The goal of time resolved fluorescence experiments is to determine the validity of a model for population (kinetics) or motion (decay of anisotropy), and to extract a suitable parameter set which quantitatively describes the sample. Ideally, this analysis requires no adjustments of an ad-hoc nature in order to obtain a good fit, and the time-resolved residuals will be uniformly random. Any problems at this point lead to questions about the model being used, either with respect to its correctness in functional form, or in the exact values recovered, and it may become difficult to extract useful information about the sample. Reaching the ideal situation, and confirming that there are no experimental problems, requires considerable care. All aspects of the hardware are examined, starting with the laser light source and sample illumination optics. Collection optics for the fluorescence are analyzed, along with polarization components. Electronics checks and optimizations are described, showing the effects of characteristic problems. Finally, apparent artifacts due to numerical analysis are shown. While the examples are given for time-correlated single-photon counting, many of the optics related problems have similar consequences in the frequency domain. Some of the experimental problems have implications for the design of multiple channel detection schemes.

We describe the design and performance of a 10 GHz harmonic-content frequency-domain fluorometer. The modulated excitation is provided by the harmonic content of a train of ps pulses. High speed and/or high frequency detection was attained with a triode-type MCP PMT from Hamamatsu, R-2566-6, with 6 micron channels. To minimize the cost of the electronic components, and to minimize the noise due to these components, the detection circuits consist of two ranges, 10 MHz to 2 CHz and 2-10 GHz. The upper frequency limit of 10 GHz is determined by the current MCP PMT, so the usual configuration includes a low noise 2-10 GHz amplifier. This amplifier is easily replaced with a 2-18 GHz amplifier which allows operation to 18 GHz and the use of faster PMTs, should they become available in the future. Measurement of known optical delays demonstrates the accuracy of the instrument. For instance, a 1.69 ps optical delay was measured as 1.7 0.4 ps from 0.5 - 10 GHz, and 1.7 0.2 ps from 2- 8 GHz, where the uncertainty indicates the maximum deviation from the expected value. The data were shown to be free of systematic errors by measurements on fluorophores with single exponential decays.

Frequency domain fluorometry provides an alternative method for the recording of fluorescence decay kinetics (1), The major advantages of this method are the time resolution and the fast measurement of lifetime values. There are several systems, where a change in lifetime occurs across the emission band. These systems include heterogeneous ground state systems, where species emit at different wavelengths; excited state reactions, with products emitting at different parts of the spectrum; and dipolar relaxation processes, where the emission spectra change with time. To characterize the decay of the systems, it is necessary to collect the emission decay at several wavelengths. Generally, this process is obtained by successive measurements using a monochromator or a series of bandpass filters. For steady-state spectra, optical multichannel analyzers (OMA) are available, which can collect the entire emission spectra at once. By gating the image intensifier, used with some of these analyzers, it is possible to obtain time windows of the order of a few nanoseconds (2), This time resolution is inadequate for most of the fluorescence substances. We have developed a new method based of frequency domain fluorometry which extends the time resolution of an OMA to the picosecond region. The entire emission decay is collected in a few seconds and the lifetime information across the entire emission band is analyzed. The same method can be applied to other array detectors such as TV and CCD cameras used in fluorescence microscopy (3).

The theory of the statistical multiplexing of single photon timing fluorescence decay data is derived for the first time and shown to be consistent with measured array data. The present theory enables single photon timing arrays to be optimised for maximum data collection rate with regard to the number of channels in an array, channel count rates and cross-channel pile-up. Both the experimental and theoretical results prove that overall data collection rates up to 35% of the source repetition rate can be achieved by means of multiplexing with a single time-to-amplitude converter.

An important class of fiber optic sensors employs evanescent wave-excited fluorescence of labeled molecules bound to the waveguide surface to transduce a molecular recognition event, typically by an antibody. Unfortunately, the monolayer character of the sample, the small surface area, and the relatively low intensity of the evanescent wave all combine to limit sensitivity. In attempting to find a normalized frequency parameter (V number) of the optical fiber that optimized sensitivity, we found that our design (similar to those in the literature) had a loss mechanism which decreases its sensitivity and utility for remote sensing. We introduce here a design feature which addresses the loss mechanism.

A microscope based time-correlated single photon counting instrument has been used to measure nanosecond fluorescence decays from single cells. The excitation source for the instrument is a frequency doubled train of picosecond pulses from the cavity dumped output of a synchronously pumped dye laser. The dye laser is pumped by a mode-locked argon ion laser. In the microscope, the sample is excited and the emission collected using epi-illumination optics before being transmitted through an adjustable diaphragm, which can be closed to 10 μm in diameter. A Hamamatsu R928 photomultiplier is used to collect the fluorescence which is then analyzed using a non-linear least squares procedure. The microscope has been used to measure the intensity decays of model probes to determine the instrument performance and sensitivity. In addition, intensity and anisotropy decays collected from fura-2 loaded into single adherent rat basophilic leukemia cells were measured to demonstrate that the nanosecond fluorescence microscope can be used to obtain information about the environment and mobility of fluorescent probes in single cells.

We describe a novel fluorescence microscope system which allows us to capture rapid cellular phenomena as sequential images. This system is based on the "pulsedlaser fluorescence microscope" which we reported two years ago in the first symposium on the same topic. In the previous microscope, only one image could be taken at a certain instance in a fast event. In order to obtain sequential images, the event had to be repeated. The new version allows us to capture several images as a time series in a single measurement without repetition; thus, irreversible fast phenomena can be time resolved. The heart of this system is a highly sensitive framing camera with with a gated-microchannel plate serving as an ultra-fast shutter. This camera is basically an image converter in which several images can be projected at different positions on the output phosphor screen. The exposure time of each image can be as short as 100 nanoseconds, whereas the interval between sequential images can be variable between 300 nanoseconds to a few seconds. The camera enables rapid multishot imaging under highly intense continuous light. Faster events can be analyzed by combining the camera with a Q-switched Nd:YAG laser producing a burst of quadruple pulses. Four 10-nanosecond images can be captured at intervals of the order of 10 microseconds. With this system, we have been able to examine the behavior of giant liposomes under an intense electric field. Several deformation patterns and the formation of a large hole(s) in the lipid membrane have been visualized as microsecond sequential images.

The folding of proteins into a (single?) compact globular structure is an important mystery in contemporary biochemistry. It has even been referred to as the "second half" of the genetic code (1), since folding often provides the ultimate limit to genetic engineering. A variety of spectroscopies have been applied over the years, and stopped-flow steady-state fluorescence has contributed its share of information. Time-resolved fluorescence, however, is a newcomer to this endeavor. Realizing that time-correlated single photon counting (TCSPC) can provide lifetime resolution with only a few thousand photons, Brand & coworkers began to examine HLADH lifetimes during denaturation nearly a decade ago (2). The arc-lamp based technology used then was adequate to resolve changes requiring many minutes, and the early inclusion of global analysis in the process then called "KINDK" (3) extended the accuracy available. It was clear then that the advent of high repetition rate sources such as the modelocked, cavity-dumped dye laser would lead to improved accuracy and faster collection times. The HLADH system was revisited as soon as such technology was inhand (4).

An experimental technique for measuring time dependent circular dichroism signals with picosecond resolution is described. The details of the experimental apparatus are presented. Theoretical modeling of the detected signal demonstrates that the circular dichroism signal can be isolated from contributions due to pump-induced linear dichroism and linear birefringence effects. The experimental apparatus is used to examine the comformation relaxation in myoglobin following the photoelimination of CO from carbonmonoxymyoglobin. Probing the circular dichroism of the N-band of the herne at 355 nm reveals a relaxation of several hundreds of picosecond, over two orders of magnitude slower than the photo-induced bond cleavage. These results are discussed in terms of the restructuring of the protein following photodissociation.

Bacteriophage T4 lysozyme in its wild type form contains three tryptophan residues (at sequence postions 126, 138 and 158). These three residues are in rather different environments in the protein: 126 and 158 are near the protein surface while residue 138 is more buried. T4 lysozyme has been genetically engineered to prepare all possible variants in which one or more of the tryptophan residues have been replaced by tyrosine. The available data supports the hypothesis that this substitution has, at most, a very minor effect on the structure of the protein. The three species with single tryptophan residues have been investigated in detail. The surface location of residue 126 compared to the buried location of residue 138 is reflected in the difference in collisional quenching observed with added potassium iodide. It is found that the spectral and radiative properties of the three proteins are very similar but that their radiationless decay properties are quite distinct. This is apparently due to short-range collisional quenching by neighboring side chains. Comparison with solution quenching measurements permits the identification of the specific quenching groups involved for each tryptophan residue and provides a semi-quantitative rationale for the radiationless decay rate. This collisional quenching interpretation is supported by mutational effects on fluorescence quantum yield. This simple picture of the behavior of these single-tryptophan proteins is clearly revealed in this particular case because of the unambiguous choice of collisional quenching groups. The time dependence of the fluorescence decay of each of these single-tryptophan proteins is quite complex. Several methods of analysis are presented and discussed in terms of their underlying physical basis. Internal collisional quenching, as suggested from the comparative studies, is expected to lead to non-exponential behavior. This is consistent with the observed time dependence. Analysis of the temporal nature of the fluorescence as a function of emission wavelength is also revealing. Such data can be used to test discrete component, distribution and relaxation models of the time decay. It is found, in agreement with previous studies for other proteins, that the average lifetime for the emission increases with increasing emission wavelength. Analysis of the overall emission wavelength dependence of the time dependent data in a global sense based on a discrete population model shows acceptable agreement with the data in only one of the three cases. Application of several continuous distribution models to this data at each emission wavelength reveals that as the emission is moved to the red, a negative component appears in the distribution of decay components. This is a characteristic feature of relaxation behavior resulting in emission from kinetic species that are not present at the time of excitation. This negative preexponential character is not revealed by discrete component analyses since these do not have sufficient flexibility to describe the underlying complexity of the relaxing distribution. Finally, examination of the three proteins containing two tryptophan residues indicates that there is energy transfer between these residues in these cases and in the wild type protein. The order of energy transfer is in accord with the variation of the magnitude of the ratio k²/R⁶ controlling the efficiency of Forster energy transfer.

The fluorescence intensity decay of variant-3 scorpion neurotoxin has an average lifetime of 510 picoseconds. The consensus of least squares, maximum likelihood, Pade-Laplace, and distributional analyses of this decay is that more than 90% of the initial intensity comes from short lifetime components with lifetimes in the range 0.05 to 0.5 nanoseconds and the remaining small fraction of initial intensity is essentially one component decaying with an average lifetime of about 2 nanoseconds. Both adiabatic mapping and combined thermodynamic perturbation and umbrella sampling simulations reveal the presence of two tryptophan-47 rotational isomers. One of the isomers corresponds closely to the crystallographic structure and the second is a new rotational isomer that is separated from the first by a rotation angle of about 220 degrees and an activation barrier of about 10 kcal/mole. We propose that the short lifetime components correspond to conformers with tryptophan-47 approximately in the crystallographic orientation and the long lifetime component is the new rotational isomer identified in the simulations.

The finely-tuned regulation of biological activity in processes such as development, differentiation, and growth is most often mediated by not one macromolecule, but by relatively large complexes involving several interacting species. Such macro-molecular assemblies include multisubunit proteins and enzymes, complexes of proteins with nucleic acids and protein membrane assemblies. The activity (enzymatic, transport, or binding) of these complexes is often modulated by small effectors such as ions, protons, sugars, etc. A detailed understanding of the underlying mechanisms of such regulation necessitates the determination of the free energy linkages between the various binding equilibria. While the ligand binding equilibria in such systems have been extensively studied, the interactions between the macromolecules themselves have not in general been well characterized. Since the subunit affinities involved in such systems range from 10^-11 to 10^-5 M, a combination of fluorescence spectroscopy and high hydrostatic pressure is necessary to observe these subunit dissociation reactions at equilibrium. Results of such studies on a number of oligomeric protein systems are reviewed and a new approach to the analysis of such data is proposed.

The aromatic amino acid tryptophan is widely used as an intrinsic fluorescent probe of the solution conformation and dynamics of peptides and proteins. However, its complex photophysics makes it difficult to interpret the fluorescence results. The biexponential fluorescence decay of the tryptophan zwitterion is presumed to be due to ground-state rotamers. Intramolecular proton and electron transfer reactions involving the excited indole ring and amino acid functional groups have been proposed to account for the lifetime differences among rotamers. Excited-state H-D exchange occurs at the C-4 position of indole. In the proposed mechanism for the photosubstitution reaction, the ammonium group loops back over the aromatic ring and assists the proton exchange.

Pulsed-laser photoacoustic methods may be used to determine the fluorescent quantum yields of fluorophores in solution. Of interest to biological spectroscopists are the fluorescent quantum yields of probes bound covalently or noncovalently to proteins. Previous studies (J.R. Small et al., Fluorescence Detection III, E.R. Menzel, ed., SPIE Proceedings 1054, pp. 26-35, 1989) have been extended to examine some common protein probes and their fluorescent quantum yields. Examples include the probes Prodan [6-propionyl-2-(dimethylamino)naphthalene] and Acrylodan [6- acryloyl-2-(dimethylamino)naphthalene] in a variety of protein and solvent environments. We have found that, at the simplest level, the pulsed-laser photoacoustic technique gives us excellent results for the fluorescent quantum yields of fluorophores free in solution, but interestingly anomalous results for the fluorophores bound to proteins. The source of the anomalous protein results has not yet been determined, but several possibilities are discussed.

We report fluorescence studies with the single trp protein, S. nucelase A, and several of its site-directed mutants. One of these mutants, PA56, which has an alanine at position 56 in place of proline, has a much lower structural stability than the wild type. This is demonstrated by the much lower Tm (30 degrees C) for PA56 than for the wild type (52 degrees C) and by a much lower (urea)_1/2 for denaturation of the mutant. Also we show that PA56 can be unfolded by relatively low hydrostatic pressure (~700 bar). The free energy for unfolding of PA56 is found to be only 1.3 kcal/mole (at 20 degrees C) by thermal, urea, quanidine and pressure unfolding. Fluorescence lifetime measurements with wild type nuclease and several of its mutants show non-exponential decay kinetics. The fluorescence decay profiles are similar for the native state of each protein and the decay data at various temperatures generally reveal differences in the Tm for the various mutants. Anisotropy decay data are analyzed in terms of two rotational correlation times, a longer one for overall rotation of the protein and a shorter one for rapid, segemental motion of the trp residue. The mutant PA56 can be easily denatured by temperature, pressure or urea, and anisotropy decay data for these various denatured forms are reported.

Polarization and lifetime studies have shown that the fluorescence from nucleic acid species is complex, both at the individual chromophore level and because of the effect of stacking interactions on the electronic states. Recent work aimed at elucidating some aspects of this behavior by decay analysis and time-resolved spectroscopy is surveyed. Experimental work has been carried out using the ACO synchrotron at LURE (France) with time-correlated single photon counting, or a frequency-doubled N2-pumped dye laser, pulse width 700 ps, with fast-gated (100 ps width) analog detection and signal averaging. Decay curves are treated by global analysis using the Marquardt non-linear least-squares algorithm (synchrotron data) or the SPLMOD program (EMBO), which carries out a non-linear leastsquares minimization using cubic splines, for the laser data. Resolution of the decay data gives a model-based estimate of the number of components and their lifetimes. This information is then used to deconvolute timewindowed spectra (time-delayed spectra) into the time-resolved spectra. It is a particular feature of the combination of delayed photon counting with the continuous wavelength distribution of pulsed synchrotron radiation that excitation spectra correlating with emissions of different lifetimes can be obtained by uninterrupted repetitive scanning over a wide range of exciting wavelengths, in the present work from 230 nm to 354 urn. Such time-delayed excitation spectra can also be deconvoluted into components corresponding to the various time-resolved emission spectra. Examples of these three types of information viz resolved lifetimes, time-resolved emission spectra and their excitation spectra are presented and discussed for the following systems. I. adenosine; 6N, 6N-dimethyladenosine; protonated adenosine; this work shows the role of rotamers in the excited state behavior of this chromophore and demonstrates the forbidden nature of the lowest excited state. II. d(AT); d(TA); we observe the sequence dependence of emission from the stacked state which has been observed previously in polarization studies and an unusual excitation spectrum. III. d(CG); poly d(CG), "B"-DNA structure; single crystal duplex d(CG)3, "Z"-DNA structure. Distinctive differences are observed between the stacked emissions from the "B" and "Z" structures which we attribute to different overlapping of the stacked bases.

We have made room-temperature measurements of transfer of electronic energy in calf thymus DNA in which 75% of the guanine residues are methylated at position N-7. The methylated residues absorb light at much longer wavelengths than the nonmethylated residues and, as a result, constitute an irreversible energy trap. For excitation at 270 nm the efficiency of intrastrand transfer is found to be 0.31. We note that this is the first measurement of the efficiency of energy transfer along the double helix of a nucleic acid. In addition, the data have been simulated by using a stochastic model in terms of Förster's theory for dipole-dipole transfer. The adequacy of the latter is discussed. Preliminary measurements of the dependence of the fluorescence spectrum and of the fluorescence anisotropy of the alternating polynucleotide poly(dA-dT).poly(dA-dT) on the excitation wavelength and temperature are interpreted in terms of emission from molecular complexes formed between A and T residues.

Modern ab initio quantum chemical methods have been employed to compute excitation energies and transition moments of the nucleic acid base monomers cytosine, uradil, thymine, adenine, and guanine for the lower ir electron excited states. Full configuration interaction is allowed among the (pi)-electrons and the orbitals are optimized individually for each excited state. The multi-configurational self-consistent field method used allows full relaxation of both the (sigma)- and the (pi)-orbitals and includes correlation between the it-electrons. Basis sets of double-zeta quality have been used in these preliminary investigations. The calculated values for oscillator strengths and transition moment directions are in good agreement with experiment, where such comparisons can be made. The calculated excitation energies are, however, often too large due to the lack of (sigma)-(pi) correlation terms in the wave function.

Polarized two-photon fluorescence excitation spectra were obtained for the nucleotides TMP, CMP, GMP, and AMP in neutral aqueous room temperature solution (0.05M). The wavelength range was 400- 600 nm, corresponding to states reached with 200-300 nm single UV photons. INDO/S-CI predictions were carried out for the corresponding methyl bases, and were somewhat useful for the 260 nm bands with certain parameterizations. Theory failed to find the strong two-photon absorption seen below 450 nm, which may be of Rydberg origin.

The dynamics of self-complementary DNA decamers containing chemically-modified recognition sequences of the Eco RI endonuclease have been investigated by temperature-dependent picosecond time-resolved fluorescence spectroscopy. The unmodified decamer, d[CTGAATTCAG], as well as the decamer with 2-aminopurine (2AP) replacing adenine in position 5, have been shown to be B-type duplexes by 2D NMR and molecular dynamics (MD) simulations. Fluorescence anisotropy decay and MD of the 2AP-containing decamer show rapid motion of the base on timescales of 10-11 to 10-10 s. The multi-exponential fluorescence decay of the fluorescence and its temperature dependence, together with the 10-ns singleexponential decay of the isolated 2AP base, suggest that the 2AP base exists in 4 or more conformational states. All of these states appear to interconvert, but only two on the timescale of the fluorescence decay. The decamer with 1-((beta)-D-2'-deoxyribosyl)-2-pyrimidinone (dK) replacing dC in position 8 is a duplex which melts at about 21° C and shows multi-exponential fluorescence decay. Fluorescence is dominated, however, by a temperature-dependent 150-200 ps decay component accounting for >90% of the decay process. The unnormalized amplitude of this component decreases with decreasing temperature, reflecting hypochromism of dK as it stacks with its neighbors. In contrast with the other two modified bases, the isolated dK base has an extremely short fluorescence lifetime in aqueous solution, about 250 ps. The decamer with the 5-methyl derivative of dK (dS) placed in position 7 appears to be single-stranded above 10° C. Fluorescence from this decamer is thermodynamically and kinetically simpler than that from the duplexes. The decay time of the isolated d5 base is about 4.0 ns, while the decamer shows a temperature-independent 4.0 +/- 0.1 ns and shorter-lived, temperature-dependent decay components. Analysis of the data shows that completely unstacked (solvent-exposed) and partially stacked states exist. The completely unstacked state is a small component at all observed temperatures.

We provide a brief summary of current problems in DNA-protein interactions and argue that measurement of the elastic properties of DNA is critical to understanding these problems We discuss the ing1et depletion technique and various improvements we have been able to Achieve. Previous results using this technique are discussed, and we will present our plans for future experiments.

Complexes of linear and supercoiled DNAs with different intercalating dyes are studied by time-resolved fluorescence polarization anisotropy using intercalated ethidium as the probe. Existing theory is generalized to take account of excitation transfer between intercalated ethidiums, and Forster theory is shown to be valid in this context. The effects of intercalated ethidium, 9-aminoacridine, and proflavine on the torsional rigidity of linear and supercoiled DNAs are studied up to rather high binding ratios. Evidence is presented that metastable secondary structure persists in dye-relaxed supercoiled DNAs, which contradicts the standard model of supercoiled DNAs.

This paper begins with a review of our measurements of core particle transitions using the steady state intrinsic tyrosine fluorescence. The transitions described here include the opening of the particle at very low ionic strength, a pH-induced transition at physiological ionic strengths near neutral pH, and the high salt dissociation. The second section of the paper describes measurements of the rotational diffusion of the core particle using fluorescence anisotropy decays of ethidium intercalated into the DNA. We find that the low salt transition consists of two main stages as the ionic strength is lowered. Just below 1 mM, rearrangements in the protein core give rise to a drop in the steady state anisotropy. This first change is not accompanied by a measurable change in the rotational correlation time. Below ~0.2 mM an irreversible change sets in, and the average correlation time jumps to a much higher value. We believe that this structure is highly elongated. The pH-induced change does not give rise to a change in the rotational correlation time. At very high ionic strengths above 0.6 M, where the protein begins to dissociate from the DNA, the average rotational correlation time also increases to much higher values. There is no indication of disruption of the core particle below this ionic strength.

The interaction of a fluorescent DNA primer:template with the Klenow fragment of DNA polymerase I has been studied in solution using time-resolved fluorescence spectroscopy. The excited-state decay behavior and internal reorientation dynamics of a dansyl sulfonamide probe connected by a propyl chain to a modified uridine base in the primer strand were very sensitive to the local probe environment and exhibited characteristic changes upon binding of Kienow fragment to the DNA and elongation of the primer strand. Between 5 and 7 bases of duplex DNA upstream of the 3' primer terminus were protected from the solvent by the Kienow fragment and the strength of DNA:protein contacts varied within this region, being strongest at the 3' primer terminus. About 5% of the substrates were bound in a second spatially distinct site on the enzyme. Site-directed mutagenesis of the Kienow fragment was consistent with this being the active site for 3'->5' exonuclease activity.

Fluorescence quenching reactions, using such quenchers as oxygen and acrylamide, have been used to obtain information about the kinetic exposure of tryptophan residues in proteins. Often, both dynamic and static quenching components can be observed. The transient term of the Smoluchowski equation has been shown by Lakowicz et al (J. Biol. Chem. 10907-10910 (1987)) to be needed to fit acrylamide and oxygen quenching lifetime data with proteins. Here we show that this transient term can also explain the apparent static quenching that can be observed in some cases. By numerical integration of an impulse-response function, which includes a transient solute quenching term, we have simultaneously fitted intensity and phase lifetime Stern-Volmer plots for the oxygen and acrylamide quenching of selected single tryptophan proteins. These fits require a single fitting parameter, the diffusion of the quencher. Using this fitting procedure we have reinvestigated the effect of bulk viscosity on the acrylamide quenching of the fluorescence of the buried tryptophan in ribonuclease T1. We show that for this protein the intensity and lifetime data sets can be better fitted by a two-step diffusion model.

In time resolved photophysics, it is possible to access to transient effects of diffusionnal origin. We present here some classical kinetic models and discuss of their limitations. Particularly, we develop the important case of excimer formation when limited by diffusion.

The optimized Rouse-Zimm model is modified leading to a theory for long time random coil polypeptide dynamics [Perico etal., J. Chem. Phys., (1987), 87, 3677, Perico, J. Chem. Phys., (1988), 88, 3996, and Biopolymers, (1989), 28, 1527]. The description necessitates the knowledge of the rotational potential energy for specific amino acid residues and their friction coefficients. Static information, such as the amino acid sequence, the length of the polypeptide chain, and the location of the probe are found to affect the rotational correlation function P₂(t) and the local persistence length markedly. Compared with the fluorescence anisotropy measurements of tryptophan containing polypeptides (of the order of nanoseconds), the theory gives a reasonable prediction for the fluorescence depolarization correlation times of random coil polypeptides, but the calculated rotational correlation function predicts a much faster initial decay and a slower final decay than is observed. Possible theoretical improvements are discussed. Molecular dynamics simulations of tryptophan reorientation are also briefly described.

The fluorescence decay of membrane lipid bilayer probes are influenced by the environment of the probe and therefore by the properties of the bilayer in which it resides. Traditionally the fluorescence decay has been analyzed in the form of a multi-exponential. Analysis in the form of continuous distributions can provide a useful alternative to this approach since it introduces a new parameter, the distributional width. The physical basis underlying the distributional width, the fluorophore "environmental heterogeneity", arises from organizational aspects, compositional diversity and solvent effects. The "sampling" of environmental heterogeneity will depend on the intrinsic fluorescence lifetime, the lipid rate of motion and fluorophore charge and shape factors. We have found a good correlation of the distributional width with a number of bilayer properties. Thus as we increase the complexity of a bilayer the distributional width becomes broader, in order, from a single phospholipid molecular species, mixture of species, mixture of classes (i.e. PC+ PE etc) and species to intact natural membranes (i.e. including proteins and bilayer organizational factors). More recently we have been able to show that the protein-lipid interface itself can act as a source of fluorophore heterogeneity. Suggesting the possibility of spectroscopically isolating a region of the membrane of particular interest in the modulation of membrane processes. There are still a number of uncertainties in assigning a distributional width to a particular bilayer property. For example in comparing bilayers they could be compared at the same temperature or at the same temperature above the respective gel-liquid phase transition temperatures. In fact with regard to the relative degree of environmental heterogeneity sampled by the fluorophore it may be better to make such types of comparisons at the same respective rates of lipid motion since the rate of lipid motion is a key factor governing the sampling of environmental heterogeneity. Using this approach in the example of increasing levels of unsaturation our results suggest that unsaturation introduces structural diversity into lipid bilayers.

The performance of global (simultaneous) analysis of multiexponential fluorescence decay surfaces using reference convolution is investigated in a systematic way using simulated and experimental data sets. It is shown that the increased model discrimination ability and the more accurate parameter recovery by global analysis as compared to single curve analysis originate from combining decay traces with differing contributions of the decay components. Simultaneous analysis of decay traces in which the contributions of the components are changed as much as possible is the most beneficial. Consequently, including more decay traces in the global decay surface does not necessarily lead to a better model distinction capability. For decay surfaces collected as a function of the emission (excitation) wavelength, the decays with minimal overlap between the emission (absorption) spectra associated with the decay components will contribute the most to the improved model discrimination and parameter recovery.

We used fiber-optic phase-modulation methods as well as time-correlated single-photon counting to characterize time-dependent migration of 580-740 nm photons in living tissue, in particular in human fingers, forearms, calfs and foreheads. The phase-modulation measurements were extended to 4.2 GHz using an internally cross-correlated gatable MCP photomultiplier. Owing to the very high modulation frequencies, accurate measurements could be accomplished even at short distances (6 mm) between the source fiber and detector fibers. In order to estimate the influence of the optical tissue properties, and of geometrical factors, we performed model experiments in scattering media. The time-dependence of the reemerged light appears to be characterized by three parameters: 1) a time delay, 2) a sample transit-time spread, and 3) a (complex) exponential decay. All three parameters depend on fiber distance, absorption coefficient, scattering coefficient, detector geometry, and photon wavelength. The phase-modulation data are unique in that the time delay of the reemerged photons results in phase angles as high as 300 degrees, and the transit-time spread results in a dependence of the modulation on frequency which is a faster-than-exponentially decreasing.

Time resolved spectroscopy of tissue makes it possible to quantify tissue hemoglobin concentrations because of the direct measurement of the optical path length for photon migration. However, the laser system is bulky and unwieldy and impractical for clinical studies. Thus, the application of the more compact and efficient phase modulation technology well known for fluorescence lifetime studies to time resolved spectroscopy of tissue offers opportunities to simplify the methodology and in addition to afford continuous readout of tissue photon propagation. This paper describes single and dual wavelength systems operating at two wavelengths in the deep red region based upon a time-sharing system. These devices have noise levels in a 2 Hz bandwidth of less than 2 ps and drifts of < 1ps/min. Applications of the noninvasive devices include measurement of hemoglobin deoxygenation in brain and hemoglobin and myoglobin deoxygenation in human skeletal muscle and animal models. Numerous applications to medical and biological problems now become available.

The transport. of photons in random medium were directly studied by measuring the angular and temporal distribution of photons after traversing through a slab of a random medium. The experiments were performed using an ultrafast laser pulse of 100 fs and streak camera detection with a temporal response of 8 ps. When the ultrafast laser pulse traverse through a slab of random medium of less than 10 scattering mean free path thickness, a coherent (ballistic) pulse was found to coexists with a incoherent (diffusive) pulse. The speed of the coherent component of the 100 fS laser pulse reduces as the concentration of the scatterer increases. This reduction in speed may arise from the coherent interference of the multiply scattered waves. The transport of photons was found to deviate from the diffusion approximation when the thickness of the slab is less than 10 transport mean free paths; photons were found to arrive significantly earlier than predicted by the diffusion theory. The theoretical results from the two-frequency coherence function in the Rytov approximation were in qualitative hut not in quantitative agreement with the experimental results. Finally, we show that the ratio of signal (coherent component) to noise (incoherent component) of the light propagating through a random medium can be significantly increased by introducing an absorbing dye. This absorption technique of increasing the signal to noise ratio can improve our capability to image through a random medium.

Fluorescence decay of aorta and its fluorophores has been studied at 380, 440, and 480 nm using 320 nm picosecond dye laser excitation and time-correlated single photon counting technique. Fluorescence decay was found to be nonexponential at all emission frequencies. Fluorescence decay at 380 nm shows biggest differences between normal and plaque tissues.

Time domain measurements of radiationless energy transfer have been employed to determine the average separation and distribution of separations for several donor-acceptor pairs of biological interest. The method can allow for the effects of heterogeneity of donor lifetime and of orientation. The systems examined include troponin C and melittin-calmodulin.

Utilizing time-resolved fluorescence techniques for the examination of the protein folding problem requires the development of some new methods of data acquisition, analysis, and experimental design. Structural information (rotational relaxation times, intramolecular distance distributions, conformational dynamics) need to be obtained on transient non-equilibrium species which may exist for as little as tens of milliseconds. In this report, the development of a stopped-flow timeresolved fluorometer is described. In addition, a unique molecular genetic system developed for the protein Stapholococcal Nuclease is utilized to assist in the placement of donor/acceptor pairs in strategic locations throughout the protein. Combining these two projects together, provides the potential for determining the structural details involved in the folding (unfolding) behavior in this protein.

Cardiac troponin I (CTnI) can be phosphorylated by a c-AMP dependent protein kinase. We have investigated the effect of the phosphorylation on the emission decay properties of its single tryptophan by using a cavity dumped and synchronously pumped dye laser system. At 20°C, τ₁~0.60 ns, τ₂~2.22 ns, and τ₃~4.72 ns. The corresponding fluorescence contributions were 7%, 47%, and 46%, respectively. Upon phosphorylation four lifetimes were observed: τ₁~0.11 ns, τ₂~0.81 ns, τ₃~1.95 ns, and τ₄~6.63 ns, and fractional contributions of the four components were 2%, 16%, 52%, and 30%, respectively. This finding indicates that the environment of the tryptophan is modified by phosphorylation. In the absence of divalent metal ions, the observed three decay times of the CTnI complexed with cardiac troponin C (CTnC) remained unchanged, and addition of Ca²⁺ or Mg²⁺ resulted in only small changes in the lifetimes. When phosphorylated CTnI was complexed with CTnC, a large increase of the longest-lived component was observed: τ₄ > 11 ns with its contribution shifted to 47%. Two rotational correlation times were observed for CTnI: φ₁~0.9 ns and φ₂~ 23.5 ns. These valves increased to t₀~l.2 ns and ~30.1 ns, respectively, for the complex CtnI CTnC. Upon phosphorylation the two correlation times were significantly reduced regardless of whether CTnI was uncomplexed or complexed with CTnC. These results suggest that phosphorylation of CTnI resulted in a significantly more compact structure and enhanced motion of the tryptophan side chain. These structural changes may play a role in the transmission of Ca²⁺ signal in cardiac muscle. Thus, the effect of phosphorylation of CTnI becomes more pronounced when the protein is complexed with CTnC. These results suggest that there was likely a fluorophore heterogeneity which may arise from differences in conformation, environment, and/or different deactivation pathways for the excited state of the fluorophore.

The enzyme bacterial luciferase is able to transform the chemical exergonicity from the oxidation of FMNH₂ and tetradecanal by oxygen, into electronic excitation of an associated fluorophore. The naturally occurring fluorophore in Photobacterium is 6,7-dimethyl-8-ribityllumazine bound to a protein that interacts in some way with the luciferase. Lumazine protein is not only raised to its excited state during the reaction but its presence alters the bioluminescence kinetics, possibly by catalyzing the decomposition of a "chemically- charged" intermediate on the luciferase. The emitter of bioluminescence in the reaction of luciferase is a highly fluorescent reaction intermediate called the "Fluorescent Transient". Energy transfer between luciferase fluorophores and lumazine protein has been studied using time correlated fluorescence and anisotropy decay techniques. Results of these and previous studies show that the energy transformation process in bacterial bioluminescence does not involve these proteins in a resonance energy transfer process.

Global methods for time-resolved and steady-state experimentation and analysis are presented. Examples considered-most of which are nucleic acid systems-include methylated and unmethylated oligo- and polynucleotides formed from alternating C and G bases, for which there is energy transfer; and polynucleotides formed from A and T bases, for which there may also be exciplex formation and dissociation. We show how time-resolved and steady-state data can be used to determine the number and nature of fluorescent components in each system, resolve species-associated absorption and emission spectra (with either time-resolved or steady-state data) , and estimate rates of energy transfer (time-resolved data) or relative quantum yields associated with energy transfer (steady-state data). Where there is an Arrhenius dependence on temperature, energies of activation can also be determined from steady-state data.

The topographic analysis of fluorescence distribution has been carried out pixel-by-pixel by one dimensional, two-dimensional microspectrofluorometry and three-dimensional confocal fluorescence microscopy. Fluorescence emission spectra of NAD(P)H and benzo(a)pyrene (or metabolites) were recorded at different excitation wavelengths. Cell bioenergetics are monitored in normal and malignant cells as well as cells with genetic defects by coenzyme responses to microinjections of substrates and modifiers from key metabolic pathways in presence and absence of inhibitors and drugs active on mitochondrial structure and function. Cooperative interactions between organelles involved in detoxification mechanisms are observed in cells treated with fluorescent cytotoxic agents. Such interactions can be directly mapped by the fluorescence of cytotoxic agents, their reaction products or vital probes such as NBD ceramide for the Golgi apparatus. To identify the organelles involved parallel electron microscopic studies are carried out in cells first treated with the cytotoxic agent and then incubated with an electron opaque material. A recently developed combined X-ray laser microscope (COXRALM) holds the promise of carrying out combined phase-fluorescence-and X-ray microscopic observations of fluorescence and ultrastructural correlations in live cell probing. As further versatility is gained in such methods it may become possible to obtain a very detailed structure and function mapping of living cells within the context of cytomatrix analysis, metabolic compartmentation and organelle interactions.

For the purpose of gaining molecular information such as molecular distances and molecular motion under the fluorescence microscope, we have developed a time-resolved microfluorimetric method, a time-resolved fluorescence spectroscopic technique combined with the optical microscopy: the excitedstate lifetime of fluorescent molecules (about 1-20 nsec) in a small spot in a single living cell is measured under the microscope using a focussed pulsed laser beam as an excitation light source (0.5 micron spot diameter, (lambda)=365 nm, FWHH= 14 psec, 4 MHz repetition rate) and a synchroscan streak camera as a detector with time-resolution. A signal-to-noise ratio better than 100 was obtained for fluorescein labelled band 3 (1 label/band 3) in an area of 2.5 micron-diameter (containing 2x105 band 3) in a single erythrocyte ghost after signal accumulation for 50 sec. The time-resolved microfluorimetry has been applied to the studies of(l) the assembly mechanism ofcell adhesion protein (E-cadherin, a calcium-dependent cell adhesion receptor protein) at the site of cellcell contact in keratinocytes in culture and (2) endosome-endosome fusion in fibroblasts in culture. (1) Observation by video microscopy indicated that the calcium-induced assembly ofE-cadherin at the sites of cell contact takes place via lateral migration of cadherin in the plasma membrane of keratinocytes. Association of cadherin molecules at the molecular level has been assessed by observing the resonance energy transfer (RET) from fluorescein (donor) to reactive red 8 (acceptor), both attached to monoclonal IgG antibody specific to E-cadherin (Occurrence of RET would decrease the lifetime of the donor). The fluorescence lifetime decay data of fluorescein at various locations in cells indicated the occurrence ofRET only at the boundary region ofkeratinocytes in the high-calcium medium, suggesting that cadherin molecules form aggregates at the sites of cell contact, where they function as a cell adhesion molecule between two cells. (2) Quantitative observation of fusion of endosomes in single cultured cells (NIHI3T3) has been made. Our method is based on the observation of internal content mixing of endosomes by detecting RET between 7-nitrobenz-2-oxa- 1 ,3-diazol-4-yl (NBD)-labelled a2-macroglobulin (energy donor) and water-soluble sulforhodamine (energy acceptor), which are sequentially endocytosed by a cell. One of the most important characteristics of the fusion assay reported here is that it allows fast quantitative measurement (measuring time, 10-30 sec) of fusion of each endocytotic vesicle discernible under the fluorescence microscope. The excited-state lifetime of the donor in each endosome was measured (1.6 micron-diameter spot) in various parts of the cell. Acceleration of the fluorescence decay of the donor (occurence of RET) was detected, the extent of which increased with an increase of the incubation time with the acceptor. This result indicates the mixing of the energy donor with the acceptor in endosomes, showing that fusion between successively formed endosomes can take place in cells. Blocking of endosome acidification with NH4C1 failed to supress endosome fusion. The cells treated with cytochalasin D at concentrations sufficient to destroy the actin filaments did not show any indications of endosome fusion, suggesting that actin filaments are required for enconter and/or fusion processes of endosomes.

Room temperature phosphorescence emitted by tryptophan residues in deoxygenated aqueous solutions of proteins is extremely sensitive to the environment of these residues and can be utilized for the detailed study of protein structure and dynamics. The long decay time of their triplet state makes the phosphorescent tryptophans suitable donors for resonance energy transfer in the rapid diffusion limit. As shown by Stryer et al. (Ann. Rev. Biophys. Bioeng. 11, 203 (1982)) proper data analysis can then yield the distance of closest approach between the donor-acceptor pair. This method can thus allow one to map the distances of phosphorescent tryptophans from the surface of the protein. In the present study a laser-based photon counting system was used to follow the room-temperature phosphorescence decays of alkaline phosphatase and horse liver alcohol dehydrogenase and to study the quenching of their triplet states by several molecules whose absorption spectra overlap the long-lived emission of these proteins. The results demonstrate the potential applicability of these measurements for the mapping of phosphorescent tryptophan residues and confirm the phosphorescence of alkaline phosphatase to originate in Trp-109. Limitations to the applicability of the energy-transfer approach arise from quenching mechanisms which compete with resonance transfer. Two such processes - electron transfer and exchange interactions - are discussed.

We have implemented a new laser microscopic method, polarized fluorescence depletion (PFD), for measuring the rotational dynamics of functional membrane proteins on individual, microscopically selected cells under physiological conditions. This method combines the long lifetimes of triplet-state probes with the sensitivity of fluorescence detection to measure macromolecular rotational correlation times from 10 microsec to > 1 ms. As examples, the rotational correlation time of Fc receptors (FcR) on the surface of 2H3 rat basophilic leukemia cells is 79.9 4.4 microsec at 4°C when labeled with eosin conjugates of IgE. This value is consistent with the known 100 kDa receptor size. When labeled with intact F4 anti-FcR monoclonal antibody, the rotational correlation time for FcER is increased about 2-fold to 170.8 +/- 6.5 microsec, consistent with receptor dimer formation on the plasma membrane and with the ability of this antibody to form FcER dimers on 2H3 cell surfaces. We have also examined the rotational diffusion of the luteinizing hormone receptor on plasma membranes of small ovine luteal cells. Luteinizing hormone receptors (LHR), when occupied by ovine luteinizing hormone (oLH), have a rotational correlation time of 20.5 +/- 0.1 microsec at 4°C. When occupied by human chorionic gonadotropin (hCG), LHR have a rotational correlation time of 46.2 +/- 0.4 microsec suggesting that binding of hCG triggers additional LHR interactions with plasma membrane proteins. Together these studies suggest the utility of PFD measurements in assessing molecular size and molecular association of membrane proteins on individual cells. Relative advantages of time- and frequency-domain implementations of PFD are also discussed.

The basic theory of a frequency domain fluorometer has been known since the original work of Gaviola (1)• With the introduction of multifrequency capabilities and the use of ultra high speed detectors, the basic understanding of the details of the instrumentation has become essential. The ultimate time resolution and separation capabilities for complex decay species critically depends on the noise characteristics of the instrument's various components. In frequency domain fluorometry, it is customary to measure the standard deviation of the phase and modulation at each individual frequency point. In the correlated single photon counting method, it is assumed that the standard deviation is given by '[i, where N is the number of counts in a time channel. In single photon counting techniques, the fluctuation of the number of counts is assumed to be intrinsic to the statistical properties of light, therefore, the only way to decrease the standard deviation of an individual channel is to increase the light intensity at the detector. A reduction of the standard deviation results in lower integration time and better instrument resolvability. The major feature emerging from the analysis of the residues and the standard deviation in the frequency domain technique is that, although they appear randomly distributed and uncorrelated from one frequency point to the next, there is a correlation between different experiments. The pattern of the residues is often reproducible. This observation questions the basic assumption used for statistical data analysis, i.e., that there is a gaussian distribution of each data point around the mean value and that the standard deviation of each point is significant for the statistical analysis of the frequency response.

Fluorescence excitation and anisotropy spectra are presented for a set of methyl and methoxyindoles at -50 degreesC in propylene glycol glass. These spectra are interpreted in terms of two overlapping (pi)->(pi)^* electronic transitions (¹L_a and ¹L_b). Semiempirical molecular orbital calculations are presented to correlate with the observed spectral changes caused by methyl and methoxy substitution.

A number of fluorescence studies, both of trp residues and bound NADH, have been reported for porcine MDH. The large number of trp residues (6) complicates the interpretation of some studies. To circumvent this we have performed studies with a two tryptophan (per subunit) MDH from Rhizobium japonicum 311B-143 bacteroids. We have performed phase/modulation fluorescence lifetime measurements, as a function of temperature and added quencher KI, in order to resolved the 1.3 ns (blue) and 6.6 ns (red) contributions from the two classes of trp residues. Anisotropy decay studies have also been performed. The binding of NADH dynamically quenches the fluorescence of both tip residues, but, unlike mammalian cytoplasmic and mitochondrial MDH, there is not a large enhancement in fluorescence of bound NADH upon forming a ternary complex with either tartronic acid or D-malonate.

To obtain high quality decay data using time-correlated single photon counting, it is essential to eliminate multiplephoton events. Events should be counted only when one photon is detected for a given excitation flash. An easy way to eliminate multiple photon events is to gate data collection, permitting an event to be registered only if the energy of the pulse is within limits appropriate for the detection of a single photon. This method, called energy windowing, was first described by Schuyler and Isenberg [R. Schuyler and I. Isenberg, Rev. Sci. Instrum. 42, 813-817 (1971)]. In this paper we present a formalism for describing the statistics of multiple photon events. This formalism gives us a means of characterizing the pulse height distribution and estimating the fraction of multiple events which leak through our energy window. It also describes how multiple photon events distort a decay function and behave as contaminant decays. We then present a method for using energy windowing with Hamamatsu microchannel photomultipliers, and examine the ability two tubes of different design to differentiate between single and multiple events.

Exploration of the effect of ligand-protein interactions on conformational substates and internal dynamics in different regions of phospholipase A2 from porcine pancreas (PLA2), was performed by combining site-directed mutagenesis and time-resolved fluorescence measurements. The single tryptophan residue (Trp-3) in the wild type protein was replaced by a phenylalanine residue, whereafter Tip was substituted either for leucine-3 1 ,located in the calcium binding ioop, or for phenylalanine-94, located at the "back side" of the enzyme, in a-helix E (Dijkstra et al., J. Mol. Biol., 147, 97-123, 1981). Analyses by the Maximum Entropy Method (MEM) of the total fluorescence intensity decays, provide in each case a distribution of separate lifetime classes, which can be interpreted as reflecting the existence of discrete conformational substates in slow exchange with respect to the time-scale of the decay kinetics. The fluorescence decay of the W94 mutant is. dominated by an extremely short excited state lifetime of ~60 ps, probably arising from the presence of two proximate disulfide bridges. Time-resolved fluorescence anisotropy studies show that the Trp residue near the NH2 terminus (Trp-3) undergoes a more limited rotational motion than the Trp-3 1 located in the calcium binding loop. The widest angular rotation is observed at position 94, in a-helix E. Calcium binding displays the strongest influence on the lifetime distribution of Trp-3 1: a major local conformation corresponding to a lifetime class with a barycenter value of -5.5 ns and contributing to ~50% of the decay is selected. The conformations giving rise to the short lifetimes (τ₁ and τ₂ lifetime classes) become less important. The contribution of the third lifetime class (c3) stays at a constant value of 30%. In the presence of calcium, the amplitude of motion is wider than without the ion. There is virtually no effect of calcium binding on the lifetime distribution of the Trp residue at the 3 or the 94 position. Binding of the monomeric substrate analog n-dodecylphosphocholine (C12PN) in the presence of calcium hardly affects neither the Trp-3 excited state population distribution, nor its rotational dynamics. The binding of C12PN monomers to the W31 mutant further increases the contribution of the τ₄ lifetime class at the expense of c2. A more restricted rotation of the Trp-3 1 residue is also induced. The binding of the micellar substrate analog n-hexadecylphosphocholine (C16PN) in the presence of calcium is very efficient in modifying the lifetime distribution of Trp-3. Essentially, one major broad lifetime population (centered at ~2.6 ns) is revealed by MEM analysis of the total intensity decay. The internal motion is slowed down and the angle of rotation is much smaller in this conformation. Neither the excited state lifetime distribution of Trp-31 nor its dynamics are affected by micelle binding relative to monomer binding. In conclusion, by placing a single Tip-residue at strategic positions along the peptide chain of PLA₂, relevant to the binding of biological ligands, an excellent model system for the study of selective perturbations of conformational substates and internal dynamics is provided.

We studied energy transfer between tryptophan (donor, D) and dansyl (acceptor, A) separated by three different hexapeptides: hexaglycine, hexaalanine and hexaproline. In each case the donor tryptphanamide was located on the C-terminus, and the dansyl acceptor on the N-terminus amino group. In the absence of acceptor, in propylene glycol at 20°C, tryptophan donors show single exponential fluorescence intensity decays with mean lifetimes about 5.2 ns. In the presence of dansyl, the lifetimes become shorter and intensity decays heterogeneous. However, the extent of heterogeneity is different for each labeled oligopeptide. The strongest heterogeneity was observed for the hexaglycine donor-acceptor pair. Only small deviations from single exponential decay were found for the proline D-A pair. These results suggested a higher degree of conformational heterogeneity for the flexible glycine D-A pair, or composed to the more rigid proline D-A pair. We analyzed the data in terms of Gaussian donor-acceptor distance distributions. The recovered average distances Ray, are 10.7, 15.9 and 23.4 A and full widths at half maximum, hw, are 18.1, 11.0 and 3.9 A for hexaglycine, hexaalanine and hexaproline chain, respectively. The results indicate that flexibility of peptide chain has a strong influence on end-to-end distance distribution. The possible effects of the orientation factor κ² are discussed, along with the limits on κ² consistent with the time-dependent anisotropy decays. These distance distributions provide the basis for comparisions and/or refinement of the rotational potential functions of the amino acids.

A simple Fourier transform method for processing the time intervals recorded in a single photon decay spectroscopy (SPDS) experiment has been developed for fluorescence signals containing more than one decay constant (multiexponential signals). The main advantage of the proposed method consists in showing clearly the existence of each of the components which constitute the decaying fluorescence. This advantage increases when we are measuring relatively long decay constants (> 1 μs) and the mean fluorescence intensity is low, since in these cases the conventional technique of SPDS requires rather long data acquisition times. A comparison between the conventional technique and the sine transform method has been made for a great variety of experimental situations by resorting to a computer simulation for fluorescence signals containing two or three decaying constants.

The small-molecule (i.e. oxygen) penetrability into proteins carries important implications for both the structural character and dynamic properties of proteins. We investigated aspects of protein dynamics by studying triplet-state quenching of Zn and Pd derivatives of myoglobin by oxygen. These two derivatives make an interesting comparison because the Zn derivative of myoglobin is five coordinated and is out-of-plane (as is deoxy myoglobin), whereas Pd porphyrin is planar (as is oxy myoglobin). Therefore relaxation of the polypeptide chain in the excited state analogues of the two functional conformations of myoglobin can be studied. In case of Znmyoglobin, the spectra did not exhibit an isosbestic point and the decay kinetics to the ground state dependtd upon the wavelength of measurement. The decay profiles could be fit by a double exponential function with a good correlation. In the case of Pd-myoglobin an isosbestic point was observed in the emission and the decay profile could be fit with a single exponential function. The comparison of the quenching constants for the two metal substituted derivatives suggests that there is no significant difference between the oxygen penetrability of the Zn and the Pd substituted myoglobins.

This paper describes the use of luminescent coenzymes and substrates in the study of two proteins, horse liver alcohol dehydrogenase (ADH) and sex steroid-binding protein (SBP). We report the phosphorescence emission and lifetimes of NAD and e-NAD in pyrazole ternary complexes with ADH. Whereas the decay of NAD is adequately described as monoexponential with a lifetime of 2.4 s, the decay of e-NAD obeys a double exponential decay law with time constants of 0.4 and 0.15 sec. Ternary complexes with NAD have the same decay kinetics as ADH by itself. However, in ternary complexes with f-NAD, it is possible to selectively excite and detect the coenzyme phosphorescence. We show that f-NAD is a more useful probe than NAD for protein structure in ADH. We also measured the triplet emission and decay lifetimes of dihydroequilenin, an equine steroid bound by SBP. We find that the phosphorescence spectra and lifetimes are pH dependent, with the protonated species dominating emission below pH 10.0, and the deprotonated form dominating at pH 10.0 and above. The acidic and basic species can be selectively excited, and the emission at pH near the pKa is characteristic of the equilibrium ground state populations. Since hydrogen bonding is implicated in SBP-steroid complexes, dihydroequilenin is a potential phosphorescent probe for the binding interaction.

Ground and excited state characteristics of substituted indole derivatives reveal a sensitivity of indoles' electronic properties to the nature and location of substitutions on the indole ring. These substitutions affect both the nature of the excited electronic state and the susceptibility of this state to non-radiative decay processes. A number of mechanisms that deactivate the excited state have been identified including intersystem crossing, electron photoejection into polar solvents, and >N-H dissociation in polar solvents (see Glasser & Lami,1986) . While the >N-H group has been implicated in non-radiative decay processes in polar solvents, covalent substitutions elsewhere on the indole molecule may modulate the importance of this site in non-radiative decay mechanisms or alternatively these substitutions may introduce new deactivation mechanisms. Additionally, complexes formed between indole derivatives and β-cyclodextrin cavities show different sensitivity to excited state deactivation mechanisms dependent upon the location and nature of the covalent substitution. We have investigated the excited states of some indole derivatives substituted at position 5, para to the >N-H group on the benzyl ring, to determine the effect of such covalent substitutions on the fluorescence emission characteristics of the indole ring as well as on its susceptibility to alternate excited state decay mechanisms.

In many ways, the analysis of waveforms from pulsed-laser photoacoustics is analogous to the analysis of timeresolved fluorescence data obtained from single-photon counting techniques. We have shown previously that the Method of Moments can be used to analyze photoacoustic data (E.W. Small et al., Fluorescence Detection III, E.R. Menzel, ed., SPIE Proceedings 1054, pp. 36-53, 1989). Difficulties arose in attempts to analyze the magnitude of very fast decays, those with lifetimes less than the digitization channel width. While these decays may often be regarded as an error such as scatter in the case of fluorescence, they represent important fast relaxation processes in the case of photoacoustics, and their magnitude must be accurately determined. We present here three approaches to making the Method of Moments more applicable to photoacoustic waveform analysis. One approach involves using the method required for measuring a scatter component in a fluorescence decay. Other approaches include using a Cheng-Eisenfeld filter to fix a lifetime in order to recover its amplitude, and analyzing for MDO (moment index displacement 0), which is intrinsically sensitive to short lifetimes. The relative merits of the techniques are evaluated, and a successful approach to analyzing noisy synthetic data of up to three components is demonstrated. Preliminary studies on the performance of the Method of Moments in handling nonrandom errors in photoacoustic data are also shown.

The intrinsic fluorescence properties of molecules in condensed phases are complicated by intermolecular interactions. In crystalline solids the fluorescence depends on the photophysics of individual molecules as well as the interactions between molecules in the solid state Even a chemically pure crystal has some lattice disorder, which alters the local environment and changes the intermolecular interactions. Sample preparation and experimental technique may influence the fluorescence results.

We have studied tryptophan derivatives in which rotations about the C_α-C_β and C_β-C_γ bonds are restricted by incorporating the a-amino group into a six-membered ring. Samples were volatilized into He flows and cooled via supersonic gas expansions. Both bare molecules and molecules complexed with polar solvent adducts have been studied via laser induced fluorescence spectra and fluorescence lifetime measurements. We compare jet cooled conformer and lifetime results with some in solution. Results obtained for bare molecule and solvent are compared with analogous data for conformationally unconstrained tryptophan derivatives studied earlier. The jet-cooled lifetime data provide indications of the presence of the ¹L_a state.

Time-resolved phosphorescence anisotropy. has been used to assess the rotational dynamics of human serum lipoproteins labeled with phosphorescent probes of high triplet yield. Covalent labeling of apolipoprotein B with erythrosin revealed the existence of segmental motion of labeled domains within the particle as well as global rotation of the particles. The binding of the low density lipoprotein to cell surface receptors resulted in a freezing of the global motion but the maintenance of faster motion of domains within the apolipoprotein. Labeling of the lipid phase of the low density lipoproteins with an eosinyl fatty acid also revealed the existence of two motions. The shorter time constant was attributed to the motion of the chromophore within the lipoprotein particle, while the longer time constant represented the global tumbling of the particle in solution. To examine the the physical state of the lipid phase in lipoproteins, the steady-state fluorescence anisotropy of n-(9-anthroyloxy) fatty acids was examined in microemulsions and phospholipid bilayers. The phase transition in the surface monolayer of the microemulsion is significantly less cooperative than the same transition in a bilayer vesicle. Moreover, the rate at which cholesterol transfers from a donor structure to an acceptor is much faster for the microemulsion than for the bilayer vesicle. These results indicate differences in the packing of phospholipids in the monolayer of the microemulsion as compared to the external leaflet of the bilayer vesicle.

fluorescence decay surfaces of excited state reactions can be globally analyzed directly in terms of reaction rates and species associated spectra [Beechem et al., Chem. Phys. Letters 120 (1985) 466]. The identifiability of two-state excited state reactions has been investigated for properly normalized decay curves assuming that the ratio of the ground state absorbances is known [Ameloot et al, Chem. Phys. Letters 129 [1986] 21 1]. It is demonstrated in this paper that the condition of proper normalization is not always required. In addition, it is shown that the ratio of the absorbances of the species in the ground state can be obtained from fluorescence decay surfaces. The required experimental design is indicated.

We present data on lifetimes of 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 fluorophores like hemoglobin. Three lifetimes are detectable in the subnanosecond range. At high frequency it is possible to detect components below 10 ps. The average lifetime of hemoglobin is ligand dependent, opening a new chapter in the investigation of the allosteric behavior of hemoglobin.

Time-resolved fluorescence intensity decay data from anthracycline anticancer drugs present in model membranes were obtained using a gigahertz frequency-domain fluorometer [Lakowicz et al. (1986) Rev. Sci. Instrum. 57, 2499-2506]. Exciting light of 290 nm, modulated at multiple frequencies from 8 MHz to 400 MHz, was used to study the interactions of Adriamycin, daunomycin and related antibiotics with small unilamellar vesicles composed of dimyristoylphosphatidylcholine (DMPC) at 28°C. Fluorescence decay data for drug molecules free in solution as well as bound to membranes were best fit by exponentials requiring two terms rather than by single exponential decays. For example, one-component analysis of the decay data for Adriamycin free in phosphate buffered saline (PBS) solution resulted in a reduced x2 value of 140 ((tau) = 0.88 ns), while a two-component fit resulted in a substantially smaller reduced x2 value of 2.6 ((tau)1 = 1.13 ns, (alpha)1 = 0.60, (tau)2 = 0.30 ns). Upon association with membranes, each of the anthracyclines studied displayed a larger r1 value while the r2 value remained the same or increased (for example, DMPC-bound Adriamycin showed r1 = 1.68 ns , a1 = 0 . 64 , r2 = 0 . 33 ns) . Analyses of the fluorescence emission decays of anthracyclines were also made assuming each decay is composed of a single Lorentzian distribution of lifetimes. Data taken on Adriamycin in PBS, when fit using one continuous component, displayed (tau), (alpha), w, and reduced x2 values of 0.68 ns, 1, 0.60 ns, and 9.1, respectively. The distribution became quite broad upon drug association with membrane (DMPCbound Adriamycin: (tau) = 0.75 ns, (alpha) = 1, w = 2.24 ns, x2 = 13). For each anthracycline studied, continuous component fits showed significant broadening in the distributions upon drug association with membrane. Relatively large shifts in lifetime values were observed for the carminomycin and 4-demethoxydaunomycin analogues upon binding model lipid membranes, making these agents good candidates to employ in future studies on anthracycline interactions with more environmentally-complex biological membranes.

Mobility of HLA Class I antigen is influenced by anti-CD--4 monoclonal antibody in lymphocyte membranes. A flow cytornetric energy transfer, fluorescence photobleaching recovery and rotational relaxation study.

Time-resolved fluorescence anisotropy of diphenylhexatriene (DPH) in various binary lipid mixtures of dimysterolphosphatidycholine (DMPC), dipalmitoylphosphatidylcholine (DPPC), distearoylphosphatidylcholine (DSPC) and dioleoylphosphatidylcholine (DOPC) has been studied. Evidence is given for phase separation in binary mixtures of DMPC/DSPC and DOPC/DSPC which is not found in DMPC/DPPC and DPPC/DSPC mixtures. A rising anisotropy decay has been observed for DPH in binary lipid mixtures containing DSPC and in DSPC liposomes at the phase transition which suggests a heterogeneous environment for the fluorophore at this temperature.

Protein Al is one of the core proteins present in eukaryotic ribonucleoprotein particles. Al binds to single-stranded (ss) nucleic acids and plays a role in RNA metabolism. Limited proteolysis generates an N-terminal domain (195 amino acid residues) termed UP1, which binds ssDNA and contains 2 putative nucleic acid binding domains. The glycine-rich C-terminal fragment (124 amino acid residues) modulates Al binding to single-stranded lattices, and may itself bind to ssRNA. Aromatic side chain interactions with nucleic acid bases are postulated to contribute to the free energy of binding of single-stranded DNA-binding (SSB) proteins. UP1 contains 1 Trp and 4 Tyr residues, while the C-terminal fragment contains 8 additional Tyr residues. Modifications of fluorophore quantum yield, emission properties and anisotropy induced by complex formation were investigated to address the occurrence of intercalation of aromatic residues of Al and related systems with the nucleic acid bases. Our results indicate that the tryptophan fluorescence properties of Al and UP1 are unchanged by the interaction of these proteins with nucleic acids, whereas tyrosine fluorescence is affected by complex formation.

We examined the time -dependent donor decays of 2 - amino purine (2 -APU) , in the presence of increasing amounts of acceptor 2-aminobenzophenine (2-ABP). As the concentration of 2-ABP increases, the frequency-responses diverge from that predicted by Forster. The data were found to be consistent with modified Forster equations, but at this time we do not state that these modified expressions provide a correct molecular description of this donor-acceptor system. To the best of our knowledge this is the first paper which reports a failure of the Forster theory for randomly distributed donors and acceptors.

We investigated the influence of end-to-end diffusion on intramolecular energy transfer between a naphthalene donor and dansyl acceptor linked by polymethylene chain. A range of viscosities of 0.6 - 200cP were obtained using propylene glycol at different temperatures (0-80°C) and methanol at 20°C. The intensity decays of naphthalene were measured in frequency-domain. Several theoretical models, including distance distributions were used to fit the data. The results indicate that end-to-end diffusion of flexible donor - acceptor pairs can be readily detected and quantified using frequency-domain fluorometry.

Phosphofructokinase (PFK) is a calmodulin (CaM) binding protein (Mayr, G.W. and Heilmeyer , L.M.G., Jr (1983) FEBS Lett. 195, 51)..We found that troponin C (TnC), which is homologous to CaM, also binds PFK. PFK titration of AEDANS-TnC showed that their apparent dissociation constant is comparable to that of PFK-CaM. Fluorescent labels were used to probe contact regions on TnC and CaM. It is likely that the C-terminal end of the connecting strand of the TnC molecule is close to PFK in the binary complex. Hydrophobic regions of TnC and CaM also possibly play roles in the binding and in the polymerization of PFK.

Lumazine protein is a 21200 Da protein containing a single tryptophan residue and a non-covalently bound, highly fluorescent ligand, 6,7-dimethyl-8-ribityllumazine. Visser et al. have reported that excitation into the region of the absorption of the tryptophan residue, around 300 nm, produces a distinct rise in fluorescence emission at 475 nm from the lumazine ligand. They analyzed a rise rate of around 1 ns and attributed this to energy transfer between the tryptophan donor and the lumazine as acceptor. This present report re-investigates this phenomenon using a ten times higher resolution (FWHM = 23 ps) . The fluorescence rise is found to be more complex and can only approximately be fitted by a sum of two exponential processes, with rise times of 0.02 and 0.6 ns. For the fluorescence of the tryptophan measured at 340 nm, no rise is detected but the decay is similarly much more complex than previously recognised with data taken at lower resolution. Global analysis of three 340 nm decay curves taken with time windows of 1.2, 4.8, and 55.6 ps/channel, results in about 5 exponential components being required for a satisfactory fit to the fluorescence decay.

Diphenylhexatrienylpropanoylhydrazyl stachyose (glyco-DPH), a new fluorescence probe, was synthesized. It inserts almost instantaneously into artificial phospholipid vesicles and biological membranes. Due to its large hydrophilic carbohydrate portion, it serves as an uncharged probe with a defined orientation within the membrane bilayer. Its usefulness to monitor lipid mobility by means of its fluorescence anisotropy could be demonstrated for dipalmitoylglycerophosphocholine at temperatures around the gel to liquid phase transition and the rigidifying effect of cholesterol on egg yolk phosphatidyicholine membranes. In addition, lipid mobility was determined for biological membrane systems such as yeast spheroplasts, yeast organelles, cultured human skin fibroblasts and compared with the 'fluidities’ of vesicles made of the corresponding lipid extracts. Bimodal Lorentzian lifetime distributions determined for glyco-DPH in vesicles of 1 -palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine showed that the label is distributed homogeneously within a phospholipid bilayer. Fluorescence microscopy of living (fibroblast) cells revealed selective labeling of the surface membrane with glyco-DPH under appropriate conditions.

The interaction of rabbit muscle aldolase with glyceraldehyde-3-phosphate dehydrogenase (GPDH) labeled with fluorescein-5-isothiocyanate (FITC) has been investigated at 25°C in Tris buffer, pH 7. 5. The addition of 5 to1 0 fold excess of aldolase to 0.1-1 micron GPDH labeled with FITC (GPDHFITC) causes a large increase in both the fluorescence and polarization of FITC over a period of several hours, reflecting the formation ofa complex between the enzymes. When GPDH-FITC is incubated with either 1 mM NAD or ADP, the fluorescence of FITC increases while the polarization decreases, indicating that these nucleotides may increase the degree of dissociation of tetrameric GPDH. The rate of approach to equilibrium during the formation of the complex between the two enzymes increases in the presence of either NAD or ADP but decreases with increasing concentrations ofGPDH. Therefore, the interaction of the enzymes may involve the dissociation of tetrameric GPDH into smaller units. Aldolase causes no changes in the fluorescence properties of probes that are attacted to the active site cysteine residues of GPDH, such as fluorescein acrylamide, indicating that when the active site of GPDH is blocked the formation of the complex between the two enzymes is prevented. When GPDH-FITC is incubated with either excess NAD or aldolase, the average fluorescence lifetime of FITC slowly increases and approaches that of free FITC. Since fluorescein is quenched by tryptophan, it is likely that the increases observed both in the fluorescence intensities and lifetimes of GPDH-FITC during its interaction with aldolase arise from the removal of tryptophan residues from the vicinity of the FITC groups as a result of changes either in the conformation or in the degree of dissociation of tetrameric GPDH.

The spontaneous transfer of pyrenedecanoyl-Iabeled phospholipids from small unilamellar vesicles to human erythrocyte ghost membranes was determined by a spectroscopic method. The fluorescence emission spectra of unilamellar vesicles containing 1 -acyl-2-pyrenedecanoyl- or 1 -O-(1'-alkenyl- -2-pyrenedecanoyl-sn-glycero-3-phospholipicjs of the choline and ethanolamine type show very high excimer to monomer fluorescence intensity ratios. Upon "dilution" of the fluorescent lipids into the erythrocyte acceptor membranes the excimer intensity decreased and the monomer intensity Increased. The time course of the monomer fluorescence increase revealed faster uptake of pyrene-labeled ether lipids (plasmalogens) as compared to the diacyl analogs. Significant amounts of pyrene-labeled diacyl phospholipids were incorporated into ghosts only in the presence of phospholipid transfer proteins. After incubation with pyrenephospholipid vesicles under appropiate conditions erythrocyte membranes showed almost exclusively pyrene monomer fluorescence Indicating random distribution of the fluorescent phospholipid probe within the membrane. Monomer fluorescence lifetimes of pyrene phospholipids In the erythrocyte membranes as determined by multifrequency phase fluorometry are very heterogeneous. In contrast, a single exponential decay ((tau)=103 ns) was observed for the monomer fluorescence of the pyrene phospholipids in unilamellar egg yolk phosphatidyicholine vesicles (label/lipid = 1/300 (mol/mol)). Thus, the fluorogenic phospholipids experience a much more heterogeneous environment in the biological membrane compared with a protein-free lipid bilayer system.

Recent studies in our laboratories have focused on using tyrosine (Tyr) fluorescence of calmodulin (CaM) and tryptophan (Trp) fluorescence of CaM-bound peptdies as intrinsic probes of structure and interactions of this Ca2+ regulatory protein. Plant CaM contains a single Tyr (Tyr.-l38) and vertebrate CaM contains two (Tyr-99 and Tyr-.l38). Neither protein contains Trp. The fluorescence properties of Tyr-138 of wheat-germ CaM is sensitive to conformational changes induced by perturbations such as Ca2+ ligation or depletion, and pH changes. Effects of these perturbations on quantum yield, lifetime and dynamic quenching of Tyr-l38 fluorescence are reported. We have also studied binding of amphiphilic peptides to wheat-germ CaM. A comparison of wheat CaM induced changes in the fluorescence properties of a single Trp of these peptides with those induced by bovine testes CaM indicate general similarities of the peptide binding surfaces of plant and mammalian CaMs. Frequency domain measurements of decay of intensity and anisotropy have suggested some orientational freedom and local motion of the Trp residue of CaM-bound peptide, independent of the overall protein motion, even when the Trp is expected to be buried in the doubly apolar protein-peptide interface. Calmodulin (CaM) is a ubiquitous calcium binding protein which is believed to regulate several different enzymes in diverse cells (Klee et al., 1982). Much of the structural work to date has been carried out on mammalian CaM. However, CaM has also been isolated from plant and invertebrate sources, and a high degree of sequence homology with vertebrate CaM has been found. The amino acid sequence of wheat germ CaM shows eleven substitutions, two insertions and one deletion compared with the 148.-residue bovine brain CaM (Toda et al., 1985). Specific differences with mammalian CaM at two sites make plant CaM attractive for fluorescence spectroscopic studies. These are: (1) The presence of a single tyrosine residue (Tyr-138) as the only unambiguous source of its intrinsic fluorescence as opposed to two (Tyr-99 and Tyr-l38) in mammalian CaM, (2) The presence of a single cysteine, Cys-27 (as opposed to none in mammalian CaM) which can be specifically labeled by extrinsic fluorescent probes (Mills et al., 1988). He have found that the fluorescence quantum yield and lifetime of Tyr-l38 as well as its accessibility to polar quenchers are especially sensitive to binding of Ca2+ to wheat germ CaM. In the presence of Ca2+ wheat CaM structure becomes significantly helical and amphiphilic peptides bind with high affinity. Ca2+ dependent binding of amphiphilic peptides to bovine CaM has been thoroughly studied and this provides a viable model for interaction of CaM with its dependent enzymes (Malencik and Anderson, 1983; McDowell et al., 1985; O'Neil et al., 1987). The data presented here suggest general similarities between the peptide binding surfaces of mammalian and plant Cam.

Exploration of the effect of ligand-protein interactions on conformational substates and internal dynamics in different regions of phospholipase A2 from porcine pancreas (PLA2), was performed by combining site-directed mutagenesis and time-resolved fluorescence measurements. The single tryptophan residue (Trp-3) in the wild type protein was replaced by a phenylalanine residue, whereafter Trp was substituted either for leucine-31 ,located in the calcium binding loop, or for phenylalanine-94, located at the "back side" of the enzyme, in a-helix E (Dijkstra et al., J. Mol. Biol., 147, 97-123, 1981). Analyses by the Maximum Entropy Method (MIEM) of the total fluorescence intensity decays, provide in each case a distribution of separate lifetime classes, which can be interpreted as reflecting the existence of discrete conformational substates in slow exchange with respect to the time-scale of the decay kinetics. The fluorescence decay of the W94 mutant is dominated by an extremely short excited state lifetime of ~60 ps, probably arising from the presence of two proximate disulfide bridges. Time-resolved fluorescence anisotropy studies show that the Trp residue near the NH2 terminus (Trp-3) undergoes a more limited rotational motion than the Trp-3 1 located in the calcium binding loop. The widest angular rotation is observed at position 94, in a-helix E. Calcium binding displays the strongest influence on the lifetime distribution of Trp-31: a major local conformation corresponding to a lifetime class with a barycenter value of ~5.5 ns and contributing to ~50% of the decay is selected. The conformations giving rise to the short lifetimes ((tau)1 and (tau)2 lifetime classes) become less important. The contribution of the third lifetime class (c3) stays at a constant value of 30%. In the presence of calcium, the amplitude of motion is wider than without the ion. There is virtually no effect of calcium binding on the lifetime distribution of the Trp residue at the 3 or the 94 position. Binding of the monomeric substrate analog n-dodecylphosphocholine (C12PN) in the presence of calcium hardly affects neither the Trp-3 excited state population distribution, nor its rotational dynamics. The binding of C12PN monomers to the W31 mutant further increases the contribution of the t4lifetime class at the expense of c2. A more restricted rotation of the Trp-31 residue is also induced. The binding of the micellar substrate analog n-hexadecylphosphocholine (C16PN) in the presence of calcium is very efficient in modifying the lifetime distribution of Trp-3. Essentially, one major broad lifetime population (centered at ~2.6 ns) is revealed by MEM analysis of the total intensity decay. The internal motion is slowed down and the angle of rotation is much smaller in this conformation. Neither the excited state lifetime distribution of Trp-31 nor its dynamics are affected by micelle binding relative to monomer binding. In conclusion, by placing a single Tip-residue at strategic positions along the peptide chain of PLA2, relevant to the binding of biological ligands, an excellent model system for the study of selective perturbations of conformational substates and internal dynamics is provided.

The fluorescence emission of the single tryptophan in bovine brain S-lOOa protein has been studied by using time-resolved laser fluorescence spectroscopy. The tryptophan fluorescence emission was isolated by a Schott 0-54 cut-off filter at right angles to the excitation direction by a Hamamatsu R955P photomultiplier. With excitation at 295 nm, the fluorescence decay of S-lOOa protein in 25 mM Tris buffer was best represented by a sum of three exponential terms regardless of solvent conditions. At 20°C and pH 7.2, the three components of lifetime for apo-S-lOOa protein were (tau)1~0.43 ns, (tau)2~1.24 ns, and (tau)3~4.05 ns. The corresponding fluorescence contributions of each component were 31%, 31% and 38%, respectively. When the protein was saturated with Mg2, the lifetimes increased slightly and the contribution of the shortest fluorescence component ((tau)1) to the total emission intensity increased slightly at the expense of the other two components. Binding of Ca2+ to S-lOOa protein resulted in a significant decrease of (tau)1, and a substantial increase of (tau)2 and (tau)3, and an increase of the average of the three lifetimes. Under this condition the fractional fluorescence contributions associated with (tau)2 and (tau)3 increased very significantly at the expense of the shortest component. The fluorescence decay behavior of each of he S-lOOa samples was relatively insensitive to the variation of temperature (4~20°C). At pH 8.2 and pH 8.4, the decay behavior of the protein in response to binding of Ca2+ and Mg2+, and changes of temperature was very similar to those observed at pH 7.2. The triple exponential decay kinetics of the single tryptophan in S-lOOa protein could be rationalized by the existence of multiple local conformers.

Distance distributions of a flexible molecule were recovered from steady-state fluorescence energy transfer measurements with the method suggested by Cantor and Pechukas In this method, the Forster distance (R₀) is varied by attaching different donor-acceptor (D-A) pairs to the linker of interest. The distance distributions are recovered from the energy transfer efficiency measurements on D-A pairs with different R0s, followed by non-linear least squares fitting to distance distribution models. Thirteen compounds were synthesized with R0s ranging from 6-32 angstroms. Each compound contained a tryptamine donor linked by an alkyl chain (~10 carbon atoms) to one of thirteen different acceptors. The distance distributions were recovered by fitting the transfer efficiencies of the D-A pairs to both Gaussian and skewed Gaussian models. Our results indicate that the shape of the distribution is best described by a skewed Gaussian model. This is the first reported measurement of a skewed Gaussian distance distribution. In addition, the results of the Cantor and Pechuka's method and analysis were found to be consistent with distance distributions recovered from time-resolved frequency-domain fluorometry.

We used the Hamamatsu model PLP-01 picosecond light pulser as a 413-nm excitation light source for frequency-domain fluorescence measurements. In comparison with sync-pumped/cavity dumped/frequency-doubled dye lasers, the 413-nm PLP-01 shows a longer FWHM (40 ps), a similar pulse repetition rate (up to 10 MHz), much less output power at a fixed wavelength (0.44 mW peak, 220 nW maximum average power), but is less expensive, small-sized, and easy to handle. Using the PLP-01 , we have been able to perform fluorescence measurements up to an upper modulation frequency of about 2000 MHz, and to resolve mixtures of fluorophores exhibiting different lifetimes. During our tests, we observed remarkable and lasting (2 h) time drifts between the optical output and the electrical trigger input or output. At present, work is in progress at Harpamatsu to eliminate these drifts.

The singlet excitation transfer and trapping kinetics in photosynthetic membranes in case of low excitation intensities is studied by spectrally selective picosecond-time- domain fluorescence spectroscopy and by numerical integration of an appropriate system of equations. The essential features of our models are spectral heterogeneity of the light- harvesting antenna, inclusion of temperature effects, nonabsolute excitation traps, correlation between spectral and spatial parmeters. A reasonably good agreement between theoretical and experimental fluorescence decay kinetics for several purple photosynthetic bacteria has been achieved. This comparison gives several interesting numerical constants characterizing microscopic excitation transfer between different light-harvesting-antenna pigment-protein complexes towards the reaction centres. Some aspects of the experiment are also discussed.

We have constructed a novel picosecond x-ray (PxR) source to be used for time resolved x-ray diffraction. The PxR source has a metallic photocathode from which photoemission is induced by ultraviolet picosecond laser.pulses. Electron bunches of over 3 nC have been produced with a tantalum photocathode excited by a frequency quadrupled pulsed modelocked Nd:YAG laser.The accelerated electron bunches strike a copper anode generating 6.2x10⁶ Cu Ka x-ray photons cm^-2.sr^-1 with a time width of less than 70 ps. A high energy Nd:YLF regenerative amplified laser was constructed to increase the repetition rate of the source to 2 kHz. Several detection methods and experimental schemes for PxR diffraction are presented

short-external-cavity (SXC) is used to obtain continuous single-mode operation over 20-50 cm^-1 with near-IR semiconductor diode lasers. When the SXC lasers are used as a source in a tunable diode laser absorption spectrometer (TDLAS), very sensitive (~3x10^-6 absorbance) trace-gas detection and spectroscopic measurements are possible.

We have set up an infrared reflection absorption grating spectrometer working under high vacuum conditions. The infrared source is made of a tantalum cavity heated to 2500 K. Its brightness at 3000 cm^-1 (emissivity > 0.7) is four times higher than that of a common silicon carbide source at 1500 K. The adsorbate absorption detection technique is based on the use of a photoelastic modulator. All those features permit a high absorbance sensitivity while avoiding the measured infrared spectrum to be affected by the absorption from residual gases in the spectrometer. The measurement of infrared spectra between 2.6 and 5 μm allowed us to specially follow the evolution of the methyl and hydroxyl stretching bands of CH3OH and CD3OD versus exposure at 80 K and versus annealing temperature for a polycrystailine platinum and a Cu(110) substrates. The strong attenuation of the OH band at low coverage indicates that, in the submonolayer coverage range, the methanol molecules are adsorbed with their OH group axes quasi parallel to both metallic surfaces. Because both the symmetric and asymmetric CH3 (or CD3) stretching modes are detected, the methyl group axis must be inclined versus the surface normals. In the multilayer coverage range, the methanol forms a thin amorphous phase which can be crystallized upon annealing. The simultaneous detection of the in-phase and out of phase components of the crystallized layer hydroxyl stretching indicates that this layer is constituted by arbitrarily oriented small crystallites.

The main characteristic of molecular crystals RbCHO4,T1C8H5O4 and KC8H504 (they be1on to isomorphous crystal) is that the struc-. urai units (milecyle or atomic group) which are Located on crystal lattices essentially keep the shape and feature which they nave in the free state.

The time and temperature dependence of the intrinsic fluorescence of double-stranded d(CTGA[2AP]TTCAG)2 DNA decamer containing a modified adenine base, 2AP, show that the bases exists in at least four distinct, interconverting conformations. The melting transition of the duplex governs the equilibrium of these states. Measurements on a purely singlestranded decamer facilitate the interpretation of data. States are identified as completely unstacked, maximally stacked, and partially stacked, according to state fluorescence lifetimes and the temperature dependence of lifetimes and populations. The equilibrium of state populations in the single-stranded decamer is described accurately by simple enthalpy and entropy differences, where linear ln [population] vs. 1/T plots are observed. Similar plots for the double-stranded show nonlinearities. The cooperative melting transition of the double-stranded decamer is reflected most strongly in the interconversion of the completely unstacked state with a partially-stacked state. This completely unstacked state is not the "unstacked" state assumed in UV absorption hypochromicity measurements.

The sequence-dependent high-frequency motion of a fluorescent base in DNA has been studied by picosecond time-resolved fluorescence depolarization measurements on a series of defined-sequence synthetic oligonucleotides of the type d[GCCXCCG] d[CGG(AP)GGC], where X = T, A, G or C and AP is the fluorescent adenine analogue 2- aminopurine. The fluorescence anisotropy contained a rapidly decaying component that was due to restricted internal rotation of AP and a slowly decaying component due to overall tumbling. The tumbling correlation time varied with X in a manner that reflected the association state of the oligonucleotide strands. The internal motion dynamics of AP depended on the nature of the base-pairing interactions between AP and x.

The technique of long-range electronic energy transfer has been used to obtain structural data on nucleic acids in solution. Fluorescent chromophores have been covalently attached via an aminoethyl linker to the 5'-terminus of a synthetic oligonucleotide. Duplexes were formed with an energy donor, fluorescein, at one end and an acceptor, tetramethyl rhodamine, at the other. The rate of energy transfer has been determined from the fluorescence decay of the donor which was measured with a picosecond photon counting system. The effects of dye:DNA interactions cause additional quenching of the donor fluorescence and must be considered in the interpretation of the energy transfer experiments. An apparent donor:acceptor distance of 40.9 angtroms was calculated for a 12 b.p. labelled duplex.

The technical simplicity and speed of cross-correlation frequency domain fluorometry has been compromised recently by the need to measure and analyse the phase and modulation responses of a sample over a wide range of excitation frequencies. Phase fluorometers currently use single frequency cross-correlation detection combined with either single frequency excitation or harmonic excitation from a pulsed source. Phase and modulation response curves are constructed by taking a series of point-by-point measurements over the frequency region of interest. This can be a time consuming process which requires the attention of an experienced operator. The Fourier transform spectrofluorometer described here combines multi-harmonic excitation with multi-harmonic cross-correlation detection and digital signal processing techniques to implement a frequency domain fluorometer which can acquire phase and modulation responses simultaneously at 40 to 50frequencies. A particularly dramatic capability of the Fourier transform fluorometer is its ability to collect multi-frequency data at millisecond rates thus allowing complex lifetime analysis of rapidly evolving fluorescence systems. Further, multi-harmonic excitation and detection is the equivalent of measuring the time course of a sample's response to impulse excitation. Therefore, time domain data is inherently available from the instrument and may be fitted to models of fluorescence decay by iterative reconvolution techniques as an alternative non-linear least squares fitting of the frequency domain data.

Endometriosis, a common disease in women in the reproductive age group, is defined pathologically by the presence of endometrial tissue (inner lining of the uterus) outside the uterus. The displaced tissue is histologically identical to endometrium. In addition to being a highly prevalent disease, this disease is associated with many distressing and debilitating symptoms. Motivated by the need to improve diagnosis by endoscopic imaging instrumentation, we have previously used several drugs to cause selective laser-induced fluorescence of active surgically induced endometriosis in the rabbit model in vivo using ultraviolet-wavelength (351.1 and 363.8 nm) excitation from an argon-ion laser. In the present study we have investigated methods of enhancing differentiation between normal and abnormal tissue by using other excitation wavelengths. In addition to an enhanced capability for detecting abnormal tissue, there are several other advantages associated with using visible-wavelength excitation, such as deeper penetration into the tissue, as well as increased equipment performance, reliability, versatility, and availability. The disadvantage is that because only wavelengths longer than the excitation wavelength can be used for detection, some of the spectral information is lost. Because human endomeiriosis samples were somewhat limited in quantity, as well as specimen size, we used normal ovarian tissue for the laser-induced-fluorescence differentiation-enhancement studies. Positive enhancement of the laser-induced- fluorescence differentiation was found in human ovarian tissue in vitro utilizing 514.5-nm excitation from an argonion laser. Additionally, preliminary verification of this concept was accomplished in active surgically induced endometriosis in the rabbit model in vivo with visible argon-ion laser excitation of two tetracycline-based drugs. Future experiments with other drug treatments and excitation/detection parameters are planned.

The ability to reversibly control fluorescence quenching by processes such as resonance energy transfer to a photochromic acceptor is of potential interest in mechanistic fluorescence studies, and might also find practical application. In this paper we report on studies of a model system comprising a europium chelate (europium tetrakis (2-naphthoyltrifluoroacetone) piperidine) and a photochromic spiropyran(SP) (1-hexadecyl-6-methoxy-3, 3'-dimethyl-8-nitro-2H-chromene-2-spiro-2'-(2,3-dihydroindole) codissolved in a bilayer of the phospholipid Dipalmitoyl-phosphatidylcholine in aqueous dispersion. Reversible quenching of the fluorescence of the europium complex is observed when the system is irradiated with long wave UV light, resulting in transient conversion of the SP into a strongly coloured merocyanine dye (MC) , which has spectral overlap with the europium emission. There are several possible modes of fluorescence quenching in this system. One possibility is formation of a ground state complex, wherein europium dissociates from its ligand and binds to the charged merocyanine. The most likely alternatives are resonance energy transfer, collisional deactivation of the long lived excited state of europium or energy transfer from the ligand on europium to the SP triplet state. It has been shown using analysis of UV absorption spectra that ground state complex formation does not occur. Both isomers of SP and merocyanine are involved in the overall quenching of europium emission, and both collisional- and resonance energy transfer are involved.

An imaging photon detector has been modified to incorporate fast timing electronics coupled to a custom built photon correlator interfaced to a RISC computer. Using excitation with intensity- muodulated light, fluorescence images can be readily obtained where contrast is determined by the decay time of emission, rather than by intensity. This technology is readily extended to multifrequency phase/demodulation fluorescence imaging or to differential polarised phase fluorometry. The potential use of the correlator for confocal imaging with a laser scanner is also briefly discussed.

The fluorescence anisotropy of ethidium bromide intercalated into the 146 base pairs of DNA on the nucleosoine decays due to a combination of rotational diffusion and torsional flexing motions of the DNA. We characterize the Internal flexing motions el the DNA by fitting the anisotropy to a mathematical model proposed by Schurr through the variation of two parameters: a torsional coeffident and a frictional coefficient. Using data collected 5° C and 20° C in water, and at 5° C in D₂O, and assuming a model of 146 base pairs clamped at both ends, we have in all cases obtained frictional and torsional coefficients that are higher than the values reported by others.

The role of membrane lipid-protein interactions in malignant cell transformation was examined with adenosine deaminase (ADA) as a representative membrane protein. ADA's activity changes dramatically in transformed cells and accordingly it is a malignancy marker. Yet, the mechanisms controlling its variable activity are unknown. We undertook the spectroscopic deciphering of its interactions with its lipidic environment in normal and malignant cells. ADA exists in two interconvertible forms, small (45 KD) and large (21OKD). The large form consists of two small catalytic subunits (55-ADA) and a dimeric complexing protein ADCP. The physiological role of ADCP was not known either. Our studies were carried out at three levels.: 1. Solution enzyme kinetics, 2. The interaction of 55-ADA with ADCP reconstituted in liposomes: Effect of cholesterol and 3. Multifrequency phase modulation spectrofluorometry of pyrene-labeled 55-ADA bound to ADCP on the membranes of normal and RSV or RSV Ts68 transformed chick embryo fibroblasts. We found: 1. ADCP has an allosteric regulatory role on 55-ADA, which may be of physiological relevance: It inhibits 55-ADA activity at low physiological adenosine concentrations but accelerates deamination at high substrate concentration. 2. When reconstituted in DMPC liposomes, it retains 55-ADA activity (in its absence the activity is lost) and upon rigidification with cholesterol, a three fold increase in 55-ADA activity is attained, contrary to ADCP's regulatory activity when free of lipids. 3. The reduced ADA activity in transformed chick embryo fibroblasts is associated with increased membrane lipid fluidity (reduced order parameter), reduced accessibility of ADCP and increase rotational dynamics of the complex. We thus obtained spectroscopic deciphering of the vertical motion of ADCP, controlled by lipid-protein interaction, resulting in variable activity of this malignancy marker.

A possible ultrafast rotational motion of the phenol ring of buried tyrosine residue 69 in lima bean trypsin/chymotrypsin inhibitor (LBI) was observed in addition to a 40-ps restricted rotational motion and the overall nanosecond protein rotation. A picosecond laser-streak camera system was used for short time resolution and measurements of the fluorescence anisotropy, r, vs. temperature, time and glycerol concentration were done by simultaneously recording both polarized tyrosine emission components. A variable-width square well was used to describe the temperature-dependent amplitude of the ultrafast process.