| The ubiquitin-proteasome pathway is the major proteolytic system responsible for the selective degradation of proteins and housekeeping functions necessary for normal cellular physiology. The proteasome recognizes and proteolytically cleaves ubiquitin-tagged proteins. Its products are then further processed into shorter peptides by aminopeptidases downstream of the proteasome. One such aminopeptidase is the largest known eukaryotic protease, tripeptidyl peptidase II (TPP II)- an unusually large macromolecular complex (∼ 6 MDa) with a unique spindle-like shape.;A serine protease of the subtilisin family, TPP II cleaves three residue segments off the N-termini of small peptides. The 150 kDa subunits of Drosophila TPP II stack into two segmented and twisted strands of defined length in vivo although its assembly has no obvious point of termination since its size appears to vary in vitro. For a mechanistic understanding of the remarkable architecture of TPP II, a high-resolution structure of the fully assembled complex is indispensable. However, the structural heterogeneity of the rather unstable holocomplex has greatly hindered successful high-resolution structure determination efforts. To overcome this difficulty, we have applied a hybrid approach: the high-resolution structure of the dimer solved by x-ray crystallography was docked into the spindle envelope obtained from single-particle cryo-electron microscopy (cryoEM) to generate a pseudo-atomic model of the fully assembled TPP II complex.;The crystal structure of the inactive dimer shows the binding of a highly conserved and flexible loop to the substrate binding cleft and the displacement of the active-site serine ∼5 A away from where it would be expected. The presence of two glutamates blocking off one end of the binding cleft likely orients substrates and acts as a "molecular ruler" determining the size of the cleavage products.;The resulting model of the holocomplex reveals a network of chambers, which compartmentalize the active sites and prevent unintended proteolysis. Furthermore, unaccounted for EM density and supporting point mutation studies suggests a repositioning of the active-site serine in the fully assembled TPP II spindle as well as a model for the activation of TPP II involving a flexible loop with a key role as an assembly-dependent activation switch. |
