The inherent tendency of fluorescent molecules to photobleach has been turned into a very powerful tool for the study of molecular trafficking. By targeting a region of interest for laser-mediated photobleaching, biologists can study how fast the molecules go in and out of the bleached zone and deduce molecular turnover rate from the fluorescence recovery. If a certain type of molecules turns over very fast within the bleached zone, the fluorescence recovery will be rapid. Conversely, if those molecules have a very strong association constant, then the fluorescence will recover very slowly. This technique, fluorescence recovery after photobleaching (FRAP) is now very commonly used due to the wide availability of galvo-controlled lasers in most commercially available confocal microscopes. However, it is also common to see biologists misinterpret the photobleaching data.
Important information that can be gleaned from a fluorescence recovery curve.
Figure 1. A typical FRAP plot. Various information, including the mobile and immobile fractions, recovery rate and the half time of equilibrium (t1/2) can be calculated from the plot as shown.
Following the acquisition of several pre-bleach images to help establish the rate of photobleachign due to regular imaging, a single, strong laser pulse is used to photobleach the region of interest. Imaging resumes immediately following the photobleaching event to collect data points for the fluorescence recovery curve. There are several pieces of information that can be readily extracted from a standard fluorescence recovery curve. The immobile fraction, which is the molecular pool that does not undergo exchange within the FRAP zone, can be obtained by subtracting the post-bleached fluorescent intensity from the pre-bleached intensity. The mobile fraction will then be the remaining fraction. The half time of equilibrium, which is the time it takes for the fluorescence intensity to recover to half of the plateau level, is usually indicated as t1/2. This is the number many understood to be the rate of recovery. But is that an accurate assumption?
Why relying on the t1/2 of the recovery curve is a bad idea.
Let’s consider the fluorescence recovery curves in Figure 2. The blue recovery curve reaches a much higher plateau than the green curve, indicative of significantly larger mobile fraction. However, a quick estimation will show that both recovery curves have the exact same t1/2, ln(2)/τ, which is 3.5 seconds.
Figure 2. Two vastly different fluorescence recovery curves. Or are they the same? Both blue and green recovery curves produce the same t1/2, yet their actual recovery rates are vastly different from one another. Why? And how is this biologically significant?
If proper curve fitting is performed on the two curves, one will obtain the recovery rates of 0.9 (blue curve) and 0.5 (green curve). These are the real initial rates of the two fluorescence recovery events. The initial rate is the most important, accurate and informative measurement of fluorescence recovery. It is accurate because it is concentration-independent. Even if your photobleaching of the FRAP zone is incomplete, it will not affect the initial rate. It is informative because it truly measures how fast the unbleached fluorescent molecules moves into the FRAP zone, as soon as the photobleaching event is over. Unlike t1/2, its true value will not be skewed by the difference in mobile fractions. Unfortunately, this simple and yet powerful numerical rate constant is not usually provided by photobleaching software programs.
In addition, it is important to correct for the inherent baseline photobleaching rate that is caused by the mere act of imaging the molecules. As shown in Figure 3, the green line indicates the linear deterioration of fluorescence signals due to imaging, data that can be easily teased out by imaging a neighboring cell. The black line shows the actual recorded fluorescent intensity recovery after photobleaching. The blue curve is the normalized curve, taking into account the inherent photobleaching rate. The Advanced Imaging Center has created a free program to help you calculate the rate constant from your FRAP data.
Figure 3. Inherent imaging-related photobleaching of the fluorophore will affect the recovery rate. Most fluorophores will bleach during the course of imaging as traced by the green line. This inherent photobleaching event will skew the fluorescence recovery rate if not corrected. To obtain the inherent bleaching rate, image a neighboring cell in the same field of view if possible. Another way is to capture a few timelapse images before introducing the strong photobleaching pulse to begin the FRAP experiment. The time-dependent fluorescent intensity can then be easily extracted from the fluorescent intensity plot.
What is the rate-limiting step in fluorescence recovery? Apparently, when one measures the fluorescence recovery rate within a FRAP zone, one measures the “On” rate – i.e., how fast the unbleached molecules come onto the bleached zone. Let’s call the “on” rate kon. However, the rate-limiting step is not how fast the unbleached molecules move in, but how fast the bleached molecules move out. Since fluorescence recovery cannot take place if the bleached molecules do not free up the binding sites. So, even though the apparent value measured in a FRAP experiment is kon, the real rate being deduced is the “off” rate, or koff. In other words, FRAP almost always measures the dissociation constant of the bleached molecules from the FRAP zone, indirectly through the measurement of fluorescence recovery. So a slow recovery rate implicates a stable complex with very little molecular exchange (low koff), and not necessarily the diffusive molecules having a slow rate in associating with the complex.
If you are interested, here is a Matlab GUI for loading and analyzing data using Fluorescence Recovery After Photobleaching (FRAP) technique. The GUI will calculate recovery fraction(s) and half-time(s). It supports the following features: 1. Single or multi-component recovery curve fit; 2. Background photobleaching correction; 3. Export results to Excel.
Note: It is compiled as an .exe file to run on 64-bit Windows machines. If you would like the original Matlab code or exe compiled for other operating systems (i.e. Mac OS or Linux), contact Jesse Aaron here.