Visualization of antigenic proteins on Western blots

ANALYTICAL

BIOCHEMISTRY

136, 180- 184 ( 1984)

Visualization

of Antigenic

DAVID A. KNECHT' Department

of Bacteriology,

Proteins on Western

ANDRANDALL

University

of Wisconsin,

Blots

L. DIMOND’ Madison,

Wisconsin

53706

Received May 2, 1983 A new technique for the detection of antibodies bound to proteins blotted onto n&cellulose paper was developed. The method is rapid, sensitive, and does not require radioactive probes. Proteins transferred to nitrocellulose paper are first reacted with primary antibody followed by reaction with an alkaline phosphatase conjugated second antibody. The phosphatase activity is then visualized using an agar gel impregnated with the histochemical phosphatase stain 5-bromo4-chloro-3-indolyl phosphate (BCIP) (J. P. Horwitz, J. Chua, M. Noel, J. T. Donatti, and J. Freisler (1966) J. Med. Gem. 9, 447; Sigma Chemical Co., Technical bulletin No. 710-EP (1978)). Antigen-antibody complexes give rise to sharp, permanent blue stained bands both on the nitrocellulose paper and in the agar overlay gel. This procedure allows detection of bands containing less than 20 ng of protein. KEY WORDS: Immunostaining; Western blots; antibody staining; immunological methods.

The development of the “Western blot” by Towbin et al. (1) was an important step in increasing our ability to analyze antigenic proteins fractionated on gels. Antibody-antigen reactions that take place on nitrocellulose paper are usually visualized using iodinated protein A (IPA) and autoradiography. We have sought to develop an alternative method for visualizing bound antibodies to eliminate the expense and hazards of working with iodinated compounds. Our experiments have used alkaline phosphatase-conjugated second antibodies rather than radioactive protein A and we have screened a number of common histochemical staining reagents for visualization of the bound ’ Present address: Department of Biology B-022, University of California at San Diego, La Jolla, California 92093.

* To whom correspondence should be addressed. 3 Abbreviations used: IPA, iodinated protein A, BCIP, 5-bromo-4-chloro-3-indolyl phosphate; ELISA, enzymelinked immunosorbent assay;SDS, sodium dodecyl sulfate; BSA, bovine serum albumin; TN, 10 mM Tris-HCl, 150 mM NaCl, pH 7.4; IgG, immunoglobulin G, DMSO, dimethyl sulfoxide.

0003-2697184

$3.00

Copyright 0 1984 by Academic Press, Inc. All r-i&s of reproduction in any form wewed.

enzymatic activity. The resulting method has the advantage of not depending on the ability of protein A to bind to the primary antibody, a matter of some consequence when working with monoclonal antibodies. While several methods gave adequate results, the BCIP method described was the simplest, most reproducible method giving good sensitivity. The staining procedure should be adaptable to any ELISA-type assay for detection of antigen-antibody complexes. METHODS

AND

MATERIALS

Antibodies and antigens. Lysosomal enzymes N-acetylglucosaminidase, /3-glucosidase, and cY-mannosidase were purified from Dictyostelium discoideum amoebae as described previously (2,3). The monoclonal antibody used in these studies cross reacts with all D. discoideum lysosomal enzymes due to the presence of a common modification antigen on these proteins. Preparation of this antibody and its properties will be described elsewhere (D. Knecht and R. Dimond, manuscript in preparation). It shows specificity

180

IMMUNOSTAINING

comparable to the rabbit antiserum described previously (4). Gel electrophoresis and blotting. Gels used in these experiments were SDS-polyacrylamide slabs of either 7.5 or 10% acrylamide (5). Gels were blotted and antibody stained according to the procedure of Bumette (6) with minor modifications. Briefly, the proteins were electrophoretically transferred overnight in the cold at 10 V/cm onto nitrocellulose paper (BA85, Schleicher and Schuell, Keene, N. H.). The paper was then incubated for 1 h at 35°C in blocking solution ( 10 mM TrisHCl, 150 mM NaCl, 0.3% BSA, 0.3% calf serum, pH 7.4). The primary antibody was bound to the paper by incubation for l-3 h using a l/ 1000 dilution of monoclonal ascites fluid in blocking buffer. The paper was then washed for 10 min in TN buffer ( 10 mM TrisHCl, 150 InM NaCl, pH 7.4), twice for 20 min in TN + 0.05% NP-40, and then for 10 min with TN. All reactions were carried out in plastic dishes just larger than the size of the nitrocellulose paper but can also be performed in sealed plastic bags to further reduce the volume of antibody required. Antibody visualization. Following primary antibody reaction and washing, the nitrocellulose sheet was reacted with alkaline phosphatase conjugated anti-mouse IgG antibody. We have used both rabbit anti-mouse whole IgG (Sigma Chemical Co., St. Louis, MO.) and affinity purified goat anti-mouse IgG (gamma chain specific-Kierkegaard and Perry Labs, Gaithersburg, Md.) with comparable results. The antibody was routinely used at a l/1000 dilution in blocking buffer (about 0.3 units/ ml final concentration). The second antibody can be used repeatedly but never for different primary antibodies as some contamination of the second antibody by first antibody has been observed. Following incubation for l-3 h the paper was washed as described above. During the final wash, the agarose gel for visualizing alkaline phosphatase was prepared. A solution containing 0.1 M Tris-HCl, pH 8.3, 1 mM MgCl*, and 1% agarose is heated to dissolve

OF

WESTERN

BLOTS

181

the agarose. The concentration and pH of the alkaline phosphatase buffer is important in order to obtain a duplicate staining pattern in the agarose gel. If no duplicate is desired, 1 M Tris-HCl, pH 8.8, stains only the nitrocellulose paper. A higher concentration and higher pH alkaline phosphatase buffer, 1 M diethanolamine-HCl, pH 9.8, resulted in high background staining with no apparent increase in sensitivity. The agarose solution was cooled to 65°C and BCIP (Sigma Chemical Co.) dissolved in a small volume of DMSO was added to a final concentration of 0.5 mg/ml. The agarose was pipetted onto a glass plate and allowed to solidify. For a 14 X 12-cm glass plate about 25 ml of gel solution was used. The washed paper was blotted completely dry and then placed with the protein side down onto the agarose gel. The gel was wrapped in Saran wrap to prevent drying and incubated at 30°C until blue bands became visible (15 min to overnight). When color had developed sufficiently, the nitrocellulose was separated from the gel and both were rinsed in water to remove uncleaved substrate and buffer. After it was air-dried, the staining of the paper seemed to fade. However, rewetting enhanced the staining to its original level. The agarose gel was saved by air drying it clamped between two sheets of wettable cellophane membrane (Bio-Rad Laboratories, Richmond, Calif.). This produced a transparent duplicate gel which was convenient for overlaying other blots and total protein gels in order to identify antigenic proteins. Alternatively, the agar gel can be dried onto filter paper. Alternate staining method. We have also visualized alkaline phosphatase stained gels using a fluorescent substrate. 4-Methylumbelliferyl phosphate (Sigma Chemical Co., St. Louis, MO.) forms a bluish white fluorescent spot upon cleavage by alkaline phosphatase as seen by illumination with 360-nm light. The substrate was dissolved in water and added to a solution containing 1% agarose, 1 M diethanolamine-HCl, pH 9.8, 1 mrvi MgC12 to a final concentration of 0.1 mg/ml. The

182

KNECHT

AND

nitrocellulose was prepared as above, incubated and then viewed using a hand-held uv light. The gels were photographed using a hand-held light source or a 360-nm light box (Fotodyne Inc., New Berlin, Wise.) with a Polaroid camera and a Tiffen UV-01 (Hauppauge, N. Y.) filter. The sensitivity of this procedure was somewhat lower than the BCIP method and required immediate uv photography of gels in order to preserve results. We have also tested a number of simultaneous coupling reagents for use in this procedure. None of the reagents tested, including cy-Napthyl phosphate or Napthol AS-MX phosphate used in conjunction with Fast Blue BB salt, Fast Blue RR salt, Fast Red TR, Fast Garnet GBC, or Variamine Blue B base, were acceptable alternatives to BCIP.

DIMOND

RESULTS

2ng

B. Agar 200

ng

DISCUSSION

To test the sensitivity of the BCIP staining procedure, various known amounts of a mixture of three purified lysosomal enzymes from D. discoideum were electrophoresed in an SDS ( 10%)-polyacrylamide gel and then transferred to nitrocellulose paper as described under Methods and Materials. The immobilized protein was reacted with a monoclonal antibody that recognizes an antigenic determinant shared by the three enzymes tested ((4) and D. Knecht and R. Dimond, manuscript in preparation). The antigen-antibody complexes were reacted with an alkaline phosphatase-conjugated second antibody, and the antigenic proteins were visualized by placing the nitrocellulose paper on an agarose gel con-

A. Nitrocellulose 20ng

AND

2nc.l

imprint 20 ng

200 ng

- Nag - Nag

- a-Man - a-Man

FIG. 1. Antigenic proteins visualized with BCIP. Each lane of the 10% SDS-polyacrylamide gel contained the indicated amounts of each of 3 purified enzyme preparations, fi-glucosidase (103,000 M,), N-acetylglucosaminidase (68,000 MI), and cY-mannosidase (58,000 and 60,000 LICK). The gels were electrophoresed, blotted, antibody stained, and visualized as described under Methods and Materials. The nitrocellulose paper was wetted prior to photography to enhance the staining. The agar gel was dried onto cellophane and photographed separately.

IMMUNOSTAINING

taining the phosphatase substrate, BCIP. The stained nitrocellulose paper is shown in Fig. 1A. In the lane with 200 ng of each enzyme, the blue bands were visible within 15 min while the 20-ng lane was visible after overnight incubation. No staining has been seen with 2 ng of protein using any of the methods tested. The detection limit varies with the antigen and antibody used and is less than 5 ng for other proteins and antibodies (D. Titus and P. Quail, personal communication). Figure 1B shows the imprint of the staining left in the agar gel after overnight incubation. We have found this transparent duplicate of the nitrocellulose staining to be useful for comparison of the antigenic proteins visualized in different samples or by different antibodies. The transparent gel can be scanned by a densitometer for quantitation of the staining. It is also convenient for overlaying against silver-stained gels of total protein samples to identify the antigenic proteins. If only the stained nitrocellulose is desired the reactions can be carried out in solution with satisfactory results. The major bands visualized by the BCIP stain correspond to the major subunits of the purified enzymes as determined by Coomassie staining of parallel gels. The minor bands correspond to antigenic contaminants or partially degraded enzyme subunits in the pure preparations. Some of these bands are not visible on silver-stained gels with 200 ng of each enzyme (data not shown). Thus this procedure is sensitive enough to allow visualization of polypeptide bands that are not detectable by silver staining. We have also analyzed antigenie proteins found in crude cell lysates that have been electrophoresed on either one or two-dimensional gels. Most of our antigenic proteins are minor cell proteins (CO. 1% of total cell protein) and cannot be detected by silver staining two-dimensional gels of crude cell lysates. These proteins can, however, be detected by BCIP visualization of Western blots (D. A. Knecht and R. L. Dimond, manuscript in preparation, and R. Mierendorf, personal communication).

OF

WESTERN

BLOTS

183

If the monoclonal antibody used above is replaced with a nonimmune antibody, no stained bands are observed (data not shown). We have occasionally seen extensive nonspecific background staining of prevalent protein bands when large amounts of crude protein extracts are examined. This artifact can be attributed to nonspecific binding of the alkaline phosphatase-conjugated second antibody and can be eliminated by including 0.05% SDS in the last set of washes (D. Titus, personal communication). The BCIP procedure is a simple and reproducible method for detecting nanogram quantities of antigenic proteins immobilized on nitrocellulose paper. It is considerably less expensive to use than iodinated probes and does not have any hazards or disposal difficulties. In addition, one is not limited to using primary antibodies that are bound by protein A. Alkaline phosphatase-conjugated second antibodies reactive with most subclasses of antibodies produced by a variety of mammals are now available. An alternative procedure using horseradish peroxidase-conjugated second antibodies instead of alkaline phosphatase has been described by Towbin et al. (1). The disadvantage of this procedure is that it does not produce a duplicate stained gel, and it requires special handling and disposal procedures for the carcinogen, o-dianiside. For several proteins that have been tested the peroxodase method was two- to fourfold more sensitive than the BCIP procedure while the iodinated protein A procedure was lo- to I OOfold more sensitive (Duncan R. Talbot, personal communication). O’Connor and Ashman have previously described a method utilizing Napthol AS-MX phosphate and Fast Red TR salt (7). We have found this procedure to give variable backgrounds and to be approximately lo-fold less sensitive than the BCIP procedure. Moreover, it cannot be used with the gel overlay. The other method described in this paper for detecting alkaline phosphatase involves a fluorescent substrate, 4-methylumbelliferyl phosphate. This proce-

184

KNECHT

AND DIMOND

dure was not as sensitive as BCIP, and required immediate photography in order to preserve a permanent record of results but was otherwise an acceptable alternative. Thus in terms of simplicity, sensitivity, economy, and safety the BCIP procedure has advantages over other methods of visualizing antigenic proteins immobilized on nitrocellulose paper.

University of Wisconsin, Madison, and by the National Institutes of Health under Grants GM 29156 and GM 31181.

REFERENCES 1. Towbin, H., Staehlin, T., and Gordon, J. (1979) Proc. Nat. Acad. Sci. USA 76, 4350. 2. Every, D., and Ashworth, J. M. (1983) Biochem. .I. 133,47.

ACKNOWLEDGMENTS We thank David Titus and Robert Mierendorf for sharing their unpublished observations with us. We also thank Larry Kahan for helping in setting up the blotting apparatus and for useful suggestions regarding the use of gel overlays for alkaline phosphatase staining. This research was supported by the College of Agricultural and Life Sciences,

3. 4.

5. 6. 7.

Dimond, R. L., and Loomis, W. F. (1976) J. Biol. Chem. 251,268O. Knecht, D. A., and Dimond, R. L. (1980) J. Biol. Chem. 254, 3564. Laemmli, U. K. (1970) Nature (London) 227, 680. Bumette, W. N. (1981) Anal. Biochem. 112, 195. O’Connor, C. G., and Ashman, L. K. (1982) J. Immunol. Methods 54, 267.