Protein conformation by EPR spectroscopy using gadolinium tags clicked to genetically encoded p-azido-l-phenylalanine

First example of gadolinium tags attached to a pair of unnatural amino acids for distance measurements by double electron–electron resonance.

Selective (15)N-labeling of the side-chain amide groups of asparagine and glutamine for applications in paramagnetic NMR spectroscopy

The side-chain amide groups of asparagine and glutamine play important roles in stabilizing the structural fold of proteins, participating in hydrogen-bonding networks and protein interactions. Selective (15)N-labeling of side-chain amides, however, can be a challenge due to enzyme-catalyzed exchange of amide groups during protein synthesis. In the present study, we developed an efficient way of selectively labeling the side chains of asparagine, or asparagine and glutamine residues with (15)NH2. Using the biosynthesis pathway of tryptophan, a protocol was also established for simultaneous selective (15)N-labeling of the side-chain NH groups of asparagine, glutamine, and tryptophan. In combination with site-specific tagging of the target protein with a lanthanide ion, we show that selective detection of (15)N-labeled side-chains of asparagine and glutamine allows determination of magnetic susceptibility anisotropy tensors based exclusively on pseudocontact shifts of amide side-chain protons.

The dengue virus NS2B-NS3 protease retains the closed conformation in the complex with BPTI

The C-terminal β-hairpin of NS2B (NS2Bc) in the dengue virus NS2B-NS3 protease is required for full enzymatic activity. In crystal structures without inhibitor and in the complex with bovine pancreatic trypsin inhibitor (BPTI), NS2Bc is displaced from the active site. In contrast, nuclear magnetic resonance (NMR) studies in solution only ever showed NS2Bc in the enzymatically active closed conformation. Here we demonstrate by pseudocontact shifts from a lanthanide tag that NS2Bc remains in the closed conformation also in the complex with BPTI. Therefore, the closed conformation is the best template for drug discovery.

Bound or free: interaction of the C-terminal domain of Escherichia coli single-stranded DNA-binding protein (SSB) with the tetrameric core of SSB

Single-stranded DNA (ssDNA)-binding protein (SSB) protects ssDNA from degradation and recruits other proteins for DNA replication and repair. Escherichia coli SSB is the prototypical eubacterial SSB in a family of tetrameric SSBs. It consists of a structurally well-defined ssDNA binding domain (OB-domain) and a disordered C-terminal domain (C-domain). The eight-residue C-terminal segment of SSB (C-peptide) mediates the binding of SSB to many different SSB-binding proteins. Previously published nuclear magnetic resonance (NMR) data of the monomeric state at pH 3.4 showed that the C-peptide binds to the OB-domain at a site that overlaps with the ssDNA binding site, but investigating the protein at neutral pH is difficult because of the high molecular mass and limited solubility of the tetramer. Here we show that the C-domain is highly mobile in the SSB tetramer at neutral pH and that binding of the C-peptide to the OB-domain is so weak that most of the C-peptides are unbound even in the absence of ssDNA. We address the problem of determining intramolecular binding affinities in the situation of fast exchange between two states, one of which cannot be observed by NMR and cannot be fully populated. The results were confirmed by electron paramagnetic resonance spectroscopy and microscale thermophoresis. The C-peptide-OB-domain interaction is shown to be driven primarily by electrostatic interactions, so that binding of 1 equiv of (dT)35 releases practically all C-peptides from the OB-domain tetramer. The interaction is much more sensitive to NaCl than to potassium glutamate, which is the usual osmolyte in E. coli. As the C-peptide is predominantly in the unbound state irrespective of the presence of ssDNA, long-range electrostatic effects from the C-peptide may contribute more to regulating the activity of SSB than any engagement of the C-peptide by the OB-domain.

Binding mode of the activity-modulating C-terminal segment of NS2B to NS3 in the dengue virus NS2B-NS3 protease

The two-component dengue virus NS2B-NS3 protease (NS2B-NS3pro) is an established drug target but inhibitor design is hampered by uncertainties about its 3D structure in solution. Crystal structures reported very different conformations for the functionally important C-terminal segment of the NS2B cofactor (NS2Bc), indicating open and closed conformations in the absence and presence of inhibitors, respectively. An earlier NMR study in solution indicated that a closed state is the preferred conformation in the absence of an artificial linker engineered between NS2B and NS3pro. To obtain direct structural information on the fold of unlinked NS2B-NS3pro in solution, we tagged NS3pro with paramagnetic tags and measured pseudocontact shifts by NMR to position NS2Bc relative to NS3pro. NS2Bc was found to bind to NS3pro in the same way as reported in a previously published model and crystal structure of the closed state. The structure is destabilized, however, by high ionic strength and basic pH, showing the importance of electrostatic forces to tie NS2Bc to NS3pro. Narrow NMR signals previously thought to represent the open state are associated with protein degradation. In conclusion, the closed conformation of the NS2B-NS3 protease is the best model for structure-guided drug design.

Intramolecular binding mode of the C-terminus of Escherichia coli single-stranded DNA binding protein determined by nuclear magnetic resonance spectroscopy

Single-stranded DNA (ssDNA) binding protein (SSB) is an essential protein to protect ssDNA and recruit specific ssDNA-processing proteins. Escherichia coli SSB forms a tetramer at neutral pH, comprising a structurally well-defined ssDNA binding domain (OB-domain) and a disordered C-terminal domain (C-domain) of ∼64 amino acid residues. The C-terminal eight-residue segment of SSB (C-peptide) has been shown to interact with the OB-domain, but crystal structures failed to reveal any electron density of the C-peptide. Here we show that SSB forms a monomer at pH 3.4, which is suitable for studies by high-resolution nuclear magnetic resonance (NMR) spectroscopy. The OB-domain retains its 3D structure in the monomer, and the C-peptide is shown by nuclear Overhauser effects and lanthanide-induced pseudocontact shifts to bind to the OB-domain at a site that harbors ssDNA in the crystal structure of the SSB–ssDNA complex. ¹⁵N relaxation data demonstrate high flexibility of the polypeptide segment linking the C-peptide to the OB-domain and somewhat increased flexibility of the C-peptide compared with the OB-domain, suggesting that the C-peptide either retains high mobility in the bound state or is in a fast equilibrium with an unbound state.

How reliable are pseudocontact shifts induced in proteins and ligands by mobile paramagnetic metal tags? A modelling study

The anisotropic component of the magnetic susceptibility tensor (Δχ tensor) associated with various paramagnetic metal ions can induce pseudocontact shifts (PCSs) and residual dipolar couplings (RDCs) in proteins, yielding valuable restraints in structural studies. In particular, PCSs have successfully been used to study ligands that bind to proteins tagged with a paramagnetic metal ion, which is of great interest in fragment-based drug design. To create easy-to-interpret PCSs, the metal ion must be attached to the protein in a rigid manner. Most of the existing methods for site-specific attachment of a metal tag, however, result in tethers with residual flexibility. Here we present model calculations to quantify the extent, to which mobility of the metal-binding tag can compromise the quality of the Δχ tensor that can be determined from the PCSs observed in the protein. Assuming that the protein can be approximated by a sphere and the tag is attached by a single tether, the results show that a single effective ∆χ tensor can describe the PCSs and RDCs of the protein spins very well even in the presence of substantial tag mobility, implying that PCSs of ligands in binding pockets of the protein can be predicted with similar accuracy. In contrast, the quality of the PCS prediction for nuclear spins positioned above the surface of the protein is significantly poorer, with implications for studies of protein-protein complexes. The simulations probed the sensitivity of the effective Δχ tensor to different parameters, including length of the tether between protein and metal ion, protein size, type and amplitude of tag motion, tensor orientation relative to the protein and direction of tag motion. Tether length and amplitude of motion were identified as two key parameters. It is shown that the amplitude of tag motions cannot be quantified by simple comparisons of the effective Δχ tensor with the alignment tensor determined from RDCs.

Three-dimensional protein fold determination from backbone amide pseudocontact shifts generated by lanthanide tags at multiple sites

Site-specific attachment of paramagnetic lanthanide ions to a protein generates pseudocontact shifts (PCS) in the nuclear magnetic resonance (NMR) spectra of the protein that are easily measured as changes in chemical shifts. By labeling the protein with lanthanide tags at four different sites, PCSs are observed for most amide protons and accurate information is obtained about their coordinates in three-dimensional space. The approach is demonstrated with the chaperone ERp29, for which large differences have been reported between X-ray and NMR structures of the C-terminal domain, ERp29-C. The results unambiguously show that the structure of rat ERp29-C in solution is similar to the crystal structure of human ERp29-C. PCSs of backbone amides were the only structural restraints required. Because these can be measured for more dilute protein solutions than other NMR restraints, the approach greatly widens the range of proteins amenable to structural studies in solution.

Proofreading exonuclease on a tether: the complex between the E. coli DNA polymerase III subunits α, epsilon, θ and β reveals a highly flexible arrangement of the proofreading domain

A complex of the three (αεθ) core subunits and the β2 sliding clamp is responsible for DNA synthesis by Pol III, the Escherichia coli chromosomal DNA replicase. The 1.7 Å crystal structure of a complex between the PHP domain of α (polymerase) and the C-terminal segment of ε (proofreading exonuclease) subunits shows that ε is attached to α at a site far from the polymerase active site. Both α and ε contain clamp-binding motifs (CBMs) that interact simultaneously with β2 in the polymerization mode of DNA replication by Pol III. Strengthening of both CBMs enables isolation of stable αεθ:β2 complexes. Nuclear magnetic resonance experiments with reconstituted αεθ:β2 demonstrate retention of high mobility of a segment of 22 residues in the linker that connects the exonuclease domain of ε with its α-binding segment. In spite of this, small-angle X-ray scattering data show that the isolated complex with strengthened CBMs has a compact, but still flexible, structure. Photo-crosslinking with p-benzoyl-L-phenylalanine incorporated at different sites in the α-PHP domain confirm the conformational variability of the tether. Structural models of the αεθ:β2 replicase complex with primer-template DNA combine all available structural data.

W-band orientation selective DEER measurements on a Gd3+/nitroxide mixed-labeled protein dimer with a dual mode cavity

Double electron-electron resonance (DEER) at W-band (95 GHz) was applied to measure the distance between a pair of nitroxide and Gd(3+) chelate spin labels, about 6 nm apart, in a homodimer of the protein ERp29. While high-field DEER measurements on systems with such mixed labels can be highly attractive in terms of sensitivity and the potential to access long distances, a major difficulty arises from the large frequency spacing (about 700 MHz) between the narrow, intense signal of the Gd(3+) central transition and the nitroxide signal. This is particularly problematic when using standard single-mode cavities. Here we show that a novel dual-mode cavity that matches this large frequency separation dramatically increases the sensitivity of DEER measurements, allowing evolution times as long as 12 μs in a protein. This opens the possibility of accessing distances of 8 nm and longer. In addition, orientation selection can be resolved and analyzed, thus providing additional structural information. In the case of W-band DEER on a Gd(3+)-nitroxide pair, only two angles and their distributions have to be determined, which is a much simpler problem to solve than the five angles and their distributions associated with two nitroxide spin labels.

A systematic study of labelling an α-helix in a protein with a lanthanide using IDA-SH or NTA-SH tags

The previously published IDA-SH and NTA-SH tags are small synthetic lanthanide-binding tags derived from cysteine, which afford site-specific lanthanide labelling by disulfide-bond formation with a cysteine residue of the target protein. Following attachment to a single cysteine in an α-helix, sizeable pseudocontact shifts (PCS) can be observed, if the lanthanide is immobilized by additional coordination to a negatively charged amino-acid side chain that is located in a neighboring turn of the helix. To identify the best labelling strategy for PCS measurements, we performed a systematic study, where IDA-SH or NTA-SH tags were ligated to a cysteine residue in position i of an α-helix, and aspartate or glutamate residues were placed in the positions i - 4 or i + 4. The largest anisotropy components of the magnetic susceptibility tensor were observed for an NTA-SH tag in position i with a glutamate residue in position i - 4. While the NTA-SH tag produced sizeable PCSs regardless of the presence of nearby carboxyl groups of the protein, the IDA-SH tag generated a good lanthanide binding site only if an aspartate was placed in position i + 4. The findings provide a firm basis for the design of site-directed mutants that are suitable for the reliable generation of PCSs in proteins with paramagnetic lanthanides.

Structural basis for 5'-end-specific recognition of single-stranded DNA by the R3H domain from human Sμbp-2

The R3H domain is a conserved sequence motif in nucleic acid binding proteins. Previously, we reported the solution structure of the R3H domain and identified a putative nucleic acid binding site composed of three conserved basic residues [Liepinsh, E., Leonchiks, A., Sharipo, A., Guignard, L. & Otting, G. (2003). Solution structure of the R3H domain from human Sμbp-2. J. Mol. Biol.326, 217-223]. Here, we determine the binding affinities of mononucleotides and dinucleotides for the R3H domain from human Sμbp-2 (Sμbp2-R3H) and map their binding sites on the protein's surface. Although the binding affinities show up to 260-fold selectivity between different nucleotides, their binding sites and conformations seem very similar. Further, we report the NMR structure of the Sμbp2-R3H in complex with deoxyguanosine 5'-monophosphate (dGMP) mimicking the 5'-end of single-stranded DNA. Pseudocontact shifts from a paramagnetic lanthanide tag attached to residue 731 in the mutant A731C confirmed that binding of dGMP brings a loop of the protein into closer proximity. The structure provides the first structural insight into single-stranded nucleic acid recognition by the R3H domain and shows that the R3H domain specifically binds the phosphorylated 5'-end through electrostatic interactions with the two conserved arginines and stacking interactions with the highly conserved histidine.