- Email: [email protected]
Effects of the blocking agents bovine serum albumin and Tween 20 in different buffers on immunoblotting of brain proteins and marker proteins
Journal of Immunological Methods, 88 (1986) 233-237 Elsevier
233
JIM 03872
Effects of the blocking agents bovine serum albumin and Tween 20 in different buffers on immunoblotting of brain proteins and marker proteins Elisabeth Wedege 1,, a n d G e r d Svenneby 2 J Department of Methodology, National Institute of Public Health, Geitmyrsv. 75, 0462 Oslo 4, and 2 Neurochemical Laboratory, Preclinieal Medicine, The Oslo University, P.O. Box 1115, Blindern, 0317 Oslo 3, Norway
(Received 22 October 1985, accepted 2 December 1985)
The effects of the blocking agents bovine serum albumin and Tween 20 in buffers at pH values 7.2 and 10.2 were compared in immunoblotting with 2 different antisera. The antisera were raised against a purified brain-specific protein fraction from human brain, soluble in perchloric acid, and phosphate-activated glutaminase from pig brain, respectively. The antigens were a crude perchloric acid-soluble brain extract, a crude brain phosphate-activated glutaminase fraction, and proteins commonly used as molecular weight markers. The binding patterns of the 2 antisera to the respective brain antigen preparations changed, depending on the blocking agent and the pH of the blocking buffer. Also, antibody binding to the molecular weight marker proteins was observed with some of the blocking buffers. Immunoblotting with Tris-saline, pH 10.2, containing 3% bovine serum albumin as blocking agent and diluting buffer for the antisera, showed negligible antibody binding to the marker proteins and most specific binding to the brain antigens. Key words: lmmunoblotting; Blocking conditions
Introduction
Various blocking agents, such as bovine serum albumin (BSA) (Towbin et al., 1979), Tween 20 (Batteiger et al., 1982; Muilerman et al., 1982), gelatin (Lin and Kasamatsu, 1983), or haemoglobin (Gershoni and Palade, 1982), have been used to saturate vacant protein binding sites on NC-
* To whom correspondence should be addressed. Abbreviations: BSA, bovine serum albumin; NC, nitrocellulose; PAG, phosphate-activated glutaminase; ab-PAG, antibodies against PAG; P2, brain-specific protein fraction; ab-P2, antibodies against P2; PBS, phosphate-buffered saline; PCA, perchloric acid; SDS, sodium dodecyl sulphate.
filters in immunoblotting procedures. We have compared the blocking effects of the cheaper reagent Tween 20 to that of BSA at pH values 7.2 and 10.2 employing 2 different antisera. The antisera were raised in rabbits against a purified PCA-soluble brain-specific protein fraction from human brain (P2) (Wedege, 1979) and against phosphate-activated glutaminase from pig brain (PAG) (Svenneby et al., 1973) respectively. The antigens were a crude PCA-soluble brain protein fraction, a crude PAG preparation and standard molecular weight proteins. Depending on the blocking conditions these antisera gave various patterns of binding to the standard proteins and to different components in the antigen preparations.
0022-1759/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
234 Materials and methods
SDS-gel electrophoresis and electrotransfer of proteins
Materials
SDS-gel electrophoresis was performed in 1.5 mm thick 12% acrylamide gel slabs (14 × 12 cm) with 4% stacking gel (Laemmli, 1970). Pyronin Y was added to the samples to orientate the NC filters after blotting. The gels were stained with Coomassie brilliant blue 250 R (Nicolas and Goodwin, 1982). Electrophoretic transfer of proteins from SDS gels to NC filters was performed in cold Tris-glycine buffer, pH 8.3, containing 20% methanol (v/v) at 300-400 mA (Towbin et al., 1979). The human brain extract, containing low M r proteins, was transferred to 0.2 ~m NC filters (Burnette, 1981) for 3 h, while PAG-containing material was transferred overnight to 0.45/.tin NC filters. The filters were stained with 0.1% Amido black (Gershoni and Palade, 1982).
Nitrocellulose filters, BA 83 and BA 85 of pore sizes 0.2 /~m and 0.45 /~m respectively, were obtained from Schleicher and Sch~ll. Peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins were from Dakopatts, Denmark. The standard proteins were phosphorylase a from rabbit muscle (EC 2.4.1.1) M r 94000; carbonic anhydrase from bovine erythrocytes (EC 4.2.1.1) M r 30000; cytochrome c, type III from horse heart, M r 12400; all from Sigma Chemical Co. The following proteins in a calibration kit were from Pharmacia Fine Chemicals: BSA, M r 67 000; ovalbumin, M r 43 000; chymotrypsinogen A, M r 25000; and ribonuclease A (EC 3.1.4.22) M r 13700. Tween 20 and BSA (fraction V), which were used as blocking agents; 2-aminoethanol; 3-amino-9-ethylcarbazole; and N,N-dimethyl-form a m i d e were from Sigma Chemical Co. Glutaraldehyde was from Serva. All other reagents were of highest purity.
Antibody preparations A PCA soluble brain-specific fraction (P2), containing 2 proteins of low molecular weight (Wedege, 1979), was purified by gel filtration from human brain (Wedege, 1973). PAG was purified (Svenneby et al., 1973), and a preparation which was approximately 50% pure, judged by SDS-gel electrophoresis, was used as antigen in the immunization procedures. Antibodies against P2 and PAG (ab-P2 and ab-PAG) were raised in rabbits. The antibody-containing sera were pooled, and the immunoglobulins isolated by ammonium sulphate precipitation (Harboe and Ingild, 1973). Ab-PAG was absorbed with pig serum.
Brain antigens A crude PCA-soluble protein extract from human brain, composed of P2 and other proteins, was prepared as described (Wedege, 1973). P2 contains 2 proteins of M r 12000 and 13000 (Wedege, 1979). Partly purified PAG, named B2 in Svenneby et al. (1973), was isolated from pig brain. Purified PAG from brain consists of identical subunits of M r 64000 (Svenneby et al., 1973).
Immunoblotting After electrotransfer the NC filters were blocked for 30-60 min with BSA or Tween 20 in phosphate-buffered saline (PBS), pH 7.2, or Tris-saline (20 mM Tris, 150 mM NaCI), pH 10.2. Thus 4 blocking buffers were used: A: 3% BSA in PBS, pH 7.2. B: 3% BSA in Tris-saline, pH 10.2. C: 0.5% Tween 20 in PBS, pH 7.2. D: 0.5% Tween 20 in Tris-saline, pH 10.2. The same blocking buffer (that is A, B, C or D) was used for blocking and dilution of the primary and secondary antibodies, whereas the washing buffers were without BSA and Tween 20. The respective blocking and washing buffers used throughout the procedure will be referred to as system A, B, C, and D respectively. The NC-filters were incubated for 18 h with the primary antibodies, washed with the respective washing buffers, and incubated for 2 h with peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins diluted 1 : 500. The immunostaining was performed with 3-amino-9-ethylcarbazole and H202 in sodium acetate buffer, pH 5.0 (Orstavik, 1981). In some experiments the filters were treated after electrotransfer with 5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, for 30 min followed by 30 min incubation with 1 M 2-aminoethanolHC1, pH 7.4. The filters were then immunoblotted as described above.
235
Results and discussion Immunoblots with ab-P2 Immunoblots of standard proteins, incubated with ab-P2 showed a distinct immunoreactive band for carbonic anhydrase in a system A (Fig. 1). A band of M r 66000, just below BSA, was also observed. In system B the staining intensity of these bands were diminished. In contrast, immunoblotting in system C and D demonstrated binding of antibodies to chymotrypsinogen A and 2 bands of M r 15000 and 14000. The 2 low M r
12
bands were probably proteolytic products or contaminants of chymotrypsinogen A, as electrophoresis and blotting of chymotrypsinogen A alone gave the same results. Similar findings were observed with chymotrypsinogen A from another manufacturer (Koch-Light Laboratories). A preimmune serum did not stain any of the standard proteins using systems A and B, whereas systems C and D gave the same results as the postimmune serum. The band of M r 66 000 just below BSA was also seen in SDS gels stained with Coomassie brilliant
12
12
12
A
B
G
12
OA-
Fig. 1. Effect of various blocking conditions on the immunostaining of standard proteins and a crude PCA-soluble protein extract from h u m a n brain detected by ab-P2 (1 : 1 000). Lane 1 shows standard proteins and lane 2 the crude h u m a n brain extract. 3 /Lg of each standard protein and 50 ktg of the brain extract were applied on the SDS gels. The arrows mark the 2 proteins of P2 of M r 12000 and 13000. The 2 lanes to the left were stained with Amido black. The blocking buffers were: A, 3% BSA in PBS, p H 7.2; B, 3% BSA in Tris-saline, pH 10.2; C, 0.5% Tween 20 in PBS, p H 7.2; D, 0.5% Tween 20 in Tris-saline, p H 10.2. Washing buffers were without BSA or Tween 20. The positions of the standard proteins are marked to the left: Ph a, phosphorylase a; BSA, bovine serum albumin; OA, ovalbumin; CA, carbonic anhydrase; Ch A, chymotrypsinogen A; RNAse, ribonuclease A; Cyt c, cytochrome c.
236
blue before or after electrotransfer, but not on NC-filters stained with Amido black. This band presumably corresponded to artifacts caused either by 2-mercaptoethanol in the sample buffer (Tasheva and Dessev, 1983) or by keratin protein contamination (Ochs, t983). The crude PCA-soluble cortical extract contained other proteins in addition to P2, as seen in Coomassie brilliant blue stained gels (not shown) and after Amido black staining of corresponding blotted NC-filter (Fig. 1). The bands marked with arrows in Fig. 1 constituted P2 of M r 12000 and 13 000, as demonstrated previously (Wedege, 1979). After incubation with ab-P2 these 2 proteins bound antibodies distinctly in system A. Some faintly stained bands of higher M r were also observed. However, these bands were negligible in system B. The Tween 20 containing systems C and D
12
12
changed the pattern of antibody binding in the low M r region of the crude brain extract. The band of M r 12000 disappeared and a band of M r 15000 was seen. The latter band was only observed as a trace after protein staining of SDS gels and NC filters. No immunoreactivity of any of the cortical proteins was observed with a preimmune serum in systems A, B, C or D. Treatment of the NC filters with glutaraldehyde and 2-aminoethanol resulted in strong antibody binding of ab-P2 to all standard proteins and all cortical proteins in the 4 systems (not shown). Thus, immunoblotting with ab-P2 using BSA in Tris-saline, pH 10.2, was the most suitable condition for detecting the 2 proteins of P2 in a crude brain extract. Only negligible antibody binding of ab-P2 to the marker proteins was observed under this condition. In contrast, immunoblotting with
12
12
12
88J OJ
A
B
C
D
Fig. 2. Effect of various blocking conditions on the immunostaining of standard proteins and a crude pig brain PAG preparation detected by ab-PAG (1 : 1013). Lane 1 shows standard proteins and lane 2 crude pig brain PAG. 3 p.g of each standard protein and 80 /*g of crude pig brain PAG were applied on the SDS gels. The experimental conditions were as in Fig. 1.
237 Tween 20 containing buffers changed the binding pattern of ab-P2 and also showed distinct antibody binding to some of the marker proteins.
Immunoblots with ab-PA G Fig. 2 shows the reaction of ab-PAG with standard proteins and a crude PAG preparation. The standard proteins reacted differently in the 4 systems. Carbonic anhydrase showed a faint band in all systems, whereas chymotrypsinogen A demonstrated variable staining in system C. Several components of crude P A G bound antibodies in system A (Fig. 2). The most intensely stained band had a M r of approximately 64000. Fewer bands were observed in system C. In systems B and D only faint bands in addition to the 64000 band could be observed, and the intensity of the 64000 band in system D was reduced compared to that in system B. It should be noted that the intensity of the immunoreactive bands increased with prolonged staining. The preimmune serum gave faint coloured bands with carbonic anhydrase in systems A and B, and with phosphorylase a in system D. When the N C blots were treated with glutaraldehyde and 2-aminoethanol prior to incubation with ab-PAG, the immunoreactivity of all proteins was reduced using system A or B. The only visible coloured band corresponded to M r 64000, whereas no bands were seen in systems C and D (not shown). The decreased immunostaining observed after pretreatment of the NC filters with glutaraldehyde and 2-aminoethanol in systems A and B, makes the ab-PAG suitable in immunocytochemistry. We have shown PAG-like immunoreactivity in slices from mouse hippocampus, apparently visualizing PAG-containing structures (Svenneby and Storm-Mathisen, 1983). In conclusion, immunoblotting with 3% BSA in Tris-saline, pH 10.2 (system B) seemed to be the best choice, whereas the cheaper reagent Tween 20 gave non-specific staining. The advantage of system B was observed with 2 sets of experiments where the electrotransfer conditions, the antigens
and the antisera differed. Gershoni and Palade (1982) suggest that false positive bands on immunoblots may be obtained as a result of hydrophobic interaction between antigens and antibodies mediated by detergents. Such interaction might explain the distinct binding of preimmune serum and ab-P2 to chymotrypsinogen A and its low M r components, and the binding of ab-P2 to another protein in the crude PCA-soluble fraction in the presence of Tween 20. Lin and Kasamatsu (1983) have also shown that the non-ionic detergent NP-40 removes protein bound to NC filters. The results presented here show that it is important to control the blocking conditions to avoid non-specific binding in immunoblotting, as also reported recently by Spinola and Cannon (1985).
References Batteiger, B., W.J. Newhall and R.B. Jones, 1982, J. Immunol. Methods 55, 297. Burnette, W.N., 1981, Anal. Biochem. 112, 195. Gershoni, J.M. and G.E. Palade, 1982, Anal. Biochem. 124, 396. Harboe, N. and A. Ingild, 1973, in: A Manual of Quantitative Immunoelectrophoresis, eds. N.H. Axelsen, J. Kroll and B. Weede (Universitetsforlaget, Oslo) p. 161. Laemmli, U.K., 1970, Nature (London) 227, 680. Lin, W. and H. Kasamatsu, 1983, Anal. Biochem. 128, 302. Muilerman, H.G., H.G.J. Ter Hart and W. Van Dijk, 1982, Anal. Biochem. 120, 46. Nicolas, R.H. and G.H. Goodwin, 1982, in: The HMG Chromosomal Proteins, ed. E.W. Johns (Academic Press, New York) p. 41. Ochs, D., 1983, Anal. Biochem. 135, 470. Orstavik, K.H., 1981, Br. J. Haematol. 48, 15. Spinola, S.M. and J.G. Cannon, 1985, J. Immunol. Methods 81, 161. Svenneby, G. and J. Storm-Mathisen, 1983, in: The Metabolic Relationship of Glutamine, Glutamate and GABA, eds. L. Hertz, E. Kvamme, E.G. McGeer and A. Schousboe(Alan R. Liss, New York) p. 69. Svenneby, G., I.A. Torgner and E. Kvamme, 1973, J. Neurochem. 20, 1217. Tasheva, B. and G. Dessev, 1983, Anal. Biochem. 129, 98. Towbin, H., T. Staehlin and J. Gordon, 1979, Proc. Natl. Acad. Sci. U.S.A. 76, 4350. Wedege, E., 1973, J. Neurochem. 21, 1487. Wedege, E., 1979, Cell. Mol. Biol. 25, 207.
233
JIM 03872
Effects of the blocking agents bovine serum albumin and Tween 20 in different buffers on immunoblotting of brain proteins and marker proteins Elisabeth Wedege 1,, a n d G e r d Svenneby 2 J Department of Methodology, National Institute of Public Health, Geitmyrsv. 75, 0462 Oslo 4, and 2 Neurochemical Laboratory, Preclinieal Medicine, The Oslo University, P.O. Box 1115, Blindern, 0317 Oslo 3, Norway
(Received 22 October 1985, accepted 2 December 1985)
The effects of the blocking agents bovine serum albumin and Tween 20 in buffers at pH values 7.2 and 10.2 were compared in immunoblotting with 2 different antisera. The antisera were raised against a purified brain-specific protein fraction from human brain, soluble in perchloric acid, and phosphate-activated glutaminase from pig brain, respectively. The antigens were a crude perchloric acid-soluble brain extract, a crude brain phosphate-activated glutaminase fraction, and proteins commonly used as molecular weight markers. The binding patterns of the 2 antisera to the respective brain antigen preparations changed, depending on the blocking agent and the pH of the blocking buffer. Also, antibody binding to the molecular weight marker proteins was observed with some of the blocking buffers. Immunoblotting with Tris-saline, pH 10.2, containing 3% bovine serum albumin as blocking agent and diluting buffer for the antisera, showed negligible antibody binding to the marker proteins and most specific binding to the brain antigens. Key words: lmmunoblotting; Blocking conditions
Introduction
Various blocking agents, such as bovine serum albumin (BSA) (Towbin et al., 1979), Tween 20 (Batteiger et al., 1982; Muilerman et al., 1982), gelatin (Lin and Kasamatsu, 1983), or haemoglobin (Gershoni and Palade, 1982), have been used to saturate vacant protein binding sites on NC-
* To whom correspondence should be addressed. Abbreviations: BSA, bovine serum albumin; NC, nitrocellulose; PAG, phosphate-activated glutaminase; ab-PAG, antibodies against PAG; P2, brain-specific protein fraction; ab-P2, antibodies against P2; PBS, phosphate-buffered saline; PCA, perchloric acid; SDS, sodium dodecyl sulphate.
filters in immunoblotting procedures. We have compared the blocking effects of the cheaper reagent Tween 20 to that of BSA at pH values 7.2 and 10.2 employing 2 different antisera. The antisera were raised in rabbits against a purified PCA-soluble brain-specific protein fraction from human brain (P2) (Wedege, 1979) and against phosphate-activated glutaminase from pig brain (PAG) (Svenneby et al., 1973) respectively. The antigens were a crude PCA-soluble brain protein fraction, a crude PAG preparation and standard molecular weight proteins. Depending on the blocking conditions these antisera gave various patterns of binding to the standard proteins and to different components in the antigen preparations.
0022-1759/86/$03.50 © 1986 Elsevier Science Publishers B.V. (Biomedical Division)
234 Materials and methods
SDS-gel electrophoresis and electrotransfer of proteins
Materials
SDS-gel electrophoresis was performed in 1.5 mm thick 12% acrylamide gel slabs (14 × 12 cm) with 4% stacking gel (Laemmli, 1970). Pyronin Y was added to the samples to orientate the NC filters after blotting. The gels were stained with Coomassie brilliant blue 250 R (Nicolas and Goodwin, 1982). Electrophoretic transfer of proteins from SDS gels to NC filters was performed in cold Tris-glycine buffer, pH 8.3, containing 20% methanol (v/v) at 300-400 mA (Towbin et al., 1979). The human brain extract, containing low M r proteins, was transferred to 0.2 ~m NC filters (Burnette, 1981) for 3 h, while PAG-containing material was transferred overnight to 0.45/.tin NC filters. The filters were stained with 0.1% Amido black (Gershoni and Palade, 1982).
Nitrocellulose filters, BA 83 and BA 85 of pore sizes 0.2 /~m and 0.45 /~m respectively, were obtained from Schleicher and Sch~ll. Peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins were from Dakopatts, Denmark. The standard proteins were phosphorylase a from rabbit muscle (EC 2.4.1.1) M r 94000; carbonic anhydrase from bovine erythrocytes (EC 4.2.1.1) M r 30000; cytochrome c, type III from horse heart, M r 12400; all from Sigma Chemical Co. The following proteins in a calibration kit were from Pharmacia Fine Chemicals: BSA, M r 67 000; ovalbumin, M r 43 000; chymotrypsinogen A, M r 25000; and ribonuclease A (EC 3.1.4.22) M r 13700. Tween 20 and BSA (fraction V), which were used as blocking agents; 2-aminoethanol; 3-amino-9-ethylcarbazole; and N,N-dimethyl-form a m i d e were from Sigma Chemical Co. Glutaraldehyde was from Serva. All other reagents were of highest purity.
Antibody preparations A PCA soluble brain-specific fraction (P2), containing 2 proteins of low molecular weight (Wedege, 1979), was purified by gel filtration from human brain (Wedege, 1973). PAG was purified (Svenneby et al., 1973), and a preparation which was approximately 50% pure, judged by SDS-gel electrophoresis, was used as antigen in the immunization procedures. Antibodies against P2 and PAG (ab-P2 and ab-PAG) were raised in rabbits. The antibody-containing sera were pooled, and the immunoglobulins isolated by ammonium sulphate precipitation (Harboe and Ingild, 1973). Ab-PAG was absorbed with pig serum.
Brain antigens A crude PCA-soluble protein extract from human brain, composed of P2 and other proteins, was prepared as described (Wedege, 1973). P2 contains 2 proteins of M r 12000 and 13000 (Wedege, 1979). Partly purified PAG, named B2 in Svenneby et al. (1973), was isolated from pig brain. Purified PAG from brain consists of identical subunits of M r 64000 (Svenneby et al., 1973).
Immunoblotting After electrotransfer the NC filters were blocked for 30-60 min with BSA or Tween 20 in phosphate-buffered saline (PBS), pH 7.2, or Tris-saline (20 mM Tris, 150 mM NaCI), pH 10.2. Thus 4 blocking buffers were used: A: 3% BSA in PBS, pH 7.2. B: 3% BSA in Tris-saline, pH 10.2. C: 0.5% Tween 20 in PBS, pH 7.2. D: 0.5% Tween 20 in Tris-saline, pH 10.2. The same blocking buffer (that is A, B, C or D) was used for blocking and dilution of the primary and secondary antibodies, whereas the washing buffers were without BSA and Tween 20. The respective blocking and washing buffers used throughout the procedure will be referred to as system A, B, C, and D respectively. The NC-filters were incubated for 18 h with the primary antibodies, washed with the respective washing buffers, and incubated for 2 h with peroxidase-conjugated swine immunoglobulins to rabbit immunoglobulins diluted 1 : 500. The immunostaining was performed with 3-amino-9-ethylcarbazole and H202 in sodium acetate buffer, pH 5.0 (Orstavik, 1981). In some experiments the filters were treated after electrotransfer with 5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, for 30 min followed by 30 min incubation with 1 M 2-aminoethanolHC1, pH 7.4. The filters were then immunoblotted as described above.
235
Results and discussion Immunoblots with ab-P2 Immunoblots of standard proteins, incubated with ab-P2 showed a distinct immunoreactive band for carbonic anhydrase in a system A (Fig. 1). A band of M r 66000, just below BSA, was also observed. In system B the staining intensity of these bands were diminished. In contrast, immunoblotting in system C and D demonstrated binding of antibodies to chymotrypsinogen A and 2 bands of M r 15000 and 14000. The 2 low M r
12
bands were probably proteolytic products or contaminants of chymotrypsinogen A, as electrophoresis and blotting of chymotrypsinogen A alone gave the same results. Similar findings were observed with chymotrypsinogen A from another manufacturer (Koch-Light Laboratories). A preimmune serum did not stain any of the standard proteins using systems A and B, whereas systems C and D gave the same results as the postimmune serum. The band of M r 66 000 just below BSA was also seen in SDS gels stained with Coomassie brilliant
12
12
12
A
B
G
12
OA-
Fig. 1. Effect of various blocking conditions on the immunostaining of standard proteins and a crude PCA-soluble protein extract from h u m a n brain detected by ab-P2 (1 : 1 000). Lane 1 shows standard proteins and lane 2 the crude h u m a n brain extract. 3 /Lg of each standard protein and 50 ktg of the brain extract were applied on the SDS gels. The arrows mark the 2 proteins of P2 of M r 12000 and 13000. The 2 lanes to the left were stained with Amido black. The blocking buffers were: A, 3% BSA in PBS, p H 7.2; B, 3% BSA in Tris-saline, pH 10.2; C, 0.5% Tween 20 in PBS, p H 7.2; D, 0.5% Tween 20 in Tris-saline, p H 10.2. Washing buffers were without BSA or Tween 20. The positions of the standard proteins are marked to the left: Ph a, phosphorylase a; BSA, bovine serum albumin; OA, ovalbumin; CA, carbonic anhydrase; Ch A, chymotrypsinogen A; RNAse, ribonuclease A; Cyt c, cytochrome c.
236
blue before or after electrotransfer, but not on NC-filters stained with Amido black. This band presumably corresponded to artifacts caused either by 2-mercaptoethanol in the sample buffer (Tasheva and Dessev, 1983) or by keratin protein contamination (Ochs, t983). The crude PCA-soluble cortical extract contained other proteins in addition to P2, as seen in Coomassie brilliant blue stained gels (not shown) and after Amido black staining of corresponding blotted NC-filter (Fig. 1). The bands marked with arrows in Fig. 1 constituted P2 of M r 12000 and 13 000, as demonstrated previously (Wedege, 1979). After incubation with ab-P2 these 2 proteins bound antibodies distinctly in system A. Some faintly stained bands of higher M r were also observed. However, these bands were negligible in system B. The Tween 20 containing systems C and D
12
12
changed the pattern of antibody binding in the low M r region of the crude brain extract. The band of M r 12000 disappeared and a band of M r 15000 was seen. The latter band was only observed as a trace after protein staining of SDS gels and NC filters. No immunoreactivity of any of the cortical proteins was observed with a preimmune serum in systems A, B, C or D. Treatment of the NC filters with glutaraldehyde and 2-aminoethanol resulted in strong antibody binding of ab-P2 to all standard proteins and all cortical proteins in the 4 systems (not shown). Thus, immunoblotting with ab-P2 using BSA in Tris-saline, pH 10.2, was the most suitable condition for detecting the 2 proteins of P2 in a crude brain extract. Only negligible antibody binding of ab-P2 to the marker proteins was observed under this condition. In contrast, immunoblotting with
12
12
12
88J OJ
A
B
C
D
Fig. 2. Effect of various blocking conditions on the immunostaining of standard proteins and a crude pig brain PAG preparation detected by ab-PAG (1 : 1013). Lane 1 shows standard proteins and lane 2 crude pig brain PAG. 3 p.g of each standard protein and 80 /*g of crude pig brain PAG were applied on the SDS gels. The experimental conditions were as in Fig. 1.
237 Tween 20 containing buffers changed the binding pattern of ab-P2 and also showed distinct antibody binding to some of the marker proteins.
Immunoblots with ab-PA G Fig. 2 shows the reaction of ab-PAG with standard proteins and a crude PAG preparation. The standard proteins reacted differently in the 4 systems. Carbonic anhydrase showed a faint band in all systems, whereas chymotrypsinogen A demonstrated variable staining in system C. Several components of crude P A G bound antibodies in system A (Fig. 2). The most intensely stained band had a M r of approximately 64000. Fewer bands were observed in system C. In systems B and D only faint bands in addition to the 64000 band could be observed, and the intensity of the 64000 band in system D was reduced compared to that in system B. It should be noted that the intensity of the immunoreactive bands increased with prolonged staining. The preimmune serum gave faint coloured bands with carbonic anhydrase in systems A and B, and with phosphorylase a in system D. When the N C blots were treated with glutaraldehyde and 2-aminoethanol prior to incubation with ab-PAG, the immunoreactivity of all proteins was reduced using system A or B. The only visible coloured band corresponded to M r 64000, whereas no bands were seen in systems C and D (not shown). The decreased immunostaining observed after pretreatment of the NC filters with glutaraldehyde and 2-aminoethanol in systems A and B, makes the ab-PAG suitable in immunocytochemistry. We have shown PAG-like immunoreactivity in slices from mouse hippocampus, apparently visualizing PAG-containing structures (Svenneby and Storm-Mathisen, 1983). In conclusion, immunoblotting with 3% BSA in Tris-saline, pH 10.2 (system B) seemed to be the best choice, whereas the cheaper reagent Tween 20 gave non-specific staining. The advantage of system B was observed with 2 sets of experiments where the electrotransfer conditions, the antigens
and the antisera differed. Gershoni and Palade (1982) suggest that false positive bands on immunoblots may be obtained as a result of hydrophobic interaction between antigens and antibodies mediated by detergents. Such interaction might explain the distinct binding of preimmune serum and ab-P2 to chymotrypsinogen A and its low M r components, and the binding of ab-P2 to another protein in the crude PCA-soluble fraction in the presence of Tween 20. Lin and Kasamatsu (1983) have also shown that the non-ionic detergent NP-40 removes protein bound to NC filters. The results presented here show that it is important to control the blocking conditions to avoid non-specific binding in immunoblotting, as also reported recently by Spinola and Cannon (1985).
References Batteiger, B., W.J. Newhall and R.B. Jones, 1982, J. Immunol. Methods 55, 297. Burnette, W.N., 1981, Anal. Biochem. 112, 195. Gershoni, J.M. and G.E. Palade, 1982, Anal. Biochem. 124, 396. Harboe, N. and A. Ingild, 1973, in: A Manual of Quantitative Immunoelectrophoresis, eds. N.H. Axelsen, J. Kroll and B. Weede (Universitetsforlaget, Oslo) p. 161. Laemmli, U.K., 1970, Nature (London) 227, 680. Lin, W. and H. Kasamatsu, 1983, Anal. Biochem. 128, 302. Muilerman, H.G., H.G.J. Ter Hart and W. Van Dijk, 1982, Anal. Biochem. 120, 46. Nicolas, R.H. and G.H. Goodwin, 1982, in: The HMG Chromosomal Proteins, ed. E.W. Johns (Academic Press, New York) p. 41. Ochs, D., 1983, Anal. Biochem. 135, 470. Orstavik, K.H., 1981, Br. J. Haematol. 48, 15. Spinola, S.M. and J.G. Cannon, 1985, J. Immunol. Methods 81, 161. Svenneby, G. and J. Storm-Mathisen, 1983, in: The Metabolic Relationship of Glutamine, Glutamate and GABA, eds. L. Hertz, E. Kvamme, E.G. McGeer and A. Schousboe(Alan R. Liss, New York) p. 69. Svenneby, G., I.A. Torgner and E. Kvamme, 1973, J. Neurochem. 20, 1217. Tasheva, B. and G. Dessev, 1983, Anal. Biochem. 129, 98. Towbin, H., T. Staehlin and J. Gordon, 1979, Proc. Natl. Acad. Sci. U.S.A. 76, 4350. Wedege, E., 1973, J. Neurochem. 21, 1487. Wedege, E., 1979, Cell. Mol. Biol. 25, 207.