ble for their higher thermodynamic stability and mechanical resistance compared to the A2 domain. In this study, the effects of type 2A mutations located near the C-terminal helix of the protein were investigated by a combination of molecular dynamics simulations and cleavage experiments. Mutations were excluded which introduced an obvious disruption of the native state, i.e., mutating a hydrophobic residue into a charged side chain or vice versa. In total, three single point mutations linked to type 2A von Willebrand Disease, L1657I, I1628T and E1638K, were selected for this study. Molecular dynamics analysis was used to characterize the effect of mutations onto the structural stability of the A2 domain with and without applied tensile force. The computational results were then validated against an ADAMTS13-induced cleavage assay 
using mutagenesis and protein engineering. A thorough structural understanding of the regulation of the A2 domain is essential to guide structure-based computational drug design and discover novel therapeutic molecules to treat von Willebrand disease or thrombogenic illnesses. Results Analysis of the native state In order to understand whether type 2A mutations alter the kinetic stability of the native state of the A2 domain, room temperature simulations were run with the wild-type and the three single point NP-031112 chemical information mutants L1657I, I1628T and E1638K. All three mutations are clinically known to cause type 2A von Willebrand disease. In total, 12 simulations were run at 300 K, three for each mutant and the wild-type. The simulations were first compared with the available crystallographic data in order to check for convergence. The Ca root mean square deviation from the initial conformation remained below 2 A during the course of all simulations at 300 K with the wild-type and the mutants. This indicates that the mutations did not significantly alter the overall fold of the A2 domain during the time course of the room temperature simulations. Also, the Ca RMSD of helix a5 and helix a6 remained generally below 1.5 A in all simulations. This is a further indication that secondary structure elements are likely conserved in the structure of the mutants. Most of the total Ca RMSD is probably accounted for by the motion of the a4 -less loop and the 3/10 helix between a3 and b4. In fact, these regions fluctuate in the simulations more than the rest of the protein as indicated by the larger Ca root mean square fluctuations in 2 Structural Basis of Type 2A VWD Type 2Aa Tensile forceb Name WT_1,2,3 c Starting structure Wild-type, PDB code 3GXB L1657I I1628T E1638K WT_1, WT_2 d L1657I_1, L1657I_2 d I1628T_1, I1628T_2 d E1638K_1, E1638K_2 d F1520A I1651A A1661G F1520A_1, F1520A_2 d I1651A_1, I1651A_2 d c c Duration 3640 L1657I_1,2,3c I1628T_1,2,3c E1638K_1,2,3 WT_pull_1,2,3c L1657I_pull_1,2,3c I1628T_pull_1,2,3c E1638K_pull_1,2,3 F1520A_1,2,3c I1651A_1,2,3c A1661G_1,2,3c F1520A_pull_1,2,3c I1651A_pull_1,2,3 c X X X X X X X X X X 3640 3640 3640 3655 3625 3625 3625 3640 3640 3640 X X X X X X 3625 3625 3625 A1661G_pull_1,2,3c A1661G_1, A1661G_2 d a The mutation induces type 2A VWD. b Tensile force was applied by pulling the C-terminus at constant velocity from the PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/22210737 N-terminus. c Three simulations were started for the wild-type and each mutant with and without applied tensile force; they are labeled with 1, 2 and 3, respectively. d The pulling runs were started from snapshots sampled during the simulations with no tensile force, more spec
