Computational design of a symmetric homodimer using β-strand assembly

Computational design of novel protein–protein interfaces is a test of our understanding of protein interactions and has the potential to allow modification of cellular physiology. Methods for designing high-affinity interactions that adopt a predetermined binding mode have proved elusive, suggesting the need for new strategies that simplify the design process. A solvent-exposed backbone on a β-strand is thought of as “sticky” and β-strand pairing stabilizes many naturally occurring protein complexes. Here, we computationally redesign a monomeric protein to form a symmetric homodimer by pairing exposed β-strands to form an intermolecular β-sheet. A crystal structure of the designed complex closely matches the computational model (rmsd = 1.0 Å). This work demonstrates that β-strand pairing can be used to computationally design new interactions with high accuracy.     

Transition states of native and drug-resistant HIV-1 protease are the same

HIV-1 protease is an important target for the treatment of HIV/AIDS. However, drug resistance is a persistent problem and new inhibitors are needed. An approach toward understanding enzyme chemistry, the basis of drug resistance, and the design of powerful inhibitors is to establish the structure of enzymatic transition states. Enzymatic transition structures can be established by matching experimental kinetic isotope effects (KIEs) with theoretical predictions. However, the HIV-1 protease transition state has not been previously resolved using these methods. We have measured primary ¹⁴C and ¹⁵N KIEs and secondary ³H and ¹⁸O KIEs for native and multidrug-resistant HIV-1 protease (I84V). We observed ¹⁴C KIEs (¹⁴V/K) of 1.029 ± 0.003 and 1.025 ± 0.005, ¹⁵N KIEs (¹⁵V/K) of 0.987 ± 0.004 and 0.989 ± 0.003, ¹⁸O KIEs (¹⁸V/K) of 0.999 ± 0.003 and 0.993 ± 0.003, and ³H KIEs (³V/K) KIEs of 0.968 ± 0.001 and 0.976 ± 0.001 for the native and I84V enzyme, respectively. The chemical reaction involves nucleophilic water attack at the carbonyl carbon, proton transfer to the amide nitrogen leaving group, and C-N bond cleavage. A transition structure consistent with the KIE values involves proton transfer from the active site Asp-125 (1.32 Å) with partial hydrogen bond formation to the accepting nitrogen (1.20 Å) and partial bond loss from the carbonyl carbon to the amide leaving group (1.52 Å). The KIEs measured for the native and I84V enzyme indicate nearly identical transition states, implying that a true transition-state analogue should be effective against both enzymes.     

Folding helical proteins in explicit solvent using dihedral-biased tempering

Using a single-trajectory-based tempering method with a high-temperature dihedral bias, we repeatedly folded four helical proteins [α₃D (PDB ID: 2A3D, 73 residues), α₃W (1LQ7, 67 residues), Fap1-NR_α (2KUB, 81 residues) and S-836 (2JUA, 102 residues)] and some of the mutants in explicit solvent within several microseconds. The lowest root-mean-square deviations of backbone atoms from the experimentally determined structures were 1.9, 1.4, 1.0, and 2.1 Å, respectively. Cluster analyses of folding trajectories showed the native conformation usually occupied the most populated cluster. The simulation protocol can be applied to large-scale simulations of other helical proteins on commonly accessible computing platforms.     

Structural dependence of HET-s amyloid fibril infectivity assessed by cryoelectron microscopy

HET-s is a prion protein of the fungus Podospora anserina which, in the prion state, is active in a self/nonself recognition process called heterokaryon incompatibility. Its prionogenic properties reside in the C-terminal “prion domain.” The HET-s prion domain polymerizes in vitro into amyloid fibrils whose properties depend on the pH of assembly; above pH 3, infectious singlet fibrils are produced, and below pH 3, noninfectious triplet fibrils. To investigate the correlation between structure and infectivity, we performed cryo-EM analyses. Singlet fibrils have a helical pitch of approximately 410 Å and a left-handed twist. Triplet fibrils have three protofibrils whose lateral dimensions (36 × 25 Å) and axial packing (one subunit per 9.4 Å) match those of singlets but differ in their supercoiling. At 8.5-Å resolution, the cross-section of the singlet fibril reconstruction is largely consistent with that of a β-solenoid model previously determined by solid-state NMR. Reconstructions of the triplet fibrils show three protofibrils coiling around a common axis and packed less tightly at pH 3 than at pH 2, eventually peeling off. Taken together with the earlier observation that fragmentation of triplet fibrils by sonication does not increase infectivity, these observations suggest a novel mechanism for self-propagation, whereby daughter fibrils nucleate on the lateral surface of singlet fibrils. In triplets, this surface is occluded, blocking nucleation and thereby explaining their lack of infectivity.     

Structural basis of coactivation of liver receptor homolog-1 by β-catenin

We report the three-dimensional structure of a β-catenin armadillo repeat in complex with the liver receptor homolog-1 (LRH-1) ligand binding domain at 2.8 Å resolution as the first structure of β-catenin in complex with any nuclear receptor. The surface of β-catenin that binds LRH-1 partly overlaps defined contact sites for peptide segments of β-catenin partners, including T-cell factor-4. The surface of LRH-1 that engages β-catenin is comprised of helices 1, 9, and 10 and is distinct from known interaction surfaces of LRH-1, including corepressor and coactivator binding sites. Targeted mutagenesis of amino acids forming both sides of the LRH-1/β-catenin interface reveals that they are essential for stable interactions between these proteins in solution. The LRH-1 binding site in β-catenin is also required for association with androgen receptor, providing evidence that the observed LRH-1/β-catenin interaction may be prototypic.     

Structural and mechanistic insight into N-glycan processing by endo-α-mannosidase

N-linked glycans play key roles in protein folding, stability, and function. Biosynthetic modification of N-linked glycans, within the endoplasmic reticulum, features sequential trimming and readornment steps. One unusual enzyme, endo-α-mannosidase, cleaves mannoside linkages internally within an N-linked glycan chain, short circuiting the classical N-glycan biosynthetic pathway. Here, using two bacterial orthologs, we present the first structural and mechanistic dissection of endo-α-mannosidase. Structures solved at resolutions 1.7–2.1 Å reveal a (β/α)₈ barrel fold in which the catalytic center is present in a long substrate-binding groove, consistent with cleavage within the N-glycan chain. Enzymatic cleavage of authentic Glc_1/3Man₉GlcNAc₂ yields Glc_1/3-Man. Using the bespoke substrate α-Glc-1,3-α-Man fluoride, the enzyme was shown to act with retention of anomeric configuration. Complexes with the established endo-α-mannosidase inhibitor α-Glc-1,3-deoxymannonojirimycin and a newly developed inhibitor, α-Glc-1,3-isofagomine, and with the reducing-end product α-1,2-mannobiose structurally define the -2 to +2 subsites of the enzyme. These structural and mechanistic data provide a foundation upon which to develop new enzyme inhibitors targeting the hijacking of N-glycan synthesis in viral disease and cancer.     

Nonplanar peptide bonds in proteins are common and conserved but not biased toward active sites

The planarity of peptide bonds is an assumption that underlies decades of theoretical modeling of proteins. Peptide bonds strongly deviating from planarity are considered very rare features of protein structure that occur for functional reasons. Here, empirical analyses of atomic-resolution protein structures reveal that trans peptide groups can vary by more than 25° from planarity and that the true extent of nonplanarity is underestimated even in 1.2 Å resolution structures. Analyses as a function of the φ,ψ-backbone dihedral angles show that the expected value deviates by ± 8° from planar as a systematic function of conformation, but that the large majority of variation in planarity depends on tertiary effects. Furthermore, we show that those peptide bonds in proteins that are most nonplanar, deviating by over 20° from planarity, are not strongly associated with active sites. Instead, highly nonplanar peptides are simply integral components of protein structure related to local and tertiary structural features that tend to be conserved among homologs. To account for the systematic φ,ψ-dependent component of nonplanarity, we present a conformation-dependent library that can be used in crystallographic refinement and predictive protein modeling.     

Crystal structure of LpxK, the 4'-kinase of lipid A biosynthesis and atypical P-loop kinase functioning at the membrane interface

In Gram-negative bacteria, the hydrophobic anchor of the outer membrane lipopolysaccharide is lipid A, a saccharolipid that plays key roles in both viability and pathogenicity of these organisms. The tetraacyldisaccharide 4′-kinase (LpxK) of the diverse P-loop–containing nucleoside triphosphate hydrolase superfamily catalyzes the sixth step in the biosynthetic pathway of lipid A, and is the only known P-loop kinase to act upon a lipid substrate at the membrane. Here, we report the crystal structures of apo- and ADP/Mg²⁺-bound forms of Aquifex aeolicus LpxK to a resolution of 1.9 Å and 2.2 Å, respectively. LpxK consists of two α/β/α sandwich domains connected by a two-stranded β-sheet linker. The N-terminal domain, which has most structural homology to other family members, is responsible for catalysis at the P-loop and positioning of the disaccharide-1-phosphate substrate for phosphoryl transfer on the inner membrane. The smaller C-terminal domain, a substructure unique to LpxK, helps bind the nucleotide substrate and Mg²⁺ cation using a 25° hinge motion about its base. Activity was severely reduced in alanine point mutants of conserved residues D138 and D139, which are not directly involved in ADP or Mg²⁺ binding in our structures, indicating possible roles in phosphoryl acceptor positioning or catalysis. Combined structural and kinetic studies have led to an increased understanding of the enzymatic mechanism of LpxK and provided the framework for structure-based antimicrobial design.     

Crystal structure of prethrombin-1

Prothrombin is the zymogen precursor of the clotting enzyme thrombin, which is generated by two sequential cleavages at R271 and R320 by the prothrombinase complex. The structure of prothrombin is currently unknown. Prethrombin-1 differs from prothrombin for the absence of 155 residues in the N-terminal domain and is composed of a single polypeptide chain containing fragment 2 (residues 156–271), A chain (residues 272–320), and B chain (residues 321–579). The X-ray crystal structure of prethrombin-1 solved at 2.2-Å resolution shows an overall conformation significantly different (rmsd = 3.6 Å) from that of its active form meizothrombin desF1 carrying a cleavage at R320. Fragment 2 is rotated around the y axis by 29° and makes only few contacts with the B chain. In the B chain, the oxyanion hole is disrupted due to absence of the I16-D194 ion pair and the Na⁺ binding site and adjacent primary specificity pocket are highly perturbed. A remarkable feature of the structure is that the autolysis loop assumes a helical conformation enabling W148 and W215, located 17 Å apart in meizothrombin desF1, to come within 3.3 Å of each other and completely occlude access to the active site. These findings suggest that the zymogen form of thrombin possesses conformational plasticity comparable to that of the mature enzyme and have significant implications for the mechanism of prothrombin activation and the zymogen → protease conversion in trypsin-like proteases.     

Accurate protein structure modeling using sparse NMR data and homologous structure information

While information from homologous structures plays a central role in X-ray structure determination by molecular replacement, such information is rarely used in NMR structure determination because it can be incorrect, both locally and globally, when evolutionary relationships are inferred incorrectly or there has been considerable evolutionary structural divergence. Here we describe a method that allows robust modeling of protein structures of up to 225 residues by combining , ¹³C, and ¹⁵N backbone and ¹³Cβ chemical shift data, distance restraints derived from homologous structures, and a physically realistic all-atom energy function. Accurate models are distinguished from inaccurate models generated using incorrect sequence alignments by requiring that (i) the all-atom energies of models generated using the restraints are lower than models generated in unrestrained calculations and (ii) the low-energy structures converge to within 2.0 Å backbone rmsd over 75% of the protein. Benchmark calculations on known structures and blind targets show that the method can accurately model protein structures, even with very remote homology information, to a backbone rmsd of 1.2–1.9 Å relative to the conventional determined NMR ensembles and of 0.9–1.6 Å relative to X-ray structures for well-defined regions of the protein structures. This approach facilitates the accurate modeling of protein structures using backbone chemical shift data without need for side-chain resonance assignments and extensive analysis of NOESY cross-peak assignments.     

Ubiquitin vinyl methyl ester binding orients the misaligned active site of the ubiquitin hydrolase UCHL1 into productive conformation

Ubiquitin carboxy-terminal hydrolase L1 (UCHL1) is a Parkinson disease-associated, putative cysteine protease found abundantly and selectively expressed in neurons. The crystal structure of apo UCHL1 showed that the active-site residues are not aligned in a canonical form, with the nucleophilic cysteine being 7.7 Å from the general base histidine, an arrangement consistent with an inactive form of the enzyme. Here we report the crystal structures of the wild type and two Parkinson disease-associated variants of the enzyme, S18Y and I93M, bound to a ubiquitin-based suicide substrate, ubiquitin vinyl methyl ester. These structures reveal that ubiquitin vinyl methyl ester binds primarily at two sites on the enzyme, with its carboxy terminus at the active site and with its amino-terminal β-hairpin at the distal site—a surface-exposed hydrophobic crevice 17 Å away from the active site. Binding at the distal site initiates a cascade of side-chain movements in the enzyme that starts at a highly conserved, surface-exposed phenylalanine and is relayed to the active site resulting in the reorientation and proximal placement of the general base within 4 Å of the catalytic cysteine, an arrangement found in productive cysteine proteases. Mutation of the distal-site, surface-exposed phenylalanine to alanine reduces ubiquitin binding and severely impairs the catalytic activity of the enzyme. These results suggest that the activity of UCHL1 may be regulated by its own substrate.     

New enzyme lineages by subdomain shuffling

Protein functions have evolved in part via domain recombination events. Such events, for example, recombine structurally independent functional domains and shuffle targeting, regulatory, and/or catalytic functions. Domain recombination, however, can generate new functions, as implied by the observation of catalytic sites at interfaces of distinct folding domains. If useful to an evolving organism, such initially rudimentary functions would likely acquire greater efficiency and diversity, whereas the initially distinct folding domains would likely develop into single functional domains. This represents the probable evolution of the S1 serine protease family, whose two homologous β-barrel subdomains assemble to form the binding sites and the catalytic machinery. Among S1 family members, the contact interface and catalytic residues are highly conserved whereas surrounding surfaces are highly variable. This observation suggests a new strategy to engineer viable proteins with novel properties, by swapping folding subdomains chosen from among protein family members. Such hybrid proteins would retain properties conserved throughout the family, including folding stability as single domain proteins, while providing new surfaces amenable to directed evolution or engineering of specific new properties. We show here that recombining the N-terminal subdomain from coagulation factor X with the C-terminal subdomain from trypsin creates a potent enzyme (fXYa) with novel properties, in particular a broad substrate specificity. As shown by the 2.15-Å crystal structure, plasticity at the hydrophobic subdomain interface maintains activity, while surface loops are displaced compared with the parent subdomains. fXYa thus represents a new serine proteinase lineage with hybrid fX, trypsin, and novel properties.     

Near-atomic resolution structural model of the yeast 26S proteasome

The 26S proteasome operates at the executive end of the ubiquitin-proteasome pathway. Here, we present a cryo-EM structure of the Saccharomyces cerevisiae 26S proteasome at a resolution of 7.4 Å or 6.7 Å (Fourier-Shell Correlation of 0.5 or 0.3, respectively). We used this map in conjunction with molecular dynamics-based flexible fitting to build a near-atomic resolution model of the holocomplex. The quality of the map allowed us to assign α-helices, the predominant secondary structure element of the regulatory particle subunits, throughout the entire map. We were able to determine the architecture of the Rpn8/Rpn11 heterodimer, which had hitherto remained elusive. The MPN domain of Rpn11 is positioned directly above the AAA-ATPase N-ring suggesting that Rpn11 deubiquitylates substrates immediately following commitment and prior to their unfolding by the AAA-ATPase module. The MPN domain of Rpn11 dimerizes with that of Rpn8 and the C-termini of both subunits form long helices, which are integral parts of a coiled-coil module. Together with the C-terminal helices of the six PCI-domain subunits they form a very large coiled-coil bundle, which appears to serve as a flexible anchoring device for all the lid subunits.     

Ultrahigh resolution protein structures using NMR chemical shift tensors

NMR chemical shift tensors (CSTs) in proteins, as well as their orientations, represent an important new restraint class for protein structure refinement and determination. Here, we present the first determination of both CST magnitudes and orientations for ¹³Cα and ¹⁵N (peptide backbone) groups in a protein, the β1 IgG binding domain of protein G from Streptococcus spp., GB1. Site-specific ¹³Cα and ¹⁵N CSTs were measured using synchronously evolved recoupling experiments in which ¹³C and ¹⁵N tensors were projected onto the ¹H-¹³C and ¹H-¹⁵N vectors, respectively, and onto the ¹⁵N-¹³C vector in the case of ¹³Cα. The orientations of the ¹³Cα CSTs to the ¹H-¹³C and ¹³C-¹⁵N vectors agreed well with the results of ab initio calculations, with an rmsd of approximately 8°. In addition, the measured ¹⁵N tensors exhibited larger reduced anisotropies in α-helical versus β-sheet regions, with very limited variation (18 ± 4°) in the orientation of the z-axis of the ¹⁵N CST with respect to the ¹H-¹⁵N vector. Incorporation of the ¹³Cα CST restraints into structure calculations, in combination with isotropic chemical shifts, transferred echo double resonance ¹³C-¹⁵N distances and vector angle restraints, improved the backbone rmsd to 0.16 Å (PDB ID code 2LGI) and is consistent with existing X-ray structures (0.51 Å agreement with PDB ID code 2QMT). These results demonstrate that chemical shift tensors have considerable utility in protein structure refinement, with the best structures comparable to 1.0-Å crystal structures, based upon empirical metrics such as Ramachandran geometries and χ¹/χ² distributions, providing solid-state NMR with a powerful tool for de novo structure determination.     

Glycosylation of the enhanced aromatic sequon is similarly stabilizing in three distinct reverse turn contexts

Cotranslational N-glycosylation can accelerate protein folding, slow protein unfolding, and increase protein stability, but the molecular basis for these energetic effects is incompletely understood. N-glycosylation of proteins at naïve sites could be a useful strategy for stabilizing proteins in therapeutic and research applications, but without engineering guidelines, often results in unpredictable changes to protein energetics. We recently introduced the enhanced aromatic sequon as a family of portable structural motifs that are stabilized upon glycosylation in specific reverse turn contexts: a five-residue type I β-turn harboring a G1 β-bulge (using a Phe–Yyy–Asn–Xxx–Thr sequon) and a type II β-turn within a six-residue loop (using a Phe–Yyy–Zzz–Asn–Xxx–Thr sequon) [Culyba EK, et al. (2011) Science 331:571–575]. Here we show that glycosylating a new enhanced aromatic sequon, Phe–Asn–Xxx–Thr, in a type I′ β-turn stabilizes the Pin 1 WW domain. Comparing the energetic effects of glycosylating these three enhanced aromatic sequons in the same host WW domain revealed that the glycosylation-mediated stabilization is greatest for the enhanced aromatic sequon complementary to the type I β-turn with a G1 β-bulge. However, the portion of the stabilization from the tripartite interaction between Phe, Asn(GlcNAc), and Thr is similar for each enhanced aromatic sequon in its respective reverse turn context. Adding the Phe–Asn–Xxx–Thr motif (in a type I′ β-turn) to the enhanced aromatic sequon family doubles the number of proteins that can be stabilized by glycosylation without having to alter the native reverse turn type.     

Crystal structure of the synergistic antibiotic pair, lankamycin and lankacidin, in complex with the large ribosomal subunit

The structures of the large ribosomal subunit of Deinococcus radiodurans (D50S) in complex with the antibiotic lankamycin (3.2 Å) and a double antibiotic complex of lankamycin and lankacidin C (3.45 Å) have been determined, in continuation of previous crystallographic studies on lankacidin-D50S complex. These two drugs have been previously reported to inhibit ribosomal function with mild synergistic effect. Lankamycin, a member of the macrolide family, binds in a similar manner to erythromycin. However, when in complex with lankacidin, lankamycin is located so that it can form interactions with lankacidin in the adjacent ribosomal binding site. When compared to the well-documented synergistic antibiotics, Streptogramins A and B, the pair of lankacidin and lankamycin bind in similar sites, the peptidyl transferase center and nascent peptide exit tunnel, respectively. Herein, we discuss the structural basis for antibiotic synergism and highlight the key factors involved in ribosomal inhibition.     

Structural basis for the wobbler mouse neurodegenerative disorder caused by mutation in the Vps54 subunit of the GARP complex

The multisubunit Golgi-associated retrograde protein (GARP) complex is required for tethering and fusion of endosome-derived transport vesicles to the trans-Golgi network. Mutation of leucine-967 to glutamine in the Vps54 subunit of GARP is responsible for spinal muscular atrophy in the wobbler mouse, an animal model of amyotrophic lateral sclerosis. The crystal structure at 1.7 Å resolution of the mouse Vps54 C-terminal fragment harboring leucine-967, in conjunction with comparative sequence analysis, reveals that Vps54 has a continuous α-helical bundle organization similar to that of other multisubunit tethering complexes. The structure shows that leucine-967 is buried within the α-helical bundle through predominantly hydrophobic interactions that are critical for domain stability and folding in vitro. Mutation of this residue to glutamine does not prevent integration of Vps54 into the GARP complex but greatly reduces the half-life and levels of the protein in vivo. Severely reduced levels of mutant Vps54 and, consequently, of the whole GARP complex underlie the phenotype of the wobbler mouse.