Davis replies:

In our recent Nature Immunology paper1 we suggested that after contact with peptide–major histocompatibility complex ligands presented by a B cell, “removal of activated TCRs from the contact interface had no appreciable effect on the amount of synapse-associated PI3K activity.” Finkel disputes this conclusion because CD3ζ can dissociate from the TCR in certain situations2.

Although CD3ζ dissociation is a legitimate concern, it seems unlikely that this is an important factor in our study for two reasons. First, in an earlier paper we found that CD3ζ-GFP expression in a T cell line colocalized with TCR staining3 and thus it seems reasonable in most circumstances to use CD3ζ and TCRs interchangeably. A second and stronger point is that in this manuscript, we showed by flow cytometry that approximately 70% of surface-localized TCR was internalized 30 minutes after initial B cell contact. This is in excellent agreement with the pool sizes of internalized and cell surface–localized CD3ζ-CFP that we quantified at this time point. Finkel notes that our flow cytometry results indicate a loss of surface TCR, but implies that this might not be due to internalization at the synapse. Only at the synapse, however, would TCRs contact their stimulatory ligands and be tagged for internalization and degradation4. Thus we have seen no evidence for any substantial separation of TCR from CD3ζ in the T cells that we have studied.