Abstract not available.

Computational studies of photophysical processes in solution require accurate representations of the potential energy function. The interactions in a polar solute-solvent system are primarily electrostatic in nature, and they are, in molecular simulations, typically mediated by partial atomic charges on the solute and the solvent atoms. We have developed procedures to create partial atomic charges for any electronic state in a solute molecule from a least squares fit to the molecular electrostatic potential. The quantum mechanical electrostatic potential for the electronically excited 1La state of 3-methylindole derived from a semiempirical INDO/S configuration interaction wavefunction is presented. Partial atomic charges for the 1La state are derived from the quantum mechanical potential, and the classical and quantum mechanical electrostatic potentials are compared. Molecular dynamics simulations have been carried out on the ground (So) and two lowest excited singlet states (1Lb, 1La) of 3-methylindole in water using potential derived partial atomic charges. Solvent-induced inversion of the gas phase excited state ordering (1Lb below 1La) is computed for the excited states. A method for introducing polarization by the solvent into the solute electronic wavefunctions, and hence into the solute partial atomic charges, is introduced. A significant increase in solute dipole moment is computed for the 1La state, leading to dramatic increases in solute-solvent interaction energies and additional preferential stabilization of the 1La state over the 1Lb state.

Supersonic gas expansion techniques can be used to prepare samples of cold gas phase analogs of tryptophan. The excited state origins of different conformers can be selectively laser excited and the fluorescence decays recorded. When possible 1La state effects on emission lifetime are recognized and eliminated. The remaining data set can be appraised in terms of conformer dependent fluorescence quenching due to election transfer.

The fluorescence properties of a set of tryptophan analogs, including several that are conformationally restricted, is reported. The conformationally restricted analogs include tetrahydrocarboline-3-carboxylic acid, 3-amino-3-carboxytetrahydrocarbazole, 3- aminotetrahydrocarbazole, and 3-carboxytetrahydrocarbazole. Steady-state and time-resolved measurements were made. The fluorescence decay of most of the conformationally restricted analogs is a bi-exponential. The fluorescence quantum yield and average lifetime of the restricted analogs is a bi-exponential. The fluorescence quantum yield and average lifetime of the restricted analogs are higher than the values for tryptophan and corresponding flexible tryptophan analogs. This indicates that the effective rate constant for intramolecular non- radiative (quenching) processes, knx, is smaller for the restricted tryptophan analogs than for the flexible analogs. For a series of flexible analogs, having a bifunctional side chain, there is synergism between the two functional groups (i.e., (alpha) -NH3+ and (alpha) -CO2- groups) in determining the effective knx. Potential mechanisms for intramolecular quenching mechanisms, which contribute to knx, include proton transfer from a side chain ammonium group to position four of the excited indole ring and charge transfer from the excited indole ring to a side chain acceptor group. By measuring the rate of photo-induced isotope exchange into position four of an analog's indole ring, we have estimated the contribution of proton transfer quenching to the total intramolecular quenching rate constant for several flexible tryptophan derivatives.

The fluorescence lifetime of N-acetyl-tryptophan-amide (NATA) was measured by multifrequency phase fluorometry, in the presence of increasing concentrations of imidazole. Two pH values were tested, pH 4.5 where imidazole is fully protonated and pH 9.0 where it is fully unprotonated. At both pH values, the inverse lifetime increases in a nonlinear way with the imidazole concentration, showing that imidazole is not a high efficiency collisional quencher. The data can be analyzed in terms of the formation of a complex with a reduced fluorescence lifetime. The rate constants for association (at 25 degree(s)C) are around 5 (+/- 0.2) X 109 M-1 s-1 and are thus diffusion controlled. The association equilibrium constant is strongly pH-dependent and is much higher than the expected value of 0.4 M-1 for a collisional complex. The intrinsic fluorescence lifetime of the complex is 1.56 (+/- 0.02) ns at pH 9.0 and 1.82 (+/- 0.03) ns at pH 4.5, as compared to 2.37 (+/- 0.03) ns for free tryptophan at pH 9.0 and 2.83 (+/- 0.05) at pH 4.5 (all at I equals 0.34). This means that at both pH values the fluorescence lifetime of tryptophan in the complex is reduced to 61% (+/- 0.05) of its value in the free state. Despite this, the protonated form of imidazole is a better quencher at low concentrations, due to a longer residence-time of the complex. At high viscosity the association equilibration is too slow and the system is described by two lifetimes. The quenching effect of His-18 on the fluorescence of the proximal Trp-94 of Barnase is discussed in terms of these findings. An extensive account of these results has appeared.

The effect of the collisional quenching on the fluorescence intensity decays has been studied by frequency-domain fluorometry. We used an efficient (CBr4) and/or inefficient (CCl4 quencher to quench the fluorescence of 1,2-benzanthracene (1,2-BA). The wide range of diffusion has been obtained by using propylene glycol at different temperatures (-40 degree(s)C to 40 degree(s)C). The measured intensity decays cannot be satisfactorily fitted either to the Smoluchowski or Collins-Kimball (RBC) model, except the case of inefficient quencher in the presence of high diffusion. In particular, we observed quenching in diffusionless conditions (-40 degree(s)C). To describe the collisional quenching of the fluorescence more correctly we propose a new model which includes a distance-dependent quenching rate (DDQ model). The DDQ simulations show that the local concentration of quencher surrounding the excited fluorophore cannot be approximated by using the RBC model, except in the case of high diffusion and low quenching rate. The DDQ model describes well all measured intensity decays of 1,2-benzanthracene in the presence of CBr4 and/or CCl4. Also, the DDQ model more correctly predicts the curvature of Stern-Volmer plots and activation energies obtained from the temperature dependent rate of quenching.

The fluorescence decay surface of tryptophan measured over a very wide pH range (1.35 to 12.38) was analyzed in a single global analysis. It is clear that the decay components remain the same over the whole pH range. This is also confirmed by the decay-associated emission spectra. At low pH the decays are biexponential and both decay components contribute to the fluorescence up to pH 11. From about pH 7.5 onwards a third component has to be taken into account which becomes the only component at pH > 11. The decay times decrease at low pH due to quenching by H3O+. At pH > 11 the decay times decline due to quenching by OH-. Decays collected at different temperatures at neutral pH indicate that the short decay time is independent of temperature, both in H2O and D2O solution. Both decay times increase by a factor of two when H2O is replaced by D2O as solvent. The decay times and their normalized pre-exponential terms at neutral pH in H2O solution are constant as a function of excitation wavelength (from 250 to 295 nm). This indicates that the absorption spectra associated with the two ground-state species overlap completely and that the contribution of both species to the total absorption remains constant over the whole absorption spectrum.

The relationship between the fluorescence quantum yield, (phi) f, radiative lifetime, (tau) r, and excited singlet state lifetime, (tau) s, is considered to be fundamental in fluorescence studies. Because of advances in instrumentation and data analysis the intensity decay times of tryptophan in proteins can be determined with a high degree of confidence and accuracy. Site directed mutagenesis allows one to place tryptophan residues in different locations within proteins. In several mutant single tryptophan containing proteins studied in our laboratory it has often been noted that there is a lack of a proportional correlation between the measured fluorescence quantum yield and the mean fluorescence decay time. This data and examples from the literature are presented and rationalizations are discussed. The intent is to stimulate experimental design and theoretical studies. This will lead to new intuitions from fluorescence studies of mutant proteins regarding the interrelationship of their structure, dynamics, and function.

The time-resolved fluorescence properties of a tryptophan residue should be useful for probing protein structure, function, and dynamics. To date, however, the non-single exponential fluorescence intensity decay kinetics for numerous peptides and proteins having a single tryptophan residue have not been adequately explained. Many possibilities have been considered and include: (1) contributions from the ¹L_a and ¹L_b states of indole; (2) excited-state hydrogen exchange; and (3) environmental heterogeneity from (chi) ¹ and (chi) ² rotamers. In addition, it has been suggested that generally many factors contribute to the decay and a distribution of probabilities may be more appropriate. Two recent results support multiple species due to conformational heterogeneity as the major contributor to complex kinetics. First, a rotationally constrained tryptophan analogue has fluorescence intensity decay kinetics that can be described by the sum of two exponentials with amplitudes comparable to the relative populations of the two rotational isomers. Second, the multiple exponentials observed for tyrosine-containing model compounds and peptides correlate with the (chi) ¹ rotamer populations independently determined by ¹H NMR. We now report similar correlations between rotamer populations and fluorescence intensity decay kinetics for a tryptophan analogue of oxytocin. It appears for this compound that either (chi) ² rotations do not appreciably alter the indole environment, (chi) ² rotations are rapid enough to average the observed dependence, or only one of two possible (chi) ² populations is associated with each (chi) ¹ rotamer.

Site-directed mutagenesis of specific residues in bacteriophage T4 lysozyme is shown to result in changes in the emission spectra of the tryptophan residues of this protein. In some cases a significant red-shift is observed. This is interpreted in terms of enhanced dielectric relaxation due to fluctuations that expose a buried residue to the aqueous solvent. For substitutions at position 146, the spectral shift is strongly correlated with the rate of a specific proteolytic digestion of the T4 lysozyme by trypsin as determined by Signor, Dalzoppo, and Schellman. In cases where a spectral shift is observed there is also an enhancement of internal mobility of a tryptophan residue as indicated by the amplitude of a short correlation time component of the anisotropy decay. All of these spectral and enzymatic susceptibility effects are reversed by introduction of a disulfide linkage spanning the two lobes of the protein. The interpretation of these results in terms of molecular dynamics is discussed. The effect of mutational changes on the average fluorescence lifetime and quantum yield of tryptophan fluorescence is also discussed in terms of collisional quenching of tryptophan residues by neighboring groups.

The authors outline some examples of the advantages found in subdividing overall quenching into heterogeneous contributions. Subdivision is accomplished by overdetermination (global) and association (DAS, decay associated spectral) methods. In some cases, the subdivision of fluorescence leads to the unique identification of different fluorophores in different sites. Alternatively, the recovered components may reflect conformational heterogeneity at each site. For intrinsic protein fluorescence, it is often noted in the literature that single Trp proteins may be multiexponential. Genetic substitution in multi-Trp proteins, however, often leads to very strong (if not complete) lifetime-to-Trp assignment. Even if a single Trp experiences two or more microenvironments, it can be a useful reporter. The linkage of multiple lifetimes and amplitudes to changes in global conformation often reveals a more `sensitive' subpopulation or lifetime component that becomes a better indicator for important conformational states than aggregate intensity can provide. This has proven useful in studying pH transitions of proteins both in solution and embedded in membranes. Energy transfer is particularly useful in differentiating sites at different distances. Further, the disclosure of heterogeneity in distance is clearly superior to the reporting of a mean distance. This report surveys several systems that have been examined via emission DAS techniques, showing how each protein is better understood when viewed in terms of discrete spectral contributions. We conclude with an overview and some details about our construction of an EDAS (excitation-DAS) instrument; i.e., how excitation scans can be incorporated into a time-resolved instrument.

The tryptophan repressor from Escherichia coli binds to the trp operator in the presence of L- tryptophan, thereby inhibiting the biosynthesis of L-tryptophan. Site-directed mutagenesis was used to change tryptophan-19 and tryptophan-99 to leucine and methionine, respectively. This mutant protein without tryptophan in its amino acid sequence has wild-type repressor activity and is a suitable model for fluorescence studies of corepressor binding. Both steady-state and time-resolved fluorescence spectroscopy have been used to compare the binding of L- tryptophan, indole-3-propionic acid, indole-3-butyric acid, and indole. In all cases, binding to the mutant aporepressor results in a large blue shift and a change in the intensity of the ligand fluorescence. The decay of the total fluorescence intensity from the complex indicates the presence of three distinct bound states of the ligand. The distribution of ligand binding modes is influenced by the substituent at the 3-position of the indole ring. The rotational correlation time of the complexes formed with L-tryptophan or indole-3-propionic acid indicate that the protein is present as a dimer, whereas with indole or indole-3-butyric acid the correlation times are much lower, suggesting that the protein is present as a monomer.

The three-dimensional structures of a number of DNA-binding proteins alone and complexed with their cognate DNA sequences have recently been reported. One of these is the trp repressor from E. coli, for which the structure in the absence and the presence of the co- repressor, tryptophan, has been deduced from x-ray crystallographic data, as well as that of the ternary complex between protein, co-repressor, and operator DNA. Rather than provide a definitive answer for the mechanism of the activation of the repressor by tryptophan binding, these structures have raised even more questions concerning the basis for the affinity changes observed. DNA binding studies of this protein using a number of biochemical techniques have brought to light the existence of repressor multimer-DNA complexes at physiological concentrations, although these have not been well characterized. Recently, we have demonstrated, using fluorescence polarization techniques, that the repressor oligomerizes in absence of DNA and that co-repressor binding destabilizes the oligomers. Clearly, in this case, as in many others, there are subtle thermodynamic relationships between protein-protein, protein-ligand, and protein-DNA interactions. The role of these energetic couplings in the allosteric regulation of the repressor by tryptophan is, however, not understood. Because of the non-equilibrium nature of filter-binding, gel mobility shift, and nuclease protection assays it is not possible to fully explore these different binding phenomena over broad concentration ranges of each component and under different solution conditions. Fluorescence methodologies provide observables of the multiple binding equilibria in solution between nanomolar and micromolar concentrations. Variables such as the protein and ligand concentration dependence of repressor-DNA interactions, as well as different solution conditions (i.e., salt concentration), have been probed using these techniques. The intrinsic tryptophan fluorescence of the protein has been used to characterize dimer stability. The fluorescence polarization of a covalently bound DNS label has been used to characterize the higher order protein-protein interactions and the effect of tryptophan and salt concentration upon them. DNS polarization has also been used to probe the protein and tryptophan concentration dependence of operator DNA binding. Finally, fluorescence quenching of coumarin covalently coupled to the trp repressor protein upon binding the operator fragment has been used to study these protein- DNA complexes. The combination of appropriate fluorescence methodologies allows for a detailed characterization of binding affinities and stoichiometries in such systems.

Time-resolved fluorescence and fluorescence anisotropy experiments in combination with molecular dynamics simulations show that the flavin chromophore bound in oxidized and reduced flavodoxins is immobilized within the protein matrix. Experimental results show that a small fraction of the flavin chromophores in oxidized flavodoxin have more rotational freedom, as indicated by significantly smaller rotational correlation times than expected for the whole protein.

Protein sequence analysis, nuclear magnetic resonance, and fluorescence dynamics have been applied in a determination of the interactions of the lumazine derivative with the amino acid residues in the proposed ligand binding site of lumazine protein. It is these interactions that `tune' the excited state properties of the bound lumazine so that it can perform its photobiological function as the emitter of bioluminescence in Photobacterium species. A three- way sequence alignment shows that lumazine protein is homologous with the yellow- fluorescent protein of Vibrio fischeri and the riboflavin synthase from Bacillus subtilis. This last enzyme is ubiquitous in procaryotes, and utilizes two of these same lumazines as substrates for the production of riboflavin. By analogy with riboflavin synthase, a short sequence in the lumazine protein has been suggested as the ligand binding site. In riboflavin synthase there is a second binding site, but this is absent in lumazine protein, thus negating any synthase activity for this protein. Hydrogen bonds to the residues in this binding domain and `freeze' the lumazine structure into the highly polar tautomer deduced from NMR evidence. This also accounts for the rigidity of binding shown by the 23 ns (2 degree(s)C) rotational correlation time of the bound ligand as well as the strong blue shift of the fluorescence maximum, from 490 nm free to 475 nm when bound.

The fibrous region of myosin, called myosin rod when isolated from myosin after proteolysis, is a two-stranded coil made of identical chains of nearly 1000 residues. Myosin from rabbit skeletal muscle has two tryptophans per chain located at identical hydrophobic d sites in the heptad repeat which forms the basis for hydrophobic dimerization. Steady-state quenching of these tryptophans by KI shows downward curvature in a classic Stern-Volmer plot; analysis using a modified S-V equation indicates that only about 80% of the tryptophans are accessible to this charged quencher. The emission intensity decays are complex and are fit to lifetimes of 5.3, 1.6, and 0.4 ns with relative amplitudes of 0.78, 0.13, and 0.09. Lifetime resolved quenching studies indicate that only the long lifetime component is quenched by iodide; the collisional rate constant for quenching this component is very similar to that calculated from the steady-state quenching data. The long lifetime component thus corresponds to a population of solvent accessible tryptophans that may be on the surface of the coil protein. These studies suggest that the tryptophans in myosin rod may be in equilibrium between accessible and inaccessible sites at the coil interface.

The multi-component fluorescence intensity decay of variant-3 scorpion neurotoxin is measured with time-correlated single photon counting as a function of guanidine hydrochloride (GuHCl) concentration. Available evidence suggests that while the NH₂-terminal (beta) - sheet strand may be denatured in GuHCl, the remaining core neurotoxin structure remains intact. We investigate this hypothesis with computer simulations of variant-3 scorpion neurotoxin with lysine-1 to tyrosine-4 deleted. Previous combination thermodynamic perturbation and umbrella sampling, adiabatic mapping and minimum perturbation mapping computer simulations of tryptophan-47 in the native neurotoxin exhibited multiple rotational isomers that might correspond to the observed fluorescence intensity decay components. The new simulations allow us to compare the number of rotational isomers, the isomer populations, the order parameters, and the transition state theory isomer interconversion rates in the native and denatured states.

Proteins exist in a predominately aqueous solvent environment. Hydration of the protein surface significantly affects many aspects of the protein's structure and function; these effects may be related to influences on the internal dynamics of the protein. We are examining the influence of hydration on the internal dynamics of hen egg white lysozyme using room temperature phosphorescence from the intrinsic tryptophan residues; since lysozyme is not phosphorescent in solution, the emission intensity serves as a probe of protein conformational fluctuations. Powders of lyophilized lysozyme are hydrated in situ (in the phosphorimeter) using a flow system that allows for continuous manipulation of relative humidity (RH) over the range from 0.5 to 100%; this system allows us to directly compare intensity differences that result from changes in hydration. Lysozyme phosphorescence intensity decreases monotonically as a function of hydration over the entire accessible RH range; the decrease is not linear but appears to occur in distinct phases. The phosphorescence intensity decays are multiexponential over the hydration range and hydration has the largest influence on the long lifetime component. These studies confirm previous work on the effect of hydration on the internal dynamics of globular proteins.

A combination of molecular relaxation spectroscopy and microwave conductivity experiments on model flavin compounds has resulted in the determination of the dipole moment difference in the ground and first excited singlet states. The results obtained with both methods were in excellent agreement. The dipole moment in the excited state is only slightly larger than the corresponding ground-state dipole moment. With the literature value of the S₀ dipole moment of 7.8 D, the S₁ dipole moment is then 8.8 D.

We report on excitation wavelength dependence of electron transfer reaction in bianthryl (BA) molecule bound by human serum albumin (HSA). The rate of this reaction is known to be controlled by the dynamics of the probe dielectric environment and, if the dynamic is slow in comparison with excited-state lifetime, by site-dependent photoselection at the red edge of BA absorption spectrum. In BA-HSA complex in the temperature range 10 - 40 degree(s)C we observe the absence of electron transfer at the main band and its high effectiveness at red-edge excitation. These results demonstrate that the distribution on protein-probe interaction energy is of substantial width and the local dynamics in the binding site is slower than a nanosecond. This distribution and slow dynamics are definitely the origins of inhomogeneous kinetics of the electron transfer reaction.

We report measurements of site-to-site diffusion in proteins, using frequency-domain measurements of time-dependent energy transfer. The possibility of such measurements is shown from simulations which demonstrate that donor-to-acceptor (D-to-A) diffusion alters the donor frequency response, and that this effect is observable in the presence of a distribution of distances. For decay times typical of tryptophan fluorescence, the simulations indicate D-to-A diffusion coefficients can be measured ranging from 10^-7 to 10^-5 cm²/s. This possibility was verified by studies of a methylene-chain linked D-A pairs in solutions of varying viscosity. D-to-A diffusion was also measured for acceptor-labeled melittin in the random coil and (alpha) -helical states. Unfolding of troponin I results in increased D-A diffusion. Surprisingly, more rapid diffusion was observed for melittin in the (alpha) -helical state, but over a limited range of distances.

Time-resolved polarized fluorescence decays of FAD bound to glutathione reductase have been obtained upon separate excitation at 457.9 nm and 514.5 nm. From the inverse Laplace transform of the fluorescence decays as obtained by the maximum entropy method, five enzyme conformers can be distinguished in solution. By red-edge and main-band excitation we demonstrate that intersubunit energy transfer occurs between the flavin prosthetic groups as well as restricted motion of flavin. From a 2-D maximum entropy analysis, it can be deduced that the observed conformers of glutathione reductase have different dynamic properties. The ability of the maximum entropy method to resolve a heterogeneous population of emitters with distinct dynamical properties is tested by simulated data. From the results, a role in catalysis is proposed to equilibrium fluctuations in glutathione reductase.

Imaging of thick tissue has been an area of active research during the past several years. Among the methods proposed to deal with the high scattering of biological tissues, the time resolution of a short light probe traversing a tissue seems to be the most promising. Time resolution can be achieved in the time domain using correlated single photon counting techniques or in the frequency domain using phase resolved methods. We have developed a CCD camera system which provides ultra high time resolution on the entire field of view. The phase of the photon diffusion wave traveling in the highly turbid medium can be measured with an accuracy of about one degree at each pixel. The camera has been successfully modulated at frequencies on the order of 100 MHz. At this frequency, one degree of phase shift corresponds to about 30 ps maximum time resolution. Powerful image processing software displays in real time the phase resolved image on the computer screen.

The application of our development of 750 nm excitation from an argon-hydrogen filled spark source to time-resolved fluorescence probe studies of lipid membranes and inverse micelles is reported. The laser dye IR-140 was studied using the single-photon counting technique both in a lipid membrane of L-(alpha) dipalmitoylphosphatidylcholine (DPPC) and in sodium sulfosuccinic acid bis (2-ethylhexyl) ester (AOT) in iso-octane. In DPPC a dramatic change in the fluorescence behavior of IR-140 is observed between the gel and liquid crystalline phases. In inverse micelles of AOT an increase in the intensity of the peak fluorescence emission and a decrease in fluorescence lifetime is noticed on increasing the water content.

Multichannel photon-counting fluorimetric systems using optical fiber dynamic memory with high time resolution and high data-gathering efficiency are presented. The principle of operation is based on the vernier chronotron technique. A measurement system using two single-mode optical fiber loops in conjunction with pulsed laser diodes and avalanche photodiodes was described in detail. The feasibility of rapid measurement was indicated. The extremely low loss, good linearity, and large available time-bandwidth of single-mode fiber allow signals to propagate large distances without significant attenuation or distortion, so that good stability as well as high time resolution of the measurement system can be constructed. In order to simplify the system and improve system performances, other modified systems using optical fiber dynamic memory are discussed.

A new technique for the simultaneous measurement of fluorescence and phosphorescence depolarization is reported. The technique is based on a newly developed photomultiplier gain suppression technique. The gain suppression characteristics of the circuit were studied by applying a constant level of incident light. To demonstrate this new technique, a solid sample of the triplet probe eosin was excited by a pulsed, frequency doubled Nd:YAG laser; the resulting orthogonally-polarized emissions were collected by two balanced photomultiplier (PMTs). The gains of the PMTs were suppressed for the duration of the fluorescence emission, and subsequently returned to normal, in order to keep the measured fluorescence and phosphorescence signals at the same level. The suppression of the gain was controlled by using a specially designed dual-channel PMT gain suppression circuit. The signals from the PMTs were recorded by two digital oscilloscopes, one set at a fast sweep time in order to measure the transient fluorescence depolarization and the other set at a slow sweep time in order to measure the slow phosphorescence depolarization. The anisotropies can then be deduced from the two sets of recordings.

The development of deconvolution techniques in pulsed-laser, time-resolved photoacoustics has opened the possibility of accurately distinguishing between processes occurring on different time scales, and has given photoacoustics better resolution in determining reaction enthalpies and quantum yields. While fluorescent signals are usually generated by a single de- excitation pathway in the fluorophore, photoacoustic signals usually arise from different sources, such as excited singlet and triplet deactivation, occurring on well-distinguished time scales. The understanding of the effect of quenching on photoacoustic signals therefore requires careful analysis of the data. In this work, a model is developed to describe the effect of fluorescence quenching on photoacoustic signals. The model takes advantage of the time resolution in pulsed-laser photoacoustics. Both static and dynamic quenching are taken into account. Important photophysical parameters (fluorescence and intersystem crossing quantum yields, the bimolecular quenching rate constant, and the volume of the sphere of action) appear in the expressions describing the dependence of photoacoustic signal on quencher concentration. Data from both steady-state fluorescence and time-resolved photoacoustic quenching measurements are analyzed simultaneously using a set of equations containing common parameters. Experimental data on the quenching of organic dyes are presented which support the validity of the model.

Using a combination of time-resolved room temperature phosphorescence and time-resolved circularly polarized phosphorescence (TR-CPP, demonstrated in proteins for the first time), we provide strong evidence supporting the existence of conformational heterogeneity in the apo form of horse liver alcohol dehydrogenase (LADH). Using extensive photon counting and correspondingly improved statistics, we show a clear biexponential decay in the phosphorescence of deoxygenated LADH in contrast to earlier studies where monoexponential decay was reported. Under extensively deoxygenated conditions, the protein shows long lifetime components of 510 and 765 milliseconds (pre-exponentials vary with conditions). TR- CPP studies of LADH show that the emission anisotropy factors (a measure of the local chirality associated with the excited state) associated with the two decay components are identical (g_em equals 0.004; where the fractional anisotropy is given by g_em equals {2^*[I_left-I_right)/I_total]}. This indicates that the chirality of the environment of the phosphorescent tryptophan is the same for both lifetimes. Upon binding the coenzyme fragment adenosine-diphosphoribose (ADPR), LADH phosphorescence decay becomes essentially monoexponential (1285 milliseconds) with a time independent g_em comparable to the g_em of apo LADH. The quenching rates of the long lifetime components of apo-LADH by the electron acceptor TEMPOL were found to be similar, suggesting that the conformational states of LADH have nearly equal susceptibility to oxidation by TEMPOL.

Alterations in plasma membrane structure and function seem to be of primary importance in the pathogenesis of cell injury, calling for more understanding of the changes in plasma membrane lipid structure (e.g., lipid order, lateral diffusion, dependence of phase states, and viscoelasticity) during the evolution of hypoxic injury in hepatocytes using multiple fluorescent spectroscopic techniques. Following hypoxic injury, fluorescence recovery after photobleaching was used to monitor plasma membrane lipid diffusion, resonance energy transfer microscopy was used to detect the lipid topography (domain formation), and the laser trapping technique was used to measure the plasma membrane viscoelasticity. The use of these different kinds of fluorescent spectroscopic techniques coupled with the authors' previous studies using digitized fluorescence polarization microscopy which was used to measure lipid order (fluidity) allowed the delineation of alterations in membrane structure during hypoxic injury and a model of membrane architecture during hypoxic injury, which could not be obtained from the use of any of these techniques alone. A model is proposed in which gel- and fluid-phase lipid islands form during hypoxic cell injury. Formation of these lipid domains promotes cell surface bleb formation, with eventual weakening of plasma membrane integrity, bleb rupture, and cell death. 11

The utility of fluorescence depletion methods for the measurement of slow protein rotational diffusion has been limited by the lack of a rigorous mathematical model to obtain, from depletion data, anisotropies directly comparable to those obtained from phosphorescence emission or triplet absorption measurements. A generalized theory to meet this need is described. The experimental method requires the acquisition of, at most, three separate measurements to calculate absorption or emission anisotropies. Each measurement is made with a different orientation of either the probe beam polarization, pump beam polarization, or emission polarizer. The results of the theory are applied to two experimental configurations. The first of these involves collecting emission at 90 degree(s) to colinear pump and probe beams. From such data we are able to calculate the absorption anisotropy, the emission anisotropy, and the interdipole angle. The second configuration represents a system where all polarization axes lie in a single plane as would occur in a microscope-based system. For this configuration we are able to calculate, given the interdipole angle derived from the 90 degree(s) case, the true absorption anisotropy, the true emission anisotropy, or both.

In frequency domain fluorometry, as well as all other spectroscopic techniques, the noise ultimately limits the sensitivity of the instrument and the precision of the measurement. The analysis of the sources of noise in different instruments has revealed that the noise is due to a number of different instrumental factors rather than photon statistics. The ultimate goal is to eliminate those factors to achieve a situation in which the limit is the detector intrinsic noise. We have developed a system, based on digital signal processing, in which the influence of several spurious noise sources has been reduced. A study of the range of cross-correlation frequencies used to obtain the best signal-to-noise ratio is presented.

The fluorescence decay analysis of intramolecular two-state excited-state processes with added quencher is discussed in terms of compartments. The kinetic equations specifying the fluorescence decay and the time-course of the two excited-state species concentrations are expressed in terms of the rate constants and the spectroscopic parameters b₁ and c₁. b₁ and c₁ are respectively the relative absorbance and the normalized spectral emission weighting factor of species 1. The report investigates what has to be known beforehand to determine all relevant parameters. The results of this identifiability study indicate that the following conditions have to be satisfied in order to make an intramolecular two-state excited-state system with added quencher identifiable. First, at least three different quencher concentrations must be used. Second, the two rate constants of quenching must be different. Third, at least one parameter must be known. This parameter can be (1) one rate constant which is not a rate constant of quenching, (2) one b₁ value or, (3) one c₁ value. In each of these cases an alternative set of system parameters is mathematically possible. A unique solution is guaranteed when the fluorescence decays of a quenched model compound are included in the compartmental analysis.

The distinguishability between models in the recovery of fluorescence lifetime distributions is determined by the distance between the estimation spaces spanned by the models in the 2N dimensional sample space, where N is the number of frequencies used in the phase-modulation fluorescence lifetime measurements. The distinguishability regions between the model of two discrete components and the Gaussian distribution model are described in this paper as an example. The use of a priori distribution functions such as Gaussian, Lorentzian, and uniform distributions without a physical basis poses a serious problem in the physical interpretation of the recovered parameters. This work introduces the application of the exponential series method (ESM) for recovering lifetime distributions in the frequency domain and distinguishing between models without using any a priori models.

A simple model is developed, based on the diffusion approximation to the linear transport equation, which gives analytic expressions that describe the coherent propagation of sinusoidally intensity-modulated light through a strongly scattering, low absorbing, homogeneous, infinite medium in terms of the interaction coefficients of the medium. Previously, the analytical multiform solutions to Laplace's equation and the Helmholtz equation were derived respectively by Sommerfeld and Carslaw in their studies of electrostatics, sound, and heat to model intensity modulated light in the presence of an absorbing or reflecting `semi-infinite' plane bounded by a straight edge that is immersed in an infinite, strongly scattering, low absorbing, homogeneous medium. The model predictions are in good agreement with the present results of experiments performed on media consisting of Intralipid^TM and skim milk emulsions containing minute quantities of black India ink and with Monte Carlo simulations. These studies provide a theoretical basis for the understanding of photon diffusion in tissues and allow the determination of conditions to obtain maximum resolution and penetration.

The attainable frequency range of ordinary heterodyning and super-heterodyning frequency- domain fluorescence instrumentation is limited by the response of the optical detector rather than by the harmonic content of the light source. The replacement of the photomultiplier detector by a 6 (mu) microchannel plate detector has improved the frequency response from 500 MHz to 10 GHz. A new method is developed to detect fast, excited state processes by extending to the frequency-domain the well known pump/probe (absorption) technique used in the time-domain. The upper frequency limit attainable with this method is limited only by the pulse width of the light sources. For picosecond pulse lasers this limit extends to hundreds of gigahertz. A theoretical determination of the basic equations is given, and data are shown for the excited state decay of a rhodamine 6G sample in ethylene glycol.

A new dimension in quantitative fluorescence microscopy may be accessed by imaging of fluorescence decay times. To obtain spatially resolved information from microscopic sample locations, one must not only have sufficient optical resolution and detection sensitivity but also the ability to exclude fluorescence photons originating from outside the focal volume of interest. This background rejection is measured by the signal-to-background ratio, which must be large if three-dimensional information is to be obtained from a thick fluorescence sample. Two-photon excitation in laser scanning microscopy has an unparalleled ability to meet these demands. The two-photon excitation of a transition normally in the ultraviolet arises from the simultaneous non-linear absorption of two red photons. Because two-photon excitation depends on the square of the incident intensity, the resulting fluorescence is limited to the focal volume where the photon density of the focused laser illumination is high. This localization limits photobleaching and any photodamage to the focal plane of the image. This property is a major advantage over widefield or confocal microscopy. Two-photon excitation provides the depth discrimination associated with confocal microscopy without a confocal spatial filter, an advantage which allows for major simplifications of the apparatus. The resolution and background rejection properties of two-photon excitation have been calculated and measured, and have been shown to be identical to an ideal confocal microscope with the same optical wavelengths. Combined, these properties provide the ideal conditions for time-resolved imaging of fluorescence decay dynamics in order to characterize the submicroscopic environment of the fluorophore molecules within the specimen. A preliminary apparatus was designed and built to test these concepts, and it was found that fluorescence decay time images of living cells can be conveniently recorded with diffraction limited resolution in a few seconds of image acquisition time.

Fluorescence lifetime imaging (FLIM) is a new methodology in which the image contrast is derived from the fluorescence lifetime, not the local concentration and/or intensity of the fluorophore, at each point in a two-dimensional image. In our apparatus, the lifetime images are created from a series of phase-sensitive images obtained with a gain-modulated image intensifier. The phase-sensitive images obtained with various phase shifts of the gain- modulation signal are used to determine the phase angle and/or modulation of the emission at each pixel, which is in essence the phase or modulation lifetime image. Pixel-to-pixel scanning is not required to obtain the images. As an example of biochemical imaging we created lifetime images of the calcium concentration based on Ca²⁺-induced lifetime changes of calcium green (CaG), which is shown to be highly sensitive to [Ca²⁺]. Importantly, the FLIM method does not require the probe to display shifts in the excitation or emission spectra, which allows Ca²⁺ imaging using Ca²⁺ probes which do not display spectral shifts. The concept of fluorescence lifetime imaging has numerous potential applications in the biosciences. Fluorescence lifetimes are known to be sensitive to numerous chemical and physical factors such as pH, oxygen, temperature, cations, polarity, and binding to macromolecules. Hence, the FLIM method allows chemical or physical imaging of macroscopic and microscopic samples.

The authors studied the use of destructive interference of two diffusive photon-density waves for localization of an absorbing body and a fluorescent probe embedded in a scattering medium. The effect of the position of the embedded objects on the magnitude and phase of the light re-emitted from the medium was evaluated theoretically and experimentally. The objectives, accomplished with an asymmetrical laser-beam arrangement, were to reduce sensitivity to absorbing bodies located in superficial layers, while maintaining sensitivity to those lying deeper; and to establish a confined region of maximum sensitivity in which the distance of an absorbing body could be determined via phase measurement. Intensity and phase data were acquired with a modified frequency-domain spectrometer at modulation frequencies up to 600 MHz. Fluorescent probes were spatially localized with a symmetrical laser-beam arrangement. Magnitude and phase images acquired with a gated intensified CCD camera further defined the probe location. Simulations and experiments show potential applications to imaging.

Theory and instrumentation of lifetime-resolved fluorescence-detected circular dichroism spectroscopy (LRFDCD) are described. Advantages of LRFDCD over the individual techniques of fluorescence lifetime determinations and circular dichroism spectroscopy are discussed. LRFDCD is implemented on a frequency-domain fluorescence lifetime instrument that is modified for fluorescence-detected circular dichroism measurements.

Brookhaven National Laboratory is designing an ultraviolet free-electron laser (UV-FEL) user facility that will provide picosecond and subpicosecond pulses of coherent ultraviolet radiation for wavelengths from 300 to 75 nm. Pulse width will be variable from about 7 ps to approximately equals 200 fs, with repetition rates as high as 10⁴ Hz, single pulse energies > 1 mJ and, hence, peak pulse power > 200 MW and average beam power > 10 W. The facility will be capable of `pump-probe' experiments utilizing the FEL radiation with: (1) synchronized auxiliary lasers, (2) a second, independently tunable FEL beam or, (3) broad- spectrum, high-repetition rate recirculating superconducting linear accelerator which feeds pulses of electrons to two magnetic wigglers. Within these two devices, photons from tunable `conventional' lasers would be frequency multiplied and amplified. By synchronously tuning the seed laser and modulating the energy of the electron beam, tuning of as much as 60% in wavelength is possible between alternating pulses supplied to different experimental stations, with Fourier transform limited resolution. Thus, up to four independent experiments may operate at one time, each with independent control of the wavelength and pulse duration. The UV-FEL will make possible new avenues of inquiry in time-resolved studies of diverse fields including chemical, surface, and solid state physics, biology, and materials science. The experimental area is scheduled to include a station dedicated to biological research. The complement of experimental and support facilities required by the biology station will be determined by the interests of the user community.

Time-resolved ratio method is described for multicomponent pattern analysis by using a fluorescence lifetime imaging system and a ratio method calculation. Nanosecond-level time- resolved fluorescence images of a sample under a pulsed light excitation were detected directly by using a gated microchannel plate (MCP) image intensifier coupled to a CCD camera. Differences in fluorescence lifetimes of components in which each component has similar fluorescence wavelength characteristics were positively utilized to estimate every component spatial distribution. Spatial deconvolution of multicomponent fluorescence data was rapidly carried out from a set of time-resolved, two-dimensional fluorescence data. An application of a multigate time-resolved ratio method for fluorescence pattern analysis was provided. From the results, it is shown that the new method can be successfully used for multicomponent fluorescence pattern analysis at a reasonable signal-to-noise ratio.

A new microchannel plate photomultiplier (MCP PMT) has been developed and applied to the time-correlated photon counting technique. The ultrafast MCP PMT with 6 micrometers capillaries, which is the previous type of MCP PMT, was modified. We introduce the new compact and ultrafast MCP PMT and the high-speed electronics designed to be coupled with the MCP PMT. The newly developed compact ultrafast MCP PMT has achieved a time response of 25 psec of transit time spread and is compact in size. In addition, the high-speed amplifier and electronics module (amplifier and constant fraction discriminator) have been developed to deal with the pulses for the MCP PMT. The performance of these devices and their application are presented in this paper.

The extension of microscope luminescence measurements into the temporal domain provides the possibility of determining time-resolved properties of microscope samples and their surrounding environments, and thereby extends the conventional steady state measurements. `Time resolved imaging microscopy' is a relatively new technique whereby fast kinetic and luminescence decay parameters (decay times and the corresponding time or phase resolved amplitudes) are directly and simultaneously measured throughout an image, pixel by pixel, in an optical microscope. Molecular rotation, solvent and matrix relaxation, quenching mechanisms, reactions, and energy transfer are examples of molecular spectroscopic processes that can be studied best by directly measuring the time dependent properties. Dynamic measurements are generally much more informative than their steady state counterparts. The goal of our work is to develop time-resolved methods that can be applied conveniently and routinely to biological material in the microscope over a wide time domain. In addition to the augmented purely spectroscopic and reaction kinetic information, simultaneous spatial and temporal resolution of an image in a microscope provides significant improvement in image contrast, probe identification and differentiation, background (light scattering and inherent luminescence) reduction, and provides additional parameters for digital image analysis. Time resolution makes it possible to recover structures in an image concealed by background luminescence with a different lifetime. Examples of these procedures are given, and the instrumentation required for the data acquisition and analysis is discussed. The technique employs phase-locked coordination between the modulation of the perturbation and the recording of the luminescence image together with a Photometrics (Tucson, Ariz.) series 200 high resolution slow-scan scientific CCD camera. A normal fluorescence microscope is used.

Most proteins are capable of emitting tryptophan phosphorescence at room temperature in deoxygenated aqueous solutions. Like fluorescence, phosphorescence intensities and lifetimes are useful for studying protein structure. Phosphorescence differs from fluorescence, however, in several ways. Phosphorescence occurs on a time-scale of msec-sec, while fluorescence decays in nanoseconds. Second, the phosphorescence decay of a single tryptophan is nearly monoexponential, making assignments of decay components to individual residues possible. Finally, phosphorescence is a more sensitive probe of the local tryptophan environment, as the lifetime can change by orders of magnitude depending on site rigidity and other factors. The authors describe applications of phosphorescence spectroscopy for protein study. In particular, tryptophan phosphorescence quenching by resonance energy transfer to freely diffusing acceptors was used to show that Trp 109 is the origin of phosphorescence in E. coli alkaline phosphatase (AP). By following changes in the emissive lifetime of this deeply buried residue, the presence of an enzymatically active but structurally modified intermediate state is detected in the unfolding of AP in high concentrations of Guanidine:HCl, and followed the kinetics of the decline in activity upon further unfolding. In addition to the new understanding of AP, the results of these experiments show that room temperature tryptophan phosphorescence is a powerful tool for the study of proteins.

As a means to study the relationship of diffusion through proteins with the protein dynamics the authors used the quenching of phosphorescence of an intrinsic tryptophan by molecules that are free to diffuse in solution. Tryptophans in rigid environments within proteins are well known to exhibit long-lived phosphorescence at room temperature. Phosphorescence can be used to study the triplet quenching reaction in this way: the reactive triplet species of tryptophan is formed by excitation to the singlet state followed by spontaneous conversion to the triplet state and the reaction of the triplet state with an external molecule is monitored by a decrease in phosphorescence lifetime. The phosphorescence lifetimes of tryptophan in proteins are often > msec, as contrasted to fluorescence lifetimes on the order of nsec, therefore, phosphorescence is very sensitive to quenching reactions. In previous studies the quenching reaction of tryptophan by molecules with > 4 atoms was examined with the conclusion they interacted with buried tryptophan by a reaction that occurs over distance, not requiring the physical diffusion of these molecules through the protein for the reaction to occur. In contrast, diatomic molecules appear to be able to diffuse through the protein matrix, although diffusion is hindered relative to aqueous medium. Little information is available about the diffusion of triatomic molecules H₂O and CO₂. The authors use their sulfur analogues, H₂S and CS₂, respectively, to approach the question whether protein fluctuations allow their penetration.

The inherent resolution of protein phosphorescence spectra and the long lifetime of the triplet state lead to unique information on the structure as well as the slow rotational and conformational dynamics of proteins. However, protein triplet states are populated through singlet excitation, so that initial phosphorescence intensities are determined by prior occurrences at the excited singlet level. The relative intensities of the tryptophan components within protein phosphorescence spectra, when coupled with time-dependent triplet-state measurements, are a source of information on singlet transfer between residues. While anomalous phosphorescence decay times derive from triplet interactions, the relative intensities of the anomalous components in time-dependent measurements reflect quenching and excitation exchange at the singlet level. Time-dependent phosphorescence measurements can be employed to distinguish between singlet quenching mechanisms involving enhanced spin-orbit coupling and energy or electron transfer.

A linear correlation between spectroscopic and thermodynamic properties of systems is rarely encountered. In triplet state ODMR studies of various DNA complexes of echinomycin, a quinoxaline-containing cyclic depsipeptide bis-intercalating antibiotic, and its biosynthesized quinoline analogs, such correlations are observed. The zero field splitting D-parameter of the intercalated quinoxaline or quinoline residue varies linearly with the free energy of drug-DNA complexing. From previous work, the DNA sequence specificity of echinomycin analogs is known to be influenced by the identity of the intercalating residue (e.g., quinoxaline vs. quinoline). The present results strongly suggest that the DNA sequence-specificity of these drugs is controlled largely by the intercalated residue, and that the energetics of the peptide- DNA interaction, although considerable, are relatively sequence independent. These conclusions run counter to the generally accepted idea that DNA recognition by sequence- seeking proteins is controlled by specific hydrogen bonding interactions. The high degree of N-methylation of the echinomycin peptide portion severely restricts these interactions, however. A simple theoretical model is presented to support the experimentally observed linear correlation between (Delta) D and (Delta) G.

The erythrocyte membrane with its underlying cytoskeleton undergoes gross deformation during passage of the cell through fine capillary networks in the circulation. The physical linkages between the cytoskeleton and the membrane must involve substantial internal flexibility or points of articulation within and between molecules in the system. We have used time-resolved phosphorescence spectroscopy to examine three aspects of this dynamic flexibility on the microsecond to millisecond time scale: (1) the rotational diffusion and internal flexibility of F-actin, (2) the intrinsic flexibility of spectrin in solution and when reconstituted to erythrocyte membranes, and (3) internal flexibility in the cytoplasmic domain of band three. The results show that the spectrin dimer displays substantial flexibility on the microsecond time scale: this flexibility is largely retained when spectrin is reassociated with the erythrocyte membrane. There is also evidence that the cytoplasmic domain of band three itself is flexible since the rotational correlation time for band three labeled through this domain (via a phosphorescently labeled monoclonal Fab) is significantly shorter than when the protein is labeled directly at the external face of its transmembrane domain. Short actin filaments possess limited torsional motion on the submicrosecond time scale and are unlikely to contribute to the flexibility of the cytoskeleton.

A spin-labeled derivative of eosin was chemically synthesized from 5-aminoeosin and the nitroxide spin label 2,2,5,5-tetramethylpyrrolin-1-oxyl-3-carboxylic acid. Following determination of the chemical identity of the spin-labeled eosin (5-SLE) by FAB mass spectroscopy, its optical and magnetic resonance spectroscopic properties were characterized in aqueous solution and compared to a diamagnetic eosin derivative, 5-acetamido eosin (5- AcE). The visible light absorption maximum of 5-SLE was 518 nm, the same as for 5-AcE. The fluorescence quantum yield of 5-SLE was only reduced by approximately 10% relative to 5-AcE, and the fluorescence lifetime was marginally reduced relative to 5-AcE. The phosphorescence lifetime and yield for 5-SLE were very similar to those for 5-AcE. The phosphorescence yield of 5-SLE bound noncovalently to BSA was reduced by approximately 60% relative to 5-AcE, and the phosphorescence lifetime reduced from approximately 2.4 msec (5-AcE) to 1.6 msec (5-SLE). Reduction of the nitroxide moiety of the 5-SLE with sodium ascorbate resulted in minimal changes in the fluorescence and phosphorescence quantum yields and lifetimes. This indicated that the unpaired electron of the nitroxide spin label did not seriously affect the optical spectroscopic characteristics of the spin-labeled eosin molecule. The quantum yields and lifetimes of 5-SLE were still quite acceptable for time- resolved fluorescence and phosphorescence studies. The electron paramagnetic resonance (EPR) spectrum of 5-SLE in aqueous solution has a lineshape consistent with a molecule the size of 5-SLE undergoing rapid rotational reorientation. When bound to BSA, the EPR spectrum of 5-SLE was broadened to a near slow motion limit for EPR, as expected for the relatively slowly rotating protein-5-SLE complex. Time-resolved phosphorescence anisotropy and saturation transfer EPR (ST-EPR) experiments with samples of 5-SLE bound to BSA in solutions of varying glycerol concentrations at 2 degree(s)C demonstrated that this combined probe is suitable for monitoring rotational dynamics of macromolecular systems over the appropriate time ranges. These studies have shown that one combined optical and EPR probe, 5-SLE, can be employed in the full range of fluorescence, phosphorescence, EPR, and ST- EPR spectroscopies.

A newly developed dual-channel time domain phosphorimeter is described in this paper. This employs a pulsed, frequency doubled Nd:YAG laser as a linear polarized pumping source. The resulting orthogonally polarized emission components are then simultaneously collected by two oppositely positioned photomultiplier tubes (PMTs), digitized by a digital storage adaptor, and then transferred to an IBM PC microcomputer where the intensity or the depolarization (anisotropy) of the emission is calculated. The range of lifetimes or rotational correlation times that can be routinely determined with this instrument extends from a few microseconds to milliseconds. The instrument has an 8 bit magnitude resolution and a signal sampling rate of 20 mesa-samples per second. The averaging function of the instrument yields a single measurement with an average of over 2 to 255 excitation transients. To demonstrate the utility of the new instrument, studies of the emission properties of the triplet probe eosin immobilized in a polymer matrix and covalently bound to a membrane protein are reported.

The photophysical and photochemical properties of sulphonated aluminum and zinc phthalocyanines have been investigated in a range of solvents and model biological systems. Anomalous effects are observed upon deuteration of the solvent and addition of fluoride ions. In D₂O the excited singlet and triplet state lifetimes and quantum yields of fluorescence and triplet state formation are increased relative to H₂O. No solvent isotope effect is observed between CH₃OH and CH₃OD. It is proposed that relaxation of the excited state involves a tunnelling type interaction in which the phthalocyanine's highly energetic metal-axial ligand stretching vibrations are coupled to the HO-H or DO-D stretching vibrations. A significant increase in triplet lifetimes of phthalocyanine sensitizers bound to protein substrates is observed which is a function of the degree of sulphonation. The implications of these results to the determination of the quantum yields of singlet oxygen formation in D₂O and lipophilic environments are discussed.

The `zinc finger' motif, found in nucleic acid-binding proteins, consists of a peptide domain which tetrahedrally coordinates a zinc ion via cysteine (sulfhydryl) and histidine (imidazole nitrogen) sidechain atoms. The CCHH class, in which zinc binds to a pair of cysteines and a pair of histidines, is commonly found in eukaryotic transcription factors. These transcription factors cannot bind DNA in the absence of metal ion, and physical studies (CD, NMR) indicate that a more defined structure is induced upon metal binding. Fluorescence energy transfer measurements were performed on a zinc finger peptide which contains a single CCHH metal-binding domain. An intrinsic conserved tryptophan, located at the midpoint of the peptide chain, serves as the energy donor to one of two dansyl acceptors (one acceptor is attached to the (alpha) -amino group and the other to the (epsilon) -amino group of a carboxy-terminal lysine). Distance distributions between the donor and acceptor were determined for zinc-bound and metal-free peptide using time-resolved frequency-domain fluorometry. The distance distributions were shorter and narrower for the zinc-bound peptide than those recovered for the zinc-free peptide. These results confirm previous experimental evidence which indicates that metal ion is required to form a well-defined solution conformation.

Fluorescence dynamics methods are used to probe the mechanism by which the chemi- energized intermediates of the bacterial luciferase catalyzed oxidation of FMNH₂ and tetradecanal are able to excite the ligand of lumazine protein to its first excited singlet state. A fluorescence dynamics study of the effect of lumazine protein on the reaction of several types of luciferase has recently been published (Biochemistry 30 6825, 1991). This present report examines the case of the Photobacterium leiognathi luciferase reaction in more detail. The fluorescence anisotropy of a mixture of this luciferase fluorescent transient mixed with lumazine protein decays rapidly with a correlation time of 5 ns, interpreted as due to energy transfer. There is no sign of a longer time corresponding to the rotation of the proteins themselves. No rise time of the lumazine (acceptor) fluorescence on exciting into the fluorescent transient (donor) absorption is measureable, so that no straightforward estimate of the energy transfer rate can be made.

Using the Forster equations we have estimated the rate of energy transfer from tryptophans to hemes in hemoglobin. We computed the orientation factors and the consequent transfer rates from the crystallographic coordinates of human oxy- and deoxyhemoglobin. The intrasubunit distances between hemes and tryptophans allow lifetimes between 5 and 15 ps per each ns of tryptophan lifetime. Lifetimes of several hundred ps would be allowed by the intersubunit distances, however, these distances become operative only when one heme per tetramer does not accept transfer. If more than one heme per tetramer does not function as an acceptor, lifetimes of several ns would be allowed.

We measure the intensity and anisotropy decays of the intrinsic tryptophan emission from hemoglobin solutions obtained using a 10 GHz frequency-domain fluorometer and a specially designed cuvette which allows front face excitation on a free liquid surface. The cuvette eliminates reflections and stray emissions, which become significant for low intensity fluorescence like in hemoglobin. Three lifetimes are detectable in the subnanosecond range. The average lifetime of hemoglobin is ligand dependent. Fluorescence anisotropy decays of oxy, deoxy, and carbonmonoxyhemoglobin can be fitted with up to three correlation times. When three components are used the floating initial anisotropy r_o is in each case higher than the steady-state anisotropy of tryptophan in vitrified solution. For deoxy hemoglobin it is close to 0.4. The data are consistent with an initial loss of anisotropy from 0.4 to about 0.2 occurring in the first two picoseconds.

Cytochrome b₅, isolated by detergent extraction from rabbit liver, has been extensively studied by fluorescence techniques, however, its fluorescence properties are complicated by the presence of three tryptophans in the membrane-binding domain. This protein has now been expressed in E. coli and a mutant form has been isolated which contains only one tryptophan (Trp-109) in the membrane-binding domain. The mutant protein does differ from the native protein in several of its spectroscopic properties, but it interacts with lipids in a similar way and shows the same functional activity. The availability of a mutant cytochrome b₅ with a single Trp in the nonpolar domain enables several frequency-domain experiments: to estimate the distribution of distances between Trp-109 and the heme; to study the incorporation of the protein into the membrane; to study the quenching of the tryptophan fluorescence by lipids with a brominated acyl chain; and to analyze the shape of the lifetime distribution.

The neurophysins (NP) are proteins which function in neurohypophyseal peptide hormone transport and storage. We have been investigating the fluorescence properties of the single tyrosine residue (Tyr⁴⁹) of bovine NPs. Since the intensity of Tyr⁴⁹ fluorescence increases on ligand binding, measurements have been made with and without small peptides which bind at the hormone binding site. Steady-state iodide quenching produces Stern-Volmer plots with downward and upward curvature for NP and the NP:Phe-PheNH₂ complex, respectively. Non-linear Stern-Volmer plots can result from multiple conformations of the solvated protein, multiple environments for the rotational isomers of the phenol side chain (rotamers), as well as from dynamic and static interactions of Tyr⁴⁹ with the quencher. To help explain the steady-state data, time-resolved fluorescence quenching studies have been performed. Both NP and the NP:Phe-PheNH₂ complex exhibit fluorescence intensity decays that can be fit by the sum of three exponentials that consist of the same time constants but with different weighting (amplitude) terms. With increasing [I^-], the decay parameters for Tyr⁴⁹ in NP are consistent with a multiple species model due to rotamers of Tyr⁴⁹, the amplitudes are quencher independent and the lifetimes result in linear Stern-Volmer plots. The quenching studies on the NP:PhePheNH₂ binary complex, however, cannot be analyzed in terms of a specific model because the fractional intensities of two of the exponential components are small.

The thermal and guanidine(DOT)HCl induced unfolding of Staphylococcal nuclease A, and its hybrid mutant, NCASG28, have been investigated by time-resolved studies of the fluorescence of the single tryptophan residue, Trp-140. The NCASG28 mutant has a much lower thermodynamic stability than the wild type. The fluorescence decay of both proteins is found to change significantly as temperature and guanidine concentration is increased. These changes appear to reflect unfolding transitions in the proteins. When analyzed in terms of a bi- exponential decay law, the amplitude ((alpha) ) of the long fluorescence decay time ((tau) ) decreases with unfolding; likewise the (alpha) for the short (tau) increases with unfolding. A global analysis of the temperature and guanidine dependence data sets was performed. The recovered pre-exponential values were then analyzed in terms of a two-state unfolding transition. The time-resolved data are consistent with such a two-state model and analysis yielded values of thermodynamic parameters, including the enthalpy change for the thermal transition and the `m' value for the guanidine-induced unfolding transition.

The energy transfer among several fluorophores when bound to linear DNA has been studied. The intercalation and groove binding of the fluorophores and relatively large persistent length of DNA makes it a good model for one dimensional energy transfer. In this case, as predicted by Foerster's theory, the donor intensity decayed with a t^1/6 time dependence. The presence of a finite volume with a restricted geometry leads to significantly different donor intensity decays from that predicted by Foerster's model. We used intercalated/groove-bound fluorophores as donors and transition metal ion complexes which only bind on the outside surface of the DNA as acceptors, to characterize energy transfer in a cylindrical geometry. Two models were used: a hard cylinder with a donor on the z-axis and acceptors on the surface, and a soft-boundary cylinder where a distribution of acceptors within a cylindrical volume was allowed. The energy transfer among intercalated/groove-bound donors and surface-bound acceptors in DNA can be described with soft-boundary cylindrical geometry with reasonable parameters.

Fluorescence energy transfer (FET) can be used to obtain structural and dynamic information on duplex and branched DNA molecules suitably labeled with donor and acceptor dyes. However, dye-DNA interaction cause excited-state quenching of the dyes and complicate the analysis of FET data. We have investigated the excited-state quenching of two common FET dyes, fluorescein and tetramethyl rhodamine, covalently linked to synthetic oligonucleotides. The rate of quenching is shown to depend on the base sequence and association state of the oligonucleotide, and also the length of the linker chain. Furthermore, the dyes can adopt more than one configuration with respect to the oligonucleotide strand and the degree of quenching is different in each configuration. The implications of these findings for FET measurements in nucleic acids are discussed.

Fluorescence lifetimes of free and DNA-intercalated ethidium bromide have been established in different studies to be 1.6 - 1.8 and 20 - 25 ns, respectively. Recently, several researchers have suggested a new DNA-binding mode of ethidium based on a third fluorescence lifetime between 10 and 15 ns. We are able to reproduce the latter lifetime in ethidium:DNA complexes only when the ligand per DNA base-pair ratio (r) exceeds 0.2. Other criteria that may affect the binding properties of ethidium, such as temperature, salt concentration, length of the DNA (end effects), and base composition of the DNA only influence the appearance of this new lifetime because of their effect on the spatial distribution of intercalated ethidium. Based on single curve and global analysis of fluorescence intensity and fluorescence anisotropy decay data, we propose that the new lifetime is due to self-quenching of intercalated ethidium. Detectable quenching occurs at average intercalation site separations of less than 10 base- pairs.

The interaction of copper(II) and its 5,7,7,12,14,14-hexamethyl-1,4,8,11-tetraazacyclo- tetradeca-4,11-diene(HTDE) complex with calf thymus DNA has been investigated by ethidium fluorescence spectroscopy. The experimental data shows that Cu(II), Cu(II)HTDE, and Cu(II)HTDE-thiol-hydrogen peroxide cause ethidium to dissociate from its host molecule and decrease the relative emission intensity of DNA-ethidium. Two different interaction modes could be deduced from the Cu(II) characteristics, the structure, steric constraint of Cu(II)HTDE, and the composition of the system. The evidence accumulated suggests that Cu(II)HTDE intercalates predominantly into the double-helical DNA in the absence of a reducing agent and hydrogen peroxide, and that it binds within the minor groove of DNA in the presence of RSH and H2O2.

Measurements of time-resolved intramolecular energy transfer in progressively stretched poly(vinyl alcohol) films were performed. The donor (tryptophan) and acceptor (dansyl) were linked with flexible polymethylene chain. Distance distributions were recovered from frequency-domain measurements of the donor decay. In isotropic PVA (in solution) a wide range of distances were detected (Gaussian full width at half maximum of about 16 angstroms) with an average distance of 13 angstroms. The donor-acceptor distance distribution became progressively more narrow when the PVA films were stretched. Four-fold or more stretching results in a single donor-acceptor distance. The maximal measured donor-acceptor distance of 23 angstroms is in excellent agreement with the computed distance for the fully stretched conformation. The possibility of partial and/or full ordering of polymethylene chains by stretching the PVA films can also be useful in the study of other distance-dependent interactions, such as electron transfer or distance-dependent quenching.

The interaction of dibucaine with model membrane systems has been studied using both equilibrium dialysis and time-resolved fluorescence methods. Binding isotherms, for the interaction of dibucaine with unilamellar phospholipid vesicles, were obtained as a function of ionic strength and as a function of initial surface charge on the vesicles. The binding of dibucaine is shown to be significantly influenced by electrostatic interactions; binding isotherms were fitted with a Guoy-Chapman-Stern model in terms of an intrinsic dissociation constant of 2.9 mM and a maximum extent of binding of approximately one drug molecule per phospholipid molecule. Various state steady-state and time-resolved fluorescence studies were performed to characterize the topography and dynamics of dibucaine molecules which are bound to the phospholipid vesicles. Quenching studies with KI show bound dibucaine to be protected from the anionic quencher. The fluorescence lifetime of free and vesicles-bound dibucaine is found to be 3.3 and approximately 2.6 ns, respectively. Anisotropy decay measurements indicate that the effective rotational correlation time of bound dibucaine is increased to approximately 4 ns, as compared to a value of 0.06 ns for free dibucaine. An infinitely long rotational correlation time (or a r_(infinity )) was not observed for bound dibucaine, indicating that the rotational motion of bound dibucaine is isotropic.

Fluorescent analogs of phosphatidylcholine (PC) and sphingomyelin (SM) were prepared and incorporated into the surface layer of human low-density lipoprotein (LDL) and lipoprotein (a) [Lp(a)]. Diphenylhexatrienylpropionic acid (DPH) was covalently linked to PC (DPH-PC) and SM (DPH-SM). Fluorescence lifetimes of the labeled lipoproteins were determined by phase and modulation fluorometry. Bimodal Lorentzian distributions for the decay times of DPH-PC and DPH-SM in LDL and Lp(a) were found. Lifetime distribution centers for labeled lipids were very similar except for DPH-PC in Lp(a) which was shifted to longer lifetimes. Thus, PC experiences a less polar environment in Lp(a) than in LDL. The distributional width of DPH-PC in Lp(a) was broader than in LDL. Accordingly, phosphatidylcholine must be localized in a much more homogeneous environment in LDL as compared with Lp(a). On the other hand, no difference in distributional widths was observed for DPH-SM in both lipoproteins, showing that SM organization in Lp(a) is unaffected by apo(a). From the obtained fluorescence data, it is proposed that apoprotein B preferentially associates with phosphatidylcholine.

Myosin-containing thick and actin-containing thin filaments generate force in vertebrate muscles. In an effort to monitor molecular dynamics and its relation to function in these filaments, we have labeled sulfhydryls actin (cys-374) and mysoin (SH1) with the triplet probe erythrosin-5-iodoacetamide. Fluorescence studies indicate that the probes are rigidly bound to the proteins and (probably) associated with the protein surface. Although the probe phosphorescence in solution is always mono-exponential, in the protein conjugates the decays are mono-exponential for actin but multiexponential for myosin at 20 degree(s)C. The steady- state anisotropy (averaged over the time window from 0.07 to 1.5 ms) of erythrosin-labeled G- actin is 0.0; in F-actin the anisotropy is 0.088 at 20 degree(s)C and increases to 0.10 when the peptide toxin phalloidin, which is known to stabilize F-actin filaments, is bound.

The Na⁺, K⁺-adenosine triphosphatase (ATPase) in microsomal membrane vesicles has been covalently labeled with the triplet probe eosin 5'- isothiocynate. Rotational mobility of the protein has been investigated by measurement of time-resolved depolarization of the emitted phosphorescence from the triplet state of eosin, induced by a laser pulse. The probe was attached non-specifically to the protein and under conditions where the eosin was attached specifically to a lysine residue located at the putative ATP binding site. The total anisotropy of the emission was found to be almost constant when measured over the temperature range 10 degree(s)C - 25 degree(s)C. The anisotropy value was relatively small when the label was bound non-specifically to the binding sites of the protein, but was markedly increased when specifically bound to the protein, suggesting that the independent motion of the probe was constrained at this site. The anisotropy decay curve obtained from the specifically labeled protein shows a clearly biphase character, and is composed of a rapidly rotating component with a rotational correlation time of 20 microsecond(s) - 5microsecond(s) and a slower rotating component with a rotational correlation time of 250 microsecond(s) - 90microsecond(s) , the temperature range over 10 degree(s)C - 25 degree(s)C. These motions are individually assigned to the segmental motion of the polypeptide chain and the whole protein rotating about its axis normal to the plane of the membrane. It was estimated that about 80% of the total anisotropy signal was contributed by the fast rotation, and the remainder from slow rotation.

Time-resolved phosphorescence anisotropy has been used to follow the rotational diffusion of band 3, the major integral membrane protein of erythrocytes. The rate of rotational diffusion is exquisitely sensitive to protein-protein interactions which increase the size of the rotating species. In membranes prepared from normal human erythrocytes, band 3 appears to exist as a heterogeneous population of aggregates with different mobilities. Restriction of band 3 motion is probably due, at least in part, to interactions with the peripheral membrane protein, ankyrin. Surprisingly, the further linkage of ankyrin to spectrin in the underlying cytoskeleton does not appear to influence band 3 mobility. We have found that band 3 has a highly restricted mobility in membranes prepared from malaria-infected erythrocytes. This decreased rotational freedom of band 3 correlates with a decrease in deformability and an altered morphology. Band 3 mobility was also assessed in the membranes of Melanesian ovalocytes. These malaria- resistant erythrocytes from donors in Papua New Guinea show a three-fold decrease in deformability. The decreased deformability and ovalocytic shape of these erythrocytes was found to be associated with a dramatic decrease in band 3 mobility.

The tryptic peptide OT-16 of oxidized ribonuclease A, which consist of the 20 C-terminal amino acid residues of the protein, is thought to contain a chain-folding-initiation-site at residues 106 - 118. Non-radiative energy transfer has been used to assess the structure and dynamics of this peptide alone, and conjugated to another fragment of the ribonuclease molecule, in solution. Preliminary work has also been carried out on the S-peptide of ribonuclease A.

One of the major aspects of fluorescence spectroscopy which differentiates this technique from many other spectroscopic approaches is the inherent multidimensional nature of the data. For instance, the basic pulsed-laser fluorescence data set is characterized by fluorescence versus: emission wavelength, polarization state (parallel and perpendicular intensities), time of emission (picoseconds to nanoseconds), and time of biological reaction (milliseconds to minutes). Usually, this six-dimensional data set is obtained piecemeal, single dimension at a time; often complete data sets are not even collected. This is especially true of the biological time scale axis. Data acquisition times for picosecond decay data are typically seconds to minutes, and, therefore, it has not been generally possible to perform this experiment in a kinetic mode. What is described in this report is the construction of a parallel multichannel time-correlated single-photon counting (TCSPC) fluorometer which is capable of simultaneous collection of: fluorescence vs. picosecond to nanosecond time vs. emission wavelength vs. polarization state vs. millisecond to second time. Use is made of two multi-anode microchannel plate detectors, each obtaining data at two different polarization states, six different emission wavelengths, along 12 independent TCSPC channels. This instrument is interfaced to a three-syringe stepper motor controlled stop-flow apparatus, and picosecond decay data along all of these channels is stored and collected by two 33 MHz 80486 computers at rates approaching 1200 - 12000 data sets per second.

The time-resolved fluorescence anisotropy of myosin S₁ covalently labeled with Eosin-5- maleimide and 1,5-I-AEDANS was measured in solution. Each probe was specifically attached at one SH-group on the S1. The two most reactive SH sites on the heavy chain of the myosin S₁ were used. The fluorescence anisotropy was measured at different excitation wavelengths. In this way, several absorption moments were utilized, each having a distinct orientation in the frame of the dye. The orientations of the transition moments in the dyes were determined in a separate experiment using an angle resolved fluorescence depolarization experiment on dyes embedded in stretched matrices of PVA polymers. The anisotropy decay curves exhibit fast (<3 ns) and slow (> 100 ns) components. The slow decay components reflect the motion of the large protein molecules. The fast anisotropy decay are attributed to a fast, but restricted, motion of the bound dye relative to the protein as experiments on free dyes in solution reveal subnanosecond anisotropy decays. The anisotropy decays have been analyzed in terms of a model which describes the restricted motion of the dye molecule relative to the protein and the overall rotation of the dye-protein complex in solution. An important element in the model is the incorporation of the orientational distribution of the dye relative to the protein. The observed anisotropy decays were analyzed using a global target approach in which the experimental data obtained at different excitation wavelengths are fitted simultaneously to the theoretical model. It is important to note that the orientational distribution of the dye relative to the protein, as well as the rotational correlation times of the motions for a dye attached to a given binding site, are independent of the excitation wavelength used. This leads to a reduction in the number of independent parameters optimized by the nonlinear least squares procedure. The orientational distribution of the dye relative to the protein obtained in this way is particularly useful for the interpretation of fluorescence depolarization data obtained from labeled muscle fibers. Indeed, knowledge of the distribution function of a dye attached to a binding site of the S1 protein is a prerequisite for a probe-independent determination of the orientational distribution of the S₁ proteins themselves in the muscle fiber.

A number of theoretical models are introduced in which donors and acceptors perform lateral or rotational motion during the time in which fluorescence energy transfer takes place. Each case is reduced to an eigenvalue problem, and for each model calculations are made of observables, such as fluorescence intensities and anisotropies, by employing matrix methods. It is shown that the observables depend on the size of the motional step only if fluorescence energy transfer occurs. This finding indicates that fluorescence energy transfer studies may reveal whether the dynamics of a system (e.g., a protein) is better described in terms of transitions between a number of discrete states or in terms of diffusion equations. Some theoretical tools are offered for analyzing fluorescence energy transfer data without restrictive assumptions for motional averaging regimes or the orientation factor.

Nonradiative energy transfer from perylene to Co²⁺ and Ni²⁺ ions has been investigated below the phase transition in small unilamellar vesicles. In the case of cobalt, the quenching of perylene fluorescence can be described well by long-range Foerster dipole-dipole energy transfer. Although fluorescence decay data for perylene quenched by nickel also fits the Foerster model, other evidence suggests a different quenching mechanism which is more akin to the shorter-range Dexter exchange interaction.

The single tyrosine, Tyr-139, of wheat germ calmodulin provides an intrinsic fluorescent probe to monitor Ca²⁺-binding domain 4, while the single cysteine, Cys-27, provides a site for the attachment of an extrinsic fluorescent label to monitor the N-terminal lobe. This has resulted in a means of comparing the response of the N- and C- terminal regions to pH, ionic strength, and Ca²⁺ level. Ca²⁺ ligation decreases the mobility sensed by Tyr-139 at neutral pH, while a shift in pH to 5.2 results in a further decrease.

The relation between the three Trp residues in barnase has been characterized by computer studies of the molecular model. Trp-35 is shown to be a lone residue. However, the distance between, and the orientation of Trp-94 and Trp-71, should allow efficient energy transfer in the two directions. The overlap integrals have been calculated from the spectra of the individual Trp residues, by subtracting the spectrum of single Trp mutants from the spectra of the wild-type. A multifrequency phasefluorometric study is performed for wild-type barnase and mutant proteins. The lifetimes of the three tryptophans in the wild-type protein have been resolved. To Trp-35 a single fluorescence lifetime is attributed which varies in the different proteins between 4.3 and 4.8 ns and is pH independent between pH 5.8 and 8.9. Trp-71 and Trp-94 are considered to behave as an energy transfer couple with both forward and reverse energy transfer. To the couple two fluorescence lifetimes are attributed: 2.42 (+/- 0.2) ns and 0.74 (+/- 0.1) ns at pH 8.9, and 0.89 (+/- 0.05) ns and 0.65 (+/- 0.05) ns at pH 5.8. In the mutant Trp-94->Phe the lifetime of Trp-71 is 4.73 (+/- 0.008) ns at high and 4.70 (+/- 0.004) at low pH. In the mutant Trp-71->Tyr the lifetime of Trp-94 is 1.57 (+/- 0.03) ns at high and 0.82 (+/- 0.025) at low pH. From these lifetimes, energy transfer efficiencies can be calculated according to Porter. At pH 8.9 a 71% efficiency was found for forward transfer (from Trp-71 to Trp-94) and 36% for reverse transfer. At pH 5.8 the transfer efficiency was found to be 86% for forward and 4% for reverse transfer (all +/- 2%). These transfer efficiencies correspond fairly well with the ones calculated according to the theory of Foerster. The Fluorescence lifetime of Trp-94, as determined in a mutant which lacks Trp-71, is found to be heavily quenched by the neighboring imidazole group of His-18. The results demonstrate the simultaneous forward and reverse energy transfer between two tryptophan residues and the quenching effect of a neighbor imidazole group.

Radiative decay rate depends on the efficiency of coupling between the emission dipole and the electromagnetic field. Therefore, in optically discontinuous and/or anisotropic environments the radiative rate is dependent on the orientation of the emission dipole. Single-bilayer phospholipid membranes are interesting systems for the study of this phenomenon because in these systems the functional dependence of the radiative rate is simple and the effect is stronger than in liquid crystals. The orientational dependence of the radiative rate results in a non-exponential total fluorescence intensity decay. Orientational distribution of the probe in the gel-phase membrane can be determined from the total intensity decay. In the liquid-phase membrane the coupling between decay and rotation leads to a polarized intensity decay behavior entirely different than that in isotropic media. Experimental data are consistent with every theoretical prediction.

A novel confocal microscope is presented using the Daresbury Synchrotron Radiation source as its light source. The broad spectrum of synchrotron radiation in combination with the UV compatible microscope allows the extension of confocal microscopy from the visible to the UV region down to about 200 nm. It is envisaged that structures separated by about 70 nm can be resolved at a wavelength of 200 nm. In addition, the tunability of synchrotron radiation affords the selective excitation of any specific fluorescent molecule at the maximum of the absorption band. This avoids the restriction of working at fixed laser lines. A further advantage of using synchrotron radiation is the realization of multiwavelength excitation. Test results using laser systems in the visible and in the UV are presented. Fluorescence images of test targets using UV excitation reveal the superior resolution of the microscope. Furthermore, images of Leydig cells incubated with a fluorescent cholesterol derivative whose maximum of absorption is at 325 nm are shown. These images cannot be produced by conventional confocal laser microscopes. Finally, promising preliminary results obtained with synchrotron radiation are presented.

The image of a circular edge was calculated as determined by a scanning confocal microscope with fully coherent illumination. In scalar theory, the quarter-intensity point locates the geometrical-optics image of a straight edge. For a circular object, however, the quarter- intensity point is displaced from the geometrical-optics image of the edge according to the diameter of the object. For example, for an object that has a diameter of 21 resolution limits the displacement error is approximately 0.01 resolution limits. The error that results from locating the quarter-intensity point for diameters as small as 1 resolution limit is given. The error is even greater if the object is scanned off-axis. For example, the error for an object whose diameter is 21 resolution limits and which is scanned 3 resolution limits off-axis is approximately 0.45 resolution limits. Finally errors are calculated for vertical lines of width as small as 1 resolution limit.

The fluorescence of polycyclic aromatic hydrocarbon-nucleic acid complexes is quenched by photoinduced electron transfer mechanisms in aqueous solutions at ambient temperatures. These effects are illustrated with the biologically important compound benzo[a]pyrene-7,8-diol-9,10-epoxide (BPDE), a mutagenic and carcinogenic metabolite of the environmental pollutant benzo[a]pyrene, which forms covalent mutagenic lesions with 2'-deoxyguanosine (dG) residues in DNA. The dependence of the fluorescence yield and fluorescence decay times of the covalent model adduct (+)-trans- BPDE-N²-dG as a function of temperature and methanol/water composition are described. Because of the sensitivity of the fluorescence of the pyrenyl residue to the polarity of the microenvironment, the magnitude of the fluorescence yield can be used to distinguish between highly hydrophobic (e.g., intercalation) and other more solvent-exposed BPDE- nucleic acid binding sites.

The time-resolved fluorescence spectroscopy of human and bovine oxy- and deoxyhemoglobins was measured in either 0.03 M phosphate buffer or 0.03 M borate buffer between pH 6.5 and 9.2. A frequency resolved fluorometer was used with bandwidth up to 10 GHz. Excitation was at 294 nm, the emission was monitored through a broad band interference filter centered at 335 nm, coupled to a cut-off filter at 316 nm. In all cases, the best simulations were obtained with two discrete exponential decays, one near 30 ps, and the other of several hundred ps. In human hemoglobin, the longer component showed a substantial lengthening upon removal of oxygen. In bovine hemoglobin, the shorter component decreased upon deoxygenation. It was possible to infer that the shorter lifetimes originated from the average intrachain distances. However, the hemes at this longer distance would become the main acceptors of energy transfer only when the energy transfer at intrasubunit distance is inhibited. It is suggested that this is due to the presence of `disordered' heme in the system.

The porphyrin in metal-free cytochrome c molecules was investigated using fluorescence line narrowing spectroscopy. An energy map is defined to illustrate the environmental effects on the vibronic structure of a chromophore. In the approximation of pure electronic distortion, the population distribution in the energy scale can be characterized by a function. An intuitive definition of this distribution function is determined for the porphyrin in cytochrome c, and the validity of this approximation is experimentally verified. In addition, numerous excited state vibrational levels suggest a method to quantify the relative absorption probabilities for (0,0)yields(1,k) transitions.

The rotational diffusion of nucleosome core particles were measured using the fluorescence anisotropy decay of ethidium intercalated into the core particle DNA as a function of ionic strength. The `native' form of the core particle (ionic strength from 0.001 to 0.1 M) has a rotational correlation time ((phi) <sub>max</sub>) of approximately 170 ns. At higher salt concentrations (phi) <sub>max</sub> rises slowly from approximately 170 ns at 0.1 M NaCl to a value of approximately 230 ns at 0.35 M NaCl, a point just above physiological ionic strength; we call this change the moderate-salt transition. (phi) <sub>max</sub> remains constant at approximately 230 ns until the onset of the `high-salt dissociation' which occurs above 0.7 M NaCl. This dissociation begins with the release of H2A-H2B dimers; increasing DNA flexibility in this salt range prohibits accurate measurement of the rotational correlation time beyond this point. Light treatment of the ore particles with trypsin to remove the N-terminal histone domains abolishes the plateau in (phi) _max between 0.35 and 0.65 M NaCl. Thus, the moderate-salt transition derives from the release of these N-terminal ends from the body of the core particle. The anisotropy decays show no evidence for DNA release from the core particle at salt concentrations below 0.65 M.

This paper examines the effect of ionic strength on the decay of intrinsic tyrosine fluorescence from chromatin core particles. All measured decays were complex, showing at least four components. At 0.1 M salt the average lifetime was approximately 0.5 ns with 67% of the emitting tyrosines having lifetimes on the order of 0.3 ns, 30% near 0.8 ns, and the remainder around 1.8 ns. The average lifetime increases continuously by about 50% as the ionic strength is decreased to very low values. The decrease was characterized by a small shift in contributions from components with approximately 0.3 ns lifetimes mainly to components with lifetimes on the order of 1.8 ns. The rate of change was greatest over the range of the low-salt transition ([Na]> 0.5 mM). As the salt concentration was increased beyond 0.1 M a large increase in average lifetime is observed as the histone proteins within the core particle dissociate from the DNA. The average lifetime increased to a maximum of approximately 1.6 ns due to a large decrease in contribution from components with approximately 0.3 ns lifetimes accompanied by the appearance of components with lifetimes mainly from approximately 1.8 to approximately 4.0 ns. Throughout the range of salt examined the changes in average decay lifetime paralleled changes in the steady-state fluorescence intensity. This agreement indicates that all tyrosines contributing to the fluorescence at very high and very low salt also contribute at intermediate (0.1 M) salt. It is not necessary, as suggested by others, to consider some tyrosines to be statically quenched in the intact core particle.

C-phycocyanin (CPC) and Allophycocyanin (APC) are pigment-protein complexes isolated from antenna systems in cyanobacteria. The crystal structure of CPC has recently been solved and APC has a similar structure. CPC and APC have a trimeric structure, monomeric subunits are composed of an (alpha) and (beta) polypeptide chain, each has a tetrapyrrole chromophore chemically bound to position 84. In CPC and APC trimers, the (alpha) 84 and (beta) 84 chromophores in adjacent monomers are in close proximity, forming relatively strong coupled pairs. Calculation of pairwise energy transfer rates using Foerster theory has suggested an extremely fast transfer (> 1 ps^-1) between the (alpha) 84 and (beta) 84 pair in CPC. A femtosecond fluorescence up-conversion apparatus was constructed which achieves subhundred femtosecond time resolution. This allows experimental observation of the fast energy transfer process between the (alpha) 84 and (beta) 84 pair in both CPC and APC. There was also a wavelength dependence of the fluorescence depolarization kinetics which is inconsistent with Foerster inductive resonance energy transfer theory.

Exploration of the role of tryptophan residues in aspartate transcarbamylase (ATCase) was performed by combining site-directed mutagenesis and time-resolved fluorescence measurements. ATCase, an allosteric enzyme of the pyrimidine pathway, is built from three dimeric regulatory subunits and two catalytic trimers. Each catalytic subunit contains two tryptophan residues in position 209 and 284. Two single tryptophan mutants, W209F and W284F were constructed. Analysis by the maximum entropy method of the total fluorescence intensity decays, provides three lifetime classes centered around 0.4 - 0.5, 1.4 - 1.6, and 2.4 - 2.6 ns, respectively, for the wild type enzyme. Analysis of the fluorescence decays permitted attribution of the shorter lifetime to tryptophan in position 284. The isolated catalytic trimers, although devoid of any cooperativity, display very similar fluorescence decays compared to the holoenzymes. In each case, the time-resolved fluorescence anisotropy studies did not evidence any internal flexibility in the nanosecond domain.

The bleaching process, which is the fading of the fluorescence of dye molecules under continuous excitation, can be used to analyze the specificity of the antibody-antigen reaction in immunological tests. The automated immunofluorescence measurement system based on the analysis of bleaching characteristics, which can be described by the formfactor beside the maximum fluorescence intensity, has been developed. These two parameters can be used in fluorescence imaging to detect rare events among nonspecific staining.

The range of excitation wavelengths for a setup of time-resolved detection of fluorescence was extended with tunable blue by the synchronous pumping of a Stilbene 3 dye laser. A conversion efficiency of over 30% was found when pumping a coherent radiation (CR) model CR590 with extended cavity length using the third harmonic output of a cw mode-locked Nd:YLF laser (CR Antares). For decreasing the repetition rate of excitation pulses, an electro- optic modulator in a dual pass configuration was used. In this way, pulses with durations of less than 4 ps were generated with energies of up to 2.6 nJ and a repetition rate of 600 kHz, fitting very well with the requirements of the time-correlated photon counting detection technique. The tuning range of Stilbene 3 was from 415 to 465 nm. A further extension of the wavelength range is planned by pumping Coumarin 151 (465 -520 nm) and a mixture of Stilbene 3 and Coumarin 7 (495 - 570 nm) using the same pump laser source.

The UV absorption and fluorescence of 3-methylindole in water at 300 K has been simulated over periods of 4 to 30 ps in length using a combination of classical molecular dynamics simulations and a spectroscopically calibrated semiempirical molecular orbital method (INDO/S-SCI). The absorption redshift is predicted to be 545 +/- 150 cm^-1 for ¹L_a, and 85 +/- 15 cm^-1 for ¹L_b. The distribution of transition energies due to solvent fluctuations suggests inhomogeneous broadening at lower temperatures of about 2000 cm^-1 for ¹L_a, and 400 cm^-1 for ¹L_b, (fwhm). The large time-dependent shift of the ¹L_a fluorescence was determined both by direct nonequilibrium response to an instantaneous charge redistribution and from the autocorrelation function of the fluctuations of the ¹L_a transition energies from the equilibrated ground state. The two methods gave similar response times, roughly fitting a double exponential with time constants of 17 and 290 fs. The magnitude of the calculated shifts with partial and complete solute polarizability are 3000 and 5000 - 6000 cm^-1, bracketing an experimental estimate of 3800 cm^-1. A connection with perturbation-free energy of solvation ((Delta) A_s) results is made which identifies (Delta) A_s as the average of absorption and fluorescence maximum shifts upon solvation, ($DELTA(nu) _a + (Delta) (nu) _f)/2.