Protein engineering of the chemokine CCL20 prevents psoriasiform dermatitis in an IL-23-dependent murine model.

Psoriasis is a chronic inflammatory skin disease characterized by the infiltration of T cell and other immune cells to the skin in response to injury or autoantigens. Conventional, as well as unconventional, γδ T cells are recruited to the dermis and epidermis by CCL20 and other chemokines. Together with its receptor CCR6, CCL20 plays a critical role in the development of psoriasiform dermatitis in mouse models. We screened a panel of CCL20 variants designed to form dimers stabilized by intermolecular disulfide bonds. A single-atom substitution yielded a CCL20 variant (CCL20 S64C) that acted as a partial agonist for the chemokine receptor CCR6. CCL20 S64C bound CCR6 and induced intracellular calcium release, consistent with G-protein activation, but exhibited minimal chemotactic activity. Instead, CCL20 S64C inhibited CCR6-mediated T cell migration with nominal impact on other chemokine receptor signaling. When given in an IL-23-dependent mouse model for psoriasis, CCL20 S64C prevented psoriatic inflammation and the up-regulation of IL-17A and IL-22. Our results validate CCR6 as a tractable therapeutic target for psoriasis and demonstrate the value of CCL20 S64C as a lead compound.

Monomeric solution structure of the prototypical 'C' chemokine lymphotactin.

Lymphotactin, the sole identified member of the C class of chemokines, specifically attracts T lymphocytes and natural killer cells. This 93-residue protein lacks 2 of the 4 conserved cysteine residues characteristic of the other 3 classes of chemokines and possesses an extended carboxyl terminus, which is required for chemotactic activity. We have determined the three-dimensional solution structure of recombinant human lymphotactin by NMR spectroscopy. Under the conditions used for the structure determination, lymphotactin was predominantly monomeric; however, pulsed field gradient NMR self-diffusion measurements and analytical ultracentrifugation revealed evidence of dimer formation. Sequence-specific chemical shift assignments were determined through analysis of two- and three-dimensional NMR spectra of (15)N- and (13)C/(15)N-enriched protein samples. Input for the torsion angle dynamics calculations used in determining the structure included 1258 unique NOE-derived distance constraints and 60 dihedral angle constraints obtained from chemical-shift-based searching of a protein conformational database. The ensemble of 20 structures chosen to represent the structure had backbone and heavy atom rms deviations of 0.46 +/- 0.11 and 1.02 +/- 0.14 A, respectively. The results revealed that human lymphotactin adopts the conserved chemokine fold, which is characterized by a three-stranded antiparallel beta-sheet and a C-terminal alpha-helix. Two regions are dynamically disordered as evidenced by (1)H and (13)C chemical shifts and [(15)N]-(1)H NOEs: residues 1-9 of the amino terminus and residues 69-93 of the C-terminal extension. A functional role for the C-terminal extension, which is unique to lymphotactin, remains to be elucidated.

N-terminal residues of plasmatocyte-spreading peptide possess specific determinants required for biological activity.

Plasmatocyte-spreading peptide (PSP) is a 23-amino acid cytokine that activates a class of insect immune cells called plasmatocytes. The tertiary structure of PSP consists of an unstructured N terminus (residues 1-6) and a well structured core (residues 7-23). A prior study indicated that deletion of the N terminus from PSP eliminated all biological activity. Alanine substitution of the first three residues (Glu(1)-Asn(2)-Phe(3)) further indicated that only replacement of Phe(3) resulted in a loss of activity equal to the N-terminal deletion mutant. Here, we characterized structural determinants of the N terminus. Adding a hydroxyl group to the aromatic ring of Phe(3) (making a Tyr) greatly reduced activity, whereas the addition of a fluorine (p-fluoro) did not. Substitutions that changed the chirality or replaced the aromatic ring of Phe(3) with a branched aliphatic chain (making a Val) also greatly decreased activity. The addition of a methylene group to Val (making a Leu) partially restored activity, whereas the removal of a methylene group from Phe (phenyl-Gly) eliminated all activity. These results indicated that a branched carbon chain with a methylene spacer at the third residue is the minimal structural motif required for activity. The deletion of Glu(1) also eliminated activity. Additional experiments identified the charged N-terminal amine and backbone of Glu(1) as key determinants for activity.

Biosynthesis of D-alanyl-lipoteichoic acid: the tertiary structure of apo-D-alanyl carrier protein.

The D-alanylation of lipoteichoic acid (LTA) allows the Gram-positive organism to modulate its surface charge, regulate ligand binding, and control the electromechanical properties of the cell wall. The incorporation of D-alanine into LTA requires the D-alanine:D-alanyl carrier protein ligase (AMP-forming) (Dcl) and the carrier protein (Dcp). The high-resolution solution structure of the 81-residue (8.9 kDa) Dcp has been determined by multidimensional heteronuclear NMR. An ensemble of 30 structures was calculated using the torsion angle dynamics approach of DYANA. These calculations utilized 3288 NOEs containing 1582 unique nontrivial NOE distance constraints. Superposition of residues 4-81 on the mean structure yields average atomic rmsd values of 0.43 +/- 0.08 and 0.86 +/- 0.09 A for backbone and non-hydrogen atoms, respectively. The solution structure is composed of three alpha-helices in a bundle with additional short 3(10)- and alpha-helices in intervening loops. Comparisons of the three-dimensional structure with the acyl carrier proteins involved in fatty acid, polyketide, and nonribosomal peptide syntheses support the conclusion that Dcp is a homologue in this family. While there is conservation of the three-helix bundle fold, Dcp has a higher enthalpy of unfolding and no apparent divalent metal binding site(s), features that distinguish it from the fatty acid synthase acyl carrier protein of Escherichia coli. This three-dimensional structure also provides insights into the D-alanine ligation site recognized by Dcl, as well as the site which may bind the poly(glycerophosphate) acceptor moiety of membrane-associated LTA.

Solution structure of a type I dockerin domain, a novel prokaryotic, extracellular calcium-binding domain.

The type I dockerin domain is responsible for incorporating its associated glycosyl hydrolase into the bacterial cellulosome, a multienzyme cellulolytic complex, via its interaction with a receptor domain (cohesin domain) of the cellulosomal scaffolding subunit. The highly conserved dockerin domain is characterized by two Ca(2+)-binding sites with sequence similarity to the EF-hand motif. Here, we present the three-dimensional solution structure of the 69 residue dockerin domain of Clostridium thermocellum cellobiohydrolase CelS. Torsion angle dynamics calculations utilizing a total of 728 NOE-derived distance constraints and 79 torsion angle restraints yielded an ensemble of 20 structures with an average backbone r.m.s.d. for residues 5 to 29 and 32 to 66 of 0.54 A from the mean structure. The structure consists of two Ca(2+)-binding loop-helix motifs connected by a linker; the E helices entering each loop of the classical EF-hand motif are absent from the dockerin domain. Each dockerin Ca(2+)-binding subdomain is stabilized by a cluster of buried hydrophobic side-chains. Structural comparisons reveal that, in its non-complexed state, the dockerin fold displays a dramatic departure from that of Ca(2+)-bound EF-hand domains. A putative cohesin-binding surface, comprised of conserved hydrophobic and basic residues, is proposed, providing new insight into cellulosome assembly.

Alanine-scanning mutagenesis of plasmatocyte spreading peptide identifies critical residues for biological activity.

Plasmatocyte spreading peptide (PSP) is a 23-amino acid cytokine that induces a class of insect immune cells called plasmatocytes to spread on foreign surfaces. The structure of PSP consists of a disordered N terminus (residues 1-6) and a well-defined core (residues 7-23) stabilized by a disulfide bridge between Cys(7) and Cys(19), hydrophobic interactions, and a short beta-hairpin. Structural comparisons also indicate that the core region of PSP adopts an epidermal growth factor (EGF)-like fold very similar to the C-terminal subdomain of EGF-like module 5 of thrombomodulin. To identify residues important for plasmatocyte spreading activity, we bioassayed PSP mutants in which amino acids were either replaced with alanine or deleted. Within the well-defined core of PSP, alanine replacement of Cys(7) and Cys(19) (C7.19A) eliminated all activity. Alanine replacement of Arg(13) reduced activity approximately 1000-fold in comparison to wild-type PSP, whereas replacement of the other charged residues (Asp(16), Arg(18), Lys(20)) surrounding Cys(19) diminished activity to a lesser degree. The point mutants Y11A, T14A, T22A, and F23A had activity identical or only slightly reduced to that of wild-type PSP. The mutant PSP-(7-23) lacked the entire unstructured domain of PSP and was found to have no plasmatocyte spreading activity. Surprisingly, E1A and N2A had higher activity than wild-type PSP, but F3A had almost no activity. We thus concluded that the lack of activity for PSP-(7-23) was largely due to the critical importance of Phe(3). To determine whether reductions in activity correlated with alterations in tertiary structure, we compared the C7.19A, R13A, R18A, and F3A mutants to wild-type PSP by NMR spectroscopy. As expected, the simultaneous replacement of Cys(7) and Cys(19) profoundly affected tertiary structure, but the R13A, R18A, and F3A mutants did not differ from wild-type PSP. Collectively, these results indicate that residues in both the unstructured and structured domains of PSP are required for plasmatocyte-spreading activity.

Two-state allosteric behavior in a single-domain signaling protein.

Protein actions are usually discussed in terms of static structures, but function requires motion. We find a strong correlation between phosphorylation-driven activation of the signaling protein NtrC and microsecond time-scale backbone dynamics. Using nuclear magnetic resonance relaxation, we characterized the motions of NtrC in three functional states: unphosphorylated (inactive), phosphorylated (active), and a partially active mutant. These dynamics are indicative of exchange between inactive and active conformations. Both states are populated in unphosphorylated NtrC, and phosphorylation shifts the equilibrium toward the active species. These results support a dynamic population shift between two preexisting conformations as the underlying mechanism of activation.

Secondary structure and calcium-induced folding of the Clostridium thermocellum dockerin domain determined by NMR spectroscopy.

Assembly of the cellulosome, a large, extracellular cellulase complex, depends upon docking of a myriad of enzymatic subunits to homologous receptors, or cohesin domains, arranged in tandem along a noncatalytic scaffolding protein. Docking to the cohesin domains is mediated by a highly conserved domain, dockerin (DS), borne by each enzymatic subunit. DS consists of two 22-amino-acid duplicated sequences, each bearing homology to the EF-hand calcium-binding loop. To compare the DS structure with that of the EF-hand helix-loop-helix motif, we analyzed the solution secondary structure of the DS from the cellobiohydrolase CelS subunit of the Clostridium thermocellum cellulosome using multidimensional heteronuclear NMR spectroscopy. The effect of Ca(2+)-binding on the DS structure was first investigated by using 2D (15)N-(1)H HSQC NMR spectroscopy. Changes in the spectra during Ca(2+) titration revealed that Ca(2+) induces folding of DS into its tertiary structure. This Ca(2+)-induced protein folding distinguishes DS from typical EF-hand-containing proteins. Sequential backbone assignments were determined for 63 of 69 residues. Analysis of the NOE connectivities and H(alpha) chemical shifts revealed that each half of the dockerin contains just one alpha-helix, comparable to the F-helix of the EF-hand motif. Thus, the structure of the DS Ca(2+)-binding subdomain deviates from that of the canonical EF-hand motif.

NMR solution structure of plastocyanin from the photosynthetic prokaryote, Prochlorothrix hollandica.

The solution structure of a divergent plastocyanin (PC) from the photosynthetic prokaryote Prochlorothrix hollandica was determined by homonuclear 1H NMR spectroscopy. Nineteen structures were calculated from 1222 distance restraints, yielding a family of structures having an average rmsd of 0.42 +/- 0.08 A for backbone atoms and 0.71 +/- 0.07 A for heavy atoms to the mean structure. No distance constraint was violated by more than 0.26 A in the structure family. Despite the low number of conserved residues shared with other PC homologues, the overall folding pattern of P. hollandica PC is similar to other PCs, in that the protein forms a two-sheet beta-barrel tertiary structure. The greatest variability among the backbone structures is seen in the loop region from residues 47-60. The differences seen in the P. hollandica PC homologue likely arise due to a small deletion of 2-4 residues compared to the PC consensus; this yields a less extended loop containing a short alpha-helix from residues Ala52-Leu55. Additionally, the protein has an altered hydrophobic patch thought to be important in binding reaction partners. Whereas the backbone structure is very similar within the loops of the hydrophobic region, the presence of two unique residues (Tyr12 and Pro14) yields a structurally different hydrophobic surface likely important in binding P. hollandica Photosystem I.

Structure of the insect cytokine peptide plasmatocyte-spreading peptide 1 from Pseudoplusia includens.

The structure of the recently identified plasmatocyte spreading peptide from the moth Pseudoplusia includens (PSP1) has been determined by NMR spectroscopy. This novel insect cytokine consists of 23 amino acid residues and a single disulfide bond. Torsion angle dynamics calculations utilizing a total of 337 distance constraints yielded an ensemble of 30 structures with an average backbone root mean square deviation for residues 7-22 of 0.18 A from the mean structure. The structure consists of a disordered N-terminal region and a well defined core that is stabilized by numerous hydrophobic interactions and a short beta-hairpin. Structural comparisons confirm that PSP1 adopts an epidermal growth factor (EGF)-like fold with close similarity to the C-terminal subdomain of EGF-like module 5 of human thrombomodulin. The combination of the three-dimensional structure of PSP1 and the extensive literature on EGF-receptor interactions should accelerate the process of identifying the specific residues responsible for receptor binding activity of this family of immunoregulatory peptides.

Structure of a transiently phosphorylated switch in bacterial signal transduction.

Receiver domains are the dominant molecular switches in bacterial signalling. Although several structures of non-phosphorylated receiver domains have been reported, a detailed structural understanding of the activation arising from phosphorylation has been impeded by the very short half-lives of the aspartylphosphate linkages. Here we present the first structure of a receiver domain in its active state, the phosphorylated receiver domain of the bacterial enhancer-binding protein NtrC (nitrogen regulatory protein C). Nuclear magnetic resonance spectra were taken during steady-state autophosphorylation/dephosphorylation, and three-dimensional spectra from multiple samples were combined. Phosphorylation induces a large conformational change involving a displacement of beta-strands 4 and 5 and alpha-helices 3 and 4 away from the active site, a register shift and an axial rotation in helix 4. This creates an exposed hydrophobic surface that is likely to transmit the signal to the transcriptional activation domain.

Determination of internuclear angles of DNA using paramagnetic-assisted magnetic alignment.

Paramagnetic ions have been used to assist the magnetic alignment of DNA. The anisotropy of the binding sites is sufficient to give rise to significant alignment of the DNA with the observed proton-carbon dipolar couplings spanning a 70-Hz range. The dipolar couplings have been used to determine the positions of the axial and rhombic alignment axes. The positions of the alignment axes relative to the positions of the binding sites of the paramagnetic europium ions have also been determined.

Practical model fitting approaches to the direct extraction of NMR parameters simultaneously from all dimensions of multidimensional NMR spectra.

A maximum likelihood (ML)-based approach has been established for the direct extraction of NMR parameters (e.g., frequency, amplitude, phase, and decay rate) simultaneously from all dimensions of a D-dimensional NMR spectrum. The approach, referred to here as HTFD-ML (hybrid time frequency domain maximum likelihood), constructs a time-domain model composed of a sum of exponentially-decaying sinusoidal signals. The apodized Fourier transform of this time-domain signal is a model spectrum that represents the 'best-fit' to the equivalent frequency-domain data spectrum. The desired amplitude and frequency parameters can be extracted directly from the signal model constructed by the HTFD-ML algorithm. The HTFD-ML approach presented here, as embodied in the software package CHIFIT, is designed to meet the challenges posed by model fitting of D-dimensional NMR data sets, where each consists of many data points (10(8) is not uncommon) encoding information about numerous signals (up to 10(5) for a protein of moderate size) that exhibit spectral overlap. The suitability of the approach is demonstrated by its application to the concerted analysis of a series of ten 2D 1H-15N HSQC experiments measuring 15N T1 relaxation. In addition to demonstrating the practicality of performing maximum likelihood analysis on large, multidimensional NMR spectra, the results demonstrate that this parametric model-fitting approach provides more accurate amplitude and frequency estimates than those obtained from conventional peak-based analysis of the FT spectrum. The improved performance of the model fitting approach derives from its ability to take into account the simultaneous contributions of all signals in a crowded spectral region (deconvolution) as well as to incorporate prior knowledge in constructing models to fit the data.

Solution structure and backbone dynamics of component IV Glycera dibranchiata monomeric hemoglobin-CO.

The solution structure and backbone dynamics of the recombinant, ferrous CO-ligated form of component IV monomeric hemoglobin from Glycera dibranchiata (GMH4CO) have been characterized by NMR spectroscopy. Distance geometry and simulated annealing calculations utilizing a total of 2550 distance and torsion angle constraints yielded an ensemble of 29 structures with an overall average backbone rmsd of 0.48 A from the average structure. Differences between the solution structure and a related crystal structure are confined to regions of lower precision in either the NMR or X-ray structure, or in regions where the amino acid sequences differ. 15N relaxation measurements at 76.0 and 60.8 MHz were analyzed with an extended model-free approach, and revealed low-frequency motions in the vicinity of the heme, concentrated in the F helix. Amide proton protection factors were obtained from H-D amide exchange measurements on 15N-labeled protein. Patterns in the backbone dynamics and protection factors were shown to correlate with regions of heterogeneity and disorder in the ensemble of NMR structures and with large crystallographic B-factors in the X-ray structures. Surprisingly, while the backbone atoms of the F helix have higher rmsds and larger measures of dynamics on the microsecond to millisecond time scale than the other helices, amide protection factors for residues in the F helix were observed to be similar to those of the other helices. This contrasts with H-D amide exchange measurements on sperm whale myoglobin which indicated low protection for the F helix (S. N. Loh and B. F. Volkman, unpublished results). These results for GMH4 suggest a model in which the F helix undergoes collective motions as a relatively rigid hydrogen-bonded unit, possibly pivoting about a central position near residue Val87.

Detailed NMR analysis of the heme-protein interactions in component IV Glycera dibranchiata monomeric hemoglobin-CO.

Complete 13C, 15N, and 1H resonance assignments have been obtained for the recombinant, ferrous CO-ligated from of component IV monomeric hemoglobin from Glycera dibranchiata. This 15642 Da myoglobin-like protein contains a large number of glycine and alanine residues (47) and a heme prosthetic group. Coupling constant information has allowed the determination of chi(1) and chi(2) torsion angles, backbone phi angles, as well as 43 of 81 possible assignments to H beta 2/beta 3 pairs. The 13C alpha, 13 beta, 13C', and 1H alpha assignments yield a consensus chemical shift index (CSI) that, in combination with NOE information and backbone torsion angles, defines seven distinct helical regions for the protein's global architecture. Discrepancies between the CSI and NOE/3JHNH alpha-based secondary structure definitions have been attributed to heme ring current shifts on the basis of calculations from a model structure [Alam et al. (1994) J. Protein Chem., 13, 151-164]. The agreement can be improved by correcting the 1H alpha chemical shifts for the ring current contributions. Because the holoprotein was assembled from isotopically enriched globin and natural isotope-abundance heme, data from 13C-filtered/13C-edited and 13C-filtered/13C-filtered 2D NOESY experiments could be used to determine complete heme proton assignments and to position the heme within the protein. The results confirm the unusual presence of Phe31 (B10) and Leu58 (E7) side chains near the heme ligand binding site which may alter the polarity and steric environment and thus the functional properties of this protein.

Protonation behavior of histidine 24 and histidine 119 in forming the pH 4 folding intermediate of apomyoglobin.

Heteronuclear NMR methods are used to study the protonation of histidine and aspartate residues in the acid-induced unfolding of recombinant sperm whale apomyoglobin. The results are combined with fluorescence and circular dichroism measurements of acid-induced unfolding of wild-type and double mutant (H24V/H119F) proteins. They are consistent with a simple model in which the failure to protonate a single buried histidine, H24, is largely responsible for the partial unfolding of native (N) wild-type apomyoglobin to the pH 4 folding intermediate (I). H24 is known to form an unusual interaction in which its side chain is buried and hydrogen-bonded to the side chain of H119. Two-dimensional 1H-15N heteronuclear NMR spectra indicate that H24 is present in the rare delta tautomeric form and remains neutral until N unfolds to I, while H119 becomes protonated before the N --> I reaction occurs. In the H24V/H119F double mutant, all histidines are protonated in N and the N --> I reaction occurs at lower pH. Therefore, the protonation of aspartate and/or glutamate residues must provide an additional driving force for the N to I reaction. Two-dimensional 1H-13C NMR experiments are used to measure the protonation of aspartates in selectively 13C-labeled apomyoglobin; the results indicate that none of the aspartate residues has a strongly depressed pKa in N, as would be expected if it forms a stabilizing salt bridge.

Evidence for oxidation-state-dependent conformational changes in human ferredoxin from multinuclear, multidimensional NMR spectroscopy.

Human ferredoxin belongs to the vertebrate ferredoxin family which includes bovine adrenodoxin. It is a small (13.8 kDa) acidic protein with a [2Fe-2S] cluster. It functions as an electron shuttle in the cholesterol side-chain cleavage reaction which is the first step of steroid hormone biosynthesis. The protein studied here was produced in Escherichia coli and doubly labeled with 13C and 15N. The diamagnetic 15N, 13C', 13C alpha, 13C beta, 1H alpha, and 1HN resonances from about 70% of the 124 amino acid residues for oxidized human ferredoxin and 80% of those for the reduced protein have been assigned primarily on the basis of results from three-dimensional, triple-resonance experiments. Secondary structure features for human ferredoxin in its oxidized and reduced states have been identified from a combination of chemical shift index and NOE data. Comparison of NMR results from the protein in its oxidized and reduced states indicates that structural changes accompany the change in the oxidation state of the [2Fe-2S] cluster. Major differences are localized at two regions: residues 29-31 and residues 109-124; the latter stretch forms the C-terminal region of the protein. The possible functional significance of these oxidation-state-dependent structural changes is discussed.

Chemical shift mapping of the RNA-binding interface of the multiple-RBD protein sex-lethal.

The Drosophila protein Sex-lethal (Sxl) contains two RNP consensus-type RNA-binding domains (RBDs) separated by a short linker sequence. Both domains are essential for high-affinity binding to the single-stranded polypyrimidine tract (PPT) within the regulated 3' splice site of the transformer (tra) pre-mRNA. In this paper, the effect of RNA binding to a protein fragment containing both RBDs from Sxl (Sxl-RBD1 + 2) has been characterized by heteronuclear NMR. Nearly complete (85-90%) backbone resonance assignments have been obtained for unbound and RNA-bound states of Sxl-RBD1 + 2. A comparison of amide 1H and 15N chemical shifts between free and bound states has highlighted residues which respond to RNA binding. The beta-sheets in both RBDs (RBD1 and RBD2) form an RNA interaction surface, as has been observed in other RBDs. A significant number of residues display different behavior when comparing RBD1 and RBD2. This argues for a model in which RBD1 and RBD2 of Sxl have different or nonanalogous points of interaction with the tra PPT. R142 (in RBD2) exhibits the largest chemical shift change upon RNA binding. The role of R142 in RNA binding was tested by measuring the Kd of a mutant of Sxl-RBD1 + 2 in which R142 was replaced by alanine. This mutant lost the ability to bind RNA, showing a correlation with the chemical shift difference data. The RNA-binding affinities of two other mutants, F146A and T138I, were also shown to correlate with the NMR observations.

Microscopic pKa values of Escherichia coli thioredoxin.

Thiol:disulfide oxidoreductases have a CXXC motif within their active sites. To initiate the reduction of a substrate disulfide bond, the thiolate form of the N-terminal cysteine residue (CXXC) of this motif performs a nucleophilic attack. Escherichia coli thioredoxin [Trx (CGPC)] is the best characterized thiol:disulfide oxidoreductase. Previous determinations of the active-site pKa values of Trx have led to conflicting interpretations. Here, 13C-NMR spectroscopy, site-specific isotopic labeling, and site-directed mutagenesis were used to demonstrate that analysis of the titration behavior of wild-type Trx requires the invocation of microscopic pKa values for two interacting active-site residues: Asp26 (7.5 and 9.2) and Cys32 (CXXC; 7.5 and 9.2). By contrast, in two Trx variants, D26N Trx and D26L Trx, Cys32 exhibits a pKa near 7.5 and has a well-defined, single-pKa titration curve. Similarly, in oxidized wild-type Trx, Asp26 has a pKa near 7.5. In CVWC and CWGC Trx, Cys32 exhibits a single pKa near 6.2. In all five enzymes studied here, there is no evidence for a Cys35 (CXXC) pKa of < 11. This study demonstrates that a comprehensive approach must be used to unravel complex titration behavior of the functional groups in a protein.