|
T-helper-cell determinants in protein antigens are preferentially located in cysteine-rich antigen segments resistant to proteolytic cleavage by cathepsin B, L, and D.
We report on a computer algorithm capable of predicting the location of T-helper-cell epitopes in protein antigen (Ag) by analysing the Ag amino acid sequence. The algorithm was constructed with the aim of identifying segments in Ag which are resistant to proteolytic degradation by the enzymes cathepsin B, L, and D. These are prominent enzymes in the endocytic pathway through which soluble protein Ag enter APC, and resistant segments in Ag may, therefore, be expected to contain more T-cell determinants than susceptible segments. From information available in the literature on the substrate specificity of the three enzymes, it is clear that a cysteine is not accepted in any of the S2, S1, S1', and S2' subsites of cathepsin B and L, and not in the S1 and S1' subsites of cathepsin D. Moreover, we have noticed that cysteine-containing T-cell determinants in a number of protein Ag are particularly rich in the amino acids alanine, glycine, lysine, leucine, serine, threonine, and valine. By searching protein Ag for clusters of amino acids containing cysteine and two of the other amino acids we were able to predict 17 out of 23 empirically known T-cell determinants in the Ag with a relatively low number of false (positive) predictions. Furthermore, we present a new principle for searching Ag for potential amphipatic alpha-helical protein segments. Such segments accord well with empirically known T-cell determinants and our algorithm produces a lower number of false positive predictions than the principle based on discrete Fourier transformations previously described.
|
