Figure 1

Interaction between DnaK and immobilized preproinsulin peptides. DnaK was allowed to bind to a cellulose membrane carrying 50 spots of overlapping 13-mer preproinsulin peptides and three spots of control peptides (C1, C2, C3) followed by electroelution and identification of eluted DnaK by specific antibody and chemiluminescence. A: The grey-scales of the spots correspond to the amounts of DnaK released from the peptide spots during the first elution cycle. B: Total DnaK released after the first (open bars), second (hatched bars) and third (solid bars) elution cycle, quantified by densitometry. The data are given in arbitrary luminescence units (LU). C: Probability scores for the binding of DnaK to preproinsulin peptides (and control peptides) as calculated by an algorithm designed to predict the affinity of 13-mer peptides to DnaK. Negative scores denote higher DnaK binding probabilities, positive scores denote low/no DnaK binding. A cut-off score of -5 would correctly predict 95% of the good binding sites. With scores above -5 the reliability of the prediction of binding decreases.