- Email: [email protected]
Detection of eosinophil cationic protein (ECP) by an enzyme-linked immunosorbent assay
285
Journal of Immunological Methods, 138 (1991) 285-290 ":;, 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 002217599100151K
JIM05899
Detection of eosinophil cationic protein (ECP) by an enzyme-linked immunosorbent assay C.M. Reimert, P. Venge, A. K h a r a z m i a n d K. B e n d t z e n l.aborato~' of Medical A Ilergology, Uni~ersity Hospital, Copenhagen, Denmark, Department of Clinical Chemistry, University Hospital, Uppsala. Sweden. Department of Clinical Microhiologo'. University ttospital, Copenhagen, Denmark, and Laborato~." of Medical Immunology', University Hospital. Copenhagen, Denmark Received 23 October 1990. revised received 18 December 1990. accepted 15 Janua~' 1991)
Eosinophil cauonic protein (ECP) is a highly basic and potent cytotoxic single-chain zinc-containing protein present in the granules of the eosinophilic granulocytes. ECP appears to be involved in defence against parasites and in the tissue damage seen in subjects with allergic and inflammatory disease. To investigate ECP release from in vitro activated human eosinophils and to study the involvement of eosinophils in health and disease, we have developed a sensitive and specific enzyme immunoassay. ECP was purified from normal human peripheral blood eosinophils and polyclonal antibodies to ECP were subsequently raised in rabbits. The ELISA utilizes the biotin/avidin method and measures ECP within the range 15-1000 ng/l. The intra- and interassay coefficients of variation were 6% and 10%, respectively, and the recoveries of 12 and 25 pg of purified ECP added to diluted serum samples were 108 + 14.5% (mean + SD, n = 12) and 107 + 7.5%, respectively. The high sensitivity, reproducibility and specificity of this ELISA makes it suitable for the determination of minute amounts of ECP in in vitro systems as well as in various biological fluids. K~W words: Eosinophil cationic protein; ELISA: Purification
Introduction
The granules of the human eosinophilic granulocytes contain several highly cationic proteins which upon activation and stimulation of the cells are secreted. One of these proteins, the
Correspondence to: C.M. Reimert. Laborators' for Medical Allergology, TTA 7542. State University Hospital. Tagensvej 20. DK-2200. Copenhagen, Denmark. Abbre~'iations: ECP. eosinophil cationic protein: CTAB, N-cctyI-N.N.N.-trimethyl-ammonium-bromide; ELISA. enzyme-linked immunosorbent assay; SDS. sodium dodecyl sulphate: HSA. human serum albumin; RIA. radioimmunoassay.
eosinophil cationic protein (ECP), is a zinc-containing, single-chain molecule with a molecular weight of 18-22 kDa and a pl > 11 (Olsson et al., 1977). Several biological functions of ECP have been described including interaction with proteins involved in blood coagulation (Venge et al., 1979), cytotoxic activity against the schistosomula of Schistosoma mansoni (MacLaren et al., 1981, 1984), epithelial damaging effects in asthmatic individuals (DeMonchy et al., 1985; Venge et al., 1988), perforin-like activity resulting in increased membrane permeability (Yong et al., 1986), inhibition of lymphocyte proliferation in mixed lymphocyte cultures and phytohaemagglutinin (PHA)-stimulated cell cultures (Peterson et al., 1986) and neu-
286
rotoxicity when injected into the brains of rabbits (Fredens et al., 1982). The mechanisms by which ECP and the other cationic granular proteins are released from the eosinophils are poorly understood. Experimental data indicate a selective release of individual granular proteins depending upon the triggering stimulus (Khalife et al., 1986: Peterson et al., 1987). To investigate ECP release from in vitro stimulated eosinophils, and to study the involvement of eosinophils in parasitic, allergic and other inflammatory diseases, we have developed a sensitive and specific enzyme immunoassay for ECP in biological fluids and cell culture supernatants.
Materials and methods
Preparation of granula extracts Granulocyte granule extracts were prepared by a modification of the previously described method (Peterson et al., 1983). Buffy coat cells from healthy blood donors were pooled and mixed with 2 vols. of 1.5% dextran T-500 to sediment the red blood cells. After 45 min at room temperature, the leucocyte rich-plasma was siphoned off and the leucocytes were isolated and washed twice in 0.15 M NaC1, and once in a solution of 0.34 M sucrose, 5 mM EDTA, 12 mM NaHCO). The cell-pellet was resuspended in 4 vols. of ice-cold phosphatebuffered saline (PBS) containing 5 mM EDTA. 12 mM N a H C O 3 and pressurized with N 2 for 30 rain at 500 psi at 4°C under constant stirring in a cell disruption bomb (Parr Instrument Company, Moline, Ik, U.S.A.). The cavitate was collected dropwise into 1 vol. of the same buffer containing 0.7 M sucrose and 0.3 M NaCI, and centrifuged at 800 × g to eliminate cell nuclei and disrupted cells. Granules were isolated from the supernatant by centrifugation at 10,000 x g for 15 min at 4o( ". After three cycles of freezing and thawing, the granula pellet was extracted with 5 vols. of 50 mM acetic acid under constant stirring for 1 h at 40( ". An equal vol of 0.4 M ammonium-acetate was then added and the extraction continued for 3 h. The extract was centrifuged at 20,000 x g for 30 min at 4°C and the supernatant containing the granular proteins was concentrated 30 times by
diafiltration using YM-10 Danvers, MA, U.S.A.).
filters
(Amicon,
Purification of ECI' ECP was purified as previously described (Peterson et al.. 1988). Briefly, the dark green granular extract was fractionated on a Sephadex G-75 superfine column (Pharmacia, Uppsala, Sweden) equilibrated with 0.2 M ammoniumacetate, pH 4. Fractions containing 16 22 kDa bands on sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) were pooled. concentrated and equilibrated with 80 mM ammonium-acetate, 10% glycerol, pH 8.4 (starting buffer 1) by diafiltration on YM-10 filter. The material was then subjected to ion exchange chromatography on a Bio Rex-70 column (Bio-Rad, Munich, F.R.G.). After the first peak had reached the baseline, the flow direction was reversed and proteins were eluted with a combined salt and pH gradient from starting buffer 1 to 0.8 M ammonium-acetate, 10% glycerol, pH 9.2. Fractions containing the last 280 nm peak were concentrated and equilibrated with 0.2 M ammoniumacetetate, 0.5 M NaC1, 10% glycerol, pH 8.0 (starting buffer 2) and applied to a Zn-chelating Sepharose 6B column (Pharmacia). After the first peak had reached the baseline, ECP was eluted as a single peak with a pH gradient from starting buffer 2 to 0.2 M ammonium-acetate, 0.5 M Na('l, 10% glycerol, pH 4.5. In another experiment ECP was purified as decribed above, except that the ion exchange step was excluded. The protein concentration of the purified ECP was determined using the extinction coefficient (k'l~,,,) = 15.45 at 280 nm (Peterson et al.. 1988). Antibodies against ECP were raised in rabbits by three multiple-site intracutaneus injections of 0.1 mg purified protein with Freund's complete adjuvant and boosted 2 weeks before every bleeding with 0.1 nag ECP in Freund's incomplete adjuvant. The immunoglobulin fraction was isolated by standard ammonium-sulphate precipitation (Harboe et al., 1983). Biotin (N-hydroxy-succinimidobiotin, Sigma, St. Louis, MO, U.S.A.) conjugation of the rabbit-antibodies was performed as described (Kendall et al., 1983). SI)S-PAGE was performed in a discontinuous system according to Laemmli (1970) with 16%
287 polyacrylamide in the separation gel and 5% in the stacking gel. The electrophoresis buffer was 25 mM Tris with 20 mM glycine and 0.1% HC1, pH 8.3. Samples were prepared by boiling for 5 min in four-fold Laemmli sample buffer with 10% mercaptoethanol. Molecular weights were determined by comparison with prestained standards covering the range 14.3-200 kDa, (Amersham International, Buckinghamshire, England, U.K.). Silver staining of gels were performed according to the manufacturer (Bio-Rad).
posed to biotin-conjugated rabbit anti-ECP Ig, (100 p-l/well) for 1.5 h at 37°C and, finally, to avidin-peroxidase (Sigma) and left at room temperature for 30 rain. The plates were incubated with enzyme substrate solution at 100 p-l/well. The enzyme reaction was stopped after 20 min. with 150 p-l/well of 2.5 M H2SO 4 and the absorbance measured at 492 nm with a 620 nm reference.
Results
Crossed immunoelectrophoresis This was performed in IEF agarose (Pharmacia) supplemented with 0.02% N-cetyl-N,N,N,-trimethyl-ammonium-bromide (CTAB) in the first dimension gel as described (Reimert and B~gHansen, in press). The electrophoresis buffer was 0.025 M sodium-barbital, 73 mM Tris pH 8.6.
Purification of ECP
EL1SA procedure Buffer. Coating buffer (A): PBS (10 mM
ECP was purified by two methods. Fig. 1 shows the S D S - P A G E of the two ECP preparations. Both preparations contained the three molecular weight forms of ECP (18.5, 20 and 22 kDa), in agreement with previous reports (Olsson et al., 1986). Moreover, it can be seen that the ECP preparation purified by the two-step procedure (Sephadex G-75 superfine and Zn-chelate chro-
NazHPO4, 0.15 NaCI) pH 7.4. Washing buffer (B): PBS. 0.1% Tween 20, pH 7.4. I n c u b a t i o n / sample buffer (C): PBS, 0.1% Tween 20, 0.1% CTAB, 0.2% human serum albumin (HSA), 20 mM EDTA, pH 7.4. Enzyme substrate solution (D): 12 mg o-phenylenediamine (Dakopatts, Glostrup, Denmark) dissolved in 15 ml 0.1 M citrate-phosphate buffer, pH 5.6 p-I 30% hydrogen peroxide were added immediately before use. The solid phase sandwich ELISA was carried out in microtitre plates (Maxisorb, Nunc, Roskilde, Denmark). Coating of plates Each well was coated overnight at 4°C with 100 p-1 rabbit anti-ECP Ig, 1.5 p-g/ml, diluted in buffer A. Coated plates could be stored at 4°C for up to at least 6 weeks. Washing of plates. Before use, and between each subsequent step in the assay, the plates were washed three times in buffer B. Standards and samples. The ECP standards ,le-/ were calibrated using E~cm of purified ECP at 280 nm = 15.45 (Peterson et al.. 1988). Standards and samples were diluted in buffer C before application to the plates. 100 p,l/well. Plates were left overnight at 4°C. The assay procedure was accelerated by incubating the plates for 2 h at 37° C. After incubation the plates were sequentially ex-
Fig. 1. SI)S-PAGE of two ECP preparations. Lane A: ECP purified by chromatography on Sephadex G-75 superfine. BioRex 70 and Zn-chelate Sepharose 6B. Lane B: ECP purified on Sephadex (.;-75 superfine and Zn-chelate Sepharose 6B. MW: molecular weight in kDa.
288
matography) appear to be just as pure as the ECP preparation purified by the three-step procedure using ion exchange chromatography before the final affinity chromatography on Zn-chelate Sepharose. Using crossed immunoelectrophoresis or radial immunodiffusion, none of the ECP preparations exhibited immunological reactivity with antibodies against eosinophil protein X. cathepsin G. myeloperoxidase or lysozyme (not shown).
~492/620
nm
2.5~
/
2.0
('2a~,o
./
//
,
/,.
F
,'2~6)o
,'4)
/
Specificity Judged by crossed immunoelectrophoresis using a crude granular extract in the first dimension gel, the rabbit antibodies showed no cross reactivity with other granular proteins (Fig. 2). Possible interference caused by non-specific binding of endogenous peroxidase in the samples was tested using semipurified MPO and EPO. Neither of these peroxidases showed any activity in the assay. Range of measurement Optimal antibody concentrations and optimal incubation conditions were obtained by checker-
S.5-
~ .
,O.C •C
.
.
.
.
....~
.
.
.
.
.
.
.
.
~lOr'K
.............. 100
~ 00.9
Fc~' ,'~,g/; Fig. 3. Standard curve obtained using the ECP-EI.ISA Insert: serum samples containing 125 ,ug/I and 2.5 # g / I applied in four dilutions to confirm parallelism with the standard cur~e. Titres are shov,n in parentheses.
board titrations. The range of measurement was 15-1000 n g / l using purified ECP as a standard. There was no interference as judged by the parallelism between the standard curve and the diluted serum samples (Fig. 3). The inclusion of the cationic detergent CTAB in the i n c u b a t i o n - / s a m p l e buffer was crucial in stabilizing the assay, probably by preventing aggregation of ECP (data not shown).
Recover)'
..
•
•
,
^
•
Fig. 2. Crossed immunoelectrophoresis of a crude granular extract developed using rabbit anti-ECP Ig in the second dimension gel. Cathode at the top. Insert: zone-electrophoresis of crude granular extract.
After adding 12 or 25 pg of purified ECP to 100 ffl diluted serum samples the mean recoveries were 108 + 14.5% and 109 + 7.5%, respectively (n = 12). To ascertain whether the assay could be used for ECP analysis of cell-free supernatants from in vitro stimulated eosinophils. 25 and 50 pg of purified ECP were added to 100 ffl of RPMI 1640 containing 20% fetal calf serum. The recoveries were 98 + 7.2% and 91.4 + 6.8% respectively (n = 6).
Reproducibilio" When ECP was measured simultaneously in quadruplicate in 10 different sera and in duplicate in 12 sera on four different occasions, the mean
289
12s1 ~0/I) RIA
,,''''.
IOO I
f;
75
.:'. 5O
%= 0.98
2",'''
25EUSA (~/,) 0-0
25
50
75
"00
125
Fig. 4. ECP content in 40 serum samples m e a s u r e d by E C P - R I A a n d E C P - E L I S A . r, = 0.98, p < 0.00001.
intra- and interassay coefficients of variation were 6% and 10%, respectively• Comparison with an ECP-RIA
To establish the accuracy of the ELISA, 40 serum samples were measured with a recently introduced ECP-RIA (Pharmacia), measuring in the range of 2-200 /tg/l, and in a suitable dilution with the ECP-ELISA. The correlation coefficient between the two assays was & = 0.98, p < 0.00001 (Fig. 4). There was no significant difference between the two assays when tested by the W i l c o x o n / P r a t t signed rank sum test (2a = 0.05, p = 0.082).
Discussion
Purification of ECP from buffy coat cells obtained from normal individuals has been described previously (Peterson et al., 1988). Due to the low eosinophii count in normal blood, leading to low amounts of ECP in the starting material, several precautions were introduced to ensure high yields. These precautions included: degassing of buffers and the inclusion of glycerol and ammoniumacetate in buffers to prevent aggregation, adherence and oxidation of ECP. The use of reversed flow for the gradient elution in ion exchange chromatography on Bio-Rex 70 was of particular importance in achieving a high yield. Despite these precautions, the recovery from the ion exchange chromatography step was relatively low compared
with the recoveries from the size exclusion and affinity chromatrography steps. However, on the basis of the silver-stained SDS-PAGE and crossed immunoelectrophoresis, the ion exchange chromatography step could be excluded without diminishing the purity of the final ECP preparation• Both preparations gave the characteristic three molecular weight forms of ECP (18.5, 20 and 22 kDa) in agreement with previous reports (Olsson et al., 1986; Peterson et al., 1988). The three molecular weight forms of ECP reflect different degrees of glycosylation. Ion-exchange chromatography has shown that the three forms differ in charge, the 18.5 kDa form being the most cationic and the 22 kDa form the least cationic (Petersson et al., 1988)• These differences in charge densities and amounts of the different weight forms are clearly seen by the size/height of the leading front and the following shoulders in crossed immunoelectrophoresis of crude granular extract using anti-ECP antibodies• These experiments also show the immunological identity between the different forms of ECP, and the absence of cross-reactivity between anti-ECP antibodies and other antigens found in the granular extract. There was a good linear correlation between the ELISA and the commercially available RIA for ECP. Generally, however, the ELISA values exceeded those obtained in the RIA when using serum samples with more than 50 n g / m l of ECP. Measurements of cell products in biological fluids usually require specific sample processing• Preliminary results (data not shown) support this notion since spontanous ECP release occurs when processing cell-containing samples• Therefore standardised sample processing with regard to time and temperature is important• The high sensitivity of the present ELISA procedure permits measurements of ECP in high dilutions of biological fluids, thus reducing non-specific binding and interference from other proteins. No interference has been detected in serum, EDTA-plasma, urine, sputum, saliva, eye or nasal lavage fluids• However, interference in citrated plasma was occasionally observed when such samples were tested at high concentrations• We conclude that the high sensitivity, reproducibility and specificity of this ELISA makes it suitable for in vitro as well as in vivo investigations•
290 Acknowledgements
W e t h a n k Ms. U l l a M i n u v a f o r s k i l l e d t e c h n i c a l assistance. The study was supported nology Programme.
by Danish
Biotech-
References
De Monchy, J.G.R., Kauffman, H.F., Venge, P., Koeter, G.H., Jansen, HM.. Sluiter, H.J. and De Vries, K. (1985) Bronchoalveolar eosinophilia during allergen-induced late astmatic reactions. Am. Rev. Respir. Dis. 131,373. Fredens. K. and Dahl, R. (1982) The Gordon phenomenon induced by the eosinophil cationic protein and eosinophil protein X. J. Allergy Clin. immunol. 70, 361. Harboe, N.M.G. and Ingild, A. (1983) Immunization, isolation of immunoglobulins and antibody titre determination. Scand. J. lmmunol. 17 (suppl. 10) 345. Kendall, C., lnoescu-Matiu, I. and Dreesman, G.R. (1983) Utilization of the biotin/avidin system to amplify the sensitivity of enzyme-linked immunosorbent assay (ELISA). J. Immunol. Methods 56, 329. Khalife, J., Capron, M., Cesbron, J.Y., Tai, P.-C., Taelman, H., Prin, L and Capron, A. (1986) Role of specific igE antibodies in peroxidase (EPO) release from human eosinophils. J. lmmunol. 137, 1659. 1.aemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage "I"4. Nature 227, 680. MacLaren, D.J., Mackean, J.R., Olsson, 1., Venge, P. and Kay, A.B. (1981) Morphological studies on the killing of schistosomula of Schistosoma mansoni by human eosinophil and neutrophil cationic proteins in vitro. Parasite lmmunol. 3, 359. MacLaren, D.J., Peterson, C.G.B. and Venge, P. (1984) Schis-
tosoma mansoni: further studies of the interaction between schistosomula and granulocyte derived cationic proteins in vitro. Parasitology 88, 491. Olsson, I., Persson, A.M. and Wingqvist. I. (1986) Biochemical properties of the eosinophil cationic protein and demonstration of its biosynthesis in vitro in marrow cells from patients with eosinophilia. Blood 67. 498. Olsson 1., Venge, P. and Spitznagel, I.K. (1977) Arginine-rich cationic proteins of human cosinophil granules. Comparison of the constituents of the eosinophilic and neutrophilic leukocytes. Lab. Invest. 36, 493. Peterson. C.G.B. and Venge, P. (1983) Purification and characterization of a new cationic protein - eosinophil protein X (EPX) - from the granules of human eosinophils. Immunology 50, 19. Peterson, C.G.B., Skoog, V. and Venge, P. (1986) Human eosinophil cationic proteins (ECP& EPX) and their suppressive effects on lymphocyte proliferation. Immunobiology 171, 1. Peterson, C.B.G., Jt~rnvall, H. and Venge, P. (1988) Purification and characterization of eosinophil cationic protein from normal human eosinophils. Eur. J. Haematol. 40. 415. Peterson, ('.G.B., Garcia. R.C., Carlson M.G.Ch. and Venge, P. (1987) Eosinophil cationic protein (ECP, eosinophil protein X (EPX) and eosinophil peroxidase (EPO): Granula distribution, degranulation and characterization of released proteins. In: C.G.B. Peterson (1987) Eosinophil granule proteins. Biochemical and functional studies. Doctorial Thesis at Uppsala University, Sweden. Venge, P., Dahl, R. and Hallgren, R. (1979) Enhancement of Factor XII dependent reactions by eosinophil cationic protein. Tromb. Res. 14, 641. Venge, P., Dahl, R., Fredens, K. and Peterson, C.G. (1988) Epithelial injury by human eosinophils. Am. Rev. Respir. Dis. 138, 54. Yong, J.D.E., Peterson C.B.G., Venge, P. and Cohn, Z.A. (1986) Mechanism of the membrane damage mediated by human cosinophil cationic protein. Nature 321,613.
Journal of Immunological Methods, 138 (1991) 285-290 ":;, 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 002217599100151K
JIM05899
Detection of eosinophil cationic protein (ECP) by an enzyme-linked immunosorbent assay C.M. Reimert, P. Venge, A. K h a r a z m i a n d K. B e n d t z e n l.aborato~' of Medical A Ilergology, Uni~ersity Hospital, Copenhagen, Denmark, Department of Clinical Chemistry, University Hospital, Uppsala. Sweden. Department of Clinical Microhiologo'. University ttospital, Copenhagen, Denmark, and Laborato~." of Medical Immunology', University Hospital. Copenhagen, Denmark Received 23 October 1990. revised received 18 December 1990. accepted 15 Janua~' 1991)
Eosinophil cauonic protein (ECP) is a highly basic and potent cytotoxic single-chain zinc-containing protein present in the granules of the eosinophilic granulocytes. ECP appears to be involved in defence against parasites and in the tissue damage seen in subjects with allergic and inflammatory disease. To investigate ECP release from in vitro activated human eosinophils and to study the involvement of eosinophils in health and disease, we have developed a sensitive and specific enzyme immunoassay. ECP was purified from normal human peripheral blood eosinophils and polyclonal antibodies to ECP were subsequently raised in rabbits. The ELISA utilizes the biotin/avidin method and measures ECP within the range 15-1000 ng/l. The intra- and interassay coefficients of variation were 6% and 10%, respectively, and the recoveries of 12 and 25 pg of purified ECP added to diluted serum samples were 108 + 14.5% (mean + SD, n = 12) and 107 + 7.5%, respectively. The high sensitivity, reproducibility and specificity of this ELISA makes it suitable for the determination of minute amounts of ECP in in vitro systems as well as in various biological fluids. K~W words: Eosinophil cationic protein; ELISA: Purification
Introduction
The granules of the human eosinophilic granulocytes contain several highly cationic proteins which upon activation and stimulation of the cells are secreted. One of these proteins, the
Correspondence to: C.M. Reimert. Laborators' for Medical Allergology, TTA 7542. State University Hospital. Tagensvej 20. DK-2200. Copenhagen, Denmark. Abbre~'iations: ECP. eosinophil cationic protein: CTAB, N-cctyI-N.N.N.-trimethyl-ammonium-bromide; ELISA. enzyme-linked immunosorbent assay; SDS. sodium dodecyl sulphate: HSA. human serum albumin; RIA. radioimmunoassay.
eosinophil cationic protein (ECP), is a zinc-containing, single-chain molecule with a molecular weight of 18-22 kDa and a pl > 11 (Olsson et al., 1977). Several biological functions of ECP have been described including interaction with proteins involved in blood coagulation (Venge et al., 1979), cytotoxic activity against the schistosomula of Schistosoma mansoni (MacLaren et al., 1981, 1984), epithelial damaging effects in asthmatic individuals (DeMonchy et al., 1985; Venge et al., 1988), perforin-like activity resulting in increased membrane permeability (Yong et al., 1986), inhibition of lymphocyte proliferation in mixed lymphocyte cultures and phytohaemagglutinin (PHA)-stimulated cell cultures (Peterson et al., 1986) and neu-
286
rotoxicity when injected into the brains of rabbits (Fredens et al., 1982). The mechanisms by which ECP and the other cationic granular proteins are released from the eosinophils are poorly understood. Experimental data indicate a selective release of individual granular proteins depending upon the triggering stimulus (Khalife et al., 1986: Peterson et al., 1987). To investigate ECP release from in vitro stimulated eosinophils, and to study the involvement of eosinophils in parasitic, allergic and other inflammatory diseases, we have developed a sensitive and specific enzyme immunoassay for ECP in biological fluids and cell culture supernatants.
Materials and methods
Preparation of granula extracts Granulocyte granule extracts were prepared by a modification of the previously described method (Peterson et al., 1983). Buffy coat cells from healthy blood donors were pooled and mixed with 2 vols. of 1.5% dextran T-500 to sediment the red blood cells. After 45 min at room temperature, the leucocyte rich-plasma was siphoned off and the leucocytes were isolated and washed twice in 0.15 M NaC1, and once in a solution of 0.34 M sucrose, 5 mM EDTA, 12 mM NaHCO). The cell-pellet was resuspended in 4 vols. of ice-cold phosphatebuffered saline (PBS) containing 5 mM EDTA. 12 mM N a H C O 3 and pressurized with N 2 for 30 rain at 500 psi at 4°C under constant stirring in a cell disruption bomb (Parr Instrument Company, Moline, Ik, U.S.A.). The cavitate was collected dropwise into 1 vol. of the same buffer containing 0.7 M sucrose and 0.3 M NaCI, and centrifuged at 800 × g to eliminate cell nuclei and disrupted cells. Granules were isolated from the supernatant by centrifugation at 10,000 x g for 15 min at 4o( ". After three cycles of freezing and thawing, the granula pellet was extracted with 5 vols. of 50 mM acetic acid under constant stirring for 1 h at 40( ". An equal vol of 0.4 M ammonium-acetate was then added and the extraction continued for 3 h. The extract was centrifuged at 20,000 x g for 30 min at 4°C and the supernatant containing the granular proteins was concentrated 30 times by
diafiltration using YM-10 Danvers, MA, U.S.A.).
filters
(Amicon,
Purification of ECI' ECP was purified as previously described (Peterson et al.. 1988). Briefly, the dark green granular extract was fractionated on a Sephadex G-75 superfine column (Pharmacia, Uppsala, Sweden) equilibrated with 0.2 M ammoniumacetate, pH 4. Fractions containing 16 22 kDa bands on sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) were pooled. concentrated and equilibrated with 80 mM ammonium-acetate, 10% glycerol, pH 8.4 (starting buffer 1) by diafiltration on YM-10 filter. The material was then subjected to ion exchange chromatography on a Bio Rex-70 column (Bio-Rad, Munich, F.R.G.). After the first peak had reached the baseline, the flow direction was reversed and proteins were eluted with a combined salt and pH gradient from starting buffer 1 to 0.8 M ammonium-acetate, 10% glycerol, pH 9.2. Fractions containing the last 280 nm peak were concentrated and equilibrated with 0.2 M ammoniumacetetate, 0.5 M NaC1, 10% glycerol, pH 8.0 (starting buffer 2) and applied to a Zn-chelating Sepharose 6B column (Pharmacia). After the first peak had reached the baseline, ECP was eluted as a single peak with a pH gradient from starting buffer 2 to 0.2 M ammonium-acetate, 0.5 M Na('l, 10% glycerol, pH 4.5. In another experiment ECP was purified as decribed above, except that the ion exchange step was excluded. The protein concentration of the purified ECP was determined using the extinction coefficient (k'l~,,,) = 15.45 at 280 nm (Peterson et al.. 1988). Antibodies against ECP were raised in rabbits by three multiple-site intracutaneus injections of 0.1 mg purified protein with Freund's complete adjuvant and boosted 2 weeks before every bleeding with 0.1 nag ECP in Freund's incomplete adjuvant. The immunoglobulin fraction was isolated by standard ammonium-sulphate precipitation (Harboe et al., 1983). Biotin (N-hydroxy-succinimidobiotin, Sigma, St. Louis, MO, U.S.A.) conjugation of the rabbit-antibodies was performed as described (Kendall et al., 1983). SI)S-PAGE was performed in a discontinuous system according to Laemmli (1970) with 16%
287 polyacrylamide in the separation gel and 5% in the stacking gel. The electrophoresis buffer was 25 mM Tris with 20 mM glycine and 0.1% HC1, pH 8.3. Samples were prepared by boiling for 5 min in four-fold Laemmli sample buffer with 10% mercaptoethanol. Molecular weights were determined by comparison with prestained standards covering the range 14.3-200 kDa, (Amersham International, Buckinghamshire, England, U.K.). Silver staining of gels were performed according to the manufacturer (Bio-Rad).
posed to biotin-conjugated rabbit anti-ECP Ig, (100 p-l/well) for 1.5 h at 37°C and, finally, to avidin-peroxidase (Sigma) and left at room temperature for 30 rain. The plates were incubated with enzyme substrate solution at 100 p-l/well. The enzyme reaction was stopped after 20 min. with 150 p-l/well of 2.5 M H2SO 4 and the absorbance measured at 492 nm with a 620 nm reference.
Results
Crossed immunoelectrophoresis This was performed in IEF agarose (Pharmacia) supplemented with 0.02% N-cetyl-N,N,N,-trimethyl-ammonium-bromide (CTAB) in the first dimension gel as described (Reimert and B~gHansen, in press). The electrophoresis buffer was 0.025 M sodium-barbital, 73 mM Tris pH 8.6.
Purification of ECP
EL1SA procedure Buffer. Coating buffer (A): PBS (10 mM
ECP was purified by two methods. Fig. 1 shows the S D S - P A G E of the two ECP preparations. Both preparations contained the three molecular weight forms of ECP (18.5, 20 and 22 kDa), in agreement with previous reports (Olsson et al., 1986). Moreover, it can be seen that the ECP preparation purified by the two-step procedure (Sephadex G-75 superfine and Zn-chelate chro-
NazHPO4, 0.15 NaCI) pH 7.4. Washing buffer (B): PBS. 0.1% Tween 20, pH 7.4. I n c u b a t i o n / sample buffer (C): PBS, 0.1% Tween 20, 0.1% CTAB, 0.2% human serum albumin (HSA), 20 mM EDTA, pH 7.4. Enzyme substrate solution (D): 12 mg o-phenylenediamine (Dakopatts, Glostrup, Denmark) dissolved in 15 ml 0.1 M citrate-phosphate buffer, pH 5.6 p-I 30% hydrogen peroxide were added immediately before use. The solid phase sandwich ELISA was carried out in microtitre plates (Maxisorb, Nunc, Roskilde, Denmark). Coating of plates Each well was coated overnight at 4°C with 100 p-1 rabbit anti-ECP Ig, 1.5 p-g/ml, diluted in buffer A. Coated plates could be stored at 4°C for up to at least 6 weeks. Washing of plates. Before use, and between each subsequent step in the assay, the plates were washed three times in buffer B. Standards and samples. The ECP standards ,le-/ were calibrated using E~cm of purified ECP at 280 nm = 15.45 (Peterson et al.. 1988). Standards and samples were diluted in buffer C before application to the plates. 100 p,l/well. Plates were left overnight at 4°C. The assay procedure was accelerated by incubating the plates for 2 h at 37° C. After incubation the plates were sequentially ex-
Fig. 1. SI)S-PAGE of two ECP preparations. Lane A: ECP purified by chromatography on Sephadex G-75 superfine. BioRex 70 and Zn-chelate Sepharose 6B. Lane B: ECP purified on Sephadex (.;-75 superfine and Zn-chelate Sepharose 6B. MW: molecular weight in kDa.
288
matography) appear to be just as pure as the ECP preparation purified by the three-step procedure using ion exchange chromatography before the final affinity chromatography on Zn-chelate Sepharose. Using crossed immunoelectrophoresis or radial immunodiffusion, none of the ECP preparations exhibited immunological reactivity with antibodies against eosinophil protein X. cathepsin G. myeloperoxidase or lysozyme (not shown).
~492/620
nm
2.5~
/
2.0
('2a~,o
./
//
,
/,.
F
,'2~6)o
,'4)
/
Specificity Judged by crossed immunoelectrophoresis using a crude granular extract in the first dimension gel, the rabbit antibodies showed no cross reactivity with other granular proteins (Fig. 2). Possible interference caused by non-specific binding of endogenous peroxidase in the samples was tested using semipurified MPO and EPO. Neither of these peroxidases showed any activity in the assay. Range of measurement Optimal antibody concentrations and optimal incubation conditions were obtained by checker-
S.5-
~ .
,O.C •C
.
.
.
.
....~
.
.
.
.
.
.
.
.
~lOr'K
.............. 100
~ 00.9
Fc~' ,'~,g/; Fig. 3. Standard curve obtained using the ECP-EI.ISA Insert: serum samples containing 125 ,ug/I and 2.5 # g / I applied in four dilutions to confirm parallelism with the standard cur~e. Titres are shov,n in parentheses.
board titrations. The range of measurement was 15-1000 n g / l using purified ECP as a standard. There was no interference as judged by the parallelism between the standard curve and the diluted serum samples (Fig. 3). The inclusion of the cationic detergent CTAB in the i n c u b a t i o n - / s a m p l e buffer was crucial in stabilizing the assay, probably by preventing aggregation of ECP (data not shown).
Recover)'
..
•
•
,
^
•
Fig. 2. Crossed immunoelectrophoresis of a crude granular extract developed using rabbit anti-ECP Ig in the second dimension gel. Cathode at the top. Insert: zone-electrophoresis of crude granular extract.
After adding 12 or 25 pg of purified ECP to 100 ffl diluted serum samples the mean recoveries were 108 + 14.5% and 109 + 7.5%, respectively (n = 12). To ascertain whether the assay could be used for ECP analysis of cell-free supernatants from in vitro stimulated eosinophils. 25 and 50 pg of purified ECP were added to 100 ffl of RPMI 1640 containing 20% fetal calf serum. The recoveries were 98 + 7.2% and 91.4 + 6.8% respectively (n = 6).
Reproducibilio" When ECP was measured simultaneously in quadruplicate in 10 different sera and in duplicate in 12 sera on four different occasions, the mean
289
12s1 ~0/I) RIA
,,''''.
IOO I
f;
75
.:'. 5O
%= 0.98
2",'''
25EUSA (~/,) 0-0
25
50
75
"00
125
Fig. 4. ECP content in 40 serum samples m e a s u r e d by E C P - R I A a n d E C P - E L I S A . r, = 0.98, p < 0.00001.
intra- and interassay coefficients of variation were 6% and 10%, respectively• Comparison with an ECP-RIA
To establish the accuracy of the ELISA, 40 serum samples were measured with a recently introduced ECP-RIA (Pharmacia), measuring in the range of 2-200 /tg/l, and in a suitable dilution with the ECP-ELISA. The correlation coefficient between the two assays was & = 0.98, p < 0.00001 (Fig. 4). There was no significant difference between the two assays when tested by the W i l c o x o n / P r a t t signed rank sum test (2a = 0.05, p = 0.082).
Discussion
Purification of ECP from buffy coat cells obtained from normal individuals has been described previously (Peterson et al., 1988). Due to the low eosinophii count in normal blood, leading to low amounts of ECP in the starting material, several precautions were introduced to ensure high yields. These precautions included: degassing of buffers and the inclusion of glycerol and ammoniumacetate in buffers to prevent aggregation, adherence and oxidation of ECP. The use of reversed flow for the gradient elution in ion exchange chromatography on Bio-Rex 70 was of particular importance in achieving a high yield. Despite these precautions, the recovery from the ion exchange chromatography step was relatively low compared
with the recoveries from the size exclusion and affinity chromatrography steps. However, on the basis of the silver-stained SDS-PAGE and crossed immunoelectrophoresis, the ion exchange chromatography step could be excluded without diminishing the purity of the final ECP preparation• Both preparations gave the characteristic three molecular weight forms of ECP (18.5, 20 and 22 kDa) in agreement with previous reports (Olsson et al., 1986; Peterson et al., 1988). The three molecular weight forms of ECP reflect different degrees of glycosylation. Ion-exchange chromatography has shown that the three forms differ in charge, the 18.5 kDa form being the most cationic and the 22 kDa form the least cationic (Petersson et al., 1988)• These differences in charge densities and amounts of the different weight forms are clearly seen by the size/height of the leading front and the following shoulders in crossed immunoelectrophoresis of crude granular extract using anti-ECP antibodies• These experiments also show the immunological identity between the different forms of ECP, and the absence of cross-reactivity between anti-ECP antibodies and other antigens found in the granular extract. There was a good linear correlation between the ELISA and the commercially available RIA for ECP. Generally, however, the ELISA values exceeded those obtained in the RIA when using serum samples with more than 50 n g / m l of ECP. Measurements of cell products in biological fluids usually require specific sample processing• Preliminary results (data not shown) support this notion since spontanous ECP release occurs when processing cell-containing samples• Therefore standardised sample processing with regard to time and temperature is important• The high sensitivity of the present ELISA procedure permits measurements of ECP in high dilutions of biological fluids, thus reducing non-specific binding and interference from other proteins. No interference has been detected in serum, EDTA-plasma, urine, sputum, saliva, eye or nasal lavage fluids• However, interference in citrated plasma was occasionally observed when such samples were tested at high concentrations• We conclude that the high sensitivity, reproducibility and specificity of this ELISA makes it suitable for in vitro as well as in vivo investigations•
290 Acknowledgements
W e t h a n k Ms. U l l a M i n u v a f o r s k i l l e d t e c h n i c a l assistance. The study was supported nology Programme.
by Danish
Biotech-
References
De Monchy, J.G.R., Kauffman, H.F., Venge, P., Koeter, G.H., Jansen, HM.. Sluiter, H.J. and De Vries, K. (1985) Bronchoalveolar eosinophilia during allergen-induced late astmatic reactions. Am. Rev. Respir. Dis. 131,373. Fredens. K. and Dahl, R. (1982) The Gordon phenomenon induced by the eosinophil cationic protein and eosinophil protein X. J. Allergy Clin. immunol. 70, 361. Harboe, N.M.G. and Ingild, A. (1983) Immunization, isolation of immunoglobulins and antibody titre determination. Scand. J. lmmunol. 17 (suppl. 10) 345. Kendall, C., lnoescu-Matiu, I. and Dreesman, G.R. (1983) Utilization of the biotin/avidin system to amplify the sensitivity of enzyme-linked immunosorbent assay (ELISA). J. Immunol. Methods 56, 329. Khalife, J., Capron, M., Cesbron, J.Y., Tai, P.-C., Taelman, H., Prin, L and Capron, A. (1986) Role of specific igE antibodies in peroxidase (EPO) release from human eosinophils. J. lmmunol. 137, 1659. 1.aemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage "I"4. Nature 227, 680. MacLaren, D.J., Mackean, J.R., Olsson, 1., Venge, P. and Kay, A.B. (1981) Morphological studies on the killing of schistosomula of Schistosoma mansoni by human eosinophil and neutrophil cationic proteins in vitro. Parasite lmmunol. 3, 359. MacLaren, D.J., Peterson, C.G.B. and Venge, P. (1984) Schis-
tosoma mansoni: further studies of the interaction between schistosomula and granulocyte derived cationic proteins in vitro. Parasitology 88, 491. Olsson, I., Persson, A.M. and Wingqvist. I. (1986) Biochemical properties of the eosinophil cationic protein and demonstration of its biosynthesis in vitro in marrow cells from patients with eosinophilia. Blood 67. 498. Olsson 1., Venge, P. and Spitznagel, I.K. (1977) Arginine-rich cationic proteins of human cosinophil granules. Comparison of the constituents of the eosinophilic and neutrophilic leukocytes. Lab. Invest. 36, 493. Peterson. C.G.B. and Venge, P. (1983) Purification and characterization of a new cationic protein - eosinophil protein X (EPX) - from the granules of human eosinophils. Immunology 50, 19. Peterson, C.G.B., Skoog, V. and Venge, P. (1986) Human eosinophil cationic proteins (ECP& EPX) and their suppressive effects on lymphocyte proliferation. Immunobiology 171, 1. Peterson, C.B.G., Jt~rnvall, H. and Venge, P. (1988) Purification and characterization of eosinophil cationic protein from normal human eosinophils. Eur. J. Haematol. 40. 415. Peterson, ('.G.B., Garcia. R.C., Carlson M.G.Ch. and Venge, P. (1987) Eosinophil cationic protein (ECP, eosinophil protein X (EPX) and eosinophil peroxidase (EPO): Granula distribution, degranulation and characterization of released proteins. In: C.G.B. Peterson (1987) Eosinophil granule proteins. Biochemical and functional studies. Doctorial Thesis at Uppsala University, Sweden. Venge, P., Dahl, R. and Hallgren, R. (1979) Enhancement of Factor XII dependent reactions by eosinophil cationic protein. Tromb. Res. 14, 641. Venge, P., Dahl, R., Fredens, K. and Peterson, C.G. (1988) Epithelial injury by human eosinophils. Am. Rev. Respir. Dis. 138, 54. Yong, J.D.E., Peterson C.B.G., Venge, P. and Cohn, Z.A. (1986) Mechanism of the membrane damage mediated by human cosinophil cationic protein. Nature 321,613.