Electrophoresis 1990. I ! . 41-45
Rapid detection of proteins by enzyme-linked immunofiltration assay
5 References Towbin, H., Staehelin, T. andGordon,J.,Proc. Natl. Acad. Sci. USA 1976,9,4350-4354. Anderson, N.L., Nance, S. L., Pearson, T. W. and Anderson, N. G., Electrophoresis 1982,3,135-142. Peltre, G.,Weyer, A,, Le Mao, J., Lapeyre, J. and David. B.. Abstracts 4~hInternationalCongressofImmunology,Paris, 1980,No.13.1.21. Peltre, G.,Lapeyre,J. andDavid,B.,in:Molina,C.(Ed.).Proceedings of the European Academy of Allergology and Clinical Immunology Meeting, Lavoisier, 1982,Vol. 1, pp. 92-97. Peltre, G., Lapeyre, J . and David, B., Immunol. Lett. 1982, 5 , 127-131. Legocki, R. P. and Verma, D. P. S., Anal. Biochem. 1981, 1 1 1 , 385-392.
Jean-Michel Pinon Dominique Puygauthier-Toubas Herd Lepan Cathy Marx Annie Bonhomme Joselyne Boulant RCgine Geers HervC Dupont Laboratoire de ParasitologieMycologie C.H.R.U., Reims
41
[71 Manabe, T. Takahashi, Y. and Okuyama, T., Anal. Biochem. 1984, 143,39-45. Lsl Miribel,L., Peltre,G., Amaud, P. and David, B.,in: Dunn.M. J.(Ed.), Proceedings of the Fifth Meeting of the International Electrophoresis Society, VCHVerlagsgesellschaft, Weinheim 1986,pp. 703-706. 191 Demeulemester, C., Peltre, G., Laurent, M. and David. B.. Electrophoresis 1987,8,71-73. lol Mecheri, S., Peltre, G. and David, B., lnt. Arch. Allergy Appl. Immunol. 1985, 78,283-289. I' Vaessen, R.T. M. J., Kreike, J. and Groot, G. S . P., FEBS Leu. 198I,
124, 193-196. [12] Bittner, M., Kupferer, P. and Morris, C. F.,Anal. Biochem. 1980,102, 459-471.
Rapid detection of proteins by enzyme-linked immunofiltrationassay after transfer onto nitrocellulose membranes Enzyme-linked immunofiltration assay (ELIFA) for labeling transferred proteins is an interesting and powerful technique for the rapid specific detection (1 5 min) of proteins immobilized on nitrocellulose or nylon membranes (0.20 and 0.45 pn).ELIFA does not require fastidious handling of the membranes. Saturation, specific labeling and washing procedures are achieved by filtration, controlled by a monitoring unit which regulates the flow rate and ensures excellent specificity, repetition and reproducibility. The recycling by closed circuit or by repetetive inversion of the flow direction offers the advantage of reducing the volumes of expensive reagents while simultaneously increasing the sensitivity of the technique. The detection limit is at least as low as 1-5 ng using directly or indirectly enzymatically labelled probes. ELIFA may be extended to the identification of glycoproteins using specific ligands such as lectins or to the immunocapture of an antigen using specific antibodies immobilized on an activated membrane. ELIFA complements fast separation, by e.g., isoelectric focusing, polyacrylamide gel electrophoresis, or sodium dodecyl sulfatepolyacrylamide gel electrophoresis and accelerated electrotransfer to membranes with rapid detection reducing the total time for separation transfer and detection to less than 2 h.
1 Introduction Described by Renart [ 11 and Towbin [21 in 1979, the transfer of proteins onto immobilizing membranes has opened up a considerable field of biological applications. At present, in most cases blotting is the intermediate step of a three-step methodology including electrophoretic separation and finally protein detection using labelled probes, ligands or specific Correspondence: Dr. J . M. Pinon, Laboratoire de Parasitologie-Mycologie, CHRU HBpital Maison Blanche, 45 rueCognacq Jay, F-5 1092Reims, France Abbreviations: 2-DE, two-dimensional electrophoresis; ELIFA, enzymelinked immunofiltration assay; IEF, isoelectric focusing; Mab, monoclonal antibody; NC, nitrocellulose: PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; Tg, Tomplasma gondii; TTBS, Tris-Tween buffered saline 0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
antibodies [3,41. This analytical approach wouldhave aneven greater impact in research or in biodiagnostics if the three steps, preferably standardized, were simple and fast, and the consumption of expensive reagents could be reduced. The recently described techniques of horizontal electrophoresis in miniaturized gels 151 and of accelerated electrotransfer 161 onto nitrocellulose (NC) membranes can easily be carried out in less than 1 h. By contrast, the last visualization step, including saturation or inactivation of the membranes, followed by specific antigen detection, has not been developed in parallel and still requires at least several hours [3,7]. Therefore, it seemed challenging to adapt the potential of rapid immunodetection, offered by the enzyme-linked immunofiltration assay (ELIFA), to immobilizing membranes. Initially, this technique was developed for the characterization of antibodies (IgG, IgM, IgA or IgE) implicated in electro-immunoprecipitating reactions carried out on cellulose acetate membranes 0 1 73-0835/90/0 101-004 I %02.50/0
42
J.-M. Pinon el al.
181. For the major part of our work, we chose the toxoplasmosis model which allowed us to evaluate the method using a heterogeneous antigenic system and different monoclonal antibodies (Mab) or human sera.
2 Materials and methods 2.1 The Toxoplasma antigen The antigen used is a soluble extract of an RH strain of Toxoplasma gondii (Tg). Briefly, tachyzoitec; were obtained from Tg cultivated on fibroblasts or recovered from mouse peritoneal fluid 72 h after infection. The parasites were lysed by successive freeze-thawing cycles followed by ultrasonication [ 81. After centrifugation, the supernatant (2-4 mg protein per mL) was aliquoted and kept at -80 OC. Culture media andperitoneal fluids, which contain soluble antigen released or excreted by Tg, were also used as a source of antigen.
Electrophoresis 1990,11,41-45
2.5 Transfer onto nitrocellulose 2.5.1 Electrophoretic transfer After PAGE, SDS-PAGE, IEF or 2-DE, the polyacrylamide gel was removed from its support using a Filmremover (LKB, Cat. No. 2 1 17-225) to allow for the electrotransfer of the proteins. Semi-dry transfer (30-45 min at 24 V) was carried out in a horizontal tank (Biolyon, No. 501 11)with glass (anode) and stainless steel (cathode) electrodes.
2.5.2 Passive transfer After IEF, the transfer can also be carried out passively by capillarity (120 min). A cellulose acetate sheet (0.45 pm; Sartorius, Cat. No. SM11200BB) is placed on top of the polyacrylamide gel, followed by an immobilizing membrane of 0.45 km and six sheets of filter paper impregnated with 1 % acetic acid. This set-up is weighted by a 250 g weight. 2.6 ELIFA
2.2 Anti-Toxoplasma antibodies Four mouse monoclonal antibodies (Clonatec, Paris) were tested in addition to the culture supernatants of the corresponding clones. These antibodies selected in our laboratory recognize different antigenic constituents. Some monoclonal antibodies, produced in mouse ascites and purified (concentration 1 mg/mL) were labelled [91 with biotin (Sigma, Cat.No. B 2643). The human sera camefrompatients with acquired or evolving congenital toxoplasmosis, diagnosed by compared immunological profiles ELKFA [81, by highly sensitive direct agglutination and by IgM or IgA immunocapture.
After protein transfer, the region of the immobilizing membrane to be studied (2.5 cm x 4 cm) was excised, while the rest was stained with Amido Black (Merck, Cat. No. 1167) or Serva Blue R (Serva, Cat. No. 35005 1). This made it possible to control the efficiency of the electrotransfer, and to determine, if necessary, the molecular weights or p1 values, using calibration kits (Pharmacia, Low or High Molecular Weight Calibration Kits, and p1 Calibration Kit 3-10). Each membrane was placed in a filtration cell, especially designed for ELIFA (Fig. 1).The active filtration was programmed by a monitoring unit. This is connected to an electrovalve and two multi-cassette pumps, and automatically ensures perfectly
2.3 Other reagents Various nylon and nitrocellulose membranes (Sartorius) with porosities of 0.20, 0.45, 0.60 and 0.80 pm were tested. Antibodies labelled with peroxidase or with alkaline phosphatase (Institut Pasteur Production, Batch No. 75200, 75061,75041; Biosys, Batch No. 2413, 2513) were used.
2.4 Electrophoretic separation techniques
Toxoplasma antigen, either in its native f'orm or denatured by sodium dodecyl sulfate (SDS, Sigma Cat. No. L57 50), in a volume of 1 pL was applied to the surface of polyacrylamide gels (PhastGel IEF, PhastGel Gradient 10-15 %; 50 x 40 x 0.35-0.45 mm; separation length 32 mm). Strips impregnated with SDS (PhastGel SDS buffer strips) were used for SDSpolyacrylamide gel electrophoresis (PAGE). Isoelectric focusing (IEF; 25 min), PAGE (60 min), SDS-PAGE (25 min) and two-dimensional electrophoresis (2-DE; 90 min) were carried out using PhastSystem, under conditions previously standardized [5], as recommended by the manufacturer (Pharmacia). In some cases, the gels were stained with Amido Black, Coomassie Brilliant Blue or with silver stain [51.
Figure 1 . Filtration cell. The NC membrane is placed at the bottom of the cell (a) between two fitting and thin window-joints which prevent damage of the membrane. The upper part of the system is the reservoir (b). The two parts are connected with automated pumps which regulate the filtration of fluid through the membrane for immunodetection.
43
Electrophoresis 199O,Il, 41-45
Rapid detection of proteins by enzyme-linked immunofiltration assay
synchronized filling and filtration of each cell (from 1to 4) with different solutions at a selected flow rate. The solutions may be recycled, either in a closed circuit, or by multiple inversions of the flow direction (top F? bottom). Two examples of experimental procedures are indicated below, depending whether the probe which recognizes the antigen is detected with or without indirect immunolabelling.
nonspecific interference between probes and matrices must be suppressed, using blocking or inactivating agents [31. However, the efficiency of such agents must not unduly hamper the later detection step, as a result either of masking of functional epitopes or of excessive reduction of hydrophobic protein interactions which may vary according to the membranes used. Because of consistently high backgroundlevels,we did not use nylon membranes in reference protocols. Rather, we chose nitrocellulose matrices (0.20 and 0.45 Fm). When higher porosities were tested (0.60 and 0.80 pm), the risk of elution during filtration necessitated a fixation step after electrotransfer, using 1 % w/v formaldehyde-glutaraldehyde (60 min). Blocking of active sites of the matrix was carried out using a solution of NaHCO, - ethanolamine containing skimmed milk. Skimmed milk was preferred to 1 % bovine serum albumin which occasionally formed an undesirable precipitate at the membrane surface. Tween 20, rather than Tween 80, was used to further inactivate the membrane and for other steps of the technique.
The first protocol is for a biotin-labelled probe (flow rate 1 mL/ min; temperature 20-22 "C; total time 15 min) and consists of the following steps. 1) Filtration (2.5 mL) using a saturation solution ( 1 NaHCO,, ~ Merck, Cat. No. 6329; 1 % ethanolamine, Merck, Cat. No. 844 and 0.2 % skimmed milk powder, pH 8.6). 2) Washing (2.5 mL) and additional inactivation using phosphate buffered saline (PBS, Biomerieux, Cat. No. 755 1-l), 1 %Tween 20 (Merck, Cat. No. 822184). 3) Filtration (2.5 mL) using biotinylated probe (0.75 % to 0.01 %in asolution composedof0.4 MNa,B,O,(Sigma,Cat. No. 9876), 0.15 M NaCl (Merck, Cat. No. 6404), 0.5 %skimmed milk powder, 0.5 % Tween 20, pH 8.2. 4) Washing (2.5 mL) using PBS, containing 0.2 %Tween 20.5) Filtration (2.5 mL) using alkaline phosphatase-labelled avidin (Sigma, Cat. No. A2527) at a concentration of 0.3 % in 0.02 M Tris (Merck, Cat. No. 8382), 0.25 M NaCl, 0.2 % Tween 20, pH 7.4 (TTBS). 6) Washing (2.5 mL) using TTBS. The membranes were then removed from the filtration cells and immersed into a developing solution, prepared immediately before use (0.05 % 5-bromo-4-chloro-3-indolyl phosphate; Sigma, Cat. No. B8503; 0.3 M 2-amino 2-methyl-1-propanoli; Prolabo, Cat. No. 21.164235, pH 10.15). The development reaction, carried out for several minutes with agitation, was stopped by washing in 0.5 M HC1 (Merck, Cat. No. 9970). In the second protocol, the probe is detected indirectly using a second antibody. The flow rate, steps 1 , 2 and 4 and the total time for ELIFA (15 min) were the same as for the first protocol. Steps 3,s and 6 were modified as follows: 3) Filtration (2.5 mL) using the probe (unconcentrated clone culture supernatants, antibodies or antiserum at a concentration of 0.25 %in PBS buffer, containing 0.5 %skimmed milk powder and 0.5 % Tween 20). 5) Filtration (2.5 mL) using labelled anti-mouse or human immunoglobulin antibodies (1 mg/mL) diluted 0.1-0.01 % in PBS. 6) Washing (2.5 mL) using PBS and 0.2 % Tween 20. If the second antibody was labelled with peroxidase, the membranes wereimmersed in three successive baths of 0.1 M Tris (Merck, Cat. No. 8382) 0.25 M HC1, pH 7.6, for 2 min. 3,3'-Diaminobenzidine (0.25 mg/mL, Sigma, Cat. No. D5637) was added to the second bath and 0.1 % H 2 0 , to the third. The reaction was stopped using a 1 % solution of acetic acid (Merck, Cat. No. 1627663).
3.2 Flow rates, volumes and duration The duration of ELIFA depends on the flow rates and on the volumes filtered. Based on many preliminary experiments aiming to simplify the procedure, we have chosen an average, constant, flow rate for all steps of ELIFA. This flow rate is 1 mL/min/cell, i. e. 1 mL/min/7 cm2 or 0.143 mL/min/cm2 of filtered surface. With higher flow rates there is a risk ofprotein elution or a reduction in the efficiency of saturation and immunodetection. In the two protocols described in Section 2.6, rest periods between the various steps did not significantly improve the results. On the contrary, they can even induce a displacement of transferred proteins. The volumes used (2.5 mL for each of the solutions filtered) were also kept constant, taking into account three interdependent factors: time, flow rates and total volumes. However, the volume can be increased if the probes are strongly diluted. Inversely, programmed repetitive change of the flow direction (or recycling by closed circuit) is an extremely interesting alternative, which allows the consumption of rare and or expensive samples (e. g. probes, labelled monoclonal antibodies) to be reduced. Under the recommended conditions, the total time of 15 min required by ELIFA, including saturation and single- or doublelayer immunodetection, was completely adequate. The procedure does not require any additional handling of the membranes. By contrast, carried out conventionally by passive contact, the same operations required at least several hours, if not all night [3,4,7], and in addition, the change of solutions at each step of saturation, inactivation, immunodetection and intermediate washings was rather fastidious.
3 Results and discussion 3.1 Standardization of parameters and choice of reagents The aim of ELIFA was to achieve rapid and sensitive immunodetection by favoring specific contact between a component initially transferred onto an immobilizing membrane and a probe brought into contact with this component by active filtration [81. The standardization of this technique required a compromise in the choice of the various parameters because several objectives had to be reconciled simultaneously. Any
3.3 Repetition and reproducibility In more than several hundred trials the repetition and reproducibility were excellent. The active filtration must be perfectly regulated and the membrane must remain submerged throughout ELIFA. If several cells are used simultaneously, the filtrations must be synchronized to avoid premature depletion and drying of any of the membranes. These three conditions are easily met using the monitoring unit, multicassette, pumps and electrovalves.
44
J.-M.Pinon el a/.
Electrophoresis 1990,ll. 41-45
Figure 2. Examples of ELIFA detection after horizontal electrophoresis, followed by electrotransfer onto 0.2 pm NC membranes. PAGE: (A), (B) Peritoneal fluid of Tg infected mice; colloidal gold total protein detection. (A) Passive procedure, 105 min, (B) filtration, 10 min. (C), (D) High Molecular Weight Calibration Kit. (C) Colloidal gold total protein detection, filtration, 10 min, (D) Amido Black staining, 100min. (E), (F)Immunodetectionof 2 immunodominant Tg antigens (66 kDa 45 kDa), ELIFA (15 min) with (E) IE,, and (F)I1 38 Mab. SDS-PAGE, Coomassie Blue PhastSystem staining. (G) Low molecular Weight Calibration Kit,(H)Tg soluble extract antigen.
3.4 Specificity and sensitivity K 04The high specificity of immunodetection by ELIFA has already been demonstrated for cellulose acetate [8,101. In addition, it has been demonstrated on nitrocellulose using: 1) anti-Tg and anti-HBS monoclonal antibodies, either unlabelled or directly coupled t o biotin or an enzyme, 2) human or mouse sera with or without toxoplasmic antibodies, and 3) anti-mouse immunoglobulins or antibodies directed against human IgG, IgM or IgA, coupled to peroxidase, alkaline phosphatase or biotin. Our tests in homologous or heterologous systems confirmed the high specificity of immunodetection by both single- and double-layer techniques (Fig. 2). A further advantage of ELIFA is the amplification of sensitivity. This can easily be demonstrated by indirect immunodetection, where the second labelled antibody can be diluted much stronger (1/30 000, i. e. 30 ng/mL) than when detection is carried out passively with similar markers [.3,4,7]. Such extreme dilution of the labelled antibody allows suppression of nonspecific background. The limit of detection of blotted proteins is at least as low as 1-5 ng and is thus at least comparable with resultsreportedpreviously [3,4,7].Itisknownthatthelimitof detection varies according to the protein blotted, the nature and concentration of the probe and the method of single or double layer detection used. It should ble noted that inversed filtration, which facilitates labelling of the membrane, increases the sensitivity of ELIFA.
3.5 Other applications 3.5.1 Detection of glycoproteins after transfer onto nitrocellulose Glycoproteins are important components of many biochemical systems. Concanavalin A (Con A) is one of the lectins capable of fixing carbohydrate residue55 [ 111. Using a complex antigen, as is the case for soluble Toxoplasma extracts, we could easily show that a constituent recognized by one of the monoclonal antibodies tested was a glycoprotein. For this
demonstration, the matrices were filtered for 10min using Con A labelled with peroxidase (0.1 % in TTBS) or with biotin. In the latter case, the final detection step was carried out using an avidin-alkaline phosphatase system.
3.5.2 Detection after immobilization of antibodies or antigens on activated membranes Based on the same principle, reverse ELIFA (antibodies immobilized on the membrane) can be applied to the search for bacterial, viral, parasitic or fungic antigens in different solutions or biological liquids (cerebrospinal or amniotic fluids, urine, sera, etc.). The technique of filtration with recycling offers two major advantages over classical techniques: the sample can be treated directly in an unconcentrated form, and the entire sample can be tested. In a model using an antitoxin, immobilized on a cellulose acetate membrane by cyanogen bromide [ 121, and alabelled second specific antibody, the limit of detection of a tetanic anatoxin was 100 pg/mL when recycled filtration was used. At last, with an alkaline-phosphatase labelled antibody, the limit of detection of immunoglobulin G immobilized on nitrocellulose membranes is 20 pg/mm2.
3.5.3 Filtration used to stain proteins transferred onto nitrocellulose We tested by filtration different methods to detect total proteins. Using the Bio-Rad Biotin-Blot Protein Detection Kit (Cat. No. 170.65 12), filtration allowed the detection of 10 ng (3 ng/mm2) in 40 min. This contrasts with the manufacturer’s recommendations, which give a detection time of 1 h 45 min, using passive agitation, and a threshold of 100 ng. With gold staining (Enhanced Colloidal Gold Total Protein Detection Kit, Bio-Rad Cat. Nr. 170.165 17), the technique can be carried out in 10 min by filtration, with a limit of detection of 100 pg without background. Using passive staining, the recommended time was 1-3 h, and the threshold was 400 pg (Fig. 2).
45
Electrophoresis 1990. I l , 41-45
Rapid detection of proteins by enzyme-linked immunofiltration assay
4 Concluding remarks
5 References
The application of ELIFA to the rapid immunodetection of proteins transferred onto immobilizing matrices seems innovative and useful by several criteria. (i) Including saturation, inactivation, detection and washing procedures, immunodetection by regulated filtration can be carried out in 15 min. This technique thus complements other fast methods of electrophoresis and electroblotting, resulting in a powerful combination with a total time of the entire procedure of less than 2 h. (ii) The detection of a blot by ELIFA, or ELIFA-blot, depends on a simple technique which can be automated. It requires no handling of the membranes. Repeated inversion of the filtration direction offers the additional advantages of reducing the volumes of expensive samples, while simultaneously increasing the sensitivity of the method. The limit of detection is at least as low as 1-5 ng, whether the probes are labelled directly or indirectly by an enzyme. (iii) The same principle of rapid detection can beused both for the identification and characterization by ligands of glycoproteins transferred onto nitrocellulose, and for the immunocapture of an antigen by an antibody immobilized onto an activated membrane.
[ I ] Renart, J., Reiser, J. and Stark, G. R., Proc. Natl. Acad. Sci. USA 1979, 76,3116-3120. 121 Towbin, H., Staehelin, T. and Gordon, J.,Proc.Natl. Acad. Sci. USA 1979, 76,4350-5354. L31 Beisiegei, U., Electrophoresis 1986, 7, 1-18. [4] Burnette, W. N., Anal. Biochem. 1981,112, 195-203. 151 Olsson, I., Axio-Fredriksson, U. B., Degerman, M. and Olsson, B., Electrophoresis 1988,9, 16-22. [61 Tovey, E. R. and Baldo, B. A., Electrophoresis 1987,8,384-387. [71 Van der Sluis, P. J., Pool, C . W. and Sluiter, A. A., Electrophoresis 1988,9,654-66 1. [Sl Pinon, J . M., Thoannes, H. and Gruson, N., J. Immunol. Methods 1985, 77, 15-23. [91 Guesdon, J. L., Ternynck, T. and Avrameas, S . , J . Histochem. Cytochern. 1979,27,113 1- 1 139. [lo1 Pinon, J . M., Gem, R., Lepan, H. and Pailler, S., Eur. J . Resp. Dis. 1987, 71,164-169. [ill Hawkes, R., Anal. Biochem. 1982,123, 143-146. 1121 Demeulemestre, C., Peltre, G., Laurent, M., Panheleux, D. and David, B., Electrophoresis 1987,8, 71-73.
Received June 1, 1989
Rapid detection of proteins by enzyme-linked immunofiltration assay
5 References Towbin, H., Staehelin, T. andGordon,J.,Proc. Natl. Acad. Sci. USA 1976,9,4350-4354. Anderson, N.L., Nance, S. L., Pearson, T. W. and Anderson, N. G., Electrophoresis 1982,3,135-142. Peltre, G.,Weyer, A,, Le Mao, J., Lapeyre, J. and David. B.. Abstracts 4~hInternationalCongressofImmunology,Paris, 1980,No.13.1.21. Peltre, G.,Lapeyre,J. andDavid,B.,in:Molina,C.(Ed.).Proceedings of the European Academy of Allergology and Clinical Immunology Meeting, Lavoisier, 1982,Vol. 1, pp. 92-97. Peltre, G., Lapeyre, J . and David, B., Immunol. Lett. 1982, 5 , 127-131. Legocki, R. P. and Verma, D. P. S., Anal. Biochem. 1981, 1 1 1 , 385-392.
Jean-Michel Pinon Dominique Puygauthier-Toubas Herd Lepan Cathy Marx Annie Bonhomme Joselyne Boulant RCgine Geers HervC Dupont Laboratoire de ParasitologieMycologie C.H.R.U., Reims
41
[71 Manabe, T. Takahashi, Y. and Okuyama, T., Anal. Biochem. 1984, 143,39-45. Lsl Miribel,L., Peltre,G., Amaud, P. and David, B.,in: Dunn.M. J.(Ed.), Proceedings of the Fifth Meeting of the International Electrophoresis Society, VCHVerlagsgesellschaft, Weinheim 1986,pp. 703-706. 191 Demeulemester, C., Peltre, G., Laurent, M. and David. B.. Electrophoresis 1987,8,71-73. lol Mecheri, S., Peltre, G. and David, B., lnt. Arch. Allergy Appl. Immunol. 1985, 78,283-289. I' Vaessen, R.T. M. J., Kreike, J. and Groot, G. S . P., FEBS Leu. 198I,
124, 193-196. [12] Bittner, M., Kupferer, P. and Morris, C. F.,Anal. Biochem. 1980,102, 459-471.
Rapid detection of proteins by enzyme-linked immunofiltrationassay after transfer onto nitrocellulose membranes Enzyme-linked immunofiltration assay (ELIFA) for labeling transferred proteins is an interesting and powerful technique for the rapid specific detection (1 5 min) of proteins immobilized on nitrocellulose or nylon membranes (0.20 and 0.45 pn).ELIFA does not require fastidious handling of the membranes. Saturation, specific labeling and washing procedures are achieved by filtration, controlled by a monitoring unit which regulates the flow rate and ensures excellent specificity, repetition and reproducibility. The recycling by closed circuit or by repetetive inversion of the flow direction offers the advantage of reducing the volumes of expensive reagents while simultaneously increasing the sensitivity of the technique. The detection limit is at least as low as 1-5 ng using directly or indirectly enzymatically labelled probes. ELIFA may be extended to the identification of glycoproteins using specific ligands such as lectins or to the immunocapture of an antigen using specific antibodies immobilized on an activated membrane. ELIFA complements fast separation, by e.g., isoelectric focusing, polyacrylamide gel electrophoresis, or sodium dodecyl sulfatepolyacrylamide gel electrophoresis and accelerated electrotransfer to membranes with rapid detection reducing the total time for separation transfer and detection to less than 2 h.
1 Introduction Described by Renart [ 11 and Towbin [21 in 1979, the transfer of proteins onto immobilizing membranes has opened up a considerable field of biological applications. At present, in most cases blotting is the intermediate step of a three-step methodology including electrophoretic separation and finally protein detection using labelled probes, ligands or specific Correspondence: Dr. J . M. Pinon, Laboratoire de Parasitologie-Mycologie, CHRU HBpital Maison Blanche, 45 rueCognacq Jay, F-5 1092Reims, France Abbreviations: 2-DE, two-dimensional electrophoresis; ELIFA, enzymelinked immunofiltration assay; IEF, isoelectric focusing; Mab, monoclonal antibody; NC, nitrocellulose: PAGE, polyacrylamide gel electrophoresis; SDS, sodium dodecyl sulfate; Tg, Tomplasma gondii; TTBS, Tris-Tween buffered saline 0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
antibodies [3,41. This analytical approach wouldhave aneven greater impact in research or in biodiagnostics if the three steps, preferably standardized, were simple and fast, and the consumption of expensive reagents could be reduced. The recently described techniques of horizontal electrophoresis in miniaturized gels 151 and of accelerated electrotransfer 161 onto nitrocellulose (NC) membranes can easily be carried out in less than 1 h. By contrast, the last visualization step, including saturation or inactivation of the membranes, followed by specific antigen detection, has not been developed in parallel and still requires at least several hours [3,7]. Therefore, it seemed challenging to adapt the potential of rapid immunodetection, offered by the enzyme-linked immunofiltration assay (ELIFA), to immobilizing membranes. Initially, this technique was developed for the characterization of antibodies (IgG, IgM, IgA or IgE) implicated in electro-immunoprecipitating reactions carried out on cellulose acetate membranes 0 1 73-0835/90/0 101-004 I %02.50/0
42
J.-M. Pinon el al.
181. For the major part of our work, we chose the toxoplasmosis model which allowed us to evaluate the method using a heterogeneous antigenic system and different monoclonal antibodies (Mab) or human sera.
2 Materials and methods 2.1 The Toxoplasma antigen The antigen used is a soluble extract of an RH strain of Toxoplasma gondii (Tg). Briefly, tachyzoitec; were obtained from Tg cultivated on fibroblasts or recovered from mouse peritoneal fluid 72 h after infection. The parasites were lysed by successive freeze-thawing cycles followed by ultrasonication [ 81. After centrifugation, the supernatant (2-4 mg protein per mL) was aliquoted and kept at -80 OC. Culture media andperitoneal fluids, which contain soluble antigen released or excreted by Tg, were also used as a source of antigen.
Electrophoresis 1990,11,41-45
2.5 Transfer onto nitrocellulose 2.5.1 Electrophoretic transfer After PAGE, SDS-PAGE, IEF or 2-DE, the polyacrylamide gel was removed from its support using a Filmremover (LKB, Cat. No. 2 1 17-225) to allow for the electrotransfer of the proteins. Semi-dry transfer (30-45 min at 24 V) was carried out in a horizontal tank (Biolyon, No. 501 11)with glass (anode) and stainless steel (cathode) electrodes.
2.5.2 Passive transfer After IEF, the transfer can also be carried out passively by capillarity (120 min). A cellulose acetate sheet (0.45 pm; Sartorius, Cat. No. SM11200BB) is placed on top of the polyacrylamide gel, followed by an immobilizing membrane of 0.45 km and six sheets of filter paper impregnated with 1 % acetic acid. This set-up is weighted by a 250 g weight. 2.6 ELIFA
2.2 Anti-Toxoplasma antibodies Four mouse monoclonal antibodies (Clonatec, Paris) were tested in addition to the culture supernatants of the corresponding clones. These antibodies selected in our laboratory recognize different antigenic constituents. Some monoclonal antibodies, produced in mouse ascites and purified (concentration 1 mg/mL) were labelled [91 with biotin (Sigma, Cat.No. B 2643). The human sera camefrompatients with acquired or evolving congenital toxoplasmosis, diagnosed by compared immunological profiles ELKFA [81, by highly sensitive direct agglutination and by IgM or IgA immunocapture.
After protein transfer, the region of the immobilizing membrane to be studied (2.5 cm x 4 cm) was excised, while the rest was stained with Amido Black (Merck, Cat. No. 1167) or Serva Blue R (Serva, Cat. No. 35005 1). This made it possible to control the efficiency of the electrotransfer, and to determine, if necessary, the molecular weights or p1 values, using calibration kits (Pharmacia, Low or High Molecular Weight Calibration Kits, and p1 Calibration Kit 3-10). Each membrane was placed in a filtration cell, especially designed for ELIFA (Fig. 1).The active filtration was programmed by a monitoring unit. This is connected to an electrovalve and two multi-cassette pumps, and automatically ensures perfectly
2.3 Other reagents Various nylon and nitrocellulose membranes (Sartorius) with porosities of 0.20, 0.45, 0.60 and 0.80 pm were tested. Antibodies labelled with peroxidase or with alkaline phosphatase (Institut Pasteur Production, Batch No. 75200, 75061,75041; Biosys, Batch No. 2413, 2513) were used.
2.4 Electrophoretic separation techniques
Toxoplasma antigen, either in its native f'orm or denatured by sodium dodecyl sulfate (SDS, Sigma Cat. No. L57 50), in a volume of 1 pL was applied to the surface of polyacrylamide gels (PhastGel IEF, PhastGel Gradient 10-15 %; 50 x 40 x 0.35-0.45 mm; separation length 32 mm). Strips impregnated with SDS (PhastGel SDS buffer strips) were used for SDSpolyacrylamide gel electrophoresis (PAGE). Isoelectric focusing (IEF; 25 min), PAGE (60 min), SDS-PAGE (25 min) and two-dimensional electrophoresis (2-DE; 90 min) were carried out using PhastSystem, under conditions previously standardized [5], as recommended by the manufacturer (Pharmacia). In some cases, the gels were stained with Amido Black, Coomassie Brilliant Blue or with silver stain [51.
Figure 1 . Filtration cell. The NC membrane is placed at the bottom of the cell (a) between two fitting and thin window-joints which prevent damage of the membrane. The upper part of the system is the reservoir (b). The two parts are connected with automated pumps which regulate the filtration of fluid through the membrane for immunodetection.
43
Electrophoresis 199O,Il, 41-45
Rapid detection of proteins by enzyme-linked immunofiltration assay
synchronized filling and filtration of each cell (from 1to 4) with different solutions at a selected flow rate. The solutions may be recycled, either in a closed circuit, or by multiple inversions of the flow direction (top F? bottom). Two examples of experimental procedures are indicated below, depending whether the probe which recognizes the antigen is detected with or without indirect immunolabelling.
nonspecific interference between probes and matrices must be suppressed, using blocking or inactivating agents [31. However, the efficiency of such agents must not unduly hamper the later detection step, as a result either of masking of functional epitopes or of excessive reduction of hydrophobic protein interactions which may vary according to the membranes used. Because of consistently high backgroundlevels,we did not use nylon membranes in reference protocols. Rather, we chose nitrocellulose matrices (0.20 and 0.45 Fm). When higher porosities were tested (0.60 and 0.80 pm), the risk of elution during filtration necessitated a fixation step after electrotransfer, using 1 % w/v formaldehyde-glutaraldehyde (60 min). Blocking of active sites of the matrix was carried out using a solution of NaHCO, - ethanolamine containing skimmed milk. Skimmed milk was preferred to 1 % bovine serum albumin which occasionally formed an undesirable precipitate at the membrane surface. Tween 20, rather than Tween 80, was used to further inactivate the membrane and for other steps of the technique.
The first protocol is for a biotin-labelled probe (flow rate 1 mL/ min; temperature 20-22 "C; total time 15 min) and consists of the following steps. 1) Filtration (2.5 mL) using a saturation solution ( 1 NaHCO,, ~ Merck, Cat. No. 6329; 1 % ethanolamine, Merck, Cat. No. 844 and 0.2 % skimmed milk powder, pH 8.6). 2) Washing (2.5 mL) and additional inactivation using phosphate buffered saline (PBS, Biomerieux, Cat. No. 755 1-l), 1 %Tween 20 (Merck, Cat. No. 822184). 3) Filtration (2.5 mL) using biotinylated probe (0.75 % to 0.01 %in asolution composedof0.4 MNa,B,O,(Sigma,Cat. No. 9876), 0.15 M NaCl (Merck, Cat. No. 6404), 0.5 %skimmed milk powder, 0.5 % Tween 20, pH 8.2. 4) Washing (2.5 mL) using PBS, containing 0.2 %Tween 20.5) Filtration (2.5 mL) using alkaline phosphatase-labelled avidin (Sigma, Cat. No. A2527) at a concentration of 0.3 % in 0.02 M Tris (Merck, Cat. No. 8382), 0.25 M NaCl, 0.2 % Tween 20, pH 7.4 (TTBS). 6) Washing (2.5 mL) using TTBS. The membranes were then removed from the filtration cells and immersed into a developing solution, prepared immediately before use (0.05 % 5-bromo-4-chloro-3-indolyl phosphate; Sigma, Cat. No. B8503; 0.3 M 2-amino 2-methyl-1-propanoli; Prolabo, Cat. No. 21.164235, pH 10.15). The development reaction, carried out for several minutes with agitation, was stopped by washing in 0.5 M HC1 (Merck, Cat. No. 9970). In the second protocol, the probe is detected indirectly using a second antibody. The flow rate, steps 1 , 2 and 4 and the total time for ELIFA (15 min) were the same as for the first protocol. Steps 3,s and 6 were modified as follows: 3) Filtration (2.5 mL) using the probe (unconcentrated clone culture supernatants, antibodies or antiserum at a concentration of 0.25 %in PBS buffer, containing 0.5 %skimmed milk powder and 0.5 % Tween 20). 5) Filtration (2.5 mL) using labelled anti-mouse or human immunoglobulin antibodies (1 mg/mL) diluted 0.1-0.01 % in PBS. 6) Washing (2.5 mL) using PBS and 0.2 % Tween 20. If the second antibody was labelled with peroxidase, the membranes wereimmersed in three successive baths of 0.1 M Tris (Merck, Cat. No. 8382) 0.25 M HC1, pH 7.6, for 2 min. 3,3'-Diaminobenzidine (0.25 mg/mL, Sigma, Cat. No. D5637) was added to the second bath and 0.1 % H 2 0 , to the third. The reaction was stopped using a 1 % solution of acetic acid (Merck, Cat. No. 1627663).
3.2 Flow rates, volumes and duration The duration of ELIFA depends on the flow rates and on the volumes filtered. Based on many preliminary experiments aiming to simplify the procedure, we have chosen an average, constant, flow rate for all steps of ELIFA. This flow rate is 1 mL/min/cell, i. e. 1 mL/min/7 cm2 or 0.143 mL/min/cm2 of filtered surface. With higher flow rates there is a risk ofprotein elution or a reduction in the efficiency of saturation and immunodetection. In the two protocols described in Section 2.6, rest periods between the various steps did not significantly improve the results. On the contrary, they can even induce a displacement of transferred proteins. The volumes used (2.5 mL for each of the solutions filtered) were also kept constant, taking into account three interdependent factors: time, flow rates and total volumes. However, the volume can be increased if the probes are strongly diluted. Inversely, programmed repetitive change of the flow direction (or recycling by closed circuit) is an extremely interesting alternative, which allows the consumption of rare and or expensive samples (e. g. probes, labelled monoclonal antibodies) to be reduced. Under the recommended conditions, the total time of 15 min required by ELIFA, including saturation and single- or doublelayer immunodetection, was completely adequate. The procedure does not require any additional handling of the membranes. By contrast, carried out conventionally by passive contact, the same operations required at least several hours, if not all night [3,4,7], and in addition, the change of solutions at each step of saturation, inactivation, immunodetection and intermediate washings was rather fastidious.
3 Results and discussion 3.1 Standardization of parameters and choice of reagents The aim of ELIFA was to achieve rapid and sensitive immunodetection by favoring specific contact between a component initially transferred onto an immobilizing membrane and a probe brought into contact with this component by active filtration [81. The standardization of this technique required a compromise in the choice of the various parameters because several objectives had to be reconciled simultaneously. Any
3.3 Repetition and reproducibility In more than several hundred trials the repetition and reproducibility were excellent. The active filtration must be perfectly regulated and the membrane must remain submerged throughout ELIFA. If several cells are used simultaneously, the filtrations must be synchronized to avoid premature depletion and drying of any of the membranes. These three conditions are easily met using the monitoring unit, multicassette, pumps and electrovalves.
44
J.-M.Pinon el a/.
Electrophoresis 1990,ll. 41-45
Figure 2. Examples of ELIFA detection after horizontal electrophoresis, followed by electrotransfer onto 0.2 pm NC membranes. PAGE: (A), (B) Peritoneal fluid of Tg infected mice; colloidal gold total protein detection. (A) Passive procedure, 105 min, (B) filtration, 10 min. (C), (D) High Molecular Weight Calibration Kit. (C) Colloidal gold total protein detection, filtration, 10 min, (D) Amido Black staining, 100min. (E), (F)Immunodetectionof 2 immunodominant Tg antigens (66 kDa 45 kDa), ELIFA (15 min) with (E) IE,, and (F)I1 38 Mab. SDS-PAGE, Coomassie Blue PhastSystem staining. (G) Low molecular Weight Calibration Kit,(H)Tg soluble extract antigen.
3.4 Specificity and sensitivity K 04The high specificity of immunodetection by ELIFA has already been demonstrated for cellulose acetate [8,101. In addition, it has been demonstrated on nitrocellulose using: 1) anti-Tg and anti-HBS monoclonal antibodies, either unlabelled or directly coupled t o biotin or an enzyme, 2) human or mouse sera with or without toxoplasmic antibodies, and 3) anti-mouse immunoglobulins or antibodies directed against human IgG, IgM or IgA, coupled to peroxidase, alkaline phosphatase or biotin. Our tests in homologous or heterologous systems confirmed the high specificity of immunodetection by both single- and double-layer techniques (Fig. 2). A further advantage of ELIFA is the amplification of sensitivity. This can easily be demonstrated by indirect immunodetection, where the second labelled antibody can be diluted much stronger (1/30 000, i. e. 30 ng/mL) than when detection is carried out passively with similar markers [.3,4,7]. Such extreme dilution of the labelled antibody allows suppression of nonspecific background. The limit of detection of blotted proteins is at least as low as 1-5 ng and is thus at least comparable with resultsreportedpreviously [3,4,7].Itisknownthatthelimitof detection varies according to the protein blotted, the nature and concentration of the probe and the method of single or double layer detection used. It should ble noted that inversed filtration, which facilitates labelling of the membrane, increases the sensitivity of ELIFA.
3.5 Other applications 3.5.1 Detection of glycoproteins after transfer onto nitrocellulose Glycoproteins are important components of many biochemical systems. Concanavalin A (Con A) is one of the lectins capable of fixing carbohydrate residue55 [ 111. Using a complex antigen, as is the case for soluble Toxoplasma extracts, we could easily show that a constituent recognized by one of the monoclonal antibodies tested was a glycoprotein. For this
demonstration, the matrices were filtered for 10min using Con A labelled with peroxidase (0.1 % in TTBS) or with biotin. In the latter case, the final detection step was carried out using an avidin-alkaline phosphatase system.
3.5.2 Detection after immobilization of antibodies or antigens on activated membranes Based on the same principle, reverse ELIFA (antibodies immobilized on the membrane) can be applied to the search for bacterial, viral, parasitic or fungic antigens in different solutions or biological liquids (cerebrospinal or amniotic fluids, urine, sera, etc.). The technique of filtration with recycling offers two major advantages over classical techniques: the sample can be treated directly in an unconcentrated form, and the entire sample can be tested. In a model using an antitoxin, immobilized on a cellulose acetate membrane by cyanogen bromide [ 121, and alabelled second specific antibody, the limit of detection of a tetanic anatoxin was 100 pg/mL when recycled filtration was used. At last, with an alkaline-phosphatase labelled antibody, the limit of detection of immunoglobulin G immobilized on nitrocellulose membranes is 20 pg/mm2.
3.5.3 Filtration used to stain proteins transferred onto nitrocellulose We tested by filtration different methods to detect total proteins. Using the Bio-Rad Biotin-Blot Protein Detection Kit (Cat. No. 170.65 12), filtration allowed the detection of 10 ng (3 ng/mm2) in 40 min. This contrasts with the manufacturer’s recommendations, which give a detection time of 1 h 45 min, using passive agitation, and a threshold of 100 ng. With gold staining (Enhanced Colloidal Gold Total Protein Detection Kit, Bio-Rad Cat. Nr. 170.165 17), the technique can be carried out in 10 min by filtration, with a limit of detection of 100 pg without background. Using passive staining, the recommended time was 1-3 h, and the threshold was 400 pg (Fig. 2).
45
Electrophoresis 1990. I l , 41-45
Rapid detection of proteins by enzyme-linked immunofiltration assay
4 Concluding remarks
5 References
The application of ELIFA to the rapid immunodetection of proteins transferred onto immobilizing matrices seems innovative and useful by several criteria. (i) Including saturation, inactivation, detection and washing procedures, immunodetection by regulated filtration can be carried out in 15 min. This technique thus complements other fast methods of electrophoresis and electroblotting, resulting in a powerful combination with a total time of the entire procedure of less than 2 h. (ii) The detection of a blot by ELIFA, or ELIFA-blot, depends on a simple technique which can be automated. It requires no handling of the membranes. Repeated inversion of the filtration direction offers the additional advantages of reducing the volumes of expensive samples, while simultaneously increasing the sensitivity of the method. The limit of detection is at least as low as 1-5 ng, whether the probes are labelled directly or indirectly by an enzyme. (iii) The same principle of rapid detection can beused both for the identification and characterization by ligands of glycoproteins transferred onto nitrocellulose, and for the immunocapture of an antigen by an antibody immobilized onto an activated membrane.
[ I ] Renart, J., Reiser, J. and Stark, G. R., Proc. Natl. Acad. Sci. USA 1979, 76,3116-3120. 121 Towbin, H., Staehelin, T. and Gordon, J.,Proc.Natl. Acad. Sci. USA 1979, 76,4350-5354. L31 Beisiegei, U., Electrophoresis 1986, 7, 1-18. [4] Burnette, W. N., Anal. Biochem. 1981,112, 195-203. 151 Olsson, I., Axio-Fredriksson, U. B., Degerman, M. and Olsson, B., Electrophoresis 1988,9, 16-22. [61 Tovey, E. R. and Baldo, B. A., Electrophoresis 1987,8,384-387. [71 Van der Sluis, P. J., Pool, C . W. and Sluiter, A. A., Electrophoresis 1988,9,654-66 1. [Sl Pinon, J . M., Thoannes, H. and Gruson, N., J. Immunol. Methods 1985, 77, 15-23. [91 Guesdon, J. L., Ternynck, T. and Avrameas, S . , J . Histochem. Cytochern. 1979,27,113 1- 1 139. [lo1 Pinon, J . M., Gem, R., Lepan, H. and Pailler, S., Eur. J . Resp. Dis. 1987, 71,164-169. [ill Hawkes, R., Anal. Biochem. 1982,123, 143-146. 1121 Demeulemestre, C., Peltre, G., Laurent, M., Panheleux, D. and David, B., Electrophoresis 1987,8, 71-73.
Received June 1, 1989
