Frequency-domain fluorescence microscopy with the LED as a light source

We describe a frequency-domain lifetime fluorometer based on a microscope and a modulated light-emitting diode (LED) excitation source (370/460 nm), which operates in the frequency range 120 Hz–250 MHz. We collected multifrequency phase and modulation fluorescence responses from cellular areas as small as 10–15 µm in diameter. We also collected fluorescence lifetime data from cells stained by a lipophilic coumarin sensitized europium fluorophore, Coum-Eu, with a millisecond lifetime, and Ru(bpy)₂phe-C₁₂, with microsecond lifetime. Nanosecond lifetimes from native nuclei stained with SYTO 14 and SYTO 16 probes were measured as well. We demonstrate that a simple LED excitation source can, for many applications, successfully replace complex and expensive laser systems, which have been used for cellular frequency-domain lifetime measurements. As the LEDs are very stable with low noise, it will be possible to image even smaller sample areas using brighter LEDs. With availability of modulated LEDs emitting at several wavelengths covering almost the entire visible spectrum it is easy to assemble a system for the fluorophore of choice. The ability to select an excitation source for a given fluorophore and low price make such an excitation source even more practical.     

Multidie LED combined with homogenizing optics to improve frequency response and intensity for FLIM applications

Application of light‐emitting diodes (LEDs) in frequency‐domain fluorescence lifetime imaging microscopy (FLIM) has been limited by the trade‐off between modulation frequency and illumination intensity of LEDs, which affects the signal‐to‐noise ratio in fluorescence lifetime measurements. To increase modulation frequency without sacrificing output power of LEDs, we propose to use LEDs with multiple dice connected in series. The LED capacitance was reduced with series connection; therefore, the frequency response of multidie LED was significantly increased. LEDs in visible light, including blue, green, amber and red, were all applicable in FLIM. We also present a homogenizing optics design, so that multidie LEDs produced uniform illumination on the same focal spot. When the homogenizing optics was combined with multicolour emitters, it provides multiple colour selection in a compact and convenient design.     

φFLIM: a new method to avoid aliasing in frequency-domain fluorescence lifetime imaging microscopy

In conventional wide‐field frequency‐domain fluorescence lifetime imaging microscopy (FLIM), excitation light is intensity‐modulated at megahertz frequencies. Emitted fluorescence is recorded by a CCD camera through an image intensifier, which is modulated at the same frequency. From images recorded at various phase differences between excitation and intensifier gain modulation, the phase and modulation depth of the emitted light is obtained. The fluorescence lifetime is determined from the delay and the decrease in modulation depth of the emission relative to the excitation. A minimum of three images is required, but in this case measurements become susceptible to aliasing caused by the presence of higher harmonics. Taking more images to avoid this is not always possible owing to phototoxicity or movement. A method is introduced, φFLIM, requiring only three recordings that is not susceptible to aliasing. The phase difference between the excitation and the intensifier is scanned over the entire 360° range following a predefined phase profile, during which the image produced by the intensifier is integrated onto the CCD camera, yielding a single image. Three different images are produced following this procedure, each with a different phase profile. Measurements were performed with a conventional wide‐field frequency‐domain FLIM system based on an acousto‐optic modulator for modulation of the excitation and a microchannel‐plate image intensifier coupled to a CCD camera for the detection. By analysis of the harmonic content of measured signals it was found that the third harmonic was effectively the highest present. Using the conventional method with three recordings, phase errors due to aliasing of up to ± 29° and modulation depth errors of up to 30% were found. Errors in lifetimes of YFP‐transfected HeLa cells were as high as 100%. With φFLIM, using the same specimen and settings, systematic errors due to aliasing did not occur.     

Multiple frequency fluorescence lifetime imaging microscopy

The experimental configuration and the computational algorithms for performing multiple frequency fluorescence lifetime imaging microscopy (mfFLIM) are described. The mfFLIM experimental set‐up enables the simultaneous homodyne detection of fluorescence emission modulated at a set of harmonic frequencies. This was achieved in practice by using monochromatic laser light as an excitation source modulated at a harmonic set of frequencies. A minimum of four frequencies were obtained by the use of two standing wave acousto‐optic modulators placed in series. Homodyne detection at each of these frequencies was performed simultaneously by mixing with matching harmonics present in the gain characteristics of a microchannel plate (MCP) image intensifier. These harmonics arise as a natural consequence of applying a high frequency sinusoidal voltage to the photocathode of the device, which switches the flow of photoelectrons ‘on’ and ‘off’ as the sinus voltage swings from negative to positive. By changing the bias of the sinus it was possible to control the duration of the ‘on’ state of the intensifier relative to its ‘off’ state, enabling the amplitude of the higher harmonic content in the gain to be controlled. Relative modulation depths of 400% are theoretically possible from this form of square‐pulse modulation. A phase‐dependent integrated image is formed by the sum of the mixed frequencies on the phosphor of the MCP. Sampling this signal over a full period of the fundamental harmonic enables each harmonic to be resolved, provided that the Nyquist sampling criterion is satisfied for the highest harmonic component in the signal. At each frequency both the phase and modulation parameters can be estimated from a Fourier analysis of the data. These parameters enable the fractional populations and fluorescence lifetimes of individual components of a complex fluorescence decay to be resolved on a pixel‐by‐pixel basis using a non‐linear fit to the dispersion relationships. The fitting algorithms were tested on a simulated data set and were successful in disentangling two populations having 1 ns and 4 ns fluorescence lifetimes. Spatial invariance of the lifetimes was exploited to improve the accuracy significantly. Multiple frequency fluorescence lifetime imaging microscopy was then successfully applied to resolve the fluorescence lifetimes and fluorescence intensity contributions in a rhodamine dye mixture in solution, and green fluorescent protein variants co‐expressed in live cells.     

Filter cubes with built-in ultrabright light-emitting diodes as exchangeable excitation light sources in fluorescence microscopy

The use of ultrabright light‐emitting diodes as a potential substitute for conventional excitation light sources in fluorescence microscopy is demonstrated. We integrated ultrabright light‐emitting diodes in the filter block of a conventional fluorescence microscope together with a collimating Fresnel lens, a holographic diffuser and emission filters. This setup enabled convenient changes between different excitation light sources and resulted in high excitation efficiencies. Quantitative comparison of image intensities of test samples revealed that light‐emitting diodes yielded intensities in the range of a mercury arc lamp depending on the wavelength. The use of ultrabright light‐emitting diodes also enabled luminescence lifetime imaging without the need for image intensification.     

Optimized protocol of a frequency domain fluorescence lifetime imaging microscope for FRET measurements

Frequency-domain fluorescence lifetime imaging microscopy (FLIM) has become a commonly used technique to measure lifetimes in biological systems. However, lifetime measurements are strongly dependent on numerous experimental parameters. Here, we describe a complete calibration and characterization of a FLIM system and suggest parameter optimization for minimizing measurement errors during acquisition. We used standard fluorescent molecules and reference biological samples, exhibiting both single and multiple lifetime components, to calibrate and evaluate our frequency domain FLIM system. We identify several sources of lifetime precision degradation that may occur in FLIM measurements. Following a rigorous calibration of the system and a careful optimization of the acquisition parameters, we demonstrate fluorescence lifetime measurements accuracy and reliability. In addition, we show its potential on living cells by visualizing FRET in CHO cells. The proposed calibration and optimization protocol is suitable for the measurement of multiple lifetime components sample and is applicable to any frequency domain FLIM system. Using this method on our FLIM microscope enabled us to obtain the best fluorescence lifetime precision accessible with such a system. Microsc. Res. Tech., 2009. © 2008 Wiley-Liss, Inc.     

The polarized AB plot for the frequency‐domain analysis and representation of fluorophore rotation and resonance energy homotransfer

The graphical representation of single‐frequency phase‐modulation fluorescence lifetime imaging data, referred to as the AB plot, is extended to take into account measurements of the polarized components of the fluorescence. For a hindered rotator model (characterized with a single excited‐state lifetime, a single rotational correlation time and limiting initial and final anisotropies) the rotational correlation time and the excited lifetime can be determined from the AB plot of any two of the following emission components: parallel, perpendicular, total emission or combinations thereof. A strategy for resolving the component hindered rotations and lifetimes for mixtures of two hindered rotators from measurements of the total, parallel and perpendicular components of the emission is developed. The analysis does not require prior knowledge of the initial limiting anisotropy or of the steady‐state anisotropy or of the excited state lifetime. Plots in polarized AB space derived for heterogeneous systems are constructed to aid interpretation of frequency‐domain dynamic depolarization imaging microscopy experiments. These plots can be used to distinguish spatially dependent rotational correlation time heterogeneity from heterogeneity in limiting anisotropies. The effects of noise and aperture depolarization are discussed. It is anticipated that the polarized AB plot will provide a useful adjunct to existing methods for visualizing and analysing dynamic polarization phenomena arising from molecular dynamics and homo‐energy transfer in single‐frequency microscopy applications.     

Fluorescence lifetime imaging microscopy of Chlamydomonas reinhardtii: non-photochemical quenching mutants and the effect of photosynthetic inhibitors on the slow chlorophyll fluorescence transient

Fluorescence lifetime‐resolved images of chlorophyll fluorescence were acquired at the maximum P‐level and during the slower transient (up to 250 s, including P‐S‐M‐T) in the green photosynthetic alga Chlamydomonas reinhardtii. At the P‐level, wild type and the violaxanthin‐accumulating mutant npq1 show similar fluorescence intensity and fluorescence lifetime‐resolved images. The zeaxanthin‐accumulating mutant npq2 displays reduced fluorescence intensity at the P‐level (about 25–35% less) and corresponding lifetime‐resolved frequency domain phase and modulation values compared to wild type/npq1. A two‐component analysis of possible lifetime compositions shows that the reduction of the fluorescence intensity can be interpreted as an increase in the fraction of a short lifetime component. This supports the important photoprotection function of zeaxanthin in photosynthetic samples, and is consistent with the notion of a ‘dimmer switch’. Similar, but quantitatively different, behaviour was observed in the intensity and fluorescence lifetime‐resolved imaging measurements for cells that were treated with the electron transport inhibitor 3‐(3,4‐dichlorophenyl)‐1,1‐dimethyl urea, the efficient PSI electron acceptor methyl viologen and the protonophore nigericin and. Lower fluorescence intensities and lifetimes were observed for all npq2 mutant samples at the P‐level and during the slow fluorescence transient, compared to wild type and the npq1 mutant. The fluorescence lifetime‐resolved measurements during the slow fluorescence changes after the P level up to 250 s for the wild type and the two mutants, in the presence and absence of the above inhibitors, were analyzed with a graphical procedure (polar plots) to determine lifetime compositions. At higher illumination intensity, wild type and npq1 cells show a rise in fluorescence intensity and corresponding rise in the species concentration of the slow lifetime component after the initial decrease following the P level. This reversal is absent in the npq2 mutant, and for all samples in the presence of the inhibitors. Lifetime heterogeneities were observed in experiments averaged over multiple cells as well as within single cells, and these were followed over time. Cells in the resting state (induced by several hours of darkness), instead of the normal swimming state, show shortened lifetimes. The above results are discussed in terms of a superposition of effects on electron transfer and protonation rates, on the so‐called ‘State Transitions’, and on non‐photochemical quenching. Our data indicate two major populations of chlorophyll a molecules, defined by two ‘lifetime pools’ centred on slower and faster fluorescence lifetimes.     

Time-gated fluorescence lifetime imaging and microvolume spectroscopy using two-photon excitation

A scanning microscope utilizing two-photon excitation in combination with fluorescence lifetime contrast is presented. The microscope makes use of a tunable femtosecond titanium:sapphire laser enabling the two-photon excitation of a broad range of fluorescent molecules, including UV probes. Importantly, the penetration depth of the two-photon exciting (infra)red light is substantially greater than for the corresponding single-photon wavelength while photobleaching is significantly reduced. The time structure of the Ti:Sa laser can be employed in a straightforward way for the realization of fluorescence lifetime imaging. The fluorescence lifetime is sensitive to the local environment of the fluorescent molecule. This behaviour can be used for example to quantify concentrations of ions, such as pH and Ca²⁺, or pO₂ and pCO₂. In the set-up presented here the fluorescence lifetime imaging is accomplished by time-gated single photon counting. The performance and optical properties of the microscope are investigated by a number of test measurements on fluorescent test beads. Point-spread functions calculated from measurements on 230-nm beads using an iterative restoration procedure compare well with theoretical expectations. Lifetime imaging experiments on a test target containing two different types of test bead in a fluorescent buffer all with different lifetimes (2.15 ns, 2.56 ns and 3.34 ns) show excellent quantitative agreement with reference values obtained from time correlated single photon counting measurements. Moreover, the standard deviation in the results can be wholly ascribed to the photon statistics. Measurements of acridine orange stained biofilms are presented as an example of the potential of two-photon excitation combined with fluorescence lifetime contrast. Fluorescence lifetime and intensity images were recorded over the whole sample depth of 100 μm. Fluorescence intensity imaging is seriously hampered by the rapid decrease of the fluorescence signal as a function of the depth into the sample. Fluorescence lifetime imaging on the other hand is not affected by the decrease of the fluorescence intensity.     

Double-pulse fluorescence lifetime measurements

It is demonstrated that fluorescence lifetimes in the nanosecond and picosecond time-scale range can be observed with the recently proposed double-pulse fluorescence lifetime imaging technique (Müller et al. , 1995, Double-pulse fluorescence lifetime imaging in confocal microscopy. J. Microsc 177, 171–179).
A laser source with an optical parametric amplifier (OPA) system is used to obtain short pulse durations needed for high time resolution, wavelength tunability for selective excitation of specific fluorophores and high pulse energies to obtain (partial) saturation of the optical transition.
It is shown that fluorescence lifetimes can be determined correctly also with nonuniform saturation conditions over the observation area.
A correction scheme for the effect on the measurements of laser power fluctuations, which are inherently present in OPA systems, is presented. Measurements on bulk solutions of Rhodamine B and Rhodamine 6G in different solvents confirm the experimental feasibility of accessing short fluorescence lifetimes with this technique.
Because signal detection does not require fast electronics, the technique can be readily used for fluorescence lifetime imaging in confocal microscopy, especially when using bilateral scanning and cooled CCD detection.     

Condenser‐free contrast methods for transmitted‐light microscopy

Phase contrast microscopy allows the study of highly transparent yet detail‐rich specimens by producing intensity contrast from phase objects within the sample. Presented here is a generalized phase contrast illumination schema in which condenser optics are entirely abrogated, yielding a condenser‐free yet highly effective method of obtaining phase contrast in transmitted‐light microscopy. A ring of light emitting diodes (LEDs) is positioned within the light‐path such that observation of the objective back focal plane places the illuminating ring in appropriate conjunction with the phase ring. It is demonstrated that true Zernike phase contrast is obtained, whose geometry can be flexibly manipulated to provide an arbitrary working distance between illuminator and sample. Condenser‐free phase contrast is demonstrated across a range of magnifications (4–100×), numerical apertures (0.13–1.65NA) and conventional phase positions. Also demonstrated is condenser‐free darkfield microscopy as well as combinatorial contrast including Rheinberg illumination and simultaneous, colour‐contrasted, brightfield, darkfield and Zernike phase contrast. By providing enhanced and arbitrary working space above the preparation, a range of concurrent imaging and electrophysiological techniques will be technically facilitated. Condenser‐free phase contrast is demonstrated in conjunction with scanning ion conductance microscopy (SICM), using a notched ring to admit the scanned probe. The compact, versatile LED illumination schema will further lend itself to novel next‐generation transmitted‐light microscopy designs. The condenser‐free illumination method, using rings of independent or radially‐scanned emitters, may be exploited in future in other electromagnetic wavebands, including X‐rays or the infrared.     

Fluorescence lifetime imaging in scanning microscopes: acquisition speed, photon economy and lifetime resolution

In this paper a detailed discussion is presented of the factors that affect the fluorescence lifetime imaging performance of a scanning microscope equipped with a single photon counting based, two‐ to eight‐channel, time‐gated detection system. In particular we discuss the sensitivity, lifetime resolution, acquisition speed, and the shortest lifetimes that can be measured. Detection systems equipped with four to eight time‐gates are significantly more sensitive than the two time‐gate system. Only minor sensitivity differences were found between systems with four or more time‐gates. Experiments confirm that the lifetime resolution is dominated by photon statistics. The time response of the detector determines the shortest lifetimes that can be resolved; about 25 ps for fast MCP‐PMTs and 300–400 ps for other detectors. The maximum count rate of fast MCP‐PMTs, however, is 10–100 times lower than that of fast PMTs. Therefore, the acquisition speed with MCP‐PMT based systems is limited. With a fast PMT operated close to its maximum count rate we were able to record a fluorescence lifetime image of a beating myocyte in less than one second.     

Real-time cellular uptake of serotonin using fluorescence lifetime imaging with two-photon excitation

The real-time uptake of serotonin, a neurotransmitter, by rat leukemia mast cell line RBL-2H3 and 5-hydroxytryptophan by Chinese hamster V79 cells has been studied by fluorescence lifetime imaging microscopy (FLIM), monitoring ultraviolet (340 nm) fluorescence induced by two-photon subpicosecond 630 nm excitation. Comparison with two-photon excitation with 590 nm photons or by three-photon excitation at 740 nm shows that the use of 630 nm excitation provides optimal signal intensity and lowered background from auto-fluorescence of other cellular components. In intact cells, we observe using FLIM three distinct fluorescence lifetimes of serotonin and 5-hydroxytryptophan according to location. The normal fluorescence lifetimes of both serotonin (3.8 ns) and 5-hydroxytryptophan (3.5 ns) in solution are reduced to approximately 2.5 ns immediately on uptake into the cell cytosol. The lifetime of internalized serotonin in RBL-2H3 cells is further reduced to approximately 2.0 ns when stored within secretory vesicles.     

Imaging enzyme activity with polarization-sensitive confocal fluorescence microscopy

We describe a technique for imaging enzyme activity through steady‐state fluorescence anisotropy measurements on a per‐pixel basis with a confocal microscope. With this method, enzyme activity is reported by changes in the fluorescence anisotropy of a fluorescently labelled substrate. Enzymatic cleavage of the substrate yields smaller labelled fragments that tumble more readily than the intact substrate and therefore yield a lower anisotropy. Anisotropy is recovered to an accuracy of 7% or better on and off the optical axis to depths of 210 µm using objective numerical apertures as high as 0.75. Enzyme imaging experiments were performed with Bodipy‐FL‐labelled bovine serum albumin (BSA) attached to sepharose beads as a substrate for trypsin and proteinase K. Anisotropy images acquired up to 1 h after enzyme addition revealed more rapid digestion of BSA with proteinase K than with trypsin, but in both cases anisotropy decreased by at least five‐fold. Fluorescence lifetime and time‐resolved anisotropy decay measurements were made on the construct in fluid solution to reveal the effects of enzyme activity. The Bodipy‐FL lifetime increased from 1.34 ns for the construct without enzyme to 5.98 ns after 1 h in the presence of proteinase K. Anisotropy decays yielded average rotational correlation times of 1.13 ns before enzymatic action and 0.27 ns after enzymatic action, consistent with the presence of smaller Bodipy‐containing protein fragments. These results suggest wide applicability of the technique in biological systems when used in conjunction with appropriately designed constructs.     

A switched supply tunable red-green-blue light emitting diode driver

This paper presents a new switched supply tunable red-green-blue (RGB) light emitting diode (LED) driver. The RGB LEDs act not only as light emitting devices but also as rectifying diodes in the presented driver circuit. The RGB LED color control is realized by controlling the switched supply voltage amplitude, frequency, and duty cycle. The driver efficiency is high since the only loss in the driver circuit is the switch and can be further reduced by direct alternate current supply.     

A light-emitting diode light standard for photo- and videomicroscopy

A light calibration system consisting of a compact light-emitting diode (LED) source with feedback control of intensity is described. The source is positioned in the focal plane of the microscope objective and produces flat-field illumination of up to 31 μW. The source can be easily used to determine the performance of microscope optics and camera response. It can also be used as a standard light source for calibration of experimental systems. Selectable light intensities are produced by controlling the LED input power via a feedback circuit consisting of a photodiode that detects output light intensity. Spectral coverage extends between 550 and 670nm using green, yellow and red LEDs mounted side by side, which are selected individually. The LED chips are encapsulated in plastic diffusers which homogenize the light, and a flat field of illumination is obtained through a thin 1-mm-diameter aperture positioned directly over each chip. Provision is made for insertion of Ronchi rulings over the aperture to enable measurements of contrast modulation in a uniform field. The light may be pulse-modulated to assess camera response times and the device can be synchronized with video frames. Narrow bandpass interference filters can be placed between the objective lens and the LED source to produce monochromatic light without affecting the spacing of controlled light intensities since emission spectra do not shift appreciably over the range of LED powers chosen in this design. Results of tests using controlled light intensity and uniform illumination are presented.     

Navigating transdermal diffusion with multiphoton fluorescence lifetime imaging

We demonstrate the potential of fluorescence lifetime imaging by time-correlated single-photon counting as a method for monitoring the transdermal diffusion pathway and diffusion rate of pharmaceuticals in human skin. The current application relies on observing subtle changes in the fluorescence lifetime of the intrinsic fluorophores present in the intracellular region between corneocytes of the stratum corneum. We have comprehensively characterized the measured fluorescence lifetimes from intracorneocyte junctions in three skin section types (dermatomed skin, epidermal membranes and stratum corneum) revealing statistically significant differences of the short lifetime component between each of the types, which we attribute to the sample preparation and imaging method. We show using epidermal membrane sections that application of a drug/solvent formulation consisting of ethinyl estradiol and spectroscopic grade ethanol to the surface gives rise to a slight but statistically significant shortening of the fluorescence lifetime of the long-lived emitting species present in the sample, from approximately 2.8 ns to 2.5 ns. The method may be useful for future studies where the kinetics and pathways of a variety of applied formulations could be investigated.     

Graphical representation and multicomponent analysis of single-frequency fluorescence lifetime imaging microscopy data

Graphical representation of fluorescence lifetime imaging microscopy data demonstrates that a mixture of two components with single exponential decays can be resolved by single frequency measurements. We derive a method based on linear fitting that allows the calculation of the fluorescence lifetimes of the two components. We show that introduction of proper error‐weighting results in a non‐linear method that is mathematically identical to a global analysis algorithm that was recently derived. The graphical approach was applied to cellular data obtained from a lifetime‐based phosphorylation assay for the epidermal growth factor receptor and yielded results similar to those obtained by a global analysis algorithm.     

A new switched-capacitor frequency modulated driver for light emitting diodes

A new type of drivers for light emitting diodes (LEDs) is introduced based on the switched-capacitor frequency modulation. In contrast to conventional constant dc current drivers, the current pulse is provided by this new switched-capacitor LED driver. In the present driver, the charging capacitor is charged and discharged through a LED and the current flow direction is controlled by a metal oxide semiconductor switch. The input current (and thus the LED brightness) is proportional to the switch clock frequency at relatively low frequencies and becomes saturated at relatively high frequencies. This new driver circuit is simple and robust and maintains high efficiency for a wide range of input powers. In addition, the dimming control is easily realized by modulating clock frequency. Finally, this LED driver consumes no dc current and thus provides inherent protection to LED in standby mode.     

Two-photon fluorescence lifetime imaging microscopy of macrophage-mediated antigen processing

Two-photon fluorescence lifetime imaging microscopy was used noninvasively to monitor a fluorescent antigen during macrophage-mediated endocytosis, intracellular vacuolar encapsulation, and protease-dependent processing. Fluorescein-conjugated bovine serum albumin (FITC–BSA) served as the soluble exogenous antigen. As a relatively nonfluorescent probe in the native state, the antigen was designed to reflect sequential intracellular antigen processing events through time-dependent changes in fluorescence properties. Using two-photon lifetime imaging microscopy, antigen processing events were monitored continuously for several hours. During this time, the initial fluorescein fluorescence lifetime of 0.5 ns increased to α 3.0 ns. Control experiments using fluorescein conjugated poly- l -lysine and poly- d -lysine demonstrated that the increase in fluorescence parameters observed with FITC–BSA were due to intracellular proteolysis since addition of the inert d -isomer did not promote an increase in fluorescence lifetime or intensity. Comparisons of intravacuolar and extracellular FITC–dextran concentration suggested active localization of dextran in the vacuoles by the macrophage. In addition, the kinetics of degradation observed using two-photon microscopy were similar to results obtained on the flow cytometer, thus validating the use of flow cytometry for future studies.