Methods for Solving Highly Symmetric De Novo Designed Metalloproteins: Crystallographic Examination of a Novel Three-Stranded Coiled-Coil Structure Containing d-Amino Acids

The core objective of de novo metalloprotein design is to define metal-protein relationships that control the structure and function of metal centers by using simplified proteins. An essential requirement to achieve this goal is to obtain high resolution structural data using either NMR or crystallographic studies in order to evaluate successful design. X-ray crystal structures have proven that a four heptad repeat scaffold contained in the three-stranded coiled coil (3SCC), called CoilSer (CS), provides an excellent motif for modeling a three Cys binding environment capable of chelating metals into geometries that resemble heavy metal sites in metalloregulatory systems. However, new generations of more complicated designs that feature, for example, a d-amino acid or multiple metal ligand sites in the helical sequence require a more stable construct. In doing so, an extra heptad was introduced into the original CS sequence, yielding a GRAND-CoilSer (GRAND-CS) to retain the 3SCC folding. An apo-(GRAND-CSL12DLL16C)3 crystal structure, designed for Cd(II)S3 complexation, proved to be a well-folded parallel 3SCC. Because this structure is novel, protocols for crystallization, structural determination, and refinements of the apo-(GRAND-CSL12DLL16C)3 are described. This report should be generally useful for future crystallographic studies of related coiled-coil designs.

Investigation of substrate specificity of creatine kinase using chromium(III) and cobalt(III) complexes of adenosine 5'-diphosphate

The specificity for substrate binding to creatine kinase for metal-nucleotide complexes of the type Cr-(H2O)4-n(NH3)nADP (where n = 0, 3, or 4) and Co-(H2O)4-m(NH3)mADP (for m = 3 or 4) has been investigated over the pH range 5.5-7.8 with the delta-alpha, beta-bidentate diastereoisomers. These inert nucleotide complexes acted as competitive inhibitors vs. MgADP over this range. In addition, the pH dependence of the V, V/K, and Km values for MgADP has been determined. Metal-nucleotide binding to the enzyme is strongest below an approximate pK of 6.45 but again becomes pH independent above pH 7. This pK is not associated with the metal-nucleotide complex. Instead, we conclude that the pK of the acid-base catalyst (thought to be histidine) is about 6.45 in the absence of nucleotide but is raised to 7.2 in its presence. This perturbation of the pK may result from a protein conformational change that allows a hydrogen bond to form between the phosphorylated nitrogen of phosphocreatine and the acid-base catalyst. The pK of the water in Cr(H2O)(NH3)3ADP has been determined to be 6.6, and by comparison of the binding affinity of this complex with that of Cr(NH3)4ADP or Cr(H2O)4ADP, it can be deduced that the hydroxo species binds more strongly than the aquo complex. In general, chromium nucleotides are bound more strongly than cobalt complexes, and binding affinity increases as water replaces ammonia in the first coordination sphere of the metal. Both trends are a result of stronger hydrogen-bond interactions between the metal complex and protein.

Thermodynamic model for the stabilization of trigonal thiolato mercury(II) in designed three-stranded coiled coils

A thermodynamic model is presented that describes the binding of Hg(II) to de novo designed peptides, Tri L9C and Baby L9C, which were derived from the Tri family. The Tri peptides are based on the parent sequence Ac-NH-G(LKALEEK)(x)()G-CONH(2) and are known to form two-stranded coiled coils at low pH (pH <4) and three-stranded coiled coils at high pH (pH >7). Tri L9C (x = 4) contains a four heptad repeat sequence with cysteine in position 9 and leucines in the other a and d positions; Baby L9C (x = 3), which also has a cysteine in position 9 but is one heptad shorter than Tri L9C, was designed to form less stable helical coiled coils in solution. The free energies of coiled coil formation for Tri, Tri L9C, Baby Tri, and Baby L9C at pH 2.5 and 8.5 were determined by guanidinium denaturation titrations; Tri L9C was observed to be highly helical in the absence of denaturant at pH 8.5 while Baby L9C contained <20% helical content at pH 8.5, indicating a weakly associated or unassociated coiled coil. Size-exclusion chromatography (SEC) verified that Baby L9C was a monomer at pH 8.5. The helicity of Baby L9C was induced by addition of HgCl(2). The subsequent formation of a trigonal thiolato Hg(II) in the interior of a three-stranded coiled coil was verified by the presence of a characteristic HgS(3) UV band at 248 nm. Titrations of Tri L9C and Baby L9C into solutions of HgCl(2) at pH values between 7 and 9 were performed to extract binding constants. Global fits to the data employed a mechanism that involved initial binding of mercury to the peptides forming a two-stranded coiled coil with linear thiolato Hg(II) at [peptide]/[Hg] <2, followed by addition of a more weakly associated third helix to generate a three-stranded coiled coil. This mechanism would require the deprotonation of the third cysteine thiol to generate the trigonal thiolato Hg(II) at pH >7.5 [the pK(a) of the cysteine thiol in the presence of Hg(II)]. Support for this mechanism was given by the observation of a three-stranded coiled coil by SEC in a solution of Tri L9C at pH 7.0.

Preparation of site-differentiated mixed ligand and mixed ligand/mixed metal metallacrowns

Assembly reactions that can prepare reliably regioselective metallamacrocyclic complexes have been a target in the development of metallacrowns. To this end, a series of mixed ligand and mixed ligand/mixed metal metallacrowns have been synthesized in high yield and structurally characterized. Two distinct connectivities have been observed in these types of metallacrowns. The monomeric, vacant metallacrown with mixed ligand composition [12-MC(Ni(II)N(Hshi)2(pko)2-4)] (1a) shows the connectivity pattern [-O-Ni-O-N-Ni-N-]2 while the other Ni metallacrowns, [12-MC(Ni(II)N(shi)2(pko)2-4)] (2a) and the coupled [12-MC(Ni(II)N(shi))2(pko)2-4)][12-MC(Ni(II)N(shi))3(pko)-4)] (3a) fused metallacrowns as well as the mixed metal Mn-Ni metallacrown [12-MC(Ni(II)Mn(III)N(shi)2(pko)2-4)] (4a), follow the pattern [-Ni-O-N-]4. Also, three distinct arrangements of the chelate rings around the metal ions have been observed. The syntheses are completely general, allowing for the substitution of different ligands into the metallacrown core. Compounds 1 and 4 show the 6-5-6-5-6-5-6-5 arrangement, compounds 2 and 3(1) the 6-6-5-5-6-6-5-5, and the 3(2) component the 6-6-5-5-6-5-6-5. The obtained structures can be rationalized by balancing the charge at each metal site in the metallacrown. Variable temperature magnetic susceptibility measurements show that exchange interactions for all the compounds are weak and dominantly antiferromagnetic (e.g., 2a gives an exchange coupling of J = -1.2 cm(-1) with g = 2.2). In solution, the metallacrowns are shown to be stable both to decomposition and ligand exchange.

Mechanism for the reduction of the mixed-valent MnIIIMnIV[2-OHsalpn]2+ complex by tertiary amines

The mixed-valent dimanganese(III/IV) complex MnIIIMnIV(2-OHsalpn)2+, 1, is cleanly reduced in acetonitrile by aliphatic tertiary amines to give the dimanganese(III) product MnIII2(2-OHsalpn)2, 2. Thorough characterization of the organic reaction products shows that tributylamine is converted to dibutylformamide and propionaldehyde. Kinetic studies and radical trapping experiments suggest that this occurs via initial single-electron transfer from the amine to 1 coupled with C-H alpha proton transfer from the oxidized amine. EPR spectroscopy and base inhibition studies indicate that coordination of the amine to 1 is a critical step prior to the electron transfer step. Rate data and its dependence on the amine indicate that the ability of the amine to reduce 1 is correlated to its basicity rather than to its reduction potential. Weakly basic amines were unable to reduce 1 irrespective of their reduction potential. This was inferred to indicate that proton transfer from the amine radical cation is also important in the reduction of 1 by tertiary amines. Comparison of the activation energy with reaction thermodynamics indicates that proton transfer and electron transfer must be concerted to explain the rapidity of the reaction. The fate of the amine radical is dependent on the presence of oxygen, and labeling studies show that oxygen in the organic products arises from dioxygen, although incorporation from trace water was also observed. These data indicate that inhibition of the hydrolytic quenching of the amine radical in an aprotic solvent results in a different fate for the amine radical when compared to amine oxidation reactions in aqueous solution. The proposed mechanism gives new insight into the ability of amines with high reduction potential to reduce metal ions of lower potential. In particular, these data are consistent with the ability of small amines and certain amine-containing buffers to inhibit manganese-dependent oxygen evolution in photosynthesis, which arises in some cases as a result of manganese reduction and its concomitant loss from the PS II reaction center.

Recent advances in the understanding of the biological chemistry of manganese

Developments in manganese biochemistry have centered on the discovery of new manganese enzymes, X-ray analysis of binuclear manganese enzymes, and the discovery of new spectroscopic signatures for the oxygen-evolving complex. Despite these gains, many questions regarding the structure, composition and redox state of the oxygen-evolving complex remain unanswered.

The role of protonation and metal chelation preferences in defining the properties of mercury-binding coiled coils

To define the delicate interplay between metal chelation, protein folding and function in metalloproteins, a family of de novo-designed peptides was synthesized that self-assemble in aqueous solution to form two and three-stranded alpha-helical coiled coils. Each peptide contains a single Cys residue at an a or d position of the heptad repeat. Peptide association thus produces a Cys-rich coordination environment that has been used to bind Hg(II) ions. These peptides display a pH-dependent association, with trimers observed above the pKa of Glu side-chains and dimers below this value. Finite-difference Poisson-Boltzmann calculations suggest that the dimeric state decreases the unfavorable electrostatic interactions between positively charged Lys side-chains (relative to the trimer). The Cys-containing peptides bind Hg(II) in a position-dependent fashion. Cys at a positions form three-coordinate Hg complexes at high pH where the trimeric aggregation state predominates, and two-coordinate complexes at lower pH. A d position Cys, however, is only able to generate the two-coordinate complex, illustrating the difference in coordination geometry between the two positions in the coiled coil. The binding of Hg(II) was also shown to substantially increase the stability of the helical aggregates.

In vitro analysis of the interaction between sucralfate and ketoconazole

In healthy volunteers, the bioavailability of ketoconazole is significantly decreased during simultaneous administration with sucralfate. In an effort to address this problem, we examined the interaction between sucralfate and ketoconazole in aqueous solutions and in simulated gastric fluid (SGF) at various initial pHs (1, 2, 3, and 6) in the presence or absence of glutamic acid hydrochloride (GA). Samples from each solution were taken 30 min and 2 h after the addition of ketoconazole to evaluate the solubility of ketoconazole over the usual time period of maximal absorption of ketoconazole in humans. The addition of GA to SGF leads to an increase in solution acidity, while the pHs of SGF at a pH of 1, 2, or 3 are markedly increased by the addition of sucralfate. There is a net decrease in acidity from initial pHs for the pH 1, 2, and 3 solutions when GA and sucralfate are combined. The concentration of ketoconazole in SGF at pHs of 1, 2, 3, 4, and 6 was evaluated in order to assess the pH-dependent solubility properties of the drug in the absence of other interacting species. Regardless of the initial pH, combinations of GA plus ketoconazole showed high concentrations of ketoconazole (approximately 100%) in solution. In contrast, significant decreases in the concentration of soluble ketoconazole were observed when sucralfate was mixed with ketoconazole, and, in some cases, soluble ketoconazole was not detectable. The addition of GA to a mixture of sucralfate and ketoconazole leads to a significant increase in the concentration of solubilized ketoconazole. Nonetheless, important sucralfate-ketoconazole interactions are still observed. After 2 h, approximately 35% of the maximal ketoconazole concentration remained in solution. Comparison of the ketoconazole concentrations at different pHs with the predicted concentrations of the three protonation species of ketoconazole [H2(ketoconazole)(2+), H(ketoconazole)(+), or ketoconazole] showed no correlation. Therefore, the decrease in ketoconazole solubility is not simply a reflection of pH perturbation associated with the dissolution of sucralfate. The observed data are most consistent with a model that has H2(ketoconazole)(2+) or H(ketoconazole)(+) forming an electrostatic interaction with the sucralfate polyanion. The findings of this study suggest that the coadministration of sucralfate with other azole antifungal agents should be investigated.

The effect of protonation on [Mn(IV)(μ2-O)] 2 complexes

The series of complexes [Mn(IV)(X-SALPN)(μ2-O)]2, 1: X=5-OCH3; 2: X=H; 3: X=5-Cl; 4: X=3,5-diCl; 5: X=5-NO2, contain [Mn2O2](4+) cores with Mn-Mn separations of 2.7 Å. These molecules can be protonated to form [Mn(IV)(X-SALPN)(μ2-O,OH)]2 (+) in which a bridging oxide is protonated. The pKa values for the series of [Mn(IV)(X-SALPN)(μ2-O,OH)]2 (+) track linearly versus the shift in redox potential with a slope of 84 mV/pKa. This observation suggests that the [Mn2O2](4+) core can be considered as a unit in which the free energy of protonation is directly related to the ability to reduce the Mn(IV) ion. The marked sensitivity of the reduction potential to the presence of protons presents a mechanism in which an enzyme can control the oxidizing capacity of an oxo manganese cluster by the degree and timing of oxo bridge protonation.

Solution structure of Fe(II) cytochrome c551 from Pseudomonas aeruginosa as determined by two-dimensional 1H NMR

The solution structure of Fe(II) cytochrome c551 from Pseudomonas aeruginosa based on 2D 1H NMR data is reported. Two sets of structure calculations were completed with a combination of simulated annealing and distance geometry calculations: one set of 20 structures included the heme-peptide covalent linkages, and one set of 10 structures excluded them. The main-chain atoms were well constrained within the two structural ensembles (1.30 and 1.35 A average RMSD, respectively) except for two regions spanning residues 30-40 and 60-70. The results were essentially the same when global fold comparisons were made between the ensembles with an average RMSD of 1.33 A. In total, 556 constraints were used, including 479 NOEs, 53 volume constraints, and 24 other distances. This report represents the first solution structure determination of a heme protein by 2D 1H NMR and should provide a basis for the application of these techniques to other proteins containing large prosthetic groups or cofactors.

Sequential 1H NMR assignments of iron(II) cytochrome c551 from Pseudomonas aeruginosa

Sequence-specific 1H NMR resonance assignments for all but the C-terminal Lys 82 are reported for iron(II) cytochrome c551 from Pseudomonas aeruginosa at 25 degrees C and pH = 6.8. Spin systems were identified by using TOCSY and DQF-COSY spectra in 2H2O and 1H2O. Sequential assignments were made by using NOESY connectivities between adjacent amide, alpha, and beta protons. Resonances from several amino acids including His 16, Gly 24, Ile 48, and Met 61 experience strong ring-current shifts due to their placement near the heme. All heme protons, including the previously unassigned propionates, have been identified. Preliminary analysis of sequential and medium-range NOEs provides evidence for substantial amounts of helix in the solution structure. Long-range NOEs indicate that the folds in solution and crystal structures are similar. For one aromatic side chain (Tyr 27) that is close to the heme group we found a transition from hindered ring rotation at low temperature to rapid rotation at high temperature.

Structural characterization of the manganese(IV) Schiff-base complex MnIV(5-Cl-SALAHP)2

Bis[3-(5-chlorosalicylideneamino)propanolato-O,N,O']manganes e(IV) methanol solvate, [Mn(C10H10ClNO2)2].CH3OH, Mr = 510.3, monoclinic, P2(1)/c, a = 11.949 (2), b = 7.530 (2), c = 25.777 (6) A, beta = 105.75 (2) degrees, V = 2232.4 (8) A3, Z = 4, Dx = 1.518 g cm-3, lambda(Mo K alpha) = 0.7107 A, mu = 7.98 cm-1, F(000) = 1502, T = 300 K, R = 0.0343, wR = 0.032 for 2113 unique reflections with (I) greater than 3 sigma(I). The title complex MnIV(5-Cl-SALAHP)2 [5-Cl-SALAHP = 3-(5-chlorosalicylideneamino)propanolato] displays a regular octahedral geometry. The 5-Cl-SALAHP ligand is tridentate, forming a meridional chelate with one phenolato oxygen (Mn-Oavg = 1.90 A), one alkoxide oxygen (Mn-Oavg = 1.85 A) and one imine nitrogen (Mn-Navg = 2.02 A) coordinated to the metal. Important angles described by the six atoms bound to manganese are all very close to either 180 or 90 degrees except the N-Mn-N angle which is 174.7 degrees. Previous studies have shown that MnIV(5-Cl-SALAHP)2 displays a rhombic EPR spectrum with well-resolved 55Mn hyperfine structure on gx, gy and gz. In contrast, Mn(SALADHP)2 [SALADHP = 2-methyl-2-(salicylideneamino)-1,3-propanediolato] shows a broad, ill-defined signal at g = 5.15 and a weak g = 2 component. The different spectral forms result from the extent of distortion of the MnIV octahedron. The reported structure is of potential importance to the understanding of the photosynthetic water-oxidizing system.

Determination of the screw sense specificity of bovine liver fructokinase

Fructokinase from beef liver showed a clear reversal in specificity when the two isomers of ATP beta S were used as substrates with Mg2+ and Cd2+, with the Sp isomer having the higher V/K value with Mg2+ and the Rp isomer the higher value with Cd2+. The delta isomer of MgATP is thus the active form of the substrate. The substitution of sulfur for oxygen in the noncoordinated position of the beta-phosphate caused a 102-fold decrease in V/K over the value seen with MgATP, while substitution in the coordinated position gave a 21-fold decrease over the V/K value seen with CdATP. The Km values were little affected by sulfur substitution, showing that the wrong screw sense isomers were nonproductively bound almost as well as the correct ones. When ADP alpha S was used as a substrate in the reverse reaction, the Sp isomer showed the highest V/K value with both Mg2+ and Cd2+, suggesting that the metal ion is not coordinated to the alpha-phosphate during transphosphorylation. The failure of CrATP to act as a substrate for fructokinase suggests that the enzyme inserts one of its side chains into the inner coordination sphere of the metal ion during the reaction.

Stability constants of Mg2+ and Cd2+ complexes of adenine nucleotides and thionucleotides and rate constants for formation and dissociation of MgATP and MgADP

Stability constants for the Mg2+ and Cd2+ complexes of ATP, ADP, ATP alpha S, ATP beta S, and ADP alpha S have been determined at 30 degrees C and mu = 0.1 M by 31P NMR. Besides being of the utmost importance for determining species distributions for enzymatic studies, these constants allow an estimation of the preference of Cd2+ for sulfur vs. oxygen coordination in phosphorothioate complexes. Stability constants for Mg2+ complexes decreases when sulfur replaces oxygen (log K: ADP, 4.11; ADP alpha S, 3.66; ATP, 4.70; ATP alpha S, 4.47; ATP beta S, 4.04) because of (a) a statistical factor resulting from the loss of one potential phosphate oxygen ligand and (b) either an alteration in the charge distribution between oxygen and sulfur or destabilization of the chelate ring structure by loss of an internal hydrogen bond between an oxygen of coordinated phosphate and metal-bound water. Cd2+ complexes with sulfur-substituted nucleotides are more stable than those without sulfur (log K: ADP, 3.58; ADP alpha S, 4.95; ATP, 4.36; ATP alpha S, 4.42; ATP beta S, 5.44) because of the preferential binding of Cd2+ to sulfur rather than oxygen, which we estimate to be approximately 60 in CdADP alpha S and CdATP beta S. The proportion of tridentate coordination is estimated to be 50-60% in MgATP and MgATP beta S, approximately 27% in MgATP alpha S, approximately 16% in CdATP or CdATP beta S, but approximately 75% in CdATP alpha S. By analysis of the data of Jaffe and Cohn [Jaffe, E. K., & Cohn, M. (1979) J. Biol. Chem. 254, 10839], we conclude that the preference for oxygen over sulfur coordination to ATP beta S is 31 000 for Mg2+, 3100-3900 for Ca2+, and 158-193 for Mn2+. Proton NMR demonstrates that bidentate Cd2+ complexes form intramolecular chelates with the N-7 of adenine while Mg2+ nucleotides and the tridenate CdATP alpha S do not. An analysis of the 31P NMR line widths shows that the rate constants for dissociation of MgADP and MgATP are both 7000 s-1 while the association rate constants are 7 X 10(7) and 4 X 10(8) M-1 s-1, respectively. The observed dependence of the line width on nucleotide concentration is best explained by a base-stacking model at nucleotide concentrations above 5 mM.

Thermodynamic binding constants for gallium transferrin

Gallium-67 is widely used as an imaging agent for tumors and inflammatory abscesses. It is well established that Ga3+ travels through the circulatory system bound to the serum iron transport protein transferrin and that this protein binding is an essential step in tumor localization. However, there have been conflicting reports on the magnitude of the gallium-transferrin binding constants. Therefore, thermodynamic binding constants for gallium complexation at the two specific metal binding sites of human serum transferrin at pH 7.4 and 5 mM NaHCO3 have been determined by UV difference spectroscopy. The conditional constants calculated for 27 mM NaHCO3 are log K1 = 20.3 and log K2 = 19.3. These results are discussed in relation to the thermodynamics of transferrin binding of Fe3+ and to previous reports on gallium binding. The strength of transferrin complexation is also compared to that of a series of low molecular weight ligands by using calculated pM values (pM = -log [Ga-(H2O)6]) to express the effective binding strength at pH 7.4.

Iron supply to Escherichia coli by synthetic analogs of enterochelin

Synthetic analogs of enterochelin (enterobactin) were tested for their ability to support the growth of Escherichia coli K-12 under iron-limiting conditions. The cyclic compound MECAM [1,3,5-N.N'; N"-tris-(2,3-dihydroxybenzoyl)-triamino-methylbenzene] and its N-methyl derivative Me3MECAM promoted growth, whereas the 2,3-dihydroxy-5-sulfonyl derivatives MECAMS and Me3MECAMS were inactive. The same results were obtained with TRIMCAM [1,3,5-tris(2,3-dihydroxybenzoylcarbamido)-benzene] and TRIMCAMS (the 2,3-dihydroxy-5-sulfonyl derivative of TRIMCAM). However, the sulfonic acid-containing linear compound LICAMS [1,5,10-N,N', N"-tris(5-sulfo-2,3-dihydroxybenzoyl)-triaza-decane] supported growth. In contrast, LIMCAMC, in which the sulfonyl groups at the five position of LICAMS are replaced by carboxyl groups at the four position, was inactive. The uptake of the active analogs required the functions specified by the fepB, fesB, and tonB genes. Surprisingly, growth promotion of mutants lacking the enterochelin receptor protein in the outer membrane was observed. Only MECAM protected cells against colicin B (which kills cells after entering at the enterochelin uptake sites) and transported Fe3+ at about half the enterochelin rate.

Ferric ion sequestering agents: kinetics of iron release from ferritin to catechoylamides

The removal of ferric ion from the iron storage protein ferritin to synthetic catechoylamide sequestering agents has been studied using visible spectroscopy at 487 nm. One ligand which has been investigated in detail is N,N',N' ',-tris(2,3-dihydroxy-5-sulfobenzoyl)-1,5,10-triazadecane (3,4-LICAMS), which octahedrally coordinates the metal ion via six phenolic oxygens. For some related catechoylamide chelates, the percentage of iron removed after 6 h has been determined. These ligands incorporate various modifications, either on the catechol moiety or on the backbone structure of the ligand. Mobilization of iron by the catechoylamide ligands alone results in very slow exchange, and virtually no iron removal after 6 h. In contrast, addition of ascorbic acid to the reaction mixture facilitates iron exchange, with the release of 7% of the available iron in the same time span. Variation of the initial rate with ascorbic acid concentration results in Michaelis-Menten kinetics with Km = 1.7 . 10(-3) M and a maximal rate of 1.28 . 10(-7) M . min-1. The ascorbic acid-mediated rate was not affected by changing the catechoylamide ligand concentration, and was only slightly affected by variation of the ligand employed. These data are consistent with a multistep process which includes diffusion of a reductant into the ferritin inner core, reduction and possible chelation of the ferrous ion, diffusion out of the protein shell, and subsequent iron exchange with the catechoylamide molecule.