Distinct protease requirements for antigen presentation in vitro and in vivo.

Asparagine endopeptidase (AEP) or legumain is a potentially important Ag-processing enzyme that introduces limited cleavages that trigger unfolding and class II MHC binding of different Ag substrates. AEP is necessary and sufficient for optimal processing and presentation of the tetanus toxin C fragment (TTCF) Ag in vitro, but its importance has not been tested in vivo. Surprisingly, virtually normal T cell and Ab responses to TTCF were mounted in AEP-deficient mice when examined 10 d after immunization. This was the case when TTCF was emulsified with CFA, adsorbed onto alum, or expressed within live Salmonella typhimurium. In addition, the dominant Ab and T cell determinants recognized in TTCF were essentially unchanged in AEP-deficient mice. These data are explained, at least in part, by the much lower levels of AEP expressed in primary murine APCs compared with immortalized B cell lines. Even so, the initial in vivo kinetics of TTCF presentation were slower in AEP-deficient mice and, as expected, boosting AEP levels in primary APCs enhanced and accelerated TTCF processing and presentation in vitro. Thus, AEP remains the protease of choice for TTCF processing; however, in its absence, other enzymes can substitute to enable slower, but equally robust, adaptive immune responses. Moreover, clear relationships between Ags and processing proteases identified from short-term in vitro processing and presentation studies do not necessarily predict an absolute in vivo dependency on those processing enzymes, not least because they may be expressed at strikingly different levels in vitro versus in vivo.