Journal:Acta Cryst D:S0907444911047251

Flexibility of the Flap in the Active Site of BACE1 as Revealed by Crystal Structures and MD Simulations

Yechun Xu, Minjun Li, Harry Greenblatt, Wuyan Chen, Aviv Paz, Orly Dym, Yoav Peleg, Tiantian Chen, Xu Shen, Jianhua He, Hualiang Jiang, Israel Silman & Joel L Sussman^[1]

Molecular Tour

is a transmembrane aspartic protease that cleaves the β-amyloid precursor protein en route to generation of the amyloid β-peptide (Aβ) that is believed to be the principal component of the amyloid plaques seen in the brains of Alzheimer’s disease. It is thus a prime target for the development of inhibitors that may serve as drugs in the treatment and/or prevention of Alzheimer’s disease. The protease domain has a conserved aspartic protease fold, with the (colored cyan and yellow, respectively). The hairpin loop in the N-terminal lobe that is known as the , as is in general the case for the aspartic protease family. Determination of the crystal structures of both apo and complexed BACE1, structural analysis of all crystal structures of BACE1 deposited in the PDB, and MD simulations of and BACE1 were used to study conformational changes in the active-site region of the enzyme (the first monomer in magenta with a red flap, the second one in green with an orange flap). Four new crystal structures were determined: that of a P6₁22 (3tpr, BACE1 in magenta and in cyan), those of a (3tpp, BACE1 in indigo and in green) and of an (3tpj, in orange), both belonging to space group C222₁, and that of a (3tpl, in blueviolet). P6₁22 (3tpr, BSIIV/BACE1) were obtained by , while those of the complex belonging to space group C222₁ were obtained by (3tpp, BSIIV/BACE1). The two apo structures belonging to different space groups were obtained by cocrystallization in the presence of another low affinity inhibitor, using two different precipitants; in both cases the inhibitor was not seen in the crystals thus obtained. It was found that the , as do the , particularly in the . This led to a systematic analysis of all the BACE1 crystal structures deposited in the PDB (both apo structures and complexes), and to perform MD simulations on the apo enzyme, so as to investigate the underlying causes for the various conformations adopted by the flap. In particular, an attempt was made to clarify whether they might be a consequence of crystal packing in different space groups, or might indicate that the active site can assume multiple low-energy conformations. The RMSFs of main-chain atoms, calculated based on all the BACE1 crystal structures deposited in the PDB, show that the . The , reveals that the (e.g. 1w50, colored deepskyblue) with two exceptions; not surprisingly, the situation with respect to the structures of the complexes is more complex. (e.g. 3tpp, BSIIV/BACE1), except in a few cases in which the position and orientation of the inhibitors sterically preclude its closing. (3tpr, BSIIV/BACE1); but a direct H-bond between Q73 and BSIIV locks the flap in a closed conformation in the co-crystallized structure (3tpp, BSIIV/BACE1). It may thus be concluded that formation of a direct H-bond between a main-chain atom of a flap residue and the bound inhibitor is a prerequisite for the flap to assume a closed conformation in a complex. Interestingly, P6₁22 except one, the flap is in an open conformation. Our soaked crystals of the BSIIV/BACE1 complex also belong to space group P6₁22. Analysis of the crystal packing pattern reveals that in this space group the flap is at a packing interface. MD simulations were carried out to check if the observed packing would influence the conformation of the flap. In the starting model for MD that includes a symmetry-related copy of BACE1, both the degree of opening and the time that the flap dwells in the open state are much larger than in the trajectories generated for the monomer. Thus, the packing pattern in crystals belonging to space group P6₁22 hinders the conformational change in the flap, which may be the main reason why in our BSIIV/BACE1 complex (3tpr) belonging to space group P6₁22 the C222₁ (3tpp, BSIIV/BACE1). Thus a systematic study of the conformational flexibility of the enzyme may not only contribute to structure-based drug design of BACE1 inhibitors and of other targets with flexible conformations, but also help to better understand the mechanistic events associated with the binding of substrates and inhibitors to the enzyme.