Six designs (OR25, OR26, OR27, OR28, OR29, OR30) for 2D [ 1 H, 15 N] HSQC screening

1. Red: Design Green: NMR, PDB 2KI8 RMSD of C 0.99Å H2 L1 F44 L41 V31 L37 Y3 L5 I7 V4 L2 I14 L28 I6 H1 NMR:green, PDB 2KI8 Design:red L69 H3 I75 A70 A66 I51 V57 L63 I55 V53 V82 L54 L56 F95 A88 H4 I92 β4 • NMR Structures of De Novo Designed “Ideal Structure” Proteins β3 β1 β1 Gaohua Liu1, Nobuyasu Koga2, Rie Koga2, Rong Xiao1, Haleema Janjua1, Keith Hamilton1, Thomas Acton1, John Everett1, David Baker2, Gaetano T. Montelione1 1Department of Molecular Biology and Biochemistry, Northeast Structural Genomics Consortium, Rutgers, The State University of New Jersey, Piscataway, NJ 08854; 2Department of Biochemistry and Howard Hughes Medical Institute, University of Washington, Seattle, WA 98195 II. NMR Screening and Structures • Background i) OR15,designed Ferredoxin-like protein, agrees with design ii) OR16, designed Rossmann2x2 fold protein,agrees with design Structural characterization of designed proteins is a critical (and often neglected) step in validating computational design methodology. Many of the groups involved in computational protein design have limited resources for 3D structure determination, and structural genomics platforms are ideally suited for collaborative projects aimed at accelerating the field. Designed proteins are often relatively small, making them especially well suited to NMR structure determination. Computational design also often yields imperfect core packing (and marginal thermal stability), which may prevent crystallization, and also rendering NMR structure determination challenging due to exchange broadening effects. None the less, NMR has traditionally been invaluable for characterizing the structures of designed proteins (Kuhlman et al., 2003). • Red: Design • Green: NMR, PDB 2KPO • RMSD of C1.06Å H2 H1 H3 H4 NMR NMR • iii) OR28 and OR36 (not shown), designed Rossmann2x3 fold proteins, disagree with design • Six designs (OR25, OR26, OR27, OR28, OR29, OR30) for 2D [1H, 15N] HSQC screening • OR28 for structure determination • OR28 NMR structure turns out to be P-loop NTPase fold structure that has two  strands swapped compared to designed model. Design OR25 OR26 OR27 CS-Rosetta • Purpose • Establish rational methods to design structures de novo. • Reveal principles of how amino acid sequence determines native 3D-structure • Utilize designed structural scaffolds to introduce functional sites • Targets • Four different folds were targeted • Methods • Protein candidates with different primary sequences were computational designed and pre-selected based on computational energy at University of Washington. • Unlabeled or 15N labeled protein samples were prepared for selected protein candidates and were further screened by NESG at Rutgers using 1D NMR or 2D [15N-1H] - HSQC. • Suitable protein candidates were then selected for structure determination by NMR or/and X-ray. RPF Precision Map β1 & β3 swapped NMR OR29 OR28 • Red: Design • Green: NMR, PDB 2L69 • RMSD of C~3.4Å OR30, no expression Rossmann2x3 (Flavodoxin-like) NMR Ferredoxin-like Rossmann2x2 Rossmann3x3 iv) OR32, designed Rossmann3x3 fold protein,disagrees with design • OR32 NMR structure turns out to be P-loop NTPase fold structure that has two  strands swapped compared to designed model. • Ten unlabeled samples (NSM1-10) for 1D 1H NMR screening • NSM5(OR31) and NSM10(OR32) for 2D [1H,15N] HSQC screening • OR32 for structure determination • CS-Rosetta Design OR15, PDB 2kl8 OR16, 2kpo OR28, 2l69 OR36, 2lci OR32, 2l82 NSM1 NSM2 NSM3 NSM4 RPF Precision Map β1 & β4 swapped NSM5 NSM6 NSM7 NSM8 • Red: Design • Green: NMR, PDB 2L82 • RMSD of C~2.9Å • NMR OR32 OR31 NSM9 NSM10 NMR • Summary • To date, solution NMR structures have been determined for five targets of four folds. The Ferredoxin-like protein (NESG ID OR15); Rossmann 2x2 fold protein OR16; Flavodoxin-like proteins OR28 and OR36, Rossmann 3x3 fold protein OR32. The experimental NMR structures of OR15 and OR16 are in excellent agreement with their designed models. However, structures of the three proteins OR28, OR36 and OR32 turn out to be P-loop NTPase fold structures that have two -strands swapped compared to designed models. These NMR experimental structures provide unique valuable information on how to improve the protein design strategies. Reference: Kuhlman, B., Dantas, G., Ireton, G.C., Varani, G., Stoddard, B.L., and Baker, D. (2003). Design of a novel globular protein fold with atomic-level accuracy. Science 302, 1364-1368.