A time-resolved fluorescence study of azurin and metalloazurin derivatives

Nickel and cobalt derivatives of Pseudomonas fluorescens (ATCC 13525) azurin were prepared and their steady-state fluorescence and time-resolved fluorescence monitored. Like the copper-containing native protein, the fluorescence decay of both metallo derivatives was best fit to a sum of three exponentials, whereas the apoazurin from which they were prepared obeyed single-exponential decay kinetics. However, comparison of the lifetimes and fractional of each of the components in these derivatives to those in the oxidized and reduced native proteins revealed significant differences. These results suggest that the presence of a metal center in azurin imparts a conformational heterogeneity which is strongly dependent on the nature of the metal center. Further, the results are used to comment on current ideas concerning the geometry of the active site in this redox protein.     

Conformational heterogeneity of the copper binding site in azurin. A time-resolved fluorescence study.

Comparison of the fluorescence spectra and the effect of temperature on the quantum yields of fluorescence of Azurin (from Pseudomonas fluorescens ATCC-13525-2) and 3-methylindole (in methylcyclohexane solution) provides substantive evidence that the tryptophan residue in azurin is completely inaccessible to solvent molecules. The quantum yields of azurin (CuII), azurin (CuI), and apoazurin (lambda ex = 291 nm) were 0.052, 0.054, and 0.31, respectively. Other evidence indicates that there is no energy transfer from tyrosine to tryptophan in any of these proteins. The fluorescence decay behavior of each of the azurin samples was found to be invariant with emission wavelength. The fluorescences of azurin (CuII) and azurin (CuI) decay with dual exponential kinetics (tau 1 = 4.80 ns, tau 2 = 0.18 ns) while that of apoazurin obeys single exponential decay kinetics (tau = 4.90). The ratio of pre-exponentials of azurin (CuII), alpha 1/alpha 2, is found to be 0.25, and this ratio increases to 0.36 on reduction to azurin (CuI). The results are interpreted as originating from different interactions of the tryptophan with two conformers of the copper-ligand complex in azurin.     

Fourier transform infrared investigation of the Escherichia coli methionine aporepressor

This study represents the first physicochemical analysis of the recently cloned methionine repressor protein (Met aporepressor) from Escherichia coli. Infrared spectrometry was used to investigate the secondary structure and the hydrogen-deuterium exchange behavior of the E. coli Met aporepressor. The secondary structure of the native bacterial protein was derived by analysis of the amide I mode. The amide I band contour was found to consist of five major component bands (at 1625, 1639, 1653, 1665, and 1676 cm-1) which reflect the presence of various substructures. The relative areas of these component bands are consistent with a high alpha-helical content of the peptide chain secondary structure in solution (43%) and a small amount of beta-sheet structure (7%). The remaining substructure is assigned to turns (10%) and to unordered (or less ordered) structures (40%). The temperature dependence of the infrared spectra of native Met aporepressor in D2O medium over the temperature interval 20-80 degrees C indicates that there are two discrete thermal events: the first thermal event, centered at 42 degrees C, is associated with the hydrogen-deuterium exchange of the hard-to-exchange alpha-helical peptide bonds accompanied by a partial denaturation of the protein, while the second event, centered around 50 degrees C, represents the irreversible thermal denaturation of the protein.     

A pH dependence study on the unfolding and refolding of apoazurin

Azurin, a small blue copper protein from the bacterial species Pseudomonas aeruginosa, is mostly a β-sheet protein arranged into a single domain. Previous folding studies have shown that the equilibrium denaturation of the holoprotein follows a two-state process; however, upon removal of the copper, the denaturation had been reported to follow a three-state process. The two unfolding transitions measured for apoazurin had been thought to arise from two different folding domains. However, in the present work, we found that the denaturation of apoazurin occurs over a single transition and we determined the folding free energy to be −27.8±2.4 kJ mol⁻¹. From this measurement along with measurements previously reported for the unfolding of the holoazurin, we were able to determine that Cu(II) and Cu(I) stabilize the native structure by 25.1±6.9 kJ/mol and 12.9±8.1 kJ/mol, respectively. It is our contention that the second transition displayed in the denaturation curves previously reported for apoazurin arise from protein heterogeneity—in particular, from the presence of Zn(II) azurin. We extended our investigation into the denaturation of Zn(II) azurin at pH 6.0 and 7.5. The equilibrium denaturation studies show that the zinc ion significantly stabilizes the native-state structure at pH 7.5 and very little at the lower pH. We attribute the decrease in the stabilizing effect of the zinc ion with decreasing pH to the protonation of two histidinyl side chains. When protonated the ligands, His 46 and His 117, are incapable of binding a metal ion. Further, comparing the denaturation curves of Zn(II) azurin measured by circular dichroism with those measured by fluorescence indicates that the denaturation of Zn(II) azurin is far less simple than the denaturation of apoazurin.     

Homogeneity and variability in the structure of azurin molecules studied by fluorescence decay and circular polarization.

The fluorescence decay of apoazurin derived from Pseudomonas aeruginosa is monoexponential. By this criterion the population of molecules of apoazurin is homogeneous. The emission anisotropy factor and the absorption anisotropy factor at the red edge of the absorption band assume similar values, showing that the tryptophan residue in apoazurin has the same asymmetric environment both in the ground and excited states. This finding suggests tight packing of the protein at the tryptophan environment. Native azurin does not decay monoexponentially. Moreover, comparison between the quantum yield calculated from the decay kinetics and the one measured directly shows that the majority of the azurin molecules are not fluorescent. There is thus variability in the structure of azurin molecules with an equilibration time that is longer than the fluorescence lifetime. Different asymmetric environment was found for the tryptophan residue in oxidized and reduced holoprotein and in apoazurin, as studied by the circular polarization of the fluorescence. D(2)O increases the fluorescence lifetime of apoazurin by 6 percent, compared to the lifetime in H(2)O solution; therefore water molecules may have access to the tryptophan residue, though the latter is situated in a hydrophobic environment.     

Confirmation that multiexponential fluorescence decay behavior of holoazurin originates from conformational heterogeneity

Homologous azurins from Pseudomonas fluorescens (ATCC 13525) and Pseudomonas aeruginosa (ATCC 10145) were examined by a number of electrophoretic techniques, and their copper to protein stoichiometry was determined by atomic absorption and amino acid analysis. Provided that the spectral ratio (A620/A280 or A625/A280) was 0.53 and there was no evidence of a Soret band in the absorption spectrum, then these criteria can be used to judge the homogeneity of the azurin sample. If the spectral ratio was less than 0.50, evidence suggested a nonreconstitutable, non-trypsin-digestible apoazurin was present. The fluorescence decay of these homogeneous holoazurins included three components, not two as previously reported [Szabo, A. G., et al. (1983) Biophys. J. 41, 233-244]. Whereas the decay times were nearly the same for the azurins from the different sources, the fractional fluorescence of each component varied with the azurin measured. The fluorescence of the corresponding apoazurins, prepared by a refined procedure, obeyed monoexponential decay kinetics. The temperature and pH effects on the fluorescence behavior of these homologous azurins are presented with the pH study suggesting an influence by a group which titrates between pH 5 and pH 7. When taken together these results confirm that the multiexponential decay behavior originates from conformational heterogeneity and not from contamination by an apo form.     

The role of the native lipids and lattice structure in bacteriorhodopsin protein conformation and stability as studied by temperature-dependent Fourier transform-infrared spectroscopy

We report the effect of partial delipidation and monomerization on the protein conformational changes of bacteriorhodopsin (bR) as a function of temperature. Removal of up to 75% of the lipids is known to have the lattice structure of the purple membrane, albeit as a smaller unit cell, whereas treatment by Triton monomerizes bR into micelles. The effects of these modifications on the protein secondary structure is analyzed by monitoring the protein amide I and amide II bands in the Fourier transform-infrared (FT-IR) spectra. It is found that removal of the first 75% of the lipids has only a slight effect on the secondary structure at physiological temperature, whereas monomerizing bR into micelles alters the secondary structure considerably. Upon heating, the bR monomer is found to have a very low thermal stability compared with the native bR with its melting point reduced from 97 to 65 degrees C, and the pre-melting transition in which the protein changes conformation in native bR at 80 degrees C could not be observed. Also, the N[bond]H to N[bond]D exchange of the amide II band is effectively complete at room temperature, suggesting that there are no hydrophobic regions that are protected from the aqueous medium, possibly explaining the low thermal stability of the monomer. On the other hand, 75% delipidated bR has its melting temperature close to that of the native bR and does have a pre-melting transition, although the pre-melting transition occurs at significantly higher temperature than that of the native bR (91 degrees C compared with 80 degrees C) and is still reversible. Furthermore, we have also observed that the reversibility of this pre-melting transition of both native and partially delipidated bR is time-dependent and becomes irreversible upon holding at 91 degrees C between 10 and 30 min. These results are discussed in terms of the lipid and lattice contribution to the protein thermal stability of native bR.     

Photophysics of metalloazurins

The fluorescence lifetimes of Cu(II), Cu(I), Ag(I), Hg(II), Co(II), and Ni(II) azurin Pae from Pseudomonas aeruginosa and Cu(II), Cu(I), and Hg(II) azurin Afe from Alcaligenes faecalis were measured at 295 K by time-correlated single-photon counting. In addition, fluorescence lifetimes of Cu(II) azurin Pae were measured between 30 and 160 K and showed little change in value. Ultraviolet absorption difference spectra between metalloazurin Pae and apoazurin Pae were measured, as were the fluorescence spectra of metalloazurins. These spectra were used to determine the spectral overlap integral required for dipole-dipole resonance calculations. All metalloazurins exhibit a reduced fluorescence lifetime compared to their respective apoazurins. Forster electronic energy transfer rates were calculated for both metalloazurin Pae and metalloazurin Afe derivatives; both enzymes contain a single tryptophyl residue which is located in a different position in the two azurins. These azurins have markedly different fluorescence spectra, and electronic energy transfers occur from these two tryptophyl sites with different distances and orientations and spectral overlap integral values. Intramolecular distances and orientations were derived from an X-ray crystallographic structure and a molecular dynamic simulation of the homologous azurin Ade from Alcaligenes denitrificans, which contains both tryptophyl sites. Assignments were made of metal-ligand-field electronic transitions and of transition dipole moments and directions for tryptophyl residues, which accounted for the observed fluorescence quenching of Hg(II), Co(II), and Ni(II) azurin Pae and Cu(II) and Hg(II) azurin Afe. The fluorescence of azurin Pae is assigned as a 1Lb electronic transition, while that of azurin Afe is 1La. The marked fluorescence quenching of Cu(II) azurin Pae and Cu(I) azurin Pae and Afe is less well reproduced by our calculations, and long-range oxidative and reductive electron transfer, respectively, are proposed as additional quenching mechanisms. This study illustrates the application of Forster electronic energy transfer calculations to intramolecular transfers in structurally well characterized molecular systems and demonstrates its ability to predict observed fluorescence quenching rates when the necessary extensive structural, electronic transition assignment, and spectroscopic data are available. The agreement between Forster calculations and quenching rates derived from fluorescence lifetime measurements suggests there are limited changes in conformation between crystal structure and solution structures, with the exception of the tryptophyl residue of azurin Afe, where a conformation derived from a molecular simulation in water was necessary rather than that found in the crystal structure.     

Structural analysis of protein inclusion bodies by Fourier transform infrared microspectroscopy

The expression of recombinant human growth hormone (h-GH) and human interferon-alpha-2b (IFN-alpha-2b) in E. coli leads to the formation of insoluble protein aggregates or inclusion bodies (IBs). The secondary structure of these IBs, their corresponding native forms and thermal aggregates were studied by Fourier Transform Infrared (FT-IR) spectroscopy and microspectroscopy. It was demonstrated that residual native-like structures were maintained within IBs at different extents depending on the level of expression, with possible implications in biotechnology. Furthermore, comparison between infrared spectra of thermal aggregates and IBs suggests new insights on the structure of protein aggregates.     

A spectroscopic and calorimetric investigation on the thermal stability of the Cys3Ala/Cys26Ala azurin mutant.

The disulfide bond connecting Cys-3 and Cys-26 in wild type azurin has been removed to study the contribution of the -SS- bond to the high thermal resistance previously registered for this protein (. J. Phys. Chem. 99:14864-14870). Site-directed mutagenesis was used to replace both cysteines for alanines. The characterization of the Cys-3Ala/Cys-26Ala azurin mutant has been carried out by means of electron paramagnetic resonance spectroscopy at 77 K, UV-VIS optical absorption, fluorescence emission and circular dichroism at room temperature. The results show that the spectral features of the Cys-3Ala/Cys-26Ala azurin resemble those of the wild type azurin, indicating that the double mutation does not affect either the formation of the protein's overall structure or the assembly of the metal-binding site. The thermal unfolding of the Cys-3Ala/Cys-26Ala azurin has been followed by differential scanning calorimetry, optical absorption variation at lambda(max) = 625 nm, and fluorescence emission using 295 nm as excitation wavelength. The analysis of the data shows that the thermal transition from the native to the denaturated state of the modified azurin follows the same multistep unfolding pathway as observed in wild type azurin. However, the removal of the disulfide bridge results in a dramatic reduction of the thermodynamic stability of the protein. In fact, the transition temperatures registered by the different techniques are down-shifted by about 20 degrees C with respect to wild type azurin. Moreover, the Gibbs free energy value is about half of that found for the native azurin. These results suggest that the disulfide bridge is a structural element that significantly contributes to the high stability of wild type azurin.     

绿脓杆菌apoazurin在溶液中存在两种天然构象

如何解释绿脓杆菌apoazurin变性过程的复杂机制是一个有争议的问题.最近的研究表明apoazurin的复杂变性机制可以归结为其天然态存在着至少两种构象.利用内源荧光发射谱和圆二色谱进一步研究了apoazurin的脲变性机制,发现其稳态脲变性符合表观的二态过程,但其动力学为双相过程.在高浓度脲中快反应在几秒钟内完成,而慢反应要经过几个小时.快反应和慢反应的mu值分别为2.24和2.45kJ·mol-1·M-1,去折叠活化能的差值为22kJ·mol-1.时间分辨的荧光发射谱和圆二色谱可以用天然态和完全变性态的谱图通过一个固定的比例参数进行重建.结果表明,过去被广泛接受的存在着变性中间体的机制是不正确的,而apoazurin在天然态存在至少两种构象的假设是合理的.     

Effects of metal ligation and oxygen on the reversibility of the thermal denaturation of Pseudomonas aeruginosa azurin.

Thermodynamic equilibrium transition models in DSC are only applicable to reversible processes, but reversibility of the thermal transitions of proteins is comparatively rare because of intermolecular aggregation of denatured proteins and the degradation that occurs at high temperatures. The cupredoxin azurin from Pseudomonas aeruginosa has previously been found to exhibit irreversible thermal denaturation, both as holo- and apoprotein [Engeseth, H. R., and McMillin, D. R. (1986) Biochemistry 25, 2448-2455]. In this study, however, we demonstrate that this beta-barrel protein of Greek key topology in fact unfolds reversibly in anaerobic solutions when nonreducible metal ions are ligated to the protein. We show that it is the metal-coordinating cysteine residue (C112) that becomes exclusively oxidized in a transition metal catalyzed oxidation reaction with dissolved O(2) at high temperatures. Both Cu(I)- and Zn(II)-coordinating wild-type azurin therefore unfold reversibly in anaerobic solutions, as well as the Zn(II)-coordinating disulfide-deficient C3A/C26A mutant. Correspondingly, apoazurin mutants C112A and C112S unfold reversibly, even in aerobic solutions, and exhibit nearly perfect two-state transitions. Unfolding of Cu(II)-coordinating azurin is, on the other hand, always irreversible due to autoxidation of the thiolate resulting in Cu(I) and a thiyl radical prone to oxidation.     

Studies of thermally induced denaturation of azurin and azurin derivatives by differential scanning calorimetry: evidence for copper selectivity

Azurin, a blue copper protein from Pseudomonas aeruginosa, and several derivatives of azurin have been studied by differential scanning calorimetry. Two well-separated, irreversible transitions are observed in a scan of apoazurin under a variety of conditions, and they are assigned to distinct steps in the denaturation process. No specific structural component can be assigned to the lower temperature transition, but a "flap" structure which is found near the metal binding site may be involved. Circular dichroic spectra suggest that melting of the beta-sheet structure, the main structural motif in the native protein, occurs during the second transition. With the exceptions of the Ni(II) and p-(hydroxymercuri)benzoate derivatives, the transitions are superposed in the metalated forms, and the enthalpies of denaturation are more endothermic. By comparison with other first-row divalent transition ions and especially Zn(II), the Cu(II) derivative exhibits the most endothermic denaturation process. Along with other data, this suggests that the binding energy is greater for Cu(II). It is postulated that the selectivity for copper over zinc arises because of the irregular binding geometry offered by the folded protein. Denaturation of the Hg(II) derivative is even more endothermic, confirming that the type 1 binding site has a very great affinity for Hg(II). Finally, when substoichiometric amounts of Hg(II) are added to the apoprotein, there is evidence that a novel mercury-bridged dimer of azurin forms.     

Redox-induced conformational changes in plastocyanin: an infrared study.

The conformational changes associated with the redox transition of plastocyanin (PC) were investigated by absorption and reaction-induced infrared spectroscopy. In addition to spectral features readily ascribed to beta and turn protein secondary structures, the amide I band shows a major component band at 1647 cm(-1) in both redox states of the protein. The sensitivity of this component to deuteration and increasing temperature suggests that PC adopts an unusual secondary structure in solution, which differs from those described for other type I copper proteins, such as azurin and halocyanin. The conformations of oxidized and reduced PC are different, as evidenced (1) by analysis of their amide I band contour and the electrochemically induced oxidized-minus-reduced difference spectrum and (2) by their different thermal stability. The redox-induced difference spectrum exhibits a number of difference bands within the conformationally sensitive amide I band that could be assigned to peptide C=O modes, in light of their small shift upon deuteration, and to signals attributable to side chain vibrational modes of Tyr residues. Lowering the pH to 4.8 induces destabilization of both redox states of the protein, more pronounced for reduced PC, without significantly affecting their secondary structure. Besides the conformational differences obtained at neutral pH, the oxidized-minus-reduced difference spectrum shows two broad and strong negative bands at 1405 and 1571 cm(-1), assigned to COO(-) vibrations, and a broad positive band at 1710 cm(-1), attributed to the C=O vibration of a COOH group(s). These bands are indicative of a protonation of (an) Asp or Glu side chain(s) upon plastocyanin oxidation at acidic pH.     

Detection of a pH-dependent conformational change in azurin by time-resolved phosphorescence.

Azurin, a blue copper protein from the bacterial species Pseudomonas aeruginosa, contains a single tryptophan residue. Previous fluorescence measurements indicate that this residue is highly constrained and unusually inaccessible to water. In the apoprotein this residue also possesses a long-lived room-temperature phosphorescence (RTP), the nonexponential decay of which can be resolved into two major components associated with lifetimes of 417 and 592 ms, which likely originate from at least two conformations of the protein. The relative weights of these two decay components change with pH in good correlation with a change in protonation of His-35, which has been studied in Cu(II) azurin. Interestingly, the structural changes characterized in earlier work have little effect on the fluorescence decay and appear to occur away from the tryptophan residue. However, in the present work, the two RTP lifetimes suggest conformations with different structural rigidities in the vicinity of the tryptophan residue. The active conformation that predominates below a pH of 5.6 has the shorter lifetime and is less rigid. Phosphorescence decays of several metal derivatives of azurin were also measured and revealed strong similarities to that of apoazurin, indicating that the structural constraints upon the metal-binding site are imposed predominately by the protein.     

The solution structure of concanavalin A probed by FT-IR spectroscopy

The secondary structural properties of various forms of concanavalin A in solution were investigated by Fourier-transform infrared spectroscopy in the Amide I region. As in the crystal, the solution structure of the native protein consists mainly of antiparallel beta-sheet. Carbohydrate binding does not produce major changes in the overall secondary structure of concanavalin A, but affects infrared bands due to loops and beta-turns. Upon demetallization, the spectrum of concanavalin A shows only a small change in the Amide I band, indicating that whereas the beta-sheet structure is conserved, the tertiary properties may be altered. There are also changes in the bands from the tyrosine residues which are compatible with local changes in structure. Confirming tertiary structural differences, the cation-depleted apoprotein is much less stable, denaturing around 63 degrees C, while the native protein denatures only at temperatures around 85 degrees C. Tetramerization proceeds without significant secondary structural change. However, aggregation of the tetramers leads to a significant decrease of the bands corresponding to beta-sheet structure, and changes in the tyrosine bands.     

Structural characterization of the C2 domain of novel protein kinase Cepsilon.

Infrared spectroscopy (IR) and differential scanning calorimetry (DSC) were used to study the biophysical properties of the PKCepsilon-C2 domain, a C2 domain that possess special characteristics as it binds to acidic phospholipids in a Ca2+-independent manner and no structural information about it is available to date. When the secondary structure was determined by IR spectroscopy in H2O and D2O buffers, beta sheet was seen to be the major structural component. Spectroscopic studies of the thermal denaturation in D2O showed a broadening in the amide I' band starting at 45 degrees C. Curve fitting analysis of the spectra demonstrated that two components appear upon thermal denaturation, one at 1623 cm(-1) which was assigned to aggregation and a second one at 1645 cm(-1), which was assigned to unordered or open loop structures. A lipid binding assay has demonstrated that PKCepsilon-C2 domain has preferential affinity for PIP2 although it exhibits maximal binding activity for phosphatidic acid when 100 mol% of this negatively charged phospholipid was used. Thus, phosphatidic acid containing vesicles were used to characterize the effect of lipid binding on the secondary structure and thermal stability. These experiments showed that the secondary structure did not change upon lipid binding and the thermal stability was very high with no significant changes occurring in the secondary structure after heating. DSC experiments demonstrated that when the C2-protein was scanned alone, it showed a Tm of 49 degrees C and a calorimetric denaturation enthalpy of 144.318 kJ x mol(-1). However, when phoshatidic acid vesicles were included in the mixture, the transition disappeared and further IR experiments demonstrated that the protein structure was not modified under these conditions.     

Native azurin and its Ni(II) derivative: a resonance Raman study.

Resonance Raman spectra are reported for native Cu(II) $\text{Pseudomonas}$$\text{aeruginosa}$ azurin and its Ni(II) substituted derivative. The spectrum of the native azurin includes a low frequency feature and bands in the first overtone region not previously reported. The spectrum of the Ni(II) derivative exhibits three major peaks in the metal-ligand stretching region shifted to lower frequency relative to the M-L peaks in the spectrum of native azurin. Resonance enhanced ligand modes are observed which indicate that at least two of the ligands in Ni(II) azurin (cysteine and at least one histidine) are the same as in native azurin. The data also suggest that the disposition of ligands about the metal may be more nearly tetrahedral in the Ni(II) derivative than in native azurin.     

Role of flavin mononucleotide in the thermostability and oligomerization of Escherichia coli stress-defense protein WrbA

WrbA is an oligomeric flavodoxin-like protein that binds one molecule of flavin mononucleotide (FMN) per monomer and whose redox activity is implicated in oxidative stress defense. WrbA thermostability and oligomerization in the presence and absence of bound FMN were investigated using complementary biophysical methods. Infrared spectroscopy indicates similar structures for apo and holoWrbA. FMN binding has a dramatic effect on WrbA thermal stability, shifting the Tm by approximately 40 degrees C. Upon denaturation, the protein forms insoluble aggregates that lack native secondary structure and have no bound FMN. Circular dichroism (CD) reveals that the thermal unfolding of apo and holoWrbA proceeds via the formation of an aggregation-prone intermediate that retains substantial secondary structure but has lost the native configuration of the active site. This intermediate persists in solution up to 100 degrees C at micromolar concentrations. A similar partially folded state is populated during chemical denaturation with guanidinium chloride, but accumulation of the intermediate is evident only in the absence of FMN. The results also suggest that WrbA maintains some interaction with FMN in its partially folded state, despite the loss of the induced CD signal of FMN. On the basis of these data, the unfolding process can be depicted as follows: native holoprotein --> holointermediate --> apointermediate --> insoluble aggregate. Mass spectrometry shows that FMN promotes WrbA association into tetramers, which are more thermoresistant than dimers or monomers, suggesting that multimerization underlies the FMN effect on WrbA thermostability. This study illustrates the utility of analyzing conformational transitions and intermolecular interactions using methods that probe the liquid, solid, and gas phases.     

Complete sequential 1H and 15N nuclear magnetic resonance assignments and solution secondary structure of the blue copper protein azurin from Pseudomonas aeruginosa.

Complete sequential 1H and 15N resonance assignments for the reduced Cu(I) form of the blue copper protein azurin (M(r) 14,000, 128 residues) from Pseudomonas aeruginosa have been obtained at pH 5.5 and 40 degrees C by using homo- and heteronuclear two-dimensional (2D) and three-dimensional (3D) nuclear magnetic resonance spectroscopic experiments. Combined analysis of a 3D homonuclear 1H Hartmann-Hahn nuclear Overhauser (3D 1H HOHAHA-NOESY) spectrum and a 3D heteronuclear 1H nuclear Overhauser 1H[15N] single-quantum coherence (3D 1H[15N] NOESY-HSQC) spectrum proved especially useful. The latter spectrum was recorded without irradiation of the water signal and provided for differential main chain amide (NH) exchange rates. NMR data were used to determine the secondary structure of azurin in solution. Comparison with the secondary structure of azurin obtained from X-ray analysis shows a virtually complete resemblance; the two beta-sheets and a 3(10)-alpha-3(10) helix are preserved at 40 degrees C, and most loops contain well-defined turns. Special findings are the unexpectedly slow exchange of the Asn-47 and Phe-114 NH's and the observation of His-46 and His-117 N epsilon 2H resonances. The implications of these observations for the assignment of azurin resonance Raman spectra, the rigidity of the blue copper site, and the electron transfer mechanism of azurin are discussed.