Erratum

Alternatively, a fraction of our class-I molecules might dissociate during the 4°C storage of the class-I preparation prior to the experiment and we do find evidence for this when examining the class-I preparations by SDS-PAGE before and after G50 spun-column chromatography (data not shown). Interpretation ofthe kinetics data becomes even more complicated when specific peptide is taken into consideration. It has been suggested that both fiim association and dissociation depends on the affinity of peptide binding [12, 13, 38, 41, 50]. We have not been able to show such peptide effects, but most likely our MHC class-I preparations have bound endogenous high affinity peptides obscuring the effects of added peptides.