GLOSARIO DE CONCEPTOS

These structures have shed much light on the binding mode of substrates and their interactions with the active site and the changes induced in both enzyme and substrate upon binding. The nature of the changes in the structures of the aspartic proteases on binding of inhibitors (and by extension substrates) and the relation of these changes to catalysis was for some time not well understood. The first high-

Figure 1.4.3: The overall fold of the monomeric aspartic proteases

resolution structures with bound inhibitors showed several interesting features of the binding. There is a stretch of antiparallel p-sheet between residues P3 and of the inhibitor and residues 217-219 of the C-terminal lobe of the enzyme. There is no equivalent structure in the N-terminal lobe, with the only hydrogen bond formed between P2’ NH and the carbonyl of residue 34. Also observed was the large movement of a region known as the f l a p .‘*24 This 3-hairpin bend structure (residues 72-81 in rhizopuspepsin) folds over the active site, shielding the active site from solvent and providing further binding interactions to the inhibitor. The P^ and P-j’ residues of the inhibitor are totally shielded from solvent as a result of this movement.^9 The main hydrogen bonding interactions between a peptidic inhibitor and the monomeric aspartic proteases are shown in Figure 1.4.4 below.

Thr 219 Gly 217 Gly 34

Asp 77 Gly 76 Ser 74

Figure 1.4.4: Hydrogen bonding pattern between the main-chain of an inhibitor and endothiapepsin'^^'^

where W3 is a water molecule.

The extended conformation of the inhibitor seen in the X-ray structures above is reflected in inhibitor conformations observed in n.m.r. studies on inhibitors bound to p e p s i n . T h e importance of the flap can clearly be seen from the above diagram as residues D-77, G-76 and 8-74 are all from the flap and provide three strong

hydrogen bonds to the inhibitor main chain. The formation of these hydrogen bonds provides part of the driving force for flap closure. The amount of movement of the flap observed on inhibitor binding varies greatly between structures. The differences arise from the different initial flap positions in the unliganded proteases, caused by different packing interactions between enzyme molecules in the crystal. The flap is observed to be highly mobile in the unliganded structures (B value around 30 Â^), becoming much more ordered on closing over the active site (B value about 8 Â2).126 The wide-ranging nature of the conformational changes is also seen in circular dichroism studies of pepsin interacting with pepstatin.

There are other structural changes observed in some structures, apart from the closure of the flap. There have been reports of independent, rigid body motion of subdomains within the structure that are triggered by inhibitor binding. The movement of residues 190-303 of endothiapepsin when complexed with a reduced peptide inhibitor observed by Sali et al.^^ was the first reported example. The same domain in pepsin was also reported to move as one rigid body, but in a very different fashion, by Abad-Zapatero etaL'^^^ Later studies by the same group on glycol-based inhibitors of p e p s in ^ sh o w a similar domain movement to that found by Sali etaL^^ This domain can be easily seen in the comparison of X-ray structures of unliganded with liganded proteases^^® where it is the point of most divergence between the different structures.

This flexible domain aligns well between various structures only if it is allowed to move independently of the rest of the protein. This division of the structure accords with the tripartite division of the aspartic protease structure discussed above, as the C-terminal residues 190-303 correspond to one rigid body and the other rigid body corresponds to the central motif and the N-terminal domain. The triggering of a rigid body movement was referred to above (Section 1.3.5) as the possible rate-limiting step in catalysis. This rigid body movement may provide the means for scissile bond distortion postulated by many of the mechanisms discussed above. The residues of the flexible domain are used to bind side-chains P^’, P2 and P4, and the otber domain binds the other side-chains. Therefore, movement of the N-terminal domain relative to the C-terminal domain, will distort one set of side-chains relative to the other. This may distort the scissile bond from planarity as the P^ and P-j’ residues

flanking the scissile bond are bound by the different domains. Curiously, a corresponding change in enzyme fluorescence is not o b s e r v e d , ^ a s would be expected for such a structural readjustment. A rigid body movement has also been postulated in the mechanism of the serine proteases.

There seems to be little effect of inhibitor binding on the structure of the active site region, the aspartate residues remaining almost completely fixed. This is consistent with the rigid active site, reflected in the low B values and extensive hydrogen bonding, discussed above. The only movement seen is some disruption of the coplanarity of the aspartates.®^