amide gel was prepared according to Laemmli (4). Sample buffer containing 0.0625 M Tris-HCl (pH 6.8), 2% (w/v) SDS, 10% (v/v) glycerol, 5% (v/v) 2-mercaptoethanol, and 0.001% (w/v) bromophenol blue was loaded into the wells of the gel and allowed to stand for 5 min. Subsequently, electrophoresis was performed in the reverse direction at a constant current of 40 mA for 5 min to recover the sample buffer, and this sample buffer was removed from the wells. Fresh sample buffer and protein standards were loaded in the wells of the gel, and then SDS-PAGE was performed using the Laemmli method (4) at a constant current of 40 mA. The gel was silver stained according to the instructions of the manufacturer.
Results and discussion. Sample buffer was loaded only into the odd-lane wells of a polyacrylamide gel. The gel was treated as described above. The result is shown in Fig. 1. Artifactual bands were observed in the even lanes, whereas few artifactual bands were observed in the odd lanes (Fig. 1). Artifactual bands were also detected in the gel used to analyze the sample buffer collected from the odd lanes after reverse electrophoresis (Fig. 2). These results indicate that the artifactual bands were derived from the polyacrylamide gel, not from sample buffer or electrode buffer, and could be eliminated effectively by this pretreatment method. The electrophoresis pattern of the protein standards also indicates that electrophoresis in the reverse direction does not affect the quality of the subsequent electrophoresis separation.
Judging from the molecular weight of the artifactual bands, they are probably skin keratins (1-3). No artifactual bands were detected in the individual reagents used for making polyacrylamide gels (data not shown), suggesting that contamination by the skin keratins occurs during the preparation of polyacrylamide gels.