Studies on ribonucleoside-diphosphate reductase from Escherichia coli. The product dCDP is a competitive inhibitor and functions as a spectroscopic probe for the substrate binding site; demonstration by enzyme kinetics and 1H NMR.

Ribonucleoside-diphosphate reductase (EC 1.17.4.1) from Escherichia coli consists of two protein subunits, R1 of 171.5 kDa and R2 of 86.8 kDa, and catalyzes the reduction of all four common ribonucleoside diphosphates. In a search for ligands that bind weakly to the enzyme active site and may be in fast exchange suitable for NMR studies, we have found that the product dCDP is a competitive inhibitor. Kinetics with CDP as substrate shows Km = 4.8 x 10(-5) M and dCDP inhibits with Ki = 1.6 x 10(-4) M. With an assumed diffusion limited binding rate approximately less than 10(9) M-1s-1, the dissociation rate of dCDP would be approximately less than 10(5) s-1. In 1H-NMR experiments studying linewidths, i.e. spin-spin relaxation, dCDP is indeed demonstrated to be in fast exchange. Enzyme subunit R1 causes a line broadening of dCDP resonances. Unexpectedly less broadening was observed when subunit R2 combined with R1. No paramagnetic interaction from the tyrosyl radical of R2 could be detected. It is concluded that dCDP is a promising NMR probe for studies of active-site properties of the enzyme.     

Solving the structure of PTB in complex with pyrimidine tracts: an NMR study of protein-RNA complexes of weak affinities

NMR spectroscopy has proven to be a powerful tool for the structure determination of protein/RNA complexes. However, the quality of these structures depends critically on the number of unambiguous intermolecular and intra-RNA nuclear Overhauser effect (NOE) constraints that can be derived. This number is often limited due to exchange phenomena that can cause signal line broadening and the fact that unambiguous NOE assignments are challenging in systems that exchange between different conformations in the intermediate to fast exchange limit. These exchange processes can include exchange between free and bound form, as well as exchange of the ligand between different binding sites on the protein. Furthermore, for the large class of RNA metabolizing proteins that bind repetitive low-complexity RNA sequences in multiple register, exchange of the protein between these overlapping binding sites introduces additional exchange pathways. Here, we describe the strategy we used to overcome these exchange processes and to reduce significantly the line width of the RNA resonances in complexes of the RNA recognition motifs (RRMs) of the polypyrimidine tract-binding protein (PTB) in complex with pyrimidine tracts and hence allowed a highly precise structure determination. This method could be employed to derive structures of other protein/single-stranded nucleic acid complexes by NMR spectroscopy. Furthermore, we have determined the affinities of the individual RRMs of PTB for pyrimidine tracts of different length and sequence. These measurements show that PTB binds preferentially to long pyrimidine tracts that contain cytosine and hence confirm the structure of PTB in complex with RNA. Furthermore, they provide quantitative insight into the question of which pyrimidine sequences within alternatively spliced pre-mRNAs will be preferentially bound by PTB.     

Strategies for the uses of lanthanide NMR shift probes in the determination of protein structure in solutio. Application to the EF calcium binding site of carp parvalbumin.

The homologous sequences observed for many calcium binding proteins such as parvalbumin, troponin C, the myosin light chains, and calmodulin has lead to the hypothesis that these proteins have homologous structures at the level of their calcium binding sites. This paper discusses the development of a nuclear magnetic resonance (NMR) technique which will enable us to test this structural hypothesis in solution. The technique involves the substitution of a paramagnetic lanthanide ion for the calcium ion which results in lanthanide induced shifts and broadening in the 1H NMR spectrum of the protein. These shifts are sensitive monitors of the precise geometrical orientation of each proton nucleus relative to the metal. The values of several parameters in the equation relating the NMR shifts to the structure are however known as priori. We have attempted to determine these parameters, the orientation and principal elements of the magnetic susceptibility tensor of the protein bound metal, by studying the lanthanide induced shifts for the protein parvalbumin whose structure has been determined by x-ray crystallographic techniques. The interaction of the lanthanide ytterbium with parvalbumin results in high resolution NMR spectra exhibiting a series of resonances with shifts spread over the range 32 to -19 ppm. The orientation and principal elements of the ytterbium magnetic susceptibility tensor have been determined using three assigned NMR resonances, the His-26 C2 and C4 protons and the amino terminal acetyl protons, and seven methyl groups; all with known geometry relative to the EF calcium binding site. The elucidation of these parameters has allowed us to compare the observed spectrum of the nuclei surrounding the EF calcium binding site of parvalbumin with that calculated from the x-ray structure. A significant number of the calculated shifts are larger than any of the observed shifts. We feel that a refinement of the x-ray based proton coordinates will be possible utilizing the geometric information contained in the lanthanide shifted NMR spectrum.     

113Cd NMR study of bovine prothrombin fragment 1 and factor X

The interaction of Cd2+ with bovine prothrombin fragment 1, prothrombin intermediate 1, factor X, and a modified (Gla-domainless) factor X has been studied with 113Cd NMR. All the 113Cd resonances observed in this study were in the chemical shift range expected for oxygen ligands, suggesting that cadmium is binding at the same sites where calcium binds. Both fragment 1 and factor X displayed two major resonances, one near 10 ppm from 113Cd2+ that did not exchange rapidly with unbound 113Cd2+ (the high-affinity, or H, resonance) and one near -15 ppm from 113Cd2+ that exchanged rapidly with unbound 113Cd2+ (the low-affinity, or L, resonance). The difference between the chemical shift of the H resonance and the chemical shift range of -90 to -125 ppm that has been reported for three other small calcium-binding proteins is postulated to be due to different coordination geometries for monocarboxylate and dicarboxylate ligands; Cd2+ binds to fragment 1 and factor X through the dicarboxylate side chains of gamma-carboxyglutamate (Gla) residues. This allows contribution of only one oxygen per carboxyl group. At least one of the first few 113Cd2+ ions bound to fragment 1 did not appear in the 113Cd NMR spectrum until a total of five 113Cd2+ had been added. This could be due to exchange broadening of initial 113Cd2+ resonances due to sharing of ligands among several sites. Filling all sites would then restrict ligand exchange. Addition of Zn2+ displaced 113Cd2+ from the H resonance sites. Factor X did not display the interactions among ion binding sites proposed for fragment 1.     

Pre-steady-state and steady-state kinetic analysis of E. coli class I ribonucleotide reductase

E. coli ribonucleotide reductase (RNR) catalyzes the conversion of nucleoside diphosphates (NDPs) to dNDPs and is composed of two homodimeric subunits: R1 and R2. R1 binds NDPs and contains binding sites for allosteric effectors that control substrate specificity and turnover rate. R2 contains a diiron-tyrosyl radical (Y(*)) cofactor that initiates nucleotide reduction. Pre-steady-state experiments with wild type R1 or C754S/C759S-R1 and R2 were carried out to determine which step(s) are rate-limiting and whether both active sites of R1 can catalyze nucleotide reduction. Rapid chemical quench experiments monitoring dCDP formation gave k(obs) of 9 +/- 4 s(-1) with an amplitude of 1.7 +/- 0.4 equiv. This amplitude, generated in experiments with pre-reduced R1 (3 or 15 microM) in the absence of reductant, indicates that both monomers of R1 are active. Stopped-flow UV-vis spectroscopy monitoring the concentration of the Y(*) failed to reveal any changes from 2 ms to seconds under similar conditions. These pre-steady-state experiments, in conjunction with the steady-state turnover numbers for dCDP formation of 2-14 s(-1) at RNR concentrations of 0.05-0.4 microM (typical assay conditions), reveal that the rate-determining step is a physical step prior to rapid nucleotide reduction and rapid tyrosine reoxidation to Y(*). Steady-state experiments conducted at RNR concentrations of 3 and 15 microM, typical of pre-steady-state conditions, suggest that, in addition to the slow conformational change(s) prior to chemistry, re-reduction of the active site disulfide to dithiol or a conformational change accompanying this process can also be rate-limiting.     

Characterization of the EGF-like module pair 3-4 from vitamin K-dependent protein S using NMR spectroscopy reveals dynamics on three separate time scales and extensive effects from calcium binding

Protein S, a cofactor of anticoagulant activated protein C, exhibits three high-affinity Ca(2+)-binding sites in a region comprising four EGF modules. The EGF 3-4 module pair constitutes the smallest fragment that retains one high-affinity Ca(2+)-binding site and is therefore useful for investigation of the structural basis of the unusually high-affinity Ca(2+) binding compared to other EGF-containing proteins characterized so far. Extensive chemical shift effects caused by Ca(2+) binding to the EGF 3-4 module pair are observed, particularly from Ca(2+) binding to the high-affinity site in EGF 4. Ca(2+) binding to the high-affinity site in EGF 4 and the low-affinity site in EGF 3 is associated with slow and fast exchange on the NMR time-scale, respectively. We show the presence of two isoforms, characterized by a cis or trans Lys 167-Pro 168 peptide bond, that do not convert on time scales that were accessible to the experiments (k(ex) < 0.2 s(-1)). Both conformers have similar Ca(2+) affinities and backbone dynamics. Further, broadening of (1)H resonances involving residues in the major beta-sheet of EGF 3 and (15)N exchange terms, primarily in the N-terminal part of the protein, indicate the presence of slow exchange on a microsecond to millisecond time scale. (15)N spin relaxation data suggest that the module pair has a well-defined relative orientation between EGF modules 3 and 4 and has a significantly anisotropic rotational diffusion tensor in solution.     

The interaction of ellipticine derivatives with nucleic acids studied by optical and 1H-nmr spectroscopy: Effect of size of the heterocyclic ring system

The DNA interaction of derivatives of ellipticine with heterocyclic ring systems with three, four, or five rings and a dimethylaminoethyl side chain was studied. Optical spectroscopy of drug complexes with calf thymus DNA, poly [(dA-dT) · (dA-dT)], or poly [(dG-dC) · (dG-dC)] showed a 10 nm bathochromic shift of the light absorption bands of the pentacyclic and tetracyclic compounds upon binding to the nucleic acids, which indicates binding by intercalation. For the tricyclic compound a smaller shift of 1–3 nm was observed upon binding to the nucleic acids. Flow linear dichroism studies show that the geometry of all complexes is consistent with intercalation of the ring system, except for the DNA and poly [(dG-dC) · (dG-dC)] complexes of the tricyclic compound, where the average angle between the drug molecular plane and the DNA helix axis was found to be 65°. One-dimensional ¹H-nmr spectroscopy was used to study complexes between d(CGCGATCGCG)₂ and the tricyclic and pentacyclic compounds. The results on the pentacyclic compound show nonselective broadening due to intermediate chemical exchange of most oligonucleotide resonances upon drug binding. The imino proton resonances are in slow chemical exchange, and new resonances with upfield shifts approaching 1 ppm appear upon drug binding, which supports intercalative binding of the pentacyclic compound. The results on the tricyclic compound show more rapid binding kinetics and very selective broadening of resonances. The data suggest that the tricyclic compound is in an equilibrium between intercalation and minor groove binding, with a preference to bind close to the AT base pairs with the side chain residing in the minor groove. © 1994 John Wiley & Sons, Inc.     

Manganese(II) as a paramagnetic probe of the tertiary structure of transfer RNA.

The effect of manganese on both the low field (10--15 ppm) and the high field (o--3 ppm) NMR spectra of unfractionated tRNA and yeast tRNAPhe has been investigated. Trace amounts of Mn2+ cause selective broadening of resonances which are assigned to specific tertiary interactions. The order in which resonances broaden is the same as the order in which they are stabilized by the addition of magnesium, namely s4U8 - A14, U33 and A58 - T54. From this we conclude that three of the strong binding sites probably are the same for both Mn2+ and Mg2+, and that these sites are located close to the tertiary interactions which are stabilized by the strongly bound metals. The broadening data, taken in conjunction with published X-ray data on yeast tRNAPhe, permit us to suggest some plausible locations for the strong binding sites.     

Characterisation of the electron self-exchange rates in hexametaphosphate-cytochrome-c aggregates measured using high-resolution 1H-NMR spectroscopy.

1H-NMR spectroscopy has been used to measure the rate of unimolecular electron exchange between cytochrome c molecules in protein aggregates stabilised by the addition of sodium hexametaphosphate. The average intracomplex electron exchange rate is measured from line broadening of hyperfine-shifted resonances of ferricytochrome c in an equimolar mixture of reduced and oxidised protein. The line-broadening due to electron exchange is significantly greater than that due to protein aggregation and reaches a maximum value between 1-2 mol hexametaphosphate/mol protein. Significantly the exchange-induced broadening is a first-order process and is directly proportional to the size of the cytochrome c oligomer. From the temperature dependence of exchange broadening the activation enthalpy was estimated to be 75.8 kJ mol-1 whereas the activation entropy was 295 J mol-1 K-1 for a dimer of cytochrome c at a hexametaphosphate/protein molar ratio of 1. Both activation parameters decrease in magnitude as the order of the cytochrome c oligomer increases. The rates of intracomplex electron exchange in Saccharomyces cerevisiae iso-2 and Candida krusei cytochromes c are lower than that of the horse protein, implying that primary sequence plays a fundamental part in determining the rate of exchange. The relevance of these observations is discussed in terms of the function of cytochrome c.     

Carbon-13 nuclear magnetic resonance study of microtubule protein: evidence for a second colchicine site involved in the inhibition of microtubule assembly

A 13C nuclear magnetic resonance study of bovine microtubule protein was carried out at 43 kG in the presence and absence of colchicine 13C labeled at the tropolone methoxy. Analysis indicated that tubulin has at least two colchicine binding sites: a quasi-irreversibly bound, high-affinity site (i.e., the KD less than 5 microM site generally accepted as the site of colchicine action) as well as a low-affinity site(s) (KD approximately 650 microM) with which free colchicine rapidly exchanges (greater than 100 s-1). The methoxy resonance is broadened to different apparent extents as a result of binding at these two sites (50- vs. 150-Hz broadening for the high- and low-affinity sites, respectively) but undergoes no change in chemical shift upon binding. The low-affinity sites are interpreted to be analogous to the sites deduced by Schmitt and Atlas [Schmitt, H., & Atlas, D. (1976) J. Mol. Biol. 102, 743-758] from labeling studies using bromocolchicine. These sites are likely to be the sites responsible for the abrupt halt in microtubule assembly ("capping") observed at high colchicine concentrations (greater than 20 microM)--a qualitatively different behavior from that observed at low colchicine concentrations [Sternlicht, H., Ringel, I., & Szasz, J. (1983) Biophys. J. 42, 255-267]. Carbon-13 spectra from the aliphatic carbons of microtubule protein consists of narrow resonances--many with line widths less than 30 Hz--superimposed on a broad background. The narrow resonances were assigned to flexible regions in nontubulin proteins [microtubule-associated proteins (MAPs)], in accord with an earlier 1H nuclear magnetic resonance study of microtubule protein [Woody, R. W., Clark, D. C., Roberts, G. C. K., Martin, S. R., & Bayley, P. M. (1983) Biochemistry 22, 2186-2192]. This assignment was supported by 13C NMR analysis of phosphocellulose-purified (MAP-depleted) tubulin as well as heat-stable MAPs. Aliphatic carbons in the MAP preparations were characterized by narrow resonances indicative of carbons with considerable motional freedom whereas the aliphatic regions of phosphocellulose-purified tubulin were, for the most part, characterized by broad resonances indicative of carbons with restricted mobility. However, a moderately narrow resonance (approximately less than 50-Hz line width) coincident with the C gamma resonance of glutamate was detected in 13C NMR spectra of tubulin which indicated that a fraction of the glutamic acid residues is relatively mobile.(ABSTRACT TRUNCATED AT 400 WORDS)     

1H, 113Cd, and 31P NMR of osteocalcin (bovine gamma-carboxyglutamic acid containing protein)

The 1H (500-MHz), 113Cd (44-MHz), and 31P (81-MHz) NMR spectra of the bovine gamma-carboxyglutamate- (Gla-) containing protein osteocalcin and its Ca(II) and Cd(II) complexes in solution have been obtained. The 1H NMR spectrum of the native protein shows narrow resonances and a highly resolved multiplet structure suggesting rotational freedom of the side chains. In comparison to the simulated 1H NMR spectrum of a random polypeptide chain of the same amino acid composition, there is moderate chemical shift dispersion, indicating some conformational restraints to be present. Ca(II) binding broadens all 1H resonances, so severely at four Ca(II) ions per molecule that few structural conclusions can be made. Cd(II) substituted for Ca(II) has the same effect, and 113Cd NMR shows the Cd(II) to be in intermediate chemical exchange on the chemical shift time scale. Estimates of the chemical exchange rates required for 1H and 113Cd line broadening suggest a range of Kd values for the metal ion complexes from 10(-6) M to as high as 10(-3) M depending on the number of metal ions bound. Alternatively, 1H line broadening could be explained by relatively slow conformational fluxes in the protein induced by labile metal ion binding to one or more sites. Cd(II) when used to form a cadmium-phosphate mineral analogous to hydroxylapatite results in a crystal lattice that removes osteocalcin from solution just as effectively as hydroxylapatite. 113Cd(II) exchange at the binding sites of osteocalcin in solution is slowed dramatically by the addition of HPO4(2-). 31P NMR shows the interaction of phosphate with the protein to require the metal ion.(ABSTRACT TRUNCATED AT 250 WORDS)     

Dynamic NMR studies of ligand-receptor interactions: design and analysis of a rapidly exchanging complex of FKBP-12/FK506 with a 24 kDa calcineurin fragment.

Dynamic NMR methods, such as differential line broadening and transferred NOE spectroscopy, are normally reserved for the study of small molecule ligand interactions with large protein receptors. Using a combination of isotope labeling and isotope edited NMR, we have extended these techniques to characterize interactions of a much larger protein/drug complex, FKBP-12/ FK506 with its receptor protein, calcineurin. In order to examine this multicomponent system by dynamic NMR methods, the 93 kDa, tightly bound FKBP-12/FK506/Cn complex was replaced with a lower affinity, rapidly exchanging system consisting of FKBP-12/FK506 (13 kDa), recombinant calcineurin subunit B (CnB) (20 kDa), and a synthetic peptide (4 kDa) corresponding to the B binding domain (BBD) of calcineurin catalytic subunit A (CnA). Analysis of 1H-13C HSQC data acquired for the FKBP-12/ 13C-FK506 and FKBP-12/13C-FK506/CnB/BBD complexes indicates that FKBP-12/FK506 and CnB/BBD are in fast exchange in the quaternary complex. Comparison of proton line widths shows significant broadening of resonances along the macrocycle backbone at 13-CH, 13-OMe, 15-OMe, 18-CH2, 20-CH, 21-CH, and 25-Me, as well as moderate broadening on the macrocycle backbone at 17-Me, 24-CH, and the pyranose 12-CH2 protons. The tri-substituted olefin and cyclohexyl groups also show moderate broadening at the 27-Me, 28-CH, and 30-CH2 positions, respectively. Unexpectedly, little line broadening was observed for the allyl resonances of FK506 in the quaternary complex, although 13C longitudinal relaxation measurements suggest this group also makes contacts with calcineurin. In addition, intermolecular transfer NOE peaks were observed for the allyl 37-CH2, 21-CH, 30-CH2, 13-OMe, 15-OMe, 17-Me, 25-Me, and 27-Me groups, indicating that these are potential sites on the FK506 molecule that interact with calcineurin.     

The pH dependence of CD2 domain 1 self-association and 15N chemical exchange broadening is correlated with the anomalous pKa of Glu41

We have previously shown using (15)N nuclear relaxation measurements that the concentration-dependent rotational correlation time and chemical exchange broadening for selected resonances of rat CD2 domain 1 (CD2d1) are consistent with a model of low-affinity self-association of the protein molecules. The exchange broadening data, which at high protein concentrations highlight selected nuclei in the major C'-C-F-G beta-sheet face of the immunoglobulin fold, implicate a surface reminiscent of the major lattice contact within crystals of the intact CD2 ectodomain. In a separate study, we have also demonstrated that the beta-strand C' surface-exposed residue Glu41 possesses an anomalously elevated acidity constant (pK(a) = 6.7 at a protein concentration of 1.2 mM). Mutagenesis studies showed that the close contact of residue Glu41 with Glu29 (beta-strand C) is the primary cause of the high pK(a). However, the measured pK(a) of Glu41 also shows a weak dependence on protein concentration, implicating Glu41 in the mechanism of CD2d1 self-association. In the study presented here, we demonstrate a correlation of the pH dependence of the chemical shift and (15)N nuclear relaxation parameters measured for wild-type and mutant forms of CD2d1 with pH and the protonation state of Glu41. Self-association of CD2d1 molecules is promoted whenever the side chain charge of residue 41 is neutralized. These observations are consistent with a model for CD2d1 self-association that corresponds to the crystal structure lattice contact where the interatomic distances are consistent with Glu41 being in the protonated state. This study reinforces the conclusion that residue-specific chemical exchange broadening of protein resonances can arise from weak self-association phenomena. In addition, the electrostatic profile of rat CD2 interfacial residues parallels that of the homologous human CD2 in a manner that suggests a rationalization of similar exchange broadening observations.     

The identification of cation-binding domains on the surface of microsomal cytochrome b5 using 1H-NMR paramagnetic difference spectroscopy.

One-dimensional and two-dimensional 1H-NMR methods and paramagnetic difference spectroscopy have defined cation binding domains on the surface of the tryptic fragment of microsomal cytochrome b5. The addition of tris(ethylenediamine) chromium(III) [Cr(en)3(3+)] to solutions of ferricytochrome b5 reveals at least three distinct sites on the surface of the protein to which highly charged cations may bind (20 mM phosphate pH 7.0, T = 300 K). Surprisingly only one of these sites is located close to the haem edge region of the protein, whilst the remaining two sites are more remote. Site I contains the exposed haem C13 propionate and a series of carboxylate residues that includes glutamates 37, 38, 43, 44, and 48. Sites II and III are located away from the haem edge region and are delineated by the broadening of aromatic resonances of histidines 26 and 80. Further investigation of the interaction between Cr(en)3(3+) and cytochrome b5 using two-dimensional double-quantum-filtered correlated spectroscopy shows that resonances assigned to Glu59, Asp60, Glu79, Asp82 and Asp83 are broadened with the distribution of these charged side chains correlating with the relaxation broadening observed from one-dimensional experiments. In a binary complex with ferricytochrome c, Cr(en3(3+) broadens many cytochrome b45 resonances including the haem propionates, His26, Ala54, Thr55 and His80. Although the pattern of line-broadening of resonances at sites II and III is unaltered by complex formation, cytochrome c shields residues at site I, the haem edge site. The results indicate that the interaction between cytochrome b5 and c in a binary complex involves multiple protein configurations.     

Two-dimensional 1H nuclear magnetic resonance studies on the gene V-encoded single-stranded DNA-binding protein of the filamentous bacteriophage IKe. II. Characterization of the DNA-binding wing with the aid of spin-labelled oligonucleotides

The DNA-binding domain of the single-stranded DNA-binding protein IKe GVP was studied by means of 1H nuclear magnetic resonance, through use of oligonucleotides of two and three adenyl residues in length, that were spin-labelled at their 3' and/or 5' termini. These spin-labelled ligands were found to cause line broadening of specific protein resonances when bound to the protein, although they were present in small quantities, i.e. of the order of 0.04 molar equivalent and less. The line broadening of protein resonances was made manifest by means of difference one and two-dimensional spectroscopy. Difference one-dimensional experiments revealed line broadening of the same protein resonances upon binding of either 3' or 5' spin-labelled oligonucleotides. Evidence in favour of the existence of a fixed 5' to 3' orientation in the binding of oligonucleotides to the protein surface was therefore not obtained from the spin-labelled oligonucleotide binding studies. Residue-specific assignments of broadened resonances could not, or could only sparsely, be derived from the difference one-dimensional spectra, because of the tremendous overlap in the aliphatic region of the spectrum. In contrast, such assignments were easily obtained from the difference two-dimensional spectra, which were recorded by means of both total correlated spectroscopy and nuclear Overhauser effect spectroscopy. Difference signals were detected for 15 spin systems; ten out of these were assigned to the residues I29, Y27, S20, G18, R16, T28, K22, Q21, V19 and S17 in the amino acid sequence of IKe GVP; the other five spin systems could be assigned to a phenylalanyl residue, an arginyl or lysyl residue, an aspartic acid or asparagyl residue, a glycyl residue and a glutamic acid or glutamyl residue. From the evaluation of the relative difference signals, it was concluded that the direct surroundings of the spin-label group of the labelled oligonucleotide in the bound state is composed of the first five residues in the former group of residues and the five residues in the latter group.(ABSTRACT TRUNCATED AT 400 WORDS)     

DNA-binding properties of gene-5 protein encoded by bacteriophage M 13. 1. The kinetics of the dissociation of gene-5-protein.polynucleotide complexes upon addition of salt

The irreversible dissociation kinetics of complexes of M13-encoded gene-5 protein with the polynucleotides poly(dA) and M13 DNA was studied by means of stopped-flow experiments. A linear decay was found for all gene-5-protein.poly(dA) complexes and for the gene-5-protein.M13 DNA complexes for which the DNA lattice was completely saturated at the beginning of the dissociation experiments. Only at the end of the dissociation curve was a deviation from linearity observed. A single-exponential decay was found for the dissociation of gene-5-protein.M13 DNA complexes when the DNA was not completely saturated initially. These results could be interpreted by assuming that dissociation of bound protein is only possible from isolated binding sites, while during the dissociation, rearrangement of bound protein clusters takes place continuously, including the formation of newly isolated bound protein. This redistribution results from a translocation of the protein along the lattice, which, for the poly(dA) complex, is fast with respect to the dissociation step, but which is slow for the M13 DNA complex. During this process the equilibrium cluster distribution predicted by the theory of McGhee and Von Hippel is not maintained. The binding of gene-5 protein to poly(dA) or poly(dT) does not result in a broadening of the nucleotide resonances in the NMR spectra of these polynucleotides, as had been observed for E. coli DNA-binding protein and interpreted as an indication for a high rate of translocation of the protein on the polynucleotide. The absence of line broadening for gene-5-protein.polynucleotide complexes is caused by the high binding cooperativity. As a consequence the majority of the protein molecules are bound in a cluster which makes the concentration of isolated bound protein very low. This results in a decrease of the signal/noise ratio at higher degrees of binding, but does not lead to line broadening while fast translocation still occurs.     

Chloride binding to photosystem II in the dark is in slow exchange

We have studied the 35Cl- NMR line broadening in the presence of photosystem II (PS II) membranes from spinach in the dark. In the presence of NH3 (which other work has shown to competitively inhibit chloride binding to PS II) we observed no decrease in 35 Cl- linewidths. We conclude that binding of Cl- to the O2 evolving center in PS II in the dark (previously demonstrated by EPR) is in slow exchange on the NMR timescale. We assign the observed line broadening to interaction with non-specific binding sites and with free paramagnetics.     

Structural signatures of the complex formed between 3-nitro-4-hydroxybenzoate and the Zn(II)-substituted R(6) insulin hexamer

3-Nitro-4-hydroxybenzoate (3N4H) is a probe of the structure and dynamics of the metal-centered His B10 assembly sites of the insulin hexamer. Each His B10 site consists of a approximately 12 A-long cavity situated on the threefold symmetry axis. These sites play an important role in the storage and release of insulin in vivo. The allosteric behavior of the insulin hexamer is modulated by ligand binding to the His B10 zinc sites and to the phenolic pockets. Binding to these sites drives transitions among three allosteric states, designated T(6), T(3)R(3), and R(6). Although a wide variety of mono anions bind to the His B10 zinc sites of R(3), X-ray structures of ligands complexed to this site exist only for H(2)O, Cl(-), and SCN(-). This work combines one- and two-dimensional (1)H NMR and UV-Vis absorbance studies of the structure and dynamics of the 3N4H complex, which establish the following: (1). relative to the NMR time scale, 3N4H exchange between free and bound states is slow, while flipping among three equivalent orientations about the site threefold axis is fast; (2). binding of 3N4H perturbs resonances within the His B10 zinc site and generates NOEs between ligand resonances and the insulin C-alpha and side chain resonances of ValB2, AsnB3, LeuB6, and CysB7; and (3).3N4H exchange for other ligands is limited by a protein conformational transition. These results are consistent with coordination of the 3N4H carboxylate to the His B10 zinc ion and van der Waals interactions with Val B2, Asn B3, Leu B6, and Cys A7.     

EPR investigation of the Mn(II) binding sites in glutamine synthetase (Escherichia coli W). II. Intermediate-affinity binding sites.

The nature of the intermediate-affinity (n2) Mn(II) binding sites in glutamine synthetase [EC 6.3.1.2] has been studied as a function of adenylylation in a variety of enzyme-metal complexes by EPR. In the absence of nucleotide the n2 Mn(II) environment is nearly isotropic, the Mn(II) bonds are highly ionic, and the interaction distance R greater than or equal to 12-14 A. Nucleotide binding at the n2 Mn(II) site renders the n2 Mn(II) signal unobservable and causes a reduction in signal amplitude (approximately 30%) and line broadening (approximately 6 G) at the high-affinity (n1) Mn(II) site. This behavior indicates that nucleotide binding induces a conformational change in the enzyme which brings the previously distant n1 and n2 sites into closer proximity (R less than or equal to 8-11 A), possibly for the purpose of activating the nucleotide for direct phosphoryl transfer to L-glutamate. In line with this suggestion, the broad, unresolved resonances in complexes containing both L-methionine SR-sulfoximine (MSOX) and nucleotide may result from the phosphorylation of MSOX. The n2 Mn(II) site is not affected by adenylylation in all the enzyme-metal complexes studied, which suggests that the regulatory effects of adenylylation may only act at the n1 Mn(II) sites.     

Changes in structure and dynamics of the Fv fragment of a catalytic antibody upon binding of inhibitor

Binding of the product inhibitor p-nitrophenol to the monoclonal esterolytic antibody NPN43C9 has been investigated by performing NMR spectroscopy of the heterodimeric variable-domain fragment (Fv) of the antibody in the presence and absence of inhibitor. Structural information from changes in chemical shift upon binding has been related to the changes in local dynamics in the active site of the catalytic antibody using NMR relaxation measurements. Significant changes in the chemical shifts of the backbone resonances upon binding extend beyond the immediate vicinity of the antigen binding site into the interface between the two associated polypeptides that form the Fv heterodimer, a possible indication that the binding of ligand causes a change in the relative orientations of the component light (V(L)) and heavy (V(H)) chain polypeptides. Significant differences in backbone dynamics were observed between the free Fv and the complex with p-nitrophenol. A number of resonances, including almost all of the third hypervariable loop of the light chain (L3), were greatly broadened in the free form of the protein. Other residues in the antigen-binding site showed less broadening of resonances, but still required exchange terms (R(ex)) in the model-free dynamics analysis, consistent with motion on a slow timescale in the active site region of the free Fv. Binding of p-nitrophenol caused these resonances to sharpen, but some R(ex) terms are still required in the analysis of the backbone dynamics. We conclude that the slow timescale motions in the antigen-binding site are very different in the bound and free forms of the Fv, presumably due to the damping of large-amplitude motions by the bound inhibitor.