Decomposition of protein tryptophan fluorescence spectra into log-normal components. I. Decomposition algorithms

Two algorithms of decomposition of composite protein tryptophan fluorescence spectra were developed based on the possibility that the shape of elementary spectral component could be accurately described by a uniparametric log-normal function. The need for several mathematically different algorithms is dictated by the fact that decomposition of spectra into widely overlapping smooth components is a typical incorrect problem. Only the coincidence of components obtained with various algorithms can guarantee correctness and reliability of results. In this paper we propose the following algorithms of decomposition: (1) the SImple fitting procedure using the root-Mean-Square criterion (SIMS) operating with either individual emission spectra or sets of spectra measured with various quencher concentrations; and (2) the pseudo-graphic analytical procedure using a PHase plane in coordinates of normalized emission intensities at various wavelengths (wavenumbers) and REsolving sets of spectra measured with various Quencher concentrations (PHREQ). The actual experimental noise precludes decomposition of protein spectra into more than three components.     

Decomposition of protein tryptophan fluorescence spectra into log-normal components. II. The statistical proof of discreteness of tryptophan classes in proteins

The physical causes for wide variation of Stokes shift values in emission spectra of tryptophan fluorophores in proteins have been proposed in the model of discrete states (Burstein, E. A., N. S. Vedenkina, and M. N. Ivkova. 1973. Photochem. Photobiol. 18:263-279; Burstein, E. A. 1977a. Intrinsic Protein Luminescence (The Nature and Application). In Advances in Science and Technology (Itogi Nauki i Tekhniki), Biophysics Vol. 7. VINITI, Moscow [In Russian]; Burstein, E. A. 1983. Molecular Biology (Moscow) 17:455-467 [In Russian; English translation]). It was assumed that the existence of the five most probable spectral classes of emitting tryptophan residues and differences among the classes were analyzed in terms of various combinations of specific and universal interactions of excited fluorophores with their environment. The development of stable algorithms of decomposition of tryptophan fluorescence spectra into log-normal components gave us an opportunity to apply two mathematically different algorithms, SImple fitting with Mean-Square criterion (SIMS) and PHase-plot-based REsolving with Quenchers (PHREQ) for the decomposition of a representative set of emission spectra of proteins. Here we present the results of decomposition of tryptophan emission spectra of >100 different proteins, some in various structural states (native and denatured, in complexes with ions or organic ligands, in various pH-induced conformations, etc.). Analysis of the histograms of occurrence of >300 spectral log-normal components with various maximum positions confirmed the statistical discreteness of several states of emitting tryptophan fluorophores in proteins.     

Decomposition fluorescence spectra of tryptophan residues in proteins based on log-normal components by a least squares method]

An algorithm of decomposition of protein tryptophan spectra into components was developed. The spectral shape of components is described by a uniparametric log-normal function. Rise of certainty and accuracy of resolution of widely overlapping smooth spectral components (a typical uncorrect reverse problem) was achieved using several regularizing factors: (i) the set of experimental spectra used were measured at several quencher concentrations; (ii) the functional being minimized, along with the root mean square residuals of intensities, the term depending on the obedience to the Stern-Volmer law; (iii) an extra information is used--the number of experimental values greatly exceeds the number of parameters to be estimated. The minimum of functional is determined by a consecutive setting of all possible combinations of component spectral maxima values, which allows to avoid sticking in the local minima of noisy functional. The real experimental noise restricts the decomposition into not more than three components. The decomposition error does not exceed the experimental one. The algorithm functioning is illustrated by resolution of tryptophan fluorescence spectra of papain into one, two, and three components.     

Analysis of log-normal components of fluorescence spectra of prodan and acrylodan bound to proteins

Steady-state fluorescence spectra of prodan and acrylodan covalently bound to cystein residue of Lys-Cys-Phe tripeptide in solvents of different polarity were analyzed. It was shown that the shape of spectral bands is well described by a log-normal function. Linear relations between three shape-determining parameters of the log-normal function (namely, the positions of spectral maximum and two half-maximum amplitudes) were revealed and evaluated for both fluorophores. This finding enabled us to present the shape of spectral bands of these fluorophores in any environment as analytical log-normal functions depending on only two parameters, the maximum position and the peak amplitude. The empirical uniparametric log-normal curve was used for the analysis of composite fluorescence spectra of prodan bound to bovine serum albumin and acrylodan covalently attached to actin or subfragment 1 of myosin.     

Time resolved fluorescence and phosphorescence properties of the individual tryptophan residues of barnase: evidence for protein-protein interactions.

Steady-state and time-resolved fluorescence, as well as phosphorescence measurements, were used to resolve the luminescence properties of the three individual tryptophan residues of barnase. Assignment of the fluorescence properties was performed using single-tryptophan-containing mutants and the results were compared with the information available from the study of wild-type and two-tryptophan-containing mutants (Willaert, Lowenthal, Sancho, Froeyen, Fersht, Engelborghs, Biochemistry 1992;31:711-716). The fluorescence and the phosphorescence emission of wild-type barnase is dominated by Trp35, although Trp71 has the strongest intrinsic fluorescence when present alone. Fluorescence emission of these two tryptophan residues is blue-shifted and pH-independent. The fluorescence decay parameters of Trp94 are pH-dependent, and an intramolecular collision frequency of 2 to 5 x 10(9) s(-1) between Trp94 and His18 is calculated. Fluorescence emission of Trp94 is red-shifted. Fluorescence anisotropy decay reveals the local mobility of the individual tryptophan residues and this result correlates well with their phosphorescence properties. Trp35 and Trp71 display a single phosphorescence lifetime, which reflects the rigidity of their environment. Surface Trp94 does not exhibit detectable phosphorescence emission. The existence of energy transfer between Trp71 and Trp94, as previously detected by fluorescence measurements, is also observed in the phosphorescence emission of barnase. Dynamic quenching causes the phosphorescence intensity to be protein-concentration dependent. In addition, fluorescence anisotropy shows concentration dependency, and this can be described by the formation of trimers in solution.     

Correlation between internal motion and emission kinetics of tryptophan residues in proteins

Time-resolved fluorescence anisotropy measurements of tryptophan residues were carried out for 44 proteins. Internal rotational motion with a sub-nanosecond correlation time (0.9 +/- 0.6 ns at 10 degrees C) was seen in a large number of proteins, though its amplitude varied from protein to protein. It was found that tryptophan residues which were almost fixed within a protein had either a long (greater than 4 ns) or short (less than 2 ns) fluorescence lifetime, whereas a residue undergoing a large internal motion had an intermediate lifetime (1.5-3 ns). It is suggested that the emission kinetics of a tryptophan residue is coupled with its internal motion. In particular, an immobile tryptophan residue emitting at long wavelength was characterized by a long lifetime (greater than 4 ns). It appears that a tryptophan residue fixed in a polar region has little chance of being quenched by neighboring groups.     

Correlation of tryptophan fluorescence intensity decay parameters with 1H NMR-determined rotamer conformations: [tryptophan2]oxytocin.

While the fluorescence decay kinetics of tyrosine model compounds [Laws, W. R., Ross, J. B. A., Wyssbrod, H. R., Beechem, J. M., Brand, L., & Sutherland, J. C. (1986) Biochemistry 25, 599-607] and the tyrosine residue in oxytocin [Ross, J. B. A., Laws, W. R., Buku, A., Sutherland, J. C., & Wyssbrod, H. R. (1986) Biochemistry 25, 607-612] can be explained in terms of heterogeneity derived from the three ground-state chi 1 rotamers, a similar correlation has yet to be directly observed for a tryptophan residue. In addition, the asymmetric indole ring might also lead to heterogeneity from chi 2 rotations. In this paper, the time-resolved and steady-state fluorescence properties of [tryptophan2]oxytocin at pH 3 are presented and compared with 1H NMR results. According to the unrestricted analyses of individual fluorescence decay curves taken as a function of emission wavelength and a global analysis of these decay curves for common emission wavelength-independent decay constants, only three exponential terms are required. In addition, the preexponential weighting factors (amplitudes) have the same relative relationship (weights) as the 1H NMR-determined chi 1 rotamer populations of the indole side chain. 15N was used in heteronuclear coupling experiments to confirm the rotamer assignments. Inclusion of a linked function restricting the decay amplitudes to the chi 1 rotamer populations in the individual decay curve analyses and in the global analysis confirms this correlation. According to qualitative nuclear Overhauser data, there are two chi 2 populations. Depending upon the degree of correlation between chi 2 and chi 1, there may be from three to six side-chain conformations for the tryptophan residue. The combined fluorescence and NMR results are consistent with a rotamer model in which either (i) the chi 2 rotations are fast compared to the fluorescence intensity decay of the tryptophan residue, (ii) environmental factors affecting fluorescence intensity decay properties are dominated by chi 1 interactions, or (iii) the chi 2 and chi 1 rotations are highly correlated.     

Intrinsic UV-fluorescence of lysozyme and microenvironment of its tryptophan residues]

The localization of tryptophan residues in hen egg-white lysozyme macromolecule was studied on the basis of the known 3D structure. The polarity and packing density of their microenvironments were evaluated. All residues that can affect the tryptophan fluorescence were revealed. It was shown that the orientation of these active groups relative to the indole ring of tryptophan plays a dramatic role in the efficiency of their influence. Tryptophan--tryptophan nonradiative energy transfer was evaluated from distances between tryptophan residues and their mutual orientation. The conformation of the side chains of tryptophan residues was determined. Special attention was paid to microenvironment of Trp108 responsible for the minor absorption band at 305 nm.     

The fluorescence decay of tryptophan residues in native and denatured proteins.

The fluorescence decay kinetics at different ranges of the emission spectrum is reported for 17 proteins. Out of eight proteins containing a single tryptophan residue per molecule, seven proteins display multiexponential decay kinetics, suggesting that variability in protein structure may exist for most proteins. Tryptophan residues whose fluorescence spectrum is red shifted may have lifetimes longer than 7 ns. Such long lifetimes have not been detected in any of the denatured proteins studied, indicating that in native proteins the tryptophans having a red-shifted spectrum are affected by the tertiary structure of the protein. The fluorescence decay kinetics of ten denatured proteins studied obey multiexponential decay functions. It is therefore concluded that the tryptophan residues in denatured proteins can be grouped in two classes. The first characterized by a relatively long lifetime of about 4 ns and the second has a short lifetime of about 1.5 ns. The emission spectrum of the group which is characterized by the longer lifetime is red shifted relative to the emission spectrum of the group characterized by the shorter lifetime. A comparison of the decay data with the quantum yield of the proteins raises the possibility that a subgroup of the tryptophan residues is fully quenched. It is noteworthy that despite this heterogeneity in the environment of tryptophan residues in each denatured protein, almost the same decay kinetics has been obtained for all the denatured proteins studied in spite of the vastly different primary structures. It is therefore concluded that each tryptophan residue interacts in a more-or-less random manner with other groups on the polypeptide chain, and that on the average the different tryptophan residues in denatured proteins have a similar type of environment.     

Resolution of fluorescence intensity decays of the two tryptophan residues in glutamine-binding protein from Escherichia coli using single tryptophan mutants.

Time correlated single photon counting measurements of tryptophan (Trp) fluorescence intensity decay and other spectroscopic studies were performed on glutamine-binding protein (GlnBP) from Escherichia coli. Using site-specifically mutated forms of the protein in which tyrosine (Tyr) and phenylalanine (Phe) substitute for the Trp residues at positions 32 and 220, we have examined whether wild-type (Wtyp) intensity decay components may be assigned to specific Trp residues. Results indicate that: (a) two exponential intensity decay components are recovered from the Wtyp protein (6.16 ns, 0.46 ns); (b) the long decay component arises from Trp-220 and comprises greater than 90% of the total fluorescence emission; (c) the short component arises from Trp-32 and is highly quenched; (d) all four single-Trp mutants exhibit multiexponential intensity decays, yet equimolar mixtures of two single-Trp mutants yield only two decay components which are virtually indistinguishable from the Wtyp protein; (e) the recovery of additional components in protein mixtures is obscured by statistical noise inherent in the technique of photon counting; (f) various spectroscopic measurements suggest that Trp-Trp interactions occur in the Wtyp protein, but the Wtyp intensity decay may be closely approximated by a linear combination of intensity decays from single-Trp mutants; and (g) inferences derived independently from fluorescence and NMR spectroscopy which pertain to the presence of Trp-Trp interactions and the relative solvent exposure of the two Trp residues are in agreement.     

Fluorescence and excitation Escherichia coli RecA protein spectra analyzed separately for tyrosine and tryptophan residues

The method for separation of emission (EM) and excitation (EX) spectra of a protein into EM and EX spectra of its tyrosine (Tyr) and tryptophan (Trp) residues was described. The method was applied to analysis of Escherichia coli RecA protein and its complexes with Mg(2+), ATPgammaS or ADP, and single-stranded DNA (ssDNA). RecA consists of a C-terminal domain containing two Trp and two Tyr residues, a major domain with five Tyr residues, and an N-terminal domain without these residues (R. M. Story, I. T. Weber, and T. A. Steitz (1992) Nature (London) 355, 374-376). Because the fluorescence of Tyr residues in the C-terminal domain was shown to be quenched by energy transfer to Trp residues, Trp and Tyr fluorescence of RecA was provided by the C-terminal and the major domains, respectively. Spectral analysis of Trp and Tyr constituents revealed that a relative spatial location of the C-terminal and the major domains in RecA monomers was different for their complexes with either ATPgammaS or ADP, whereas this location did not change upon additional interaction of these complexes with ssDNA. Homogeneous (that is, independent of EX wavelength) and nonhomogeneous (dependent on EX wavelength) types of Tyr and Trp fluorescence quenching were analyzed for RecA and its complexes with nucleotide cofactors and ssDNA. The former was expected to result from singlet-singlet energy transfer from these residues to adenine of ATPgammaS or ADP. By analogy, the latter was suggested to proceed through energy transfer from high vibrational levels of the excited state of Trp and Tyr residues to the adenine. In this case, for correct calculation of the overlap integral, Trp and Tyr donor emission spectra were substituted by the spectral function of convolution of emission and excitation spectra that resulted in a significant increase of the overlap integral and gave an explanation of the nonhomogeneous quenching of Trp residues in the C-terminal domain.     

Intrinsic fluorescence of globular actin. Features of localization of tryptophan residues

The contribution of individual Trp residues to alpha-actin fluorescence was evaluated by means of an analysis of their microenvironment, which was done on the basis of PIR-International protein sequence database information. The contribution of Trp79 and Trp86 was shown to be low due to an effective nonradiating energy transfer according to the inductive resonance mechanism between the Trp residues and the fluorescence quenching of Trp86 by S gamma of Cys10, an efficient fluorescence quencher. The intrinsic fluorescence of actin was found to be determined mainly by Trp340 and Trp356, which are internal, inaccessible to solvent, and have a high density microenvironment formed mainly by nonpolar groups of protein. It is possible that the side chain conformation of Trp340 (t-isomer; chi 1 190 degrees, chi 2 89 degrees), aromatic rings of Tyr and Phe residues, and Pro residues in the microenvironment of Trp340 and Trp356 substantially contribute to the short-wavelength fluorescence spectrum of actin.     

The use of site-directed mutagenesis and time-resolved fluorescence spectroscopy to assign the fluorescence contributions of individual tryptophan residues in Bacillus stearothermophilus lactate dehydrogenase

Site-directed mutagenesis has been used to generate two mutant Bacillus stearothermophilus lactate dehydrogenases: in one, Trp-150 has been replaced with a tyrosine residue and, in the other, both Trp-150 and -80 are replaced with tyrosines. Both enzymes are fully catalytically active and their affinities for substrates and coenzymes, and thermal stabilities are very similar to those of the native enzyme. Time-resolved fluorescence measurements using a synchrotron source have shown that all three tryptophans in the native enzyme fluoresce. By comparing the mutant and native enzymes it was possible, for the first time, to assign, unambiguously, lifetimes to the individual tryptophans: Trp-203 (7.4 ns), Trp-80 (2.35 ns) and Trp-150 (less than 0.3 ns). Trp-203 is responsible for 75-80% of the steady-state fluorescence emission, Trp-80 for 20%, and Trp-150 for less than 2%.     

The reactivity of tryptophan residues in proteins. Stopped-flow kinetics of fluorescence quenching.

The quenching of tryptophan fluorescence by N-bromosuccinamide, studied by the fluorescence stopped-flow technique, was used to compare the reactivities of tryptophan residues in protein molecules. The reaction of N-bromosuccinamide with the indole group of N-acetyltryptophanamide, a model compound for bound tryptophan, followed second-order kinetics with a rate constant of (7.8 +/- 0.8) . 10(5) dm3 . mol-1 . s-1 at 23 degrees C. The rate does not depend on ionic strength or on the pH near neutrality. The non-fluorescent intermediate formed from N-acetyltryptophanamide on the reaction with N-bromosuccinamide appears to be a bromohydrin compound. The second-order rate constant for fluorescence quenching of tryptophan in Gly-Trp-Gly by N-bromosuccinamide was very similar, (8.8 +/- 0.8) . 10(5) dm3 . mol-1 . s-1. Apocytochrome c has the conformation of a random coil with the single tryptophan largely exposed to the solvent. The rate constant for the fluorescence quenching of the tryptophan in apocytochrome c by N-bromosuccinamide was (3.7 +/- 0.3) . 10(5) dm3 . mol-1 . s-1. The fluorescence quenching by N-bromosuccinamide of the tryptophan residues incorporated in alpha-chymotrypsin at pH 7.0 showed three exponential terms from which the following rate constants were derived: 1.74 . 10(5), 0.56 . 10(5) and 0.11 . 10(5) dm3 . mol-1 . s-1. This protein is known to have eight tryptophan residues in the native state, six residues at the surface, and two buried. Three of the surface tryptophans have the indole rings protruding out of the molecule and may account for the fastest kinetic phase of the quenching process. The intermediate phase may be due to three surface tryptophans whose indole rings point inwards, and the slowest to the two interior tryptophan residues.     

The tryptophan residues of aspartate transcarbamylase: site-directed mutagenesis and time-resolved fluorescence spectroscopy.

Aspartate transcarbamylase (EC 2.1.3.2) contains two tryptophan residues in position 209 and 284 of the catalytic chains (c) and no such chromophore in the regulatory chains (r). Thus, as a dodecamer [(c3)2(r2)3] the native enzyme molecule contains 12 tryptophan residues. The present study of the regulatory conformational changes in this enzyme is based on the fluorescence properties of these intrinsic probes. Site-directed mutagenesis was used in order to differentiate the respective contributions of the two tryptophans to the fluorescence properties of the enzyme and to identify the mobility of their environment in the course of the different regulatory processes. Each of these tryptophan residues gives two independent fluorescence decays, suggesting that the catalytic subunit exists in two slightly different conformational states. The binding of the substrate analog N-phosphonacetyl-L-aspartate promotes the same fluorescence signal whether or not the catalytic subunits are associated with the regulatory subunits, suggesting that the substrate-induced conformational change of the catalytic subunit is the essential trigger for the quaternary structure transition involved in cooperativity. The binding of the substrate analog affects mostly the environment of tryptophan 284, while the binding of the activator ATP affects mostly the environment of tryptophan 209, confirming that this activator acts through a mechanism different from that involved in homotropic cooperativity.     

Resolution of the fluorescence decay of the two tryptophan residues of lac repressor using single tryptophan mutants.

We have studied the time-resolved intrinsic tryptophan fluorescence of the lac repressor (a symmetric tetramer containing two tryptophan residues per monomer) and two single-tryptophan mutant repressors obtained by site-directed mutagenesis, lac W201Y and lac W220Y. These mutant repressor proteins have tyrosine substituted for tryptophan at positions 201 and 220, respectively, leaving a single tryptophan residue per monomeric subunit at position 220 for the W201Y mutant and at position 201 in the W220Y mutant. It was found that the two decay rates recovered from the analysis of the wild type data do not correspond to the rates recovered from the analysis of the decays of the mutant proteins. Each of these residues in the mutant repressors displays at least two decay rates. Global analysis of the multiwavelength data from all three proteins, however, yielded results consistent with the fluorescence decay of the wild type lac repressor corresponding simply to the weighted linear combination of the decays from the mutant proteins. The effect of ligation by the antagonistic ligands, inducer and operator DNA, was similar for all three proteins. The binding of the inducer sugar resulted in a quenching of the long-lived species, while binding by the operator decreased the lifetime of the short components. Investigation of the time-resolved anisotropy of the intrinsic tryptophan fluorescence in these three proteins revealed that the depolarization of fluorescence resulted from a fast motion and the global tumbling of the macromolecule. Results from the simultaneous global analysis of the frequency domain data sets from the three proteins revealed anisotropic rotations for the macromolecule, consistent with the known elongated shape of the repressor tetramer. In addition, it appears that the excited-state dipole of tryptophan 220 is alighed with the long axis of the repressor.     

Quenching of tryptophan fluorescence in human antithrombin III by iodide ion.

Iodide is an efficient quencher of antithrombin III intrinsic tryptophan fluorescence. The quenching pattern indicates that about 60% of the tryptophyl fluorescence originates from exposed residues in the multitryptophan-containing protein. In denaturing media all of the tryptophyls are solvent-exposed. The binding of heparin to antithrombin III influences the number of solvent-exposed tryptophan residues. By studying the dependence of the quenching on pH, information regarding the presence of charged residues adjacent to tryptophyls was obtained.     

Triplet state of tryptophan in proteins. 2. Differentiation between tryptophan residues 62 and 108 in lysozyme.

We have used optically detected magnetic resonance (ODMR) to characterize the degree of solvent availability of the tryptophan residues in lysozyme that are likely to be responsible for the observed phosphorescence. From the phosphorescence spectra, ODMR zero-field splittings (zfs), and ODMR line widths, we concur with the X-ray structure [Blake, C. C., Mair, G. A., North, A. C. T., Phillips, D. C., & Sarma, V. R. (1967) Proc. R. Soc. London, ser. B 167, 365-377] that Trp-62 behaves as an exposed residue and Trp-108 is buried. In addition, we present evidence that ODMR can be used in conjunction with conventional phosphorescence to evaluate the degree of order in the microenvironments of tryptophan in a protein containing several tryptophans. By the specific modification of residues Trp-62 and Trp-108, we have identified those portions of the ODMR lines in the native enzyme that are due to those specific residues. Barring major enzyme conformational changes in the vicinity of unmodified tryptophan residues when Trp-62 or Trp-108 are selectively modified, we find that Trp-108 dominates both the phosphorescence and the ODMR signals in native lysozyme. The results are discussed in view of previous fluorescence findings.     

Nanosecond segmental mobilities of tryptophan residues in proteins observed by lifetime-resolved fluorescence anisotropies.

Steady-state and lifetime-resolved fluorescence anisotropy measurements of protein fluorescence were used to investigate the depolarizing motions of tryptophan residues in proteins. Lifetime resolution was achieved by oxygen quenching. The proteins investigated were carbonic anhydrase, carboxypeptidase A, alpha-chymotrypsin, trypsin, pepsin, and bovine and human serum albumin. When corrected for overall protein rotation, the steady state anisotropies indicate that, on the average, the tryptophan residues in these proteins rotate 29 degrees +/- 6 degrees during the unquenched excited state lifetimes of these proteins, which range from 1.7 to 6.1 ns. The lifetime-resolved anisotropies reveal correlation times for these displacements ranging from 1 to 12 ns. On the average these correlation times are tenfold shorter than that expected for overall protein rotation. We conclude that the tryptophan residues in these proteins display remarkable freedom of motion within the protein matrix, which implies that these matrices are highly flexible on the nanosecond time scale.