Immunochemical detection of proteins biotinylated on nitrocellulose replicas

Journal of Immunological Methods, 92 (1986) 65-71

65

Elsevier JIM 04016

Immunochemical detection of proteins biotinylated on nitrocellulose replicas Williarn J. LaRochelle and Stanley C. Froehner Department of Biochemistry, Dartmouth Medical School, Hanover, NH, U.S.A.

(Received18 March1986, accepted4 April 1986)

A sensitive method for staining proteins after transfer from polyacrylamide gels to nitrocellulose paper is described. Transferred proteins are first derivatized by reaction of the nitrocellulose replica with sulfosuccinimidobiotin and are then reacted sequentially with streptavidin, rabbit anti-streptavidin, and horseradish peroxidase-conjugated goat anti-rabbit IgG antibody. Application of the enzyme substrate, alpha-chloronaphthol, produces dark protein bands against a white background. The binding of streptavidin to the proteins is dependent on biotin derivatization and is inhibited by biotinylated bovine serum albumin or 10 nM biotin. The procedure permits detection of less than 5 ng of transferred protein in a single band and is thus 5-10 times more sensitive than horseradish peroxidase-conjugated avidin alone. For bovine serum albumin, the method is comparable in sensitivity to silver staining of protein in polyacrylamide gels. Key words: Immunochemicalmethod;Avidin-biotinsystem;Anti-streptavidinantibody;Proteinblotting; Nitrocellulosereplica

Introduction

Since the initial publication by Towbin et al. (1979) on the preparation of replicas of polyacrylamide gel patterns, commonly called protein blotting, the technique of transferring proteins from inaccessible gel matrices to solid supports such as nitrocellulose and nylon membranes has gained in popularity. Recently, several inclusive reviews describing the uses of this technique have been published (Gershoni and Palade, 1983; Bets and Garfin, 1985). Detection of proteins on blots ranges from the Abbreviations: sulfo-NHS-biotin, sulfosuccinimidobiotin; GAR-AP, alkaline phosphatase-conjugated goat anti-rabbit IgG; GAR-HRP, horseradish peroxidase-conjugated goat anti-rabbit IgG; PBS,10 mM sodiumphosphate,0.15 M NaC1, pH 7.4; SDS,sodiumdodecylsulfate;HRP-avidin,horseradish peroxidase-conjugatexlavidin; BSA,bovineserumalbumin.

specific visualization of the protein(s) of interest to the general staining of total protein. Specific proteins are typically detected with probes such as antibodies or toxins that are radioactive or in conjunction with enzyme-conjugated second antibodies. General protein detection is usually based on physical staining methods with amido black (Bio-Rad Laboratories, 1983) or India ink (Hancock and Tsang, 1983), or on immunological procedures that require derivatization of proteins with a reactive hapten and detection with anti-hapten antibody and enzyme-conjugated second antibody (Wojtkowiak et al., 1983; Kittler et al., 1984). The high affinity interactions between avidin and biotin (Gyorgy et al., 1941) or streptavidin and biotin (Chaiet and Wolf, 1964) are well characterized and have been used extensively in signal generating systems for immunofluorescence and enzyme-linked immunoassays. Here we exploit this interaction in a technique for staining proteins on

0022-1759/86/$03.50 © 1986 ElsevierSciencePublishers B.V.(BiomedicalDivision)

66 nitrocellulose replicas. Using horseradish peroxidase-conjugated avidin (HRP-avidin) as the signal generator, we first established the concentration and time-dependent parameters of biotinylation of proteins on nitrocellulose blots with sulfosuccinimidobiotin. Then, we developed a highly sensitive technique based on sequential incubations with streptavidin, rabbit anti-streptavidin, and horseradish peroxidase-conjugated goat anti-rabbit antibody. This amplified method is capable of detecting less than 5 ng of protein in a single band and is comparable in sensitivity to silver staining of proteins in polyacrylamide gels.

Materials and methods

General Sulfosuccinimidobiotin (sulfo-NHS-biotin) was obtained from Pierce Chemical Co., Rockford, IL. Horseradish peroxidase-conjugated avidin (avidinHRP), horseradish peroxidase-conjugated goat anti-rabbit antibody (GAR-HRP), and alkaline phosphatase-conjugated goat anti-rabbit antibody (GAR-AP) were purchased from Cappel Laboratories, Malvern, PA. Streptavidin was obtained from Bethesda Research Laboratories, Gaithersburg, MD. Nitrocellulose paper (0.45 /~M) and alpha-chloronaphthol were purchased from BioRad, Richmond, CA. Bovine serum albumin (BSA) (96-99%) was obtained from Sigma Chemical Co., St. Louis, MO. SDS microgels were silver stained by methods described by Wray et al. (1981). Protein concentrations of Torpedo acetylcholine receptor-rich membranes were measured by a modification of the method of Lowry et al. (1951). BSA concentrations were determined assuming A28o = 0.60 for a 1 mg/ml solution. Staining of proteins on replicas with aoidin-HRP All procedures were performed at room temperature. In some cases, Torpedo acetylcholine receptor-rich membranes (150 /~g) (Porter and Froehner, 1983) were separated by SDS gel electrophoresis on 9% polyacrylamide microslab gels lacking wells (Laemmli, 1970; Matsudaira and Burgess, 1978). Otherwise, the indicated amounts of protein were applied to a standard

microgel with wells. Electrophoretic transfer of proteins to nitrocellulose paper was performed as previously described (Froehner et al., 1983) with the following modifications. After electrophoresis, the gel was equilibrated for 30 min in 25 mM sodium phosphate, 1 mM dithiothreitol, pH 6.5. Proteins were then electrophoretically transferred to nitrocellulose membrane by blotting for 90 min at 250 mA in 25 mM sodium phosphate, pH 6.5, in a Bio-Rad transblot apparatus. The nitrocellulose paper was either cut into approximately seventeen 6 cm x 0.5 cm strips (if prepared from a gel lacking wells) or used intact. Replicas were rinsed with 0.1 M sodium bicarbonate, pH 8.0, and then soaked in the same buffer for 5 min. Derivatization of the proteins on the replica was accomplished by a modification of the general protein biotinylation method of Bayer et al. (1976). The strips were placed in 3 ml of 0.1 M sodium bicarbonate, pH 8.0 in 12 x 75 mm plastic capped test tubes and the desired volume of sulfo-NHSbiotin solution (4.43 mg/ml) was added. After 1 h the reaction was quenched by the addition of 30 /~1 of 1 M glycine, pH 6.5. Intact replicas were biotinylated in the same manner in volumes corresponding to 0.25-0.50 ml/cm 2 of nitrocellulose paper. The replicas were then washed twice with 10 mM sodium phosphate, 0.15 M NaC1, pH 7.4 (PBS) and incubated with PBS containing 5% newborn calf serum, 3% BSA for 30 min. AvidinHRP (5 /~g/ml) was incubated with the replicas for 1 h in PBS containing 1% BSA and 0.05% Tween 80. They were then washed three times for 15 min each with PBS containing 0.05% Tween 80. Protein bands were visualized by immersing the strips in 3 ml alpha-chloronaphthol (0.6 mg/ml in PBS containing 0.01% hydrogen peroxide). Color was allowed to develope for 30 min. The replicas were rinsed with distilled water and dried between two sheets of dialysis membrane. Preparation and purification of antibodies to streptavidin Prior to preparation for injection, a solution of streptavidin (0.4 mg/ml) was incubated for 2 h with a 2-fold molar excess of sulfo-NHS-biotin that had boen reacted with glycine. An emulsion was prepared with equal volumes of streptavidinbiotin complex and complete Freund's adjuvant.

67 Each rabbit was injected subcutaneously and intradermally at several spots on the back with 1 ml of emulsion containing approximately 100 gg streptavidin. Three weeks and 5 weeks after the initial injection, the rabbits were injected with the same amount of antigen emulsified with incomplete Freund's adjuvant. One week after the third injection, the rabbits were injected intravenously with 50 gg of the streptavidin-biotin complex. Antibody activity to streptavidin was monitored with a solid-phase assay. Except for the washing steps, all volumes were 50 gl. Wells of microtiter plates were coated with 300 ng of biotinylated BSA and blocked with PBS containing 0.02% azide and 4% BSA. Streptavidin (350 ng) was then added to each well, incubated 2 h, and washed with PBS, 0.02% azide, 0.05% Tween 80. Antiserum (serial dilutions beginning at 1/200 made in PBS, 0.02% azide, 0.05% Tween 80, 1% BSA) was added to the wells, incubated 4 h, and washed. Bound antibodies were detected by incubation for 2 h with GAR-AP (1/100). After the wells were washed, 65 gl of p-nitrophenyl phosphate (1 m g / m l in 50 mM bicarbonate buffer, p H 9.8, 1 mM magnesium chloride) was added, and after approximately 30 rain, the reaction was terminated with 100 gl of 1 M sodium hydroxide. The absorbance was read at 405 nm with a Dynatech Minireader II. Prior to use in staining replicas, anti-streptavidin antibodies were purified from the antiserum. A total IgG fraction was isolated by chromatography on Protein A-Sepharose by the method of Ey et al. (1978). For affinity purification, 2 ml of this IgG fraction (8.6 mg total in PBS) were incubated overnight with an Affigel 10-streptavidin column (2 ml of Affigel containing 5 mg of streptavidin). The column was then washed with PBS and eluted with 100 mM glycine-HC1, p H 2.5. Fractions (1 ml) were collected into tubes containing 100 gl of 1.5 M Tris-HC1, p H 8.8, dialyzed against PBS, and stored in aliquots at - 70°C.

Staining of proteins on replicas by the streptavidinamplified method Biotinylation of the replicas and the subsequent incubation to block non-specific protein binding sites were performed as described above. The replicas were then incubated sequentially with the

following reagents in a PBS solution containing 1% BSA and 0.05% Tween 80: streptavidin (1 ~ g / m l for 1 h), affinity-purified anti-streptavidin (0.5/~g/ml overnight), and G A R - H R P (4 g g / m l for 4 h). Following each of the incubations, the replicas were washed three times for 15 min each with PBS containing 0.05% Tween-80. Color was then developed as described above. The concentrations of reagents and incubation times were determined empirically and were chosen to give maximum staining sensitivity.

Results and discussion

Derivatization of proteins after transfer to nitrocellulose paper avoids problems associated with alterations in electrophoretic mobility that

6

T f~ ~ct

1

2

3

,4

5

6

7

Fig. 1. Determination of the optimal sulfo-NHS-biotin concentration for the biotinylation of proteins on blots. Nitrocellulose replicas of acetylcholine receptor-rich membrane proteins were reacted for 60 min with the followingconcentrations of sulfo-NHS-biotin, incubated with avidin-HRP, and developed: lane 1=50 #M; lane 2 =25 #M; lane 3~10 gM; lane 4 = 5 #M; lane 5 = 1 /~M; lane 6 = 500 nM. Lane 7 shows a Coomassie blue stained gel of acetylcholinereceptor-richmembranes prior to blotting. Greek letters on the right indicate the positions of the receptor subunits with the followingapparent molecular weights: a, 40000; r, 50000; -/, 60000; 8, 65000.

68

16

m7

aaO~

1

2

3

4

5

6

7

8

9

10

Fig. 2. Time course of protein biotinylation on blots. Acetylcholine receptor-rich membranes were reacted with l 0 / ~ M sulfo-NHSbiotin for the following time periods, then incubated with avidin-HRP and developed. Times of biotinylation were: lane 1 = 0 min; lane 2 = 5 min; lane 3 = 10 min; lane 4 = 15 min; lane 5 = 20 rain; lane 6 = 30 min; lane 7 = 40 min; lane 8 = 50 rain; lane 9 = 60 min. Lane 10 is a Coomassie blue stained gel of membrane proteins prior to blotting. Greek letters indicate the positions of the receptor subunits.

BSA l l

-8 nC~

1

2

3

4

5

1

2

3

4

5

6

Fig. 3. Detection sensitivity of avidin-HRP system. A: The following quantifies of BSA were applied to an SDS gel, blotted, reacted with 10 ~M sulfo-NHS-biotin for 45 min and stained with avidin-HRP: lane 1 = 10 ng; lane 2 = 20 rig; lane 3 = 30 ng; lane 4 = 50 ng; lane 5 = 100 ng, B: Acetylcholine receptor-rich membranes were processed in a~ identical manner. The amount of membranes appfied to the gel was: lane 1 = 150 ng; lane 2 = 300 ng; lane 3 = 600 ng; lane 4 = 1.25/~g; lane 5 = 2.50 ?Lg; lane 6 = 5.00 #g. The subunit was estimated to be approximately 10% of the total protein applied per well.

69

- - I~ SA

1

2

3

4

5

1

2

3

4

5

Fig. 4. Comparison of sensitivity of silver staining and streptavidin/anti-streptavidin detection system. Identical SDS gels were either (A) silver stained or (B) blotted, reacted with sulfo-NHS-biotin and stained with the streptavidin-anti-streptavidin system. Quantities of BSA loaded per well were: lane 1 = 25 ng; lane 2 = 10 ng; lane 3 = 5 ng; lane 4 ~ 2.5 ng; lane 5 = 1 ng.

~m

7--

#--

ctm

may occur as a consequence of biotinylation prior to electrophoresis. Thus, we first sought to establish the optimal sulfo-NHS-biotin concentrations and reaction times for derivatization of proteins on blots. Replicas of Torpedo postsynaptic membrane proteins separated by SDS gel electrophoresis were reacted with buffered solutions containing various concentrations of sulfo-NHS-biotin and then stained with avidin-HRP (Fig. 1). For incubation times of 60 rain, concentrations of 5-10 /~M sulfo-NHS-biotin were optimal: The staining of protein bands diminished significantly at concentrations greater than 10 /~M. We have found that at higher concentrations of sulfo-NHS-biotin, postsynaptic membrane proteins do not remain tightly bound to nitrocellulose paper over the 60 rain biotinylation period (unpublished results). 40 rain reaction times were found to be optimal when concentrations of 10 #M sulfo-NHS-biotin were used (Fig. 2). Shorter incubation times may be sufficient if higher concentrations are used. While these conditions are optimal for the proteins present in these membranes, the properties of other proteins may require alterations in reac-

1

2

3

4

Fig. 5. Inhibition of streptavidin binding to biotinylated proteins on replicas. Replicas of acetylcholine receptor-rich membranes were reacted with sulfo-NHS-biotin. Subsequently, they were incubated with streptavidin solution containing the following additions: lane 1, no additions; lane 2, 1.7 ~ g / m l biotinylated BSA; lane 3, 5/,tg/ml biotinylated BSA; lane 4, 10 n M biotin. The strips were then processed for staining by the amplified procedure.

tion time and reagent concentration to achieve maximal staining. The sensitivity of the avidin-HRP procedure was examined on replicas containing Torpedo postsynaptic membrane proteins or BSA. As little as 20 ng per band of BSA or acetylcholine receptor subunit could be visualized with this technique (Fig. 3). This sensitivity is comparable to, or possibly slightly higher than, that reported for the

70

Bio-Rad staining kit (Bio-Radiations, 1985). We found that the method is more sensitive for nitrocellulose paper replicas than for proteins transferred to Biodyne nylon membranes by approximately ten-fold due to higher background staining (data not shown). Using the same derivatization procedure, we have achieved greater sensitivity of the staining method with an amplified detection system. After biotinylation of proteins on replicas, the blots were reacted sequentially with streptavidin, rabbit anti-streptavidin, and GAR-HRP. Streptavidin rather than adivin was used since the former has fewer non-specific interactions due to its neutral isoelectric point and absence of glycosylation (Chaiet and Wolf, 1964). As shown in Fig. 4, the amplified system is comparable in sensitivity to silver staining of proteins in polyacrylamide gels; it detects less than 5 ng of BSA on replicas. The staining can be attributed to the binding of streptavidin to biotinylated proteins, since the immunostaining of acetylcholine receptor-rich membranes may be abolished by inclusion of 10 nM biotin or biotinylated-BSA in the streptavidin reaction (Fig. 5). Several factors affect the sensitivity of this amplified staining technique. The efficiency of elution of the protein from the gel and its affinity for the nitrocellulose matrix are important parameters. Also, the extent of biotinylation will be directly related to the number of amino groups in a protein. Our results caution against over-derivatization, however, which may reduce staining intensity. Since this study used BSA, an acidic protein with a pI of 4.7-4.9 (Radola, 1973), the technique may be more sensitive for basic proteins. The sensitivity might be further enhanced by utilization of the peroxidase anti-peroxidase method (Sternberger, 1979) and by derivatization of the replica with the long chain analogue of succinimidobiotin, sulfosuccinimidyl 6-(biotinamido) hexanoate (Hofmann et al., 1982). Finally, the very low background staining of the paper achieved with both the avidin-HRP protocol and the amplified procedure permit visualization of dark protein bands against a white background. Since overstaining is easily avoided, dark or cloudy background staining which often occurs in gels subjected to silver staining is not a limitation.

The amplified detection method utilizing streptavidin and anti-streptavidin that we have described here for replica staining may have other uses as well. By substituting the appropriate enzyme conjugate of anti-rabbit IgG, the procedure may be adapted to provide greater sensitivity in solid-phase well assays or in immunofluorescence localization studies. A particularly useful property of the anti-streptavidin antibodies is their lack of reactivity with avidin. In solid-phase well assays, we have found that non-specific background levels, apparently due to recognition of endogenous biotin-containing enzymes by streptavidin, can be substantially reduced by preincubation of samples with avidin (LaRochelle and Froehner, unpublished results). We expect that the sensitivity and specificity of any detection system utilizing biotin-avidin interactions, including nucleic acid probes synthesized with biotin analogues of nucleotides (Langer et al., 1981), may be improved by this procedure.

Acknowledgements This work was supported by grants from the National Institutes of Health (NS-14871) and the Muscular Dystrophy Association.

References Bayer E.A., M. Wilchek and E. Skutelsky, 1976, FEBS Lett. 68, 240. Bers, G. and D. Garfin, 1985, Biotechniques 3, 276. Bio-Radiations, 1985, 56, 6. Bio-Rad Laboratories, 1983, Bulletin 1080, 1. Chalet, L. and F.J. Wolf, 1964, Arch. Biochem. Biophys. 106, 1. Ey, P.L., S.J. Prowse and C.R. Jenkin, 1978, Imrnunochemistry 15, 429. Froehner, S.C., K. Douville, S. Klink and W.J. Culp, 1983, J. Biol. Chem. 258, 7112. Gershoni, J.M. and G.E. Palade, 1983, Anal. Biochern. 124, 396. Gyorgy, P., C.S. Rose, R.E. Eakin, E.E. Snell and R.J. Williams, 1941, J. Biol. Chem. 140, 535. Hancock, K. and V.C.W. Tsang, 1983, Anal. Biochem. 133, 157. Hofmann, K., G. Titus, J.A. Montibeller and F.M. Finn, 1982, Biochemistry 21, 978. Kittler, J.M., N.T. Meisler, D. Viceps-Madore, J.A. Cidlowsky and J.W. Thanassi, 1984, Anal. Biochem. 137, 210.

71 Laemmli, U.K., 1970, Nature 227, 680. Langer, P.R., A.A. Waldrop and D.C. Ward, 1981, Proc. Natl. Acad. Sci. U.S.A. 78, 6633. Lowry, O.H., N.J. Rosebrough, A.L. Farr and R.J. Randall, 1951, J. Biol. Chem. 193, 265. Matsudaira, P.T. and D.R. Burgess, 1978, Anal. Biochem. 87, 387. Porter, S. and S.C. Froehner, 1983, J. Biol. Chem. 258, 10034. Radola, B.J., 1973, Biochim. Biophys. Acta 295, 412.

Stemberger, L.A., 1979, in: Immunochemistry, ed. L.A. Stemberger (Prentice-Hall, Engelwood Cliffs, N.I) pp. 104-169. Towbin, H., T. Staehelin and J. Gordon, 1979, Proc. Natl. Acad. Sci. U.S.A. 76, 4350. Wojtkowiak, Z., R.C. Briggs and L.S. Hnilica, 1983, Anal. Biochem. 129, 486. Wray, W., T. Boulikas, V.P. Wray and R. Hancock, 1981, Anal. Biochem. 118, 197.