Quantitative Fluorescence

A very well known method in chemistry and physics is the relaxation method [1]. The signal which is measured – in our case the fluorescence intensity – is constant in equilibrium. After the perturbation – which occurs as a pulse or a jump of a parameter, the intensity will change and reach a new equilibrium. The kinetics of the change reflect details of the mechanism that rule the competing pathways. This scheme offers an elegant way to measure the mobility of fluorescently labeled molecules. If a fraction of the field is bleached by applying intense light (pulsed perturbation), the fluorescence will decrease. In case the molecules are not mobile, the bleached area will stay black. If the fluorescing molecules can move, either by diffusion or by active processes, the bleached area will recover fluorescence again. Therefore, this protocol is called Fluorescence Recovery after Photobleaching (FRAP). The underlying mechanisms are revealed by fitting model curves to the relaxation part of the signal. And the fraction of molecules that are not moving (immobile fraction) is derived from the ratio of the equilibrium signal levels [^typo3/2].

FRAP is commonly used to measure diffusion processes in cells. Besides the classical FRAP, derivatives are sometimes used: FLAP, which is Fluorescence Loss After Photobleaching. Here, bleaching occurs continuously in an area that has no intersection with the bleached area. iFRAP indicates a method where the whole field except a small region is bleached. More acronyms are found for bleach methods, but these have no practical relevance.

Technically similar, but based on a completely different mechanism, are activation experiments with fluorescent proteins. These proteins are fluorescent and can be switched on and off [3] by a different illumination color. Alternatively, the fluorescence emission spectrum changes [4] upon the switch illumination. The advantage is that one can detect the fluorescence against a black or differently colored background and follow the diffusion or active transport of the molecule in the cell or tissue (Fluorescent Protein Activation FPA).

These methods are not only applicable for mobility measurements, but also for kinetic measurements of concentration changes. Well known are ion indicators, e.g. for calcium. Moder biosensors, often based on FRET-FLIM, allow a wide variety of cellular parameters to be monitored, including metabolite concentrations and membrane potentials.

When a photon is emitted from a fluorescence molecule, the wavelength is stokes-shifted to the red, i.e. the energy is less than the excitation. If a second fluorescent molecule (Acceptor A) with an excitation energy in the range of the emission from the first molecule (Donor D) is in close contact, the energy is transferred with a probability that increases with the 6^th power of the distance without emission and absorption (radiation-free). This phenomenon was analyzed and first correctly described by Theodor Förster and is hence called Förster Resonance Energy Transfer (FRET) [5].

A direct application of FRET is the analysis of fluorescence emitted by the acceptor upon illumination solely of the donor (Sensitized Emission SE). Alternatively, the FRET efficiency is also measured by bleaching the acceptor and measuring the donor fluorescence before and after bleaching. If FRET occurred, the donor will shine more brightly after acceptor bleaching, as no energy is flowing to the acceptor and consequently all energy is converted into donor emission (Acceptor Bleaching AB) [6typo3/]. Such experiments are conducted to prove interactions of proteins by their relative distance.

The FRET phenomenon is also utilized in modern biosensors. These are (often complex) molecules that contain a donor and an acceptor. Upon binding or sensing the target, the molecule will undergo a conformational change that alters the relative position of the D and A  parts and consequently changes the FRET efficiency. These sensors are very sensitive due to the 1/r⁶ dependence of the Förster-Wechselwirkung.

Upon excitation by a photon of appropriate energy (color), the electronic system of a molecule or atom will assume the excited state. From that state, the system will relax down to the ground state by emitting a fluorescence photon. Obviously, the system must linger for some time in the excited state. This time is called the fluorescence lifetime. In an isolated molecule, the mean lifetime is ruled by quantum physics and is hence a property of the molecule. The actual lifetime is statistically distributed about the mean life span. Typical mean life spans range from 0.5 to 10 nanoseconds.

In reality, the fluorescent molecule interacts with other molecules in the microenvironment and with radiation. These parameters can cause the excited state to relax more quickl< [7typo3/]. A well known phenomenon is fluorescence quenching by dynamic interaction with other molecules. Here, the energy is lost as heat. A second pathway is the Förster -Wechselwirkung as described above. If the fluorescent molecule (donor) interacts with the acceptor, the mean fluorescence lifespan is decreased due to the fact that the FRET process is an additional competing pathway to the fluorescence emission pathway. Changes in fluorescence lifetime hence indicate FRET occurrence – and this is the measured parameter in FLIM-FRET experiments with biosensors [8typo3/]. Finally, the excited state may interact with photons, which causes further conversion to other states, or initiates return to the ground state by emitting an additional photon (stimulated emission [9typo3/]).

Measurement of the actual mean life span of fluorochromes is done by pulse excitation and subsequent measurement of the declining emission or by modulated excitation and measurement of the phase and amplitude modulation of the emission. The most exact and versatile measurement is time-correlated single-photon counting (TCSPC) which employs measurement of the time delay from excitation to emission of the first photon. The result is a distribution curve (usually an exponential decay), which allows the lifetime to be extracted by curve fitting. This method is applied to all pixels of an image, finally producing the fluorescence lifetime image (also called a "tau-map"). The values in those images are not grey intensities, but times.

Besides the advantage of sensing the environment, FLIM is independent of dye concentration and absorption artifacts, e.g. in deeper layers of the sample.

Confocal microscopes observe the light that is emitted from a tiny diffraction-limited spot. The measurement is the light intensity as a function of time. By scanning the spot over the sample and chopping the signal into short pieces, these bits can be assigned to and stored as pixels in a frame store. If triggered correctly, the frame store contains the intensity distribution as a two-dimensional function of space (image).

On the other hand, it is also possible to just analyze the signal and its variations – with or without moving the spot over the sample. This approach allows a variety of measurements. Firstly, the mean intensity will usually decrease due to bleaching phenomena. Also, triplet state kinetics are measurable by switching the illumination between different levels. Most commonly, not the mean signal, but the variations (fluctuations) of the signal are analyzed. If many molecules are in the observed sample, the fluctuations in relation to the signal will be very small and hence immeasurable. The tiny confocal focus has the beneficial feature of containing only a few molecules (depending on the optical parameters and on the dye concentration, of course). The number of molecules is small enough to detect variations that occur due to molecular blinking (molecules assume dark states and return back to excitable states) and alterations of the number due to diffusion (or any other transport mechanism) in and out of the focal volume [10].

The analysis of the variations by means of correlation methods allows measurement of dye concentrations (local variations in the sample) and transport dynamics. By fitting model systems to the measured correlation functions, classifications of the underlying molecular transport mechanisms are possible. Qualitative measurements are also possible – for example diffusion coefficients in various compartments.

A variety of derivative methods have been developed to separate specific correlations (FCCS) or improve the reliability and precision of the results (FLCS [11]).