| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
LETTERS TO THE EDITOR |
Institut für Medizinische Immunologie Charité, Universitätsmedizin Berlin, Campus Charité Mitte, Berlin, Germany
In a recently published article (1) we criticized the way Elkington et al. used MHC/peptide-binding predictions to assign presenting HLA molecules to CMV-derived T cell-stimulating peptides in a publication in 2003 (2). Experimental binding data presented in the study argued against some of the listed associations of peptides and HLA-A*0201 (or HLA-A2, since only low-resolution typing was used for the donors). Moreover, the responses to some of these peptides in HLA-A2-positive donors were infrequent and small. In his reply, the senior coauthor of the publication argues that a peptide can be presented by HLA-A*0201 despite the fact that it fails to stabilize this molecule on T2 cells. To show this, however, he refers to the peptide, AVGGAVASV, also discussed in the same article (2), which had been predicted to bind to HLA-A*0201, but failed to stabilize this molecule on T2 cells. Nevertheless, it sensitized HLA-A2 positive PHA-blasts for lysis by an HLA-A2-positive T cell line (2). Clearly, CTL assays can be very helpful in determining the presenting HLA molecule of a given stimulating peptide, but only if the target cells are chosen in such a way that they share only certain HLA molecules with the effector cells, but not all. Meanwhile, according to the quoted publication, the PHA-blasts used as target cells, the T cell line, and the LCLs used to expand this T cell line, all originated from the same donor, i.e., they had an identical set of HLA molecules (see Ref.2 , Fig. 2). Consequently, assigning HLA-A2 as the presenting HLA molecule was not possible based on these experiments. In any case, our reference to this work was solely intended to caution against the uncritical use of predictions when assigning peptides to presenting HLA molecules, not to be depreciative of it as such.
We fully agree with R. Khanna that the absence of T cell responses to selected peptides from a protein does not prove this protein is not a T cell target, and our discussion should not have conveyed this impression. The strategy for identifying T cell epitopes used by both Elkington et al. (2) and us may be successful in some situations, but not in others.
In contrast, competition of peptides used at equimolar concentrations within the same pool has not been shown as a cause for the absence of T cell responses to a stimulating peptide, at least not for this type of assay. In regard to CTL assays, however, it is known that a massive excess of optimum binders is required to fully out-compete a recognized epitope (3).
The figure accompanying the letter by R. Khanna correlates the predictive "score" for MHC/peptide-binding with the measured increase in fluorescence intensity in the HLA-A*0201-stabilization assay. The presence of positive responses in ELISPOT/CTL assays, meanwhile, does nothing but indicate that some of the peptides were stimulating. If designed in the same way as in the criticized publication, these assays say little about the presenting allomorphs for these peptides.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
