FLIM

(fluorescence-lifetime imaging microscopy)

Remark:

This article about FLIM applications for our products was inspired by inspired by Dr. Lisa Hirvonen's and Prof Klaus Suhling's paper in Frontiers in Physics, who kindly gave us permission to use the image material shown. We would like to give visitors to our website who are interested in such topics a little insight. We ourselves are primarily concerned with the data acquisition requirements of our customers and are not chemists, biologists, or physicians, nor do we build mass spectrometers or imaging systems. However, the following content is not to be understood as a scientific treatise but also reflects subjective impressions.

Fluorescence lifetime imaging microscopy (FLIM) is a very gentle imaging method of microscopy in which the sample is excited with a short light pulse and the exponential decay rate of the fluorescence is determined. The low intensity of the light pulse is one of the major advantages of the FLIM method, which can thus be counted among the minimally invasive screening methods. It is based on the measurement of the different lifetimes of the excited states of fluorescent molecules, which are typically only a few nanoseconds. Since the respective fluorescence lifetimes strongly depend on the conditions in the microenvironment of these molecules or on the proximity of other fluorescent molecules, both structural and functional information on these environmental conditions can be obtained using this imaging technique.

As the name suggests, fluorescence lifetime is the key evaluation parameter in FLIM measurements. This decay time describes how long a molecule remains in the excited state and emits photons. In addition, FLIM also allows conclusions to be drawn about the intensity (fluorescence intensity), which results from the number of photons emitted per unit time in a specific wavelength range. FLIM is a form of photon-counting imaging and is now a well-established imaging technique in which an image is composed of individual photon emissions whose position is recorded during the detection process. The result of fluorescence lifetime microscopy is thus images of the sample in which each image pixel represents the fluorescence lifetime with respect to its corresponding position.

The development of confocal laser scanning microscopes enabled FLIM based on TCSPC for the first time, opening up new applications in biological research. FLIM has since been increasingly used in many scientific fields, including biological and medical research in particular. As described later in this article, FLIM can now be used to visualize and monitor the cellular microenvironment, including the interaction between proteins in their natural habitat.

A wide variety of FLIM methods are currently undergoing a strong development process, which is constantly generating new innovations, especially with regard to data acquisition. Therefore, we limit ourselves here to an introduction to this technique and make no claim to being complete.


Applications

Besides applications in chemistry and materials research, FLIM is mainly used in molecular biophysics, i.e. for the study of biological processes and materials. Medical research in particular benefits from the ever-improving imaging capabilities at the molecular level. FLIM allows the spatial visualization of samples in terms of pH value, viscosity, hydrophobicity, and temperature. Conclusions can be drawn, for example, about the size of a protein, the binding of enzymes or substrates. FLIM can be used as an imaging technique in confocal microscopy, two-photon excitation microscopy, and multi-photon tomography. By combining FLIM with Förster resonance energy transfer (FLIM-FRET), it is also possible to image the folding state of proteins. To measure complex multi-exponential decays or to detect subtle changes in living samples, rapid acquisition of photon arrival timing is essential.

Before we go into more detail about the principle of the measurement method, let us first look at the result of a representative FLIM measurement.

FLIM-Measurement example: Targeted fluorescence lifetime probes reveal responsive organelle viscosity and membrane fluidity. A: Representative fluorescence intensity and FLIM images of control and nystatin-treated HeLa cells. The lifetime range is 2100-3500 ps. B: Representative average fluorescence lifetime histograms from images A-B, control in green, and nystatin-treated in red. C) Box plot showing the average lifetimes of cells in the control condition (n = 140, 1μM FMR-1) and nystatin-treated condition (n = 33, 1μM FMR-1). The average control viscosity was found to be 274.7 ± 54.4 cP, and the average nystatin-treated viscosity was found to be 367 ± 78.1 cP. The difference is statistically significant (p < 0.01.). D) Representative fluorescence intensity decays from a control cell (in green) and a cell treated with nystatin (in red), and a scaled instrument response function (IRF, in black). E) Calibration plot for FMR-1 based on lifetimes in methanol-glycerol fractions. F) Control data for polarity, ionic strength, and temperature plotted on top of the calibration data, in the relevant region of the plot.

Principles


To understand the principles of FLIM, it is helpful to gain an understanding of fluorescence. Fluorescence occurs when certain materials (fluorophores) are illuminated with light of a certain wavelength and re-emit some of the light. The effect refers to the ability of some atoms and molecules to absorb light at a particular excitation wavelength and then to emit it again in a short-lived emission of light of a longer wavelength. This occurs because a fluorophore excited by a photon falls back to its ground state with a certain probability (based on decay rates) via a number of different (radiative and/or non-radiative) decay pathways. One of these pathways is the spontaneous emission of a photon - called fluorescence. In other words, spontaneous photon emission (emission of a fluorescent photon) occurs when an excited system transitions to a lower energy state. Due to the splitting of the emitted energy into heat and light, the emitted fluorescence light is always longer wavelength, i.e. with lower energy, compared to the excitation light.

FLIM aims to detect these emitted photons both spatially and in a time or frequency-based domain, allowing two- or even three-dimensional imaging. In biological imaging, especially when examining living cells, the amount of light introduced in the process is kept as low as possible in order to preserve the sample. Therefore, both, the quantum efficiency of the detectors and the fluorescence yield are of crucial importance. The latter depends not least on the fluorophores used.


A prerequisite for fluorescence measurements, e.g. on a cell sample, is its coloration. This can be done by treating the whole cell with a fluorescent dye (labeling). Alternatively, a special dye is introduced into the cell, which specifically docks to certain structures. In addition, it is possible to introduce a specific DNA vector that causes the cell to synthesize a fluorescent protein (transient transfection).


Detectors and downstream electronics for data acquisition are used to record the fluorescence decays, which allows two- or three-dimensional imaging. For this purpose, the detectors should have the highest possible gain factor and the shortest possible response times. Nevertheless, when selecting a detector for FLIM, there is no simple answer to the question of the „best“ choice. As explained in the paragraphs below, this depends on the chosen measurement method, the setup of the microscope used, and ultimately on the sample itself. The individual quality criteria lie in the different characteristics of the detectors in terms of time-of-flight dispersion, photon yield, signal-to-noise ratio, ease of use, and instrument durability.

Frequency domain method vs time domain method


Fluorescence lifetime itself can be measured either in the time domain (pulsed excitation and measurement of the temporal fluorescence decay) or in the frequency domain (steady-state, using a phase modulation technique). Both measurements, in the time domain, as well as those in the frequency domain, offer unique challenges in various FLIM scenarios, including imaging with a low photon budget, imaging within a high dynamic range, or imaging with high time resolution.


In frequency-domain methods, imaging is performed using intensity-modulated excitation of the sample. This can be implemented using ultrafast laser sources with high power density, e.g. modern pulsed diode lasers or supercontinuum sources. The modulation is performed at a high frequency of up to 500 MHz, e.g. in the form of a sine wave. Due to the decay time, the fluorescence is phase shifted with reference to the excitation photon. An image of the fluorescence lifetimes is then reconstructed from this phase shift. For detection, this method uses image sensors such as avalanche photodiode arrays or ICCD cameras. In this process, the time during which the detectors are sensitive is precisely controlled. The FLIM image is appropriately colored to different values of phase shift. A disadvantage of the frequency domain method is that the modulation must be very precise to minimize fluctuations between modulation and the detection of the phase-shifted decay times.


Time-domain measurements


To perform time-correlated single photon counting (TCSPC), the sample is excited by a short excitation pulse and the arrival times of photons are accumulated in a histogram. For this purpose, the excitation intensity is reduced to such an extent that, if possible, only one photon is detected per excitation pulse. If several fluorescent species are present, all species are usually summarized in a single histogram. This arrival time histogram is the fluorescence decay which can be fitted to extract the fluorescence lifetime.

Time-domain lifetime measurement techniques can be generally divided into single-point measurements and wide-field measurements. In single-point measurements, the excitation source scans over the sample and determines the corresponding fluorescence lifetime values for each point in a consecutive sequence. This measurement method can produce both 2D and 3D images of the sample. Wide-field FLIM, on the other hand, excites the entire sample at once and the photon arrival time and position is determined individually for each pixel position of the detector. Modern FLIM instruments that are based on time-domain measurements provide in many cases a combination of both measurement methods so that users can choose between the fast wide-field FLIM and the particularly accurate single-point detection for a specific measurement.


Single-point Measurements (scanning FLIM, Confocal Laser Scanning Microscope FLIM, CLSM)


Probably the most widely used and highly developed method for determining fluorescence lifetime is scanning FLIM measurement via time-correlated single photon counting. In this method, released photons are counted after periodic excitation with short light pulses in the picosecond range. Here, the light pulses are in any case significantly shorter than typical fluorescence lifetimes, with the lifetime of a typical fluorophore usually being less than 10 nanoseconds. The fluorescence lifetime is then calculated from the time-dependent, exponential decrease in fluorescence intensity. In the simplest case, the fluorescence decay follows a mono-exponential function, but mixtures of fluorescent species that exhibit multi-exponential decay functions within a fluorescent system are often studied. In all cases, a histogram is generated from many such individual measurements for each image pixel, which typically represents the fluorescence lifetime by its color value.

This image shows a typical fluorescence decay curve recorded for a single pixel. The shape of the curve is obtained by histogramming the emission times determined by TCSPC, thus showing how often each emission time occurred. The resolution of such TCSPC measurements is limited by various factors, the sum of which is referred to as the instrument response function (IRF).
This image shows a typical fluorescence decay curve recorded for a single pixel. The shape of the curve is obtained by histogramming the emission times determined by TCSPC, thus showing how often each emission time occurred. The resolution of such TCSPC measurements is limited by various factors, the sum of which is referred to as the instrument response function (IRF).


In confocal laser scanning microscopes, for FLIM, the specimen is scanned incrementally by a focused low-power laser beam to excite fluorescence pixel by pixel. The respective photon arrival time is measured separately for each pixel using a fast detector and fast data acquisition electronics. As a result, precise photon detection times are available for each pixel. For high-quality imaging, a histogram is created for each pixel of the image which allows exact analysis of the decay times. The time values obtained in this way are then transferred into a color spectrum so that ultimately the color of each pixel represents a corresponding value for the fluorescence lifetime.

An advantage of this method is that it is independent of variations in excitation levels.

Highly simplified scheme of a scanning FLIM measurement. A: The sample is prepared on a x-y-z stage (here featuring 3 focal planes on the z-axis) and is scanned point by point with an expanded excitation beam. Please note the detector pinhole aperture. B: On the software side, the respective fluorescence lifetimes are converted into a color scheme by means of TCSPC. C: The time values obtained in this way are then transferred into a color spectrum so that ultimately the color of each pixel represents a corresponding value for the fluorescence lifetime at it's 3 dimensional position.
Highly simplified scheme of a scanning FLIM measurement. A: The sample is prepared on a x-y-z stage (here featuring 3 focal planes on the z-axis) and is scanned point by point with an expanded excitation beam. Please note the detector pinhole aperture. B: On the software side, the respective fluorescence lifetimes are converted into a color scheme by means of TCSPC. C: The time values obtained in this way are then transferred into a color spectrum so that ultimately the color of each pixel represents a corresponding value for the fluorescence lifetime at it's 3 dimensional position.

The point-by-point scanning of the specimen can be obtained, for instance, by deflecting the laser beam horizontally and vertically using a pair of galvanometric scanners (XY scanners): The beam is thereby focused on the excitation point in the specimen by means of an objective. If a three-dimensional image is to be acquired, images are created in succession in different focal planes for this purpose. Either the specimen or the objective can be moved in height.

The confocal pinhole


The decisive feature of a CLSM compared to a conventional light microscope is a confocal aperture (pinhole). Only light components that have passed through this pinhole are detected by the detector. The diameter of the pinhole is variable and determines the extent to which the emitted light from object points outside the focal plane is masked out. The masked regions thus do not generate fluorescence, consequently, the number of photons reaching the detector depends not only on the fluorescence yield but also on the diameter of the confocal pinhole. The confocal pinhole should always have a certain minimum diameter to allow a minimum amount of radiation to reach the detector. The chosen diameter and the resulting photon yield are usually a compromise between noise ( which depends on the intensity of the excitation light), spatial resolution, and depth discrimination.


Note: There is also an excitation pinhole, but in modern CLSMs, this is replaced by coupling the excitation laser via optical fibers. The narrow light-carrying fiber core has similar optical limiting properties as the pinhole aperture.


In scanning FLIM, the pulse rate of the excitation light is basically adapted to the fluorescence lifetime to be measured. This is because the sample needs time to decay to the ground state before a new excitation pulse arrives. As a result the repetition rate and the time window need to fit the decay.

The detection rate is usually limited to about 1% of the excitation repetition rate at the brightest pixel to avoid pile-up (photon stacking). This applies to any repetition rate and time window length, and depends on the intensity (count rate) of the fluorescence.



For CLSM, mainly PMTs, SPADs, or hybrid detectors are used because the requirements for their sensitivity, the reproducibility of the time measurements, the quantum efficiency, and the photon counting capability are particularly high. High time resolution capability, low dark counts, and a high signal-to-noise ratio are as much required from the detectors as they are from the electronics in the data acquisition stage.


CLSM offers several advantages over wide-field FLIM. In addition to the possibility of three-dimensional imaging, the measurement method is always favorable when, for example, particularly thick samples (such as biological cells in their tissue composite) are examined. CLSM provides optical sectioning, reduced background blur, and reduced photobleaching outside the focal point. It is also relatively easy to combine with spectral detection or with polarization. The out-of-focus suppression methods of confocal laser scanning microscopes provide higher contrast and spatial resolution than wide-field FLIM systems.


Wide-Field FLIM

Wide-field fluorescence lifetime measurements are particularly useful for 2D lifetime contrast imaging and determination of sample homogeneity. In this measurement method, the image for all object locations is created simultaneously by illuminating the entire sample with a collimated excitation light beam for exciting fluorescence. In this process, photons from the entire field of view are collected in parallel by position-imaging detectors to calculate the decay times. Wide-field FLIM is therefore popular for rapid imaging of large sample areas.

Detectors used in wide-field FLIM are typically micro-channel plate detectors (MCP). In rare cases, detectors based on single-photon avalanche diode (SPAD) arrays, and ICCD cameras or superconducting detector technology are also used.

Wide-field FLIM has the advantage of higher frame rates compared to CLSM, making it suitable for coarse, highly automated sample investigations. However, the sensitivity and signal-to-noise ratio of the detectors used in CLSM is not matched by most camera-based detectors. This results in lower axial resolution. In addition, as a result of light diffusion, each camera pixel simultaneously detects scattered light from all other pixels in the illuminated area. This is why the temporal and spatial coordinates are mixed in wide-field FLIM.

Read more about wide-field FLIM in this review.

Highly simplified scheme of a wide-field FLIM measurement. A: Fluorescence emitted by the labeled specimen (red) is focused on the detector by the same objective that is used for the excitation light. The beam path from excitation light (blue) is reflected by a dichroic mirror onto the sample. The photons emitted by fluorescence (red) pass through the same dichroic mirror and their arrival times are registered by the position-sensitive detector. B: On the software side, the respective fluorescence lifetimes are converted into a color scheme.
Highly simplified scheme of a wide-field FLIM measurement. A: Fluorescence emitted by the labeled specimen (red) is focused on the detector by the same objective that is used for the excitation light. The beam path from excitation light (blue) is reflected by a dichroic mirror onto the sample. The photons emitted by fluorescence (red) pass through the same dichroic mirror and their arrival times are registered by the position-sensitive detector. B: On the software side, the respective fluorescence lifetimes are converted into a color scheme.

Data acquisition & quality of imaging


Since cronologic products are of particular interest for high-resolution FLIM single-point measurements in the time domain, we would like to focus on TCSPC data acquisition at this point. The quality of imaging in such single-point measurements depends on a variety of factors. Figuratively speaking, the weakest link in the chain determines the quality of the result. We will describe some of these „chain links“ in the following.


The specific requirements for detectors and data acquisition in FLIM applications depend on different factors depending on the area of application. Since the display on the screen is done by the pixel-wise output of the image data from a digital image memory, the pixel size is of course one of the most important factors. In confocal LSM, this is determined by the interaction of the confocal aperture (pinhole) and the digitization of the object information adapted to it. The selected zoom factor affects both the overall magnification and the resolution characteristics of the system.


The dosage and timing of the excitation light, including the reduction of laser noise or shot noise, is also a crucial factor. The properties of the excitation light itself are not only adapted to the sample but must also be matched to the timing resolution on the detector side as well as to the electronics and the transmission rates during data acquisition.


As mentioned above, for each image pixel, the arrival time of a photon at the detector is recorded with respect to a particular reference. Typically, this reference is the excitation laser pulse. A collection of the measured time differences follows to build up a histogram. In this process, a „time bin“ is usually assigned to each photon. The size of this time bin ultimately determines how clearly the given image represents the decay of fluorescence in that pixel. Therefore, in particular, the type of A/D conversion used is also an intrinsic factor to the quality of the sampled image. Consequently, the fast and efficient data acquisition concepts of modern TDCs (Time to Digital Converters), which convert the arrivals of the signals directly into digital timestamps, fully exploit their advantages in terms of short dead times and low noise. Plus, their high time resolution can be used to measure the lifetimes of particularly fast decay components.


Supplement: FRET Applications


In the course of our little insight into fluorescence lifetime microscopy, we would not like to miss mentioning an important application of this technique: The measurement of a physical process of energy transfer called Förster Resonance Energy Transfer (FRET).


In protein biochemistry, and thereby also in cell biology and pharmacology, Förster resonance energy transfer is analyzed to detect protein-protein interactions and the interaction of proteins with other substances in real-time with time resolution in the millisecond range and higher. For example, protein complexes, DNA-binding proteins, and enzyme-substrate interactions can be studied. The principle of these measurements is based on labeling molecules or molecular components with two different fluorescent dyes, each of which binds to different structures.


In this process, energy is transferred from one excited dye (donor) to a second dye (acceptor) without radiation, i.e. without emitting photons. The excitation light excites only electrons of the donor, which, when they fall back to their original level, do not release the energy as fluorescence if possible, but transfer it primarily to the acceptor. This energy transfer, on the other hand, excites electrons of the acceptor, whereupon the latter now emits the absorbed energy as fluorescence.


However, the energy transfer between donor and acceptor only takes place when the distance between donor and acceptor is less than 5 nm. If an interaction of the labeled molecules leads to a spatial convergence below this distance, this proximity results in fluorescence on the acceptor side. This is assuming that the donor and acceptor have different absorption spectra and that the emission spectrum of the donor overlaps with the absorption spectrum of the acceptor. FLIM-based FRET measurements are relatively insensitive to the concentration of fluorophores and detect a shortened fluorescence decay of the donor fluorophore, because FRET provides a non-radiative decay path which shortens the fluorescence lifetime.



Also recommended:

- Special Issue on FLIM - From Fundamentals to Applications in the journal “Methods and Applications in Fluorescence”.

- FLIM review - Basic concepts and some recent developments on Science Direct

Image Source

FLIM-Measurement example: Taken from PLOS ONE Collection by courtesy of Prof. Dr. Klaus Suhling