BACKGROUND OF THE INVENTION Techniques involving antibody recognition of a cellular product (antigen), most often a protein, for which the antibody is specific, and a subsequent visualization technique to detect the bound antibody/antigen complex have been used for many years to detect proteins of interest specifically and sensitively. Immunoblotting (or Western analysis) is one such method. This procedure is used to detect the level of a specific protein in a cell or tissue extract after SDS-PAGE and electrophoretic transfer of the proteins onto a nitrocellulose or PVDF filter (reviewed by Harper et al., 1990, J. Virolog. Meth., 30: 25-40; see also Egger and Bienz, 1994,Mol.Biotechnol., 1: 289-305).

Many applicable visualization protocols use secondary antibodies conjugated directly to enzymes or linked to amplification systems which use various substrates to produce a detectable product. Both colorimetric and chemiluminescent detection of horseradish peroxidase or alkaline phosphatase activity commonly are used to generate a signal.

Horseradish peroxidase (HRP) catalyzes the transfer of electrons from soluble 3,3'- diaminobenzidine (DAB), which is thereby oxidized. DAB is a particularly suitable substrate for use with HRP for both light- and electron immunohistochemical applications, since upon oxidation, DAB polymerizes to form an intensely-colored, brown precipitate which is insoluble in aqueous or organic solvents (Graham and Karnovsky, 1966, J. Histochem.

Cvtochem., 14:291-301).

Attempts to optimize immunohistochemical detection systems have shown that lowering the pH of the buffer to approach that at which HRP is maximally active (Trojanowski et al., 1983, J. Histochem. Cvtochem., 31:1217-1223) have resulted in improved visualization of the detected product. Imidazole has been suggested to induce formation of a third electron transfer site in horseradish peroxidase, thereby increasing its activity (references cited in Simionescu et al., 1975, J. Cell Biol., 64: 586-607) and, when it is added to the DAB reaction, has been demonstrated to result in improved visualization of the detected antigen (Tu et al., 1968, Experienfla, 15:219-221;

Simionescu et al., 1975, supra; Straus, 1980, J. Histochem. Cvtochem., 28: 645-652). Finally, it has been shown that the addition of metals, e.g. nickel, copper, silver or cobalt (Adams, 1977, Neuroscience, 2: 141-652; Hsu and Soban, 1982, J. Histochem. Cvtochem. 30: 1079-1082; Scopsi and Larsson, 1986, Histochemistrv, 84: 221-230) can provide for more sensitive detection of the target antigen than does the standard DAB protocol. The nature of the interaction between DAB and metallic ions is not definitively known, but may be similar to the deposition of osmium in DAB polymers (Hsu et al., 1982, supra; Seligman et al., 1968, J. Cell Biol., 38: 1-14).

Improved antibody detection in Western analysis by the use of metals (e.g. nickel, cobalt, copper and silver) in the HRP/DAB reaction has been reported, but sensitivity of signal development has still been low (Li et al., 1993, Thromb. Haemost. 69:331-334). There is a need in the art for increased sensitivity of cellular product detection in immunoblotting techniques.

SUMMARY OF THE INVENTION The invention provides an improved method for the detection of a target antigen in a sample, in which the amount of product formed after the reaction of a detection enzyme with 3,3'- diaminobenzidine (DAB) can be dramatically increased by the inclusion of imidazole and a metal, thereby enhancing sensitivity, wherein a complex is formed between the target antigen, an antibody specific for that target antigen and a secondary antibody and the complex is contacted with a signal-development reagent comprising DAB such that the bound antibody is detected, the improvement comprising the step of further developing the signal by contacting the complex with imidazole and a metal simultaneous with contacting the complex with DAB.

As used herein, "sample" is defined as an isolated protein, which may be synthetic, recombinant or naturally produced, or, alternatively, as a partially-purified or crude cell lysate or a fraction thereof, but not as a histological sample, such as a tissue slice, and not as any other sample comprising whole cells, such as isolated cells or cultured cells.

It is preferred that the signal development reagent comprises horseradish peroxidase and DAB.

It is also preferred that, prior to formation of the complex, the sample containing the target antigen is contacted with a substrate, such that the antigen is immobilized. The substrate

may be a filter paper; preferably, the substrate is a nitrocellulose- or PVDF membrane.

Preferably, the metal is selected from the group consisting of the salts of nickel (Ni), cobalt (Co), copper (Cu), silver (Ag), gold (Au), platinum (Pt), palladium (Pd), rhenium (Rh), iridium (Ir), osmium (Os), ruthenium (Ru), iron (Fe) and manganese (Mn); more preferably, the metal is selected from the group consisting of salts of Ni and Co.

Preferably, the signal is a color change.

The invention thus preferably encompasses an improved immunological protein detection method in which the combination of DAB, imidazole and metal are used in the color development step in order to obtain a significant increase in the colored signal product, the presence of which indicates the presence of the target antigen and the amount of which is proportional to the abundance of the target antigen in the protein sample.

It is preferred according to the invention that the metal is selected from the group consisting of salts of nickel (Ni), cobalt (Co), copper (Cu), silver (Ag), gold (Au), platinum (Pt), palladium (Pd), rhenium (Rh), iridium (Ir), osmium (Os), ruthenium (Ru), iron (Fe) and manganese (Mn) and more preferred that the metal is selected from the group consisting of salts ofNi and Co.

The invention provides for significant enhancement of a technique which results in a signal via the formation of an intensely colored reaction product. The amount of signal product obtained by the inventive procedure is significantly greater (i.e., 25%, 50%, 75%, or two-fold and preferably five- to ten-fold greater) than the amount of signal product obtained using standard immunological protein detection procedures that employ a detection enzyme and DAB. The inventive method is extremely sensitive and easily performed, uses inexpensive reagents, does not require radioactive reagents and does not necessitate use of darkroom procedures.

Further features and advantages of the invention will become more fully apparent in the following description of the embodiments and drawings thereof, and from the claims.

BRIEF DESCRIPTION OF DRAWINGS Figure 1 presents a series of immunoblots in which the color development step is performed using a variety of reaction mixture components.

DETAILED DESCRIPTION The invention is illustrated by the following nonlimiting examples wherein the following materials and methods are employed. The entire disclosure of each of the literature references cited hereinafter are incorporated by reference herein.

The inventive method, in which the combination of imidazole and metal substantially improves the signal development, and results in very sensitive detection of proteins of interest by production of an intense brown or blue/grey colored reaction product, with a high signal-to-noise ratio. The inventive method is applicable to a standard protocol in which DAB is used.

Immunoblotting Protein samples are obtained and their concentration is measured by standard procedures, such as a Bradford assay (Bradford, 1976, Anal. Biochem. 72: 248-254) or the BCA protein assay (Pierce Scientific, Rockford, IL). Samples are then electrophoresed on 10% SDS- polyacrylamide gels and electroblotted onto support filters, such as 0.2 Zm nitrocellulose membranes (Schleicher & Schuell, Keene, NH) in a 25 mM Tris, 192 mM glycine, 25% methanol transfer buffer. Filters are stained with 0.1% Ponceau S (0.1% Ponceau S in 5% acetic acid) for 5 minutes to visualize transferred proteins and then are cut to yield three sets of identically-treated filters. Filters are then soaked in 2.0% calf serum, 0.5% bovine serum albumin (BSA) in Tris- buffered saline (TBS) for 1 hour, both to equilibrate them in antibody binding buffer and to inhibit non-specific binding.

After a brief rinse in 0.1% BSA/TTBS (TBS plus 0.1% TWEEN-20; Atlas Chemical Industries, Inc.), filters are incubated overnight with a primary antibody in 0.1% BSA/TTBS. A typical antibody dilution is 1:2000; however, the specific activity of antibody preparations varies from one to the next, and it is necessary to assay each antibody individually using serial dilutions on a test filter to determine the appropriate dilution factor, which may range from 1:100 for a weak antiserum to 1:10,000 for a preparation with a high antibody titer. Typically, crude antisera require less extensive dilution than do ammonium sulfate precipitates or affinity-purified preparations. Filters are then washed 3 times for 5 min. in 0.5% BSA/TTBS and then incubated for 2 hours with a biotinylated secondary antibody (0.5 llg/ml in 0.1% BSA, 1% goat serum/TTBS), after which they are washed once in 0.5% BSA/TTBS, and twice in TTBS.

Horseradish peroxidase is applied as an avidin/biotin conjugate, such as is supplied in the VECTASTAIN ABC detection system (Vector Laboratories, Burlingame, CA). Filters are washed 3 times in TTBS and once in TBS.

Standard DAB visualization As previously described (Adams, 1977, supra), DAB is dissolved at a concentration of 0.5 mg/ml in 10 ml of 0.1 M Tris (pH 7.5) just prior to visualization. Immediately before use, 67 ,u1 of fresh 3% H202 is added. The DAB solution is added to the washed filters and the reaction is allowed to proceed for 3-5 min. After signal development, the filters are washed several times in H2O and allowed to dry.

Imidazole and metal enhancement protocol DAB is dissolved at a concentration of 0.5 mg/ml in 10 mls of 0.1 M imidazole (pH 7.0).

For metal intensification, NiCI2 and CoCl2 are made as stock solutions of 0.2 M in H2O. 1 ml of either NiCl2 or CoCI2 is added to 10 mls of imidazolelDAB solutions for a final metal concentration of 10 mM NiCI2 or 2.5 mM CoCI2. Immediately before use, 67 ,u1 of fresh 3% H202 is added, the solution is added to filters and the reaction is carried out for 10-30 sec. After signal development, the filters are washed in H20 and allowed to dry.

Chemiluminescent visualization protocol Filters are incubated in primary antibody as described above, rinsed twice with washing buffer (10 mM tris HCI, pH 7.5, 100 mM NaCI, 0.1% Tween-20) and washed twice for 5 minutes each, before incubation with a horseradish-peroxidase-conjugated secondary antibody, e.g. an HRP-conjugate goat anti-mouse IgG supplied by Transduction Laboratories (Lexington, KY) in blocking buffer (10 mM Tris, pH 7.5, 150 mM NaCI, 0.5% non-fat dry milk, 1% heat-treated normal goat serum) at room temperature for 1 hour. The filter then undergo twelve 5-minute washes in washing buffer to minimize signal background, before application of an HRP substrate that is processed to form a luminescent signal product. One such system is the ECL (Enhanced Chemiluminescence) kit of Amersham. When this system is employed, each filter is incubated with the mixed ECL reagents for one minute in accordance with the manufacturer's recommendation. The protocol is summarized briefly as follows: Equal volumes of ECL detection solutions 1 and 2 are mixed to yield sufficient total

volume to cover the filter. Excess wash buffer is drained from the filter, which is then placed on a sheet of plastic film, such a food wrap (e.g. SARAN WRAP; Dow Chemical Co., Indianapolis, IN), protein side up. The detection reagent is then placed on the filter and left there for precisely 1 minute, after which it is drained, and the membrane is covered by- or wrapped in a clear, plastic wrap (again e.g. SARAN WRAP). Air pockets are smoothed out, and the wrapped filter is placed in a film cassette, protein side up. The signal is visualized by exposing the filters to x-ray film, e.g. X-OMAT AR film (Kodak), for 10 minutes under an intensifying screen. Film is then developed and the signals quantified as described above.

Alternative antibody incubation and film exposure conditions are supplied by the manufacturer. Generation of a protein sample, electophoresis and transfer to a support membrane are all as described above. Incubation with the primary antibody takes place for one hour at room temperature in either TBS-T (0.1% TWEEN-20 in 20mM Tris, pH 7.6, 137mM NaCI) or PBS-T (0.1% TWEEN-20 in 80mM Na,HPO4, 20mM NaH2PO4, 100mM NaCI, final pH 7.5), and the blocking step that precedes it is performed using 5% ECL blocking reagent in TBS-T or PBS-T for one hour at room temperature with gentle agitation. Following exposure of the filter to the primary antibody, excess antibody is washed away using either TBS- T or PBS-T, horseradish peroxidase (HRP) is applied via one of two routes. Either the filter is incubated with a secondary antibody conjugated to HRP under the conditions used for incubation with the primary antibody, or a biotinylated secondary antibody is incubated with the filter for 45 to 60 minutes at room temperature, in which case the filter is later incubated with an avidin/HRP conjugate for 45 to 60 minutes, also at room temperature. Regardless of the method used, excess antibody and/or HRP are washed away using either TBS-T or PBS-T (once for 15 minutes, and twice for 5 minutes each) and the ECL reagents are applied to the filter as above. After wrapping, the filter is exposed to HYPERFI:LM-ECL (Amersham) for 15 seconds.

Signal quantitation.

After antibody binding and signal development, the amount of colorimetric reaction product can be quantified by scanning the filter bearing the signal precipitate using either a densitometer or any of a number of widely-available desktop optical scanners, e.g. the Argus II flatbed scanner (AGFA); in the latter case, scanned images are quantified using an image

processing program, e.g. the public-domain NIH Image software (version 1.5, Wayne Rosband, National Institutes of Health, Bethesda, MD), on a personal computer or workstation.

Chemiluminescent products are detected by autoradiography followed by densitometric quantitation or optical scanning, as above, or by phorphorimaging, in which the amount of signal in each band is computer quantified as the fluorescence data on the phosphorimaging screen is recorded.

EXAMPLE 1 To compare visualization procedures, standard immunoblotting procedures were used.

Protein samples were isolated from cultured rat vascular smooth muscle cells and their concentration was measured by the BCA protein assay (Pierce, Rockford, IL). Aliquots containing 6, 3 or 1.2 llg of total protein were run on SDS-PAGE and transferred by electroblotting onto 0.2 ym nitrocellulose membranes (Schleicher & Schuell, Inc., Keene, NH).

Filters were incubated with mouse monoclonal anti-ERK1 primary antibody which recognizes MAP kinase proteins of 42 and 44 kD (Santa Cruz Biotechnology, Santa Cruz, CA). For colorimetric reactions, filters were incubated in a biotinylated anti-mouse secondary antibody followed by incubation with an avidin/biotin/horseradish-peroxidase complex (ABC) signal amplification system. Figure 1 presents the results of signal development under four different reaction conditions, DAB in 0.1 M Tris pH 7.5 (Fig. 1B), DAB in 0.1 M imidazole (Fig. 1C), and DAB/imidazole with either nickel (Fig. 1D) or cobalt (Fig. 1E), demonstrating that horseradish peroxidase oxidation of DAB is dramatically enhanced by incubation with imidazole and metal.

Ponceau S staining of a duplicate filter to indicate the level of total protein present is shown in Fig. 1A. After the different DAB solutions were added to the filters, reaction product formation was monitored visually and the reactions stopped when signal to noise ratio was maximum. The standard DAB procedure gave a brown reaction product, and detected a signal at only the highest protein concentration (Fig. 1B). A slight background staining ofthe nitrocellulose was observed.

Imidazole dramatically increased the intensity of the DAB signal, lowered the sensitivity to the mid-range protein level and significantly lowered the reaction time from 3-5 min. to 10-20 sec., producing virtually no background staining (Fig. 1C). Inclusion of NiCl2 (resulting in a grey reaction product) or CoCI2 (resulting in a blue reaction product) improved the sensitivity to the

lowest protein level. DAB/imidazole combined with CoCl2 gave the greatest sensitivity (i.e., to detect lowest amount of protein), but there was some background staining of the nitrocellulose (Fig. 1E). DAB/imidazole used in conjunction with Nick, gave a high sensitivity and low background (Fig. 1D).

An identical filter was treated with the same primary antibody and a secondary antibody conjugated to horseradish peroxidase. Chemiluminescent visualization was carried out using the ECL kit (Amersham, catalogue no. RPN 2106), according to manufacture recommendation (see above), and the signal visualized by exposure of X-OMAT AR film (Kodak) to the filter and subsequent film development. The sensitivity of the signal after visualization with chemiluminescent procedure was similar to that seen after DAB/imidazole with either NICK, or CoCI2 (Figure 1F).

EXAMPLE 2 Testing of metals for use with imidazole in enhancing HRP activity on a DAB substrate.

a. Generation of a test protein sample.

In order to perform a comparative analysis of reaction components for HRP-mediated color development in immunoblotting using a DAB substrate, it is first necessary to generate a series of equivalent target protein samples against which to test various candidate substances for their ability to enhance signal production. In order to make a meaningful comparison, it is necessary to ensure equivalent protein content across samples being compared. Test samples comprising known quantities of a protein affixed to a solid substrate by the investigator provide the most reliable basis for comparison of candidate reaction components to each other and to a standard reaction mixture (wherein "standard" refers to a mixture in which an HRP/DAB chromogenic reaction will occur in the absence of metals, whether with or without imidazole, such as was described in Example 1, above).

Test protein samples are generated as follows: A multi-well SDS-PAGE gel is prepared.

A protein sample to be electophoresed on the gel may comprise a known quantity of a purified protein against which a primary antibody is to be directed during the analysis. Alternatively, a crude or partially-purified cellular lysate known to contain the target protein such that it is uniformly distributed and at a concentration sufficient for detection under standard reaction

conditions, as defined above, given the physical limitations on the volume of sample that can be loaded into a well of the gel. Note that since there is no requirement for use in testing immunological detection components other than that a protein be sufficiently antigenic to provoke an immune response, any target protein against which a plentiful and inexpensive source of antibody is commercially available is recommended for use.

Typically, three groups of three lanes each are loaded with a set of protein samples consisting of a dilution series of a cellular protein extract, in the amounts of 6-, 3- and 1.2 g of total protein in a convenient volume. The selected protein is combined with an appropriate loading buffer/dye mixture, loaded onto the gel, electophoresed and transferred to nitrocellulose, PVDF or other suitable membranous matrix by electroblotting, all as described above in Example 1. After transfer, the filter upon which the protein has been immobilized is divided into multiple, equivalent parts and the filter is cut longitudinally (vertically) into strips. In that the protein may be underrepresented at the two side edges of the filter, these are discarded. A proper guide for the distance from either side to the start of the protein sample is the line of xylene cyanol (present in the loading/dye mixture) which still should be visible on the filter; alternatively, the filter may be temporarily stained with Ponceau S, as described above, to reveal the location of the protein on the gel, to visualize molecular weight markers and to allow one to ensure that the horizontal distribution of the protein is approximately even across the filter. Protein test strips are cut to a width of at least 4 mm; preferably, such strips are 6, 8 or even 10 mm wide, since artifactual color development occurs at- and within about 1 mm of the cut edges of the membrane, particularly near or within a detected protein band. In that the depth of color development must be compared among test strips assayed under different reaction conditions, it is necessary to have sufficient area within a band for comparison that is clear of such edge effects, for example a 2- or, preferably, 4-, 6- or even 8-mm width, after 1 mm at either edge is subtracted.

Alternatively, known quantities of protein may be spotted onto a filter and tested by the same procedures as those employed in the assay using electrophoresed and electroblotted samples.

b. Antibodv binding and color development.

Antibody binding and washes are performed as described in Example 1. All protein test

strips to be used in a given experiment are sequentially co-incubated with the primary and secondary antibodies to ensure that such variations in color development as are later observed result from the choice of metal included in the reaction buffer, rather than from unequal binding efficiencies and/or wash conditions. Following the removal of excess secondary antibody by washing, test strips are individually processed in their several color-development buffers. Each set of assays, regardless of the number of metals to be tested, requires a strip to be processed using DAB in standard reaction buffer, as defined above, and one to be processed using DAB in standard reaction buffer plus 0.1 M imidazole or, alternatively, a 0.1 M imidazole buffer may be used. For each candidate metal, two more strips are required: One strip is processed using DAB in standard buffer plus the candidate metal, while the other is processed under 'complete' conditions (DAB in standard buffer plus both 0.1 M imidazole and the candidate metal or, alternatively, DAB in a 0.1 M imidazole buffer plus the candidate metal). Appropriate metal concentrations range from 0.1 to 100 mM; preferably, candidate metals are assayed in the concentration range of 1 to 20 mM; for any given metal, several initial concentrations are examined, for example, 1, 10 and 50 mM.

After antibody binding and color development, the amount of reaction product is quantified as described above. A significant increase in signal product release in reactions carried out in the presence of both imidazole and the candidate metal relative to those performed using either or neither of these components is indicative that the metal and imidazole combination is useful according to the methods of the present invention.

EXAMPLE 3 The invention also is applicable to standard ELISA assays (Fundamental ImmunoioN. 3rd Ed., W.E. Paul, ed., Raven Press Ltd., New York, 1993, chapter 12, p. 438) with the modification that the signal is developed using an imidazole/metal combination according to the invention in the amounts described herein. The reaction is read as absorbance units.

OTHER EMBODIMENTS Other embodiments will be evident to those of skill in the art. It should be understood that the foregoing detailed description is provided for clarity only and is merely exemplary. The spirit and scope of the present invention are not limited to the above examples, but are encompassed by the following claims.