A protease draws first blood

8 intact protein antigens. But, presumably, lysosomal proteases must work in concert to deliver a carefully orchestrated series of cuts if the correct peptide is to be generated. Antigen-presenting cells from knockout mice lacking cathepsins D, B or L show no obvious defects in antigen presentation, indicating that other proteases are involved in both removing Ii and converting the antigen into peptides. One might assume, perhaps naively, that many lysosomal proteases are functionally redundant in view of their broad substrate specificities. Indeed — and consistent with such redundancy — deficiencies in lysosomal proteases are quite rare. ToulouseLautrec’s short stature was attributed to a deficiency in cathepsin K, a minor lysosomal protease involved in bone remodelling. But of the cathepsin knockouts studied so far, only animals lacking cathepsin D are sick, dying at around three weeks of age. Yet the fact that some lysosomal proteases have a restricted tissue distribution (such as cathepsin S, which is involved in proteolytic removal of Ii in professional antigenpresenting cells) raises the ante for the identification of further proteases in antigen presentation. Consequently, the search has been on to implicate new proteases, many of them lurking in databases as cathepsin homologues. One of these has now been identified by Manoury et al. in B cells as the cysteine protease AEP. A direct relative of the legumains, which were first discovered in plants, AEP is likely to be identical to the recently cloned legumain homologue from mammalian tissue. The authors believe that AEP draws first blood and executes initial digestion of a protein antigen, tetanus toxin. It could help to unfold its target antigen to allow further digestion by the cathepsins. To work out how a specific protease is involved in antigen processing, we need to be able to modulate its activity selectively. Inhibitors — often modified peptides that mimic an enzyme’s substrates — transiently or permanently modify the active-site nucleophile, leading to a loss of activity. Manoury and co-workers used high concentrations of natural peptide substrates to block the activity of AEP in vitro and in vivo. But although the inhibitors were active against purified AEP, showing little or no effect on other cysteine proteases of the cathepsin family, caution is in order when transposing in vitro inhibitor studies to living cells. Many lysosomal cathepsins have similar substrate specificities in vitro, and inhibitors that are specific for single proteases are hard to find. So, small-molecule inhibitors that not only block activity, but also allow the molecular targets to be identified, are high on the wish-list. It is anyone’s guess how many proteases with functional properties news and views