Effects of the modification of transfer buffer composition and the renaturation of proteins in gels on the recognition of proteins on western blots by monoclonal antibodies

ANALYTICAL

BIOCHEMISTRY

157, 144-153 (1986)

Effects of the Modification of Transfer Buffer Composition and the Renaturation of Proteins in Gels on the Recognition of Proteins on Western Blots by Monoclonal Antibodies’ D. Dm2

STANLEY

Department of Biochemistry, University of Western Ontario, London, Ontario, Canada N6A 5Cl Received February 3, 1986 Two modifications to Western blots which enhance immunochemical recognition have been developed. The first is transfer in carbonate buffer at pH 9.9, rather than the more commonly used Tris-glycine buffer at pH 8.3. This alteration improved the recognition of four of the five subunits of Escherichia coli F,-ATPase by monoclonal antibodies, the smaller subunits showing the greatest effects. Recognition of dinitrophenyl groups attached to the subunits by polyclonal antibodies was improved by the carbonate buffer only for the smallest ATPase subunit, e. The second mod&cation was incubation of the gel in mild buffers, designed to promote the renaturation of proteins, before the electrophoretic transfer step. The most effective buffer was 20% glycerol in 50 mM Tris-HCl, pH 7.4. Improvements in the signal obtained with monoclonal antibodies to all the subunits of ATPase were obtained by this procedure. As the subunits vary markedly in size, isoelectric point, and other properties, this method should be useful for most proteins. The fate of the 15,000-Da c subunit, labeled with “‘1, was followed through a blotting experiment. As long as no sodium dodecyl sulfate was added to the transfer buffer, t was bound to nitrocelhtlose efficiently in either Tris-glycine or carbonate buffer. However, the c was retained much more strongly during the subsequent incubation steps if the transfer was done in the carbonate buffer. The binding oft to the nitrocellulose was even more stable when the gel had been treated with the buffered glycerol solution before transfer. These results indicate that the conditions under which c subunit first encounters the nitrocellulose markedly affect the stability of binding during subsequent steps. The F,-ATPase was partially fragmented by treatment with proteases and then run on a gel and either transferred immediately in Tris-glycine buffer or else treated with the buffered glycerol solution and transferred in the carbonate buffer. The second blot gave stronger recognition of residual a subunit and fragments by an anti-a monoclonal antibody, with the largest improvement for the smaller fragments. This result suggests that the modified procedure may be particularly useful in enhancing the detection of small proteins. 8 1986 Academic ~n$hc. KEY WORDS: monoclonal antibodies; immunochemical mehods; Western blots; blotting to nitrocellulose; Fi-ATPase.

Since the first description of the electrophoretie transfer of proteins from an SDS3-polyacrylamide slab gel onto nitrocellulose with subsequent immunochemical probing (l), ’ This work was supported by Grant MA-7906 from the Medical Research Council of Canada. ’ Recipient of a Medical Research Council of Canada Scholarship. 3 Abbreviations used: SDS, sodium dodecyl sulfate; MSI, Micron Separations Inc.; S&S, S&&her & Schuell; DNP, 2,4dinitrophenyl; BSA, bovine semm albumin; PAGE, polyacrylamide gel electrophoresis. 0003-2697186 $3.00 cowrigbt Q 1986 by Academic F’nza, Inc. AS rights of nproduction in any form rexwed.

144

now known as the Western blot (2), this technique has become widely used in biological research and many modifications have been formulated (3,4). The proteins in the SDSpolyacrylamide gel are definitely in a denatured state, but their structure after transfer to nitrocellulose is not known. The state of folding of the proteins is particularly important when monoclonal antibodies, which recognize only a single epitope, are used, because many antibodies recognize assembled topographic determinants (5) which do not exist on de-

WESTERN

BLOTS OF RENATURED

natured proteins. Partial renaturation might occur during the transfer if no denaturant is present in the transfer buffer. However, inclusion of a low concentration of SDS in the transfer buffer has been suggested because it improves the movement of proteins out of the gel (6,7). Initial studies of the ability of 21 monoclonai antibodies against the subunits of the Escherichia coli F,-ATPase (8) to recognize their antigen after transfer to nitrocellulose in Tris-glycine buffer lacking SDS (1) revealed that all 10 of the antibodies to the larger subunits, a! and B, of M, 55,000 and 50,000, respectively, worked well on such blots. However, only one of three antibodies to y, M, 3 1,000, worked well, while none of those against 6, M, 20,000, or c, M, 15,000, recognized their antigenic subunit well on the blots. This apparent size dependence suggests that smaller proteins may generally be detected less efficiently than larger proteins. It has been reported that smaller proteins tend to bind to nitrocellulose less efficiently than larger proteins (7,9). I have explored the possibility that exposing proteins in the gel to conditions which favor renaturation before blotting and modifying the transfer buffer would result in enhanced recognition of small proteins by monoclonal antibodies. The results indicate that while exposing the gel to conditions conducive to protein renaturation prior to blotting can improve recognition of antigen by some antibodies, a much larger effect can be obtained by transferring the proteins in a more alkaline blot buffer without SDS. MATERIALS

AND METHODS

Materials. F,-ATPase was prepared from E. coli strain AN 1460 (lo), an overproducer of ATPase, as described previously (8). Membranes used in the blots presented below were taken from an ATPase preparation at the stage just before the extraction of ATPase from the membranes. About 6% of the protein in these membranes is ATPase. Membranes were sol-

PROTEINS

145

ubilized in preparation for SDS-polyacrylamide gel electrophoresis by heating at 100°C for 5 min in SDS sample buffer containing 125 mM Tris-HCI, pH 6.8,2% SDS, 10% sucrose, 50 mM dithiothreitol, and 0.01% bromphenol blue. The preparation of e subunit (11) and rabbit anti-rat IgG (8) has been described previously. Goat anti-rabbit IgG was a gift from Dr. Eric Ball of this department. Proteins were labeled with 12’1 by the IODO-GEN method ( 12). The isolation, growth, and properties of hybridomas which produce antibodies to the ATPase have been described (8). Hybridoma culture media containing antibodies were collected aseptically and stored at 4°C. The brand of nitrocellulose used in most experiments, Micron Separations Inc. (MSI) E02HYOOO 10 with a pore size of 0.22 pm, was obtained through Fisher Scientific. Where noted, Schleicher & Schuell (S&S) BA-83 nitrocellulose, with a pore size of 0.2 pm, was substituted. Both of these membranes are pure nitrocellulose rather than mixed esters. Anti-DNP-BSA serum was from Miles. Staphylococcus aweus V-8 protease was from Sigma, Papain was from Boehringer. Zwittergent 3- 14 was from Calbiochem. IODO-GEN was from Pierce. Gels and blots. The buffer system of Laemmli (14) was used with 0.75-mm-thick, 12% polyacrylamide gels. In some experiments, the sample was applied across the entire top surface of the stacking gel in a preparative manner, taking care that the surface was level, and the sample zone of uniform thickness across the gel. In other cases, individual wells were used. Electrophoresis was terminated when the bromphenol blue of the tracking dye had penetrated 9 cm (about 80% of the length) into the separating gel. Portions of the gel were stained with Coomassie blue R-250, while others were prepared for blotting to nitrocellulose. Electrophoretic blotting was performed in a Bio-Rad Transblot cell equipped with a glass cooling coil. The current was 0.8 to 1.OA, and the temperature was maintained below 35°C by running cold water through the cooling coil.

146

STANLEY

Some blots were carried out for 2 h, but later time-course experiments indicated that 1 h was sufficient for maximal results, so this time was adopted in later experiments. Two different buffers were used in this work. The first, designated “Tris-glycine blot buffer,” was the relatively standard 25 mM Tris, 192 IIIM glycine, pH 8.3, in 20% methanol (1). When this buffer was used the voltage was around 150 V. The second, designated “carbonate blot buffer,” contained 10 mM NaHC03, 3 mM Na2C03, pH 9.9, in 20% methanol. When this buffer was used, the voltage was around 100 V. Buffers were used only once. After blotting, some portions of the nitrocellulose were stained with India ink ( 15) while the rest were blocked by treatment overnight at 4“C with 3% BSA in “blot rinse buffer” (8) which contains 10 mM Tris-HCl, pH 7.4,O. 15 M NaCl, 1 mM EDTA, 0.1% Tween 20, and 0.04% sodium azide. The next day strips or sections of nitrocellulose were incubated for 4 h at 22°C with solutions of hybridoma culture media or sera in blot rinse buffer containing 0.3% BSA. The nitrocellulose was rinsed for 20 min in blot rinse buffer and then incubated for 2 h at 22°C in blot rinse buffer containing 0.3% BSA and 1251-labeled species-specific antibody to the first antibody (50,000 dpm/ml). Finally the nitrocellulose was washed for 20 min in blot rinse buffer, dried, and radioautographed using an intensifying screen. Dinitr~p~en~luti~n of ATPase subunits. Dinitrophenyl groups were introduced into the subunits of ATPase by a method designed to introduce a single DNP moiety into each molecule. First, sulfhydryl groups were blocked by treatment of ATPase with 5 mM N-ethylmaleimide for 2 h at 22°C in 10 mM TrisHCl, pH 7.4, containing 8 M urea. The protein was dialyzed into 1~ sodium acetate, pH 5.5, containing 8 M urea, CuS04 was added to 5 mM, sodium glyoxylate was added to 100 mM, and the sample was incubated at 22°C for 30 min to allow N-terminal transamination to occur (13). Following overnight dialysis into 0.2% SDS, trifluoroacetic acid was added to

D. DUNN

0.2% and the sample was mixed with an equal volume of 10 mM 2,4dinitrophenylhydmzine in dimethyl sulfoxide. The solution was incubated at 37°C for 1 h to allow hydrazone formation, the pH was adjusted to 8 with sodium tricine buffer, and 1 mg of sodium borohydride was added per milliliter of solution. After 30 min at 22”C, the protein was dialyzed exhaustively into 0.1% SDS and dried under reduced pressure. RESULTS

Eficts of transfer bujkrs. To make subsequent comparison of blots as valid as possible, the samples, reagents, and exposure times were uniform in all of the work shown in Figs. 1 to 3. The sample was a crude preparation of E. coli membranes in SDS gel sample buffer. Each panel of Fii. 1 to 3 shows an experiment performed under different conditions. The first three lanes of each panel show the efficiency of protein transfer from the gel to the nitrocellulose. Lane 1 contains a sample of gel stained with Coomassie blue R-250, and should be identical in all panels. Lane 2 shows a piece of the same gel after blotting, and lane 3 shows the blot after staining with India ink. It is evident that the protein patterns seen with the two types of stain are very different. The India ink appears to have very different afhnities for different proteins. Lanes 4 through 11 show the radioautograph obtained after strips of the blots were treated with rat monoclonal antibodies c- 1,” c-4,&2,6-4, y- 1, (u-3, p-2, and /3-1, respectively, followed by ‘251-labeled rabbit anti-rat IgG. Figure 1A shows that electrophoretic transfer in the Tris-glycine blot buffer containing 0.01% SDS (7) results in e5cient transfer of protein out of the gel (compare lanes 1 and 2) and good binding to the nitrocellulose (lane 3). However, none of the anti-t or anti-6 antibodies recognized their antigens on this blot, while the antibodies to (Y, & and y all gave 4 The monoclonal antibodies are designated by the subunit recognized and an arabic numeral.

WESTERN

12

3

4567891011

1

BLOTS OF RENATURED

2

3

4

5

6

7

8

91011

147

PROTEINS

1

2

3

4

5

6

7

891011

FIG. 1. Effects of transfer buffers on protein blotting efficiency and immunochemical recognition. E. cob membrane proteins (250 pg in SDS sample buffer) were electrophoresed on 16.5-cm-wide, preparativestyle gels. After electrophoresis, a portion of gel was stained with Coomassie blue (lane 1 of each panel). The rest of each gel was blotted electrophoretically for 2 h. After blotting, a sample of gel was again stained as a test of transfer efficiency (lane 2). A strip of the nitrocellulose was stained with India ink (lane 3). The rest of the nitrocellulose was used for immunochemical detection with various monoclonal antibodies, as described under Materials and Methods. The hybridoma culture supematants and dilutions used were as follows: lane 4, c-l, l/100, lane 5, t-4, l/100; lane 6,8-2, l/l@, lane 7,6-4, l/10; lane 8, y-l, l/10; lane 9, a-3, l/l@ lane 10, &2, l/100; lane 11, B-1, l/ 100. The transfer buffers used were (A) Tris-glycine blot buffer containing 0.0 1% SDS; (B) plain Tris-glycine blot buffer, (C) carbonate blot buffer. Radioautographs were exposed for 4 h with an intensifying screen.

detectable bands on this standard 4-h exposure. On an 18-h exposure, t-l gave faint reaction (not shown). In the blot shown in Fig. 1B the SDS was omitted from the blot buffer. The major effects of this omission were less efficient transfer of protein out of the gel, but nevertheless improved reactivity with the antibodies, particularly t- 1, y- 1, and p-2. On longer exposures, faint bands were observed with c-4 and 6-4 (not shown). The less efficient transfer of protein observed in lanes 1 to 3 could reflect both decreased solubility of membrane proteins in the absence of SDS and also lower electrophoretic mobility due to the decrease in negative charge accompanying the loss of SDS. Modification of the buffer to one of higher pH was tried because it should allow more efficient transfer of proteins by increasing their net charge. The blot buffer used in many of the subsequent tests contained 10 mM

NaI-ICOj ,3 mM Na2C03, and 20% methanol, with a final pH of 9.9 measured by a pH meter. When this buffer (Fig. 1C) was substituted for the standard 25 mM Tris, 192 mM glycine, 20% methanol, pH 8.3, buffer (l), the amount of protein left in the gel hardly changed, but the nitrocellulose stained more strongly. More importantly, the immunological recognition of all subuits except (Ywas increased. Here E1, t-4, and 6-4 all gave bands on a 4-h exposure, and 6-2 showed up on the 18-h exposure (not shown). Although most of the results presented here were from transfers performed for 2 h, transfer of proteins to the nitrocellulose reached its maximal level after 1 h. Treatment of the gel before transfer. I also tested the effect of treating the gel with various buffers before carrying out the transfer procedure to see if conditions which have been used to renature enzymes in SDS gels could further improve the reactivity of the blotted

148

STANLEY

proteins. Incubation of the gel in 50 ItIM TrisHCl, pH 7.4, for 1 h at room temperature, followed by transfer in the carbonate blot buffer, led to further improvements in reactivity (Fig. 2A, compare to Fig. 1C), which are perhaps most notable here for the increases in intensity of the bands due to the three larger subunits of ATPase rather than the smaller ones. Several additions to the renaturation buffer were tested to see if any further improvement could be obtained. As shown in Fig. 2B, inclusion of 20% glycerol ( 16,17) in the renaturation buffer caused a notable improvement in the detection of all subunits except possibly 0, compared to plain buffer (panel A). As the efficiency of transfer was decreased for some proteins (Fig. 2B, lane 2), this result suggests that the glycerol did cause improvements in renaturation. Increasing the ionic strength of the renaturation buffer by adding 0.2 M NaCl (16) improved the reactivity of the blotted proteins about as well as the glycerol, except that (Ywas slightly weaker (not shown). In contrast, inclusion of 25% isopropanol

1 2 3

4 5 6 7 8 91011

12

3

4

D. DUNN

in the renaturation buffer, as suggested by Blank and co-workers for renaturing ribonuclease in gels (18), led to highly inefficient transfer of most of the proteins (Fig. 2C). No detectable reaction was observed with (Yor y, but the reactivity of 6 and 6 may have been improved to some extent compared to plain buffer. The use of 10% isopropanol gave intermediate results (not shown). It would appear that while isopropanol does help the renaturation of smaller proteins, it leads to the fixation of many larger proteins in the gel. Several other water-miscible organic solvents were also tested for effects during the renaturation incubation (data not shown). Methanol at 10 or 20% gave results similar to those obtained with isopropanol, although fixation of proteins was not as severe. Dimethyl sulfoxide at 20% improved the reactivity of 6, but did not enhance the recognition of (Yand y. The effects of nonionic detergents were also tested, as suggested by the results of Manrow and Dottin (16) and Mandrell and Zollinger (19). Either 1% Triton X-100 or 0.1% Zwit-

567891011

1

2

3

4

5

6

7

8

91011

FIG. 2. Effects of renaturing treatments on protein-blotting efficiency and immunochemical recognition. E. coli membrane proteins were electrophoresed on gels as described in Fig. 1, and the gels were incubated in 200 ml of various butkrs with gentle agitation for 1 h at 22°C. Every 15 min, the buffer was removed and fresh buffer added. Portions of the gel were then stained or blotted for 2 h in carbonate blot buffer. The lanes of each panel correspond exactly to those in the panels of Fig. I. The renaturation buffers used were (A) 50 mu Tris-HCl, pH 7.4; (B) 50 mu Tris-HCl, pH 7.4, containing 20% glycerol; (C) 50 mM Trk-HCI, pH 7.4, containing 25% isopropanol. Badioautographs were exposed for 4 h with an intensifying screen.

WESTERN

BLOTS OF RENATURED

tergent 3- 14 was included in the 50 mM TrisHCl, pH 7.4, for the l-h renaturation incubation, and the gel was then given a further 30-min incubation with plain 50 mM TrisHCI, pH 7.4, to remove these detergents which might interfere with the binding of proteins to the nitrocellulose (9). The Triton X-100 (Fig. 3A) had a slightly beneficial effect on transfer, but decreased the reactivity observed with (Y, t, and possibly 6 (compare with Fig. 2A). The Zwittergent 3-14 (Fig. 3B) reduced the efficiency of transfer and caused definite reductions in the strength of the bands observed for all subunits. Bowen and co-workers (20) have suggested the use of 4 M urea to aid in the renaturation process. Exposure of the gel to 4 M urea in 50 mM Tris-HCI, pH 7.4, for 1 h, followed by 30 min in plain Tris-HCI, resulted in less efficient transfer of proteins and slightly decreased immunoreactivity of all subunits, except for S (Fig. 3C). When the renaturation procedure using 20% glycerol was carried out on a gel which was

12

3

4567891011

1

2

3

4

5

149

PROTEINS

transferred in the Tris-glycine buffer without SDS, the strength of recognition was about the same as if no renaturation had been performed (not shown), suggesting that the blot buffer was more important than the renaturation step. Eflect of blot conditions on recognition of dinitrophenyl groups by polyclonal antibodies. To see if the improvements obtained were specific for assembled topographic determinants, DNP-labeled ATPase subunits were prepared as described under Materials and Methods, run on gels, and blotted by various methods. Probing these blots with commercial polyclonal anti-DNP antibodies (Fig. 4) revealed that the presence of 0.01% SDS in the Tris-glycine blot buffer was generally detrimental (compare lane 1, in which SDS was used, to lane 2, in which it was not). The carbonate buffer (lane 3) gave better results than Tris-glycine for the small subunit, t, only, while the 6 subunit was not detected in any treatment. Eflect of transfer bu$er on the retention of c subunit by nitrocellulose. To determine if im-

6 7 8 91011

123

4567891011

FIG. 3. Effects of urea and nonionic detergents on protein-blotting ethciency and immunochemical recognition. E. coli membrane proteins were electrophoresed on gels as described in Fig. 1. One gel (A) was gently agitated with 1% Triton X-100 in 50 mM Tris-HCl, pH 7.4, for 1 h, with buffer changes as described in Fig. 2 and then for 30 min with plain 50 mM Tris-HCl, pH 7.4. After this, the gel was blotted with carbonate blot buffer for 2 h, and all subsequent aspects of the experiment are identical to those of Figs. 1 and 2. A second blot (B) was treated just the same except that 0.1% Zwittergent 3-14 was substituted for the Triton X- 100. A third gel (C) was treated the same except that 4 M urea was used in place of the nonionic detergents. Radioautographs were exposed for 4 h with an intensifying screen.

150

STANLEY

I

2

3

FIG. 4. Effects of blotting conditions on the recognition of dinitrophenyl groups on ATPase subunits. Samples of DNP-ATPase, prepared as described under Materials and Methods, were electrophoresed on gels (7 a of protein per lane) and then blotted to nitrocellulose under the following conditions: lane 1, Tris-glycine blot buffer containing 0.01% SDS, lane 2, plain Tris-glycine blot buffer, lane 3, carbonate blot buffer. All blots were blocked with BSA in the usual manner, and the DNP groups were detected by probing the blots with rabbit anti-DNP-BSA serum used at a concentration of 0.1 &ml, followed by ‘z51-labeled goat anti-rabbit IgG. The radioautograph was exposed for 14 h with an intensifying screen.

proved binding of protein to n&cellulose was responsible for the improved recognition of antigens after blotting in the carbonate buffer, 1251-labeled 6 was mixed with the membrane extract and then the mixture was electrophoresed on a gel at a protein concentration similar to that used in all the other experiments. The fate of the 1251-labeled c was then followed during blot and incubation procedures similar to those used in an immunochemical detection experiment (Table 1). In Experiment 1, it can be seen that blotting in the Tris-glycine buffer containing 0.01% SDS resulted in initial binding of only 47% of the applied t and near total loss of the labeled protein during the blocking and incubation steps. Omission of the SDS (Experiment 2) led to good binding and improved retention, although still only 6% of the labeled e detected on the gel was retained by the end of the experiment.

D. DUNN

Use of the carbonate buffer (Experiment 3) resulted in a large improvement in the retention of the protein during the blocking and incubation stages. Renaturation for 1 h in the glycerol buffer before blotting (Experiments 4 and 6) allowed the loss of about 40% of the labeled e from the gel before the blot was started. The t which remained in the gel, however, was transferred efficiently and retained well by the nitrocellulose. Thus despite the loss of protein during the renaturation phase, the MS1 nitrocellulose used in Experiment 4 had the most label left at the end of the incubation. To be sure that the results were not a peculiarity of the MS1 nitrocellulose, S&S BA83 was used in Experiments 5 and 6. Again, a large improvement was seen in changing from the Tris-glycine buffer to the carbonate buffer alter renaturation. Differences between the two brands were small, although the Schleicher & Schuell product did not retain the labeled c quite as well as the MS1 nitrocellulose after the renaturation blot. Taken together, the data in Table 1 indicated that the retention of t subunit during the incubation steps was greatly influenced by the conditions under which it was transferred to the nitrocellulose. Recognition of proteolytic fragments. Distinguishing the epitopes recognized by mono clonal antibodies is frequently done by Westem blots of gels containing samples which have been subjected to limited digestion by proteases or chemical cleavage agents (2 l-23). The decreased relative reactivity of small products is a serious problem in this application because differences between epitopes are more likely to be observed by analyzing smaller fragments rather than, larger ones. The usefulness of the renaturation-carbonate blot procedure in this regard is illustrated in Fig. 5. Fragments of ATPase were produced by either light cleavage by V-8 protease or heavy cleavage by papain. After separation by SDS-PAGE, the fragments were transferred with the carbonate buffer after renaturation (panel A) or with the Tris-glycine buffer without renaturation (panel B) and then analyzed

WESTERN

BLOTS OF RENATURED TABLE

151

PROTEINS

1

INFLUENCEOFBLOTSYSTEMONRETENTIONOF~~~I-LABELED

c BYNITROCELLULOSE

Percentage of dpm in gel retained by nitrocellulose Nitrocellulose brand

No treatment

Alter blocking

After blocking and incubating

Expt

Blot system

1

Tris-Glycine (0.01% SDS)

MS1

47

4

1

2

Tris-Glycine

MS1

83

14

6

3

Carbonate

MS1

85

22

14

4

Carbonate after renaturation

MS1

50 (84)

28 (46)

19 (32)

5

Tris-Glycine

S&S

88

15

6

Carbonate after renaturation

S&S

52 (86)

22 (37)

6 15 (25)

Note. The E. coli membrane extract in SDS sample buffer was supplemented with a trace amount of 1z51-labeled t. Ten micrograms of protein, containing 15,000 dpm, was applied to each lane. Following electrophoresis, a portion of gel was briefly stained to locate the lanes and the band pattern characteristic of the region where t migrates. A l-cmlong portion of each lane centered around e was cut out, placed in a plastic tube, dried, and counted with a gamma counter. In Experiments 4 and 6, the gel was then treated with 50 mM Tris-HCl, pH 7.4, containing 20% glycerol for 1 h, again a portion was stained, and the area around e was cut out, dried, and counted. The remaining portions of the gels were blotted for 1 h using either the Tris-glycine blot buffer or else the carbonate blot buffer. In some cases, Schleicher & &hue11 (S&S) BA-83 nitrocellulose was substituted for tbe MS1 product. Portions of each blot were dried immediately, while others were blocked overnight at 4°C with blot rinse buffer containing 3% BSA, and others were further incubated for 8 h at 22°C in blot rinse buffer containing 0.3% BSA. The sections of nitrocelhtlose were radioautographed to determine the location of the ‘251-labeled c, and then the spots were cut out and counted. Two or three lanes were used for each determination. The numbers in parentheses are percentages relative to that which remained in the gel after the renaturation step. In no case did more than 1% of the radioactivity remain in the gel after the transfer.

for recognition by monoclonal antibody (w-3. Examination of Fig. 5 reveals the stronger recognition after blotting in the carbonate buffer, with the greatest improvement for the smaller fragments. This improvement parallels that obtained for t subunit (see Figs. IB and 2B), supporting the suitability of E subunit as a model for small proteins. DISCUSSION

The results presented above show that treatment of a gel before blotting and modification of the transfer buffer can make an impressive difference in the outcome of the procedure. The cumulative improvement ob-

tained by incubation of the gel in buffered glycerol and transfer in the carbonate buffer is readily visible in Fig. 5. This improvement is apparently due to two separate effects: improved retention of smaller proteins on the nitrocellulose during the incubation steps and enhanced folding of the proteins to structures which more closely resemble the native state. As demonstrated by the ‘2sI-labeled c experiments (Table 1) and the DNP-subunit studies (Fig. 4), the modifications introduced have large effects on the retention of c subunit on the nitrocellulose during the immunological probing procedure. Thus the nature or strength of the binding is influenced by the conditions under which the protein first en-

--._--__

152

Yl-ANLEY

__,~, ,, “A _‘,

-

-_--_

D. DUNN

ognize such assembled determinants5 Inclusion of an incubation step in buffer conducive to renaturation did result in improvements in a-~ monoclonal antibody recognition. The most effective addition to the Tris-HCl buffer was 20% glycerol. Alcohols tended to cause hxation of the proteins in the gel, preventing their transfer. Nonionic detergents were not generally helpful, nor was urea. It is notable that in many of the modifications tested (e.g., isopropanol), the various subunits of ATPase behaved very differently. However, the use of glycerol in the renaturation buffer resulted in enhanced antibody binding to all the subunits, I 2 3 123 suggesting that it will be useful for other proPIG. 5. Improved recognition of proteolytic fragments tein antigens. of (Yafter blotting in carbonate buffer. ATPase at a conAlthough many conditions were tried in centration of I mg/ml was denatured by heating in boiling hopes of improving detection by improving water in the presence of 0.5% SDS. Aliquots were treated with either no protease (lane l), 5’. uureus V-8 protease at renaturation, the effects obtained were small compared to those obtained by switching to 10 &ml (lane 2), or papain at 10 wg/rnl (lane 3). Proteolysis was allowed to proceed for 40 mitt at 37°C and then the carbonate blot buffer. Thus, one might terminated by addition of SDS sample buffer and heating omit the buffered glycerol renaturation step to in boiling water. Ten-microgram samples were applied to compromise sensitivity for speed and simpliclanes of a 12% polyacrylamide gel, and electrophoresed as ity. The relatively alkaline carbonate buffer usual. (A) The gel was incubated with 50 mu Tris-HCl, was originally formulated to allow more effipH 7.4, in 20% glycerol for 1 h and then blotted in carcient transfer of the y subunit, which has an bonate blot buffer. (B) The gel was blotted immediately, using Tris-glycine blot buffer. (Yand fragments containing isoelectric point of about 8.9 (24). Szewcyck the o-3 epitope were detected by incubating the blots with and Kozloff (25) have recently shown that a l/100 dilution of a-3 hybridoma culture medium, foltransfer buffers at pH values higher than 9 do lowed by ‘2sI-labeled rabbit anti-rat IgG. The radioautoindeed allow more efficient transfer of basic graph was exposed for 14 h with an intensifying screen. ribosomal proteins. I found that the recognition of other subunits, which are not basic counters the nitrocellulose. Recognition of proteins, by monoclonal antibodies was also dinitrophenyl groups attached to the larger enhanced by this modification. The large loss of e subunit observed during subunits was the same after blotting in either the blocking and incubation steps may be due the Tris-glycine buffer or the carbonate buffer, in part to inclusion of the nonionic detergent however, suggesting that a difference in retenTween 20 in the buffer used in these steps. Lin tion of protein was not the major factor for and Kasamatsu (9) have demonstrated that the larger subunits. another nonionic detergent, NP-40, can cause The state of folding of proteins must also loss of protein from nitrocellulose. Neverthebe important, as many antibodies to native less, Tween 20 is commonly used by researchproteins appear to recognize assembled toers because it reduces nonspecific binding (26). pographic determinants rather than strictly The detergent may also aid in the removal of sequential determinants. The difficulty we have had in detecting epitopcs for the anti-FlATPase monoclonal antibodies on protein 5 R. G. Tozer and S. D. Dunn, unpublished results. fragments produced by a variety of cleavage These experiments were carried out by competitive enmethods suggests that these antibodies rec- zyme-linked immunosorbent assay.

WESTERN

BLOTS OF RENATURED

any residual SDS from proteins, and thereby help in renaturation of proteins on the nitrocellulose. At the same time, however, if the binding of SDS-protein complexes to the nitrocellulose is mediated by the SDS, then removal of that SDS could lead to destabilization of the binding and loss of protein. In conclusion, use of the carbonate blot buffer, with or without a prior renaturation step, can enhance the sensitivity of Western blots while retaining the simplicity which has made nitrocellulose the favored blotting material. The improvements appear to result from better retention of small proteins and restoration of a more nearly native state for the larger proteins. In addition, the relatively high pH ensures good transfer of moderately basic proteins such as the y subunit. ACKNOWLEDGMENTS I thank Dr. Tony DAmore, Dr. Eric Ball, Dr. George Cates, and Mr. Richard Tozer for helpful discussions and Varden Zadorozny for technical assistance.

REFERENCES 1. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76,43X)-4354. 2. Burnette, W. N. (1981) Anal. B&hem. 112, 195-203. 3. Gershoni, J. M., and Palade, G. E. (1983) Anal. Biochem. 131, l-15. 4. Towbin, H., and Gordon, J. (1984) J. Immunol. Methods 72,3 13-340. 5. Benjamin, D. C., Berzofsky, J. A., East, I. J., Curd, F. R. N., Hannum, C., Leach, S. J., Margoliash, E., Michael, J. G., Miller, A., Prager, E. M., Reichlin, M., Sercarz, E. E., Smith-GiJl, S. J., Todd, P. E., and Wilson, A. C. (1984) Annu. Rev. Immunol. 2,67-101.

6.

7. 8. 9. 10. 11. 12. 13.

14.

PROTEINS

153

Erickson, P. F., Minier, L. N., and Lasher, R. S. (1982) J. Immunol. Methods 51,241-249. Nielsen, P. J., Manchester, K. L., Towbin, H., Gordon, J., and Thomas, G. (1982) J. Biol. Chem. 257, 12,316-12,321. Dunn, S. D., Tozer, R. G., Antczak, D. F., and Heppel, L. A. (1985) J. Biol. Chem. 260, 10,418-10,425. Lin, W., and Kasamatsu, H. (1983) Anal. B&hem. 128,302-3 11. Downie, J. A., Langman, L., Cox, G. B., Yanofsky, C., and Gibson, F. (1980) J. Bacterial. 143, 8-17. Dunn, S. D. (1982) J. Biol. Chem. 257, 7354-7359. Fraker, F. J., and Speck, J. C. (1978) Biochem. Biophys. Res. Commun. 80,849-857. Dixon, H. B. F., and Fields, R. (1972) in Methods Enzymology (Hits, C. H. N., and Timasheff, S. N., eds.) Vol. 25, pp. 409-419, Academic Press, New York. Laemmli, U. K. (1970) Nature (London) 227, 680685.

15. Han&k, K., and Tsang, V. C. W. (1983) Anal. Biochem. 133, 157-162. 16. Manrow, R. E., and Dottin, R. P. (1982) Anal. B&hem. 120, 181-188. 17. Copeland, B. R., Richter, R. J., and Furlong, C. E. (1982) J. Biof. Chem. 257, 15,065-15,071. 18. Blank, A., Sugiyama, R. H., and Dekker, C. A. ( 1982) Anal. Biochem. 120,267-275. 19. Mandre.11, R. E., and Zollinger, W. D. ( 1984) J. Immunol. Methods 67, I- 11. 20. Bowen, B., Steinberg, J., Laemmli, U. K., and Weintraub, H. (1980) Nucleic Acids Res. 8, l-20. 21. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, U. K. (1977) J. Biol. Chem. 152, 1102-l 106. 22. Yurchenco, P. D., Speicher, D. W., Morrow, J. S., Knowles, W. J., and Matchesi, T. (1982) J. Biol. Chem. 257,9102-9107. 23. Clark, W. A. (1983) in Bio-Radiations, No. 45EG, p. 1, Bio-Rad Laboratories, Richmond, Calif. 24. Fayle, D. R. H. (1978) Ph. D. thesis, Australian National University. 25. Szewcyck, B., and Kozloff, L. M. (1985) Anal. B&hem. 150,403-407. 26. DeBlas, A. L., and Cherwinski, H. M. (1983) Anal. Biochem. 133,2 14-2 19.