2 (}-tMJ
FU
30
30
20
10
o 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
c
HVA (mMJ
20
D
40
60
80
100
120
minutes
Fig. 1. A: EPO- and MPO-mediated HVA oxidation as a function of pH. The peroxidase activity of both an eosinophil (e) and a neutrophil extract (.) was diluted in the incubation mixture to 0.2 au jml. The buffer solutions used were the following: for pH 5 and 6, sodium acetate buffer 0.1 M; for pH 7.4, sodium phosphate buffer 0.1 M; for pH 8, Tris-HCI buffer 0.1 M; for pH 9, 10 and 10.5. glycine-NaOH buffer 0.1 M. The assay mixtures. containing 100 I-lM H 202 and 0.8 mM HVA, were incubated for 120 min at 37 a C. B: EPO- and MPO-mediated HVA oxidation as a function of H 202 concentration. The peroxidase activity of both an eosinophil (white columns) and a neutrophil (black columns) extract was diluted to 0.2 GU /ml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 0.8 mM HV A. The assay mixtures were incubated for 120 min at 37 a C. C: EPO- and MPO-mediated HVA oxidation as a function of HVA concentration. The peroxidase activity of both an eosinophil (e) and a neutrophil (.) extract was diluted to 0.2 GUjml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 10 JLM H 202 . The assay mixtures were incubated for 120 min at 37 a C. D: time course of EPO- and MPO-mediated HVA oxidation at 37 a C. The peroxidase activity of both an eosinophil (e) and a neutrophil (...) extract was diluted to 0.2 GUjml in 0.1 M glycine-NaOH buffer, pH 10.5, containing 10 I-lM H 202 •
60 TABLE I
TA BLE II
INHIBITORS OF EPO-MEDIATED HVA OXIDATION
EF F ECT OF MPO ON EPO·MED IATED HVA OXIDATION
The assay was performed in 0.1 M glycine-NaOH buffer, pH 10.5. The mixtures were incubated for 120 min at 37 0 C in the presence of 10 IlM H 20 2 and 0.8 mM HV A. Flu ores cence was detected as described in the materials and methods section. Values are the means of two dupli cate expe riments. Comp ounds added
FU
EPO b EPO + NaN 3 EPO + catalase EPO + boiled catalase
8.00 1.50 1.15 7.50
% inhi bi tion
a
81.2 85.6 6.2
FU = fluorescence units. EPO (eosinophil extract) = 0.1 GU /ml; N aN) = 2 mM; catalase = 4000 U / ml. a
b
FU 25
20
/
15 I
I
•
""
"
" .... .... "
""
.... "
,,"
The assay was performed in 0.1 M glycine-NaOH buffer, pH 10.5. The assay mixtures were incuba ted for 120 min at 37 0 C in the prese nc e of 10 }lM H 20 2 and 0.8 mM HVA . Fluorescence was de tected as de scribed in the mate rials and meth ods sec tion. Valu es are the mean of two representative dup lica te experi ments. Peroxidas e activity (G U / ml) in the assay
FU •
EPO b EPO EPO EPO MPO c M PO M PO MP O EP O EPO EPO EPO EPO
1.20 3.47 7.15 15.40 0.10 0.05 0.00 0.00 0.07 1.15 3.35 7.20 14.80
a b C
0.025 0.050 0.100 0.200 0.200 0.175 0.150 0.100 0.000 + MPO 0.025 + MP O 0.050 + MPO 0.100 +MPO 0.200 + MPO
0.200 0.175 0.150 0.100 0.000
FU = fluorescence un its. EPO = eosinophil extract. MPO = neutrophil extract. FU 10
•
8
I
10
,
•
I
/
6
I
I
5
•
4
2 0.1
•
•
I
0 .2
0.3
0 .4
0 .5
GU/ml
I
•
•
0 10
Fig. 2. EPO - and MPO -med iated H VA oxid at ion as a fun ction of peroxidase concentration . Purified prepa rations of EPO and MPO were obtained as describ ed in the materials and methods section. Eosinophil (e--e) and neutrophil (. - - . ) extracts, and purified EPO (e - -- - - - e ) and MPO (.------.) were diluted to different concentra tions in 0.1 M glycine-NaOH buffer, pH 10.5, con taining 10 IlM H P 2 and 0.8 mM HVA. The mixtures were incuba ted for 120 min at 37 0 C.
•
••• •
20 EOSINOPHILS (x10
40
50
3)
Fig. 3. HVA oxidation activit ies of granulocyte populations containing various amounts of eosinophils. Granulocyte suspensions (containing 1.5 to 12 % cos ino phils) were sonicated and their peroxid ase activity was dilu ted to 0.2 Gl.Jy'ml in 0.1 M glycine-NaO H bu ffer, pH 10.5, co n taining 10}lM H 20 2, 0.8 mM HVA and 0.05 % T riton X-100.
61 TABLE III EPO ACTIVITY IN GRANULOCYTE POPULATIONS ISOLATED FROM NORMAL AND EPO-DEFICIENT SUBJECTS Aliquots of each granulocyte population, corresponding to a final peroxidase activity of 0.2 GU/mI, were diluted in 0.1 M gtycine-NaOH buffer, pH 10.5, containing 10 iLM H 2 0 2 , 0.8 mM HVA and 0.05% Triton X-l00. The assay mixtures were incubated for 120 min at 37 0 C. The percentage of eosinophils in the granulocyte populations was determined by differential counts carried out on Wright-Giemsa stained smears. Fluorescence was determined as described in the materials and methods section.
% eosinophils in cell suspensions Normal subjects
L.c.
F.V. Z.F. Z.M. (mother) Z.Fa. (brother) Z.E. (sister) a b
3.8±0.9 4.8 3.9 8.3 3.2 5.2 3.0
b
FU "/10 6 eosinophils 169.1 ± 15.8 b
