Welcome to the web
page of  Peter Nagy

Click here to edit subtitle

FRET (Förster-type or fluorescence resonance energy transfer)

FRET has become one of the most widely used applications for measuring the clustering of proteins in biological systems. Since FRET is manifested in many measurable changes, a lot of different approaches have been established to analyze protein clustering based on FRET. If you are interested in the principles of FRET and you would like to get a brief introduction to many of the methods used to measure it, you can find resources below.
Book chapters

PDF version of a Powerpoint tutoria
l
A couple of my papers describing novel FRET-based approaches
Topic of paper
 Title of paper with link to PDF on this site
 Bibliographic specification with DOI-based links
A method for the determination of parameter alpha (describing how well an excited acceptor molecule can be detected in the FRET channel compared to an excited donor molecule in the donor channel)
A method based on the measurement of homo-FRET to analyze the size (i.e. the number of proteins/cluster) of protein homoassociations
A method based on bleaching of acceptor molecules by exciting the donor to measure their coclustered fraction
A method based on maximum likelihood estimation to evaluate FRET at low intensities and in the presence of outlier pixels
Description of a Matlab program for analysing intensity-based, microscopic FRET measurements
Improved method for measuring FRET at high excitation intensities in microscopy

FRET at high excitation intensities

FRET measurements in microscopy are often carried out at excitation intensities leading to fluorophore saturation. Fluorophore saturation is a phenomenon when most fluorophores are in the excited state breaking down the following "trivial" relationships:

  • Fluorescence intensity will not be linearly proportional to excitation intensity any more
  • FRET-induced quenching of the donor will not follow the following simple equation: F(DA)=F(D)*(1-E), where F(DA) and F(D) are the fluorescence intensities of the donor in the presence and absence of the acceptor, respectively, and E is the FRET efficiency.

These effects invalidate most approaches used for intensity-based FRET calculations.This is a serious issue since fluorophores are often saturated at commonly used excitation powers in confocal microscopy. We developed an approach for determining the FRET efficiency under such conditions. The method is described 

The problem arising on the donor side in the absence of the acceptor is briefly summarized in the scheme above. Due to fluorophore saturation the fluorescence intensity is not proportional to excitation intensity. This effect is quantitated by Dsat expressing the fraction of fluorophores in the excited state. Dsat can be converted to a form in which fluorophore saturation is compared to the saturation of enzymes expressed by the Michaelis-Menten constant.

This effect is quite considerable at excitation intensities commonly used in confocal microscopy. A photon flux of 1E24 1/(cm2 s), is achieved at 15% laser power on our Zeiss LSM810 microscope.

If a donor-acceptor system in considered, donor saturation leads to an apparent decrease in the FRET efficiency calculated according to the conventional equation set.

The magnitude of this effect can be very significant. AlexaFluor488 was assumed to be a donor characterized by five different theoretical FRET values and the apparent FRET efficiency was calculated according to the conventional donor quenching equation: 
E = 1 - Ida / Id
These apparent FRET values were normalized to the theoretical FRET values.

The method, briefly described above, has been implemented in rFRET, a Matlab-based application introduced below.

rFRET: a Matlab application to evaluate intensity-based (ratiometric) FRET measurements

One of the easiest ways to measure FRET is a ratiometric or intensity-based experiment in which a sample labeled with donor and acceptor fluorophores is measured in three fluorescence channels:

Channel
 Excitation
 Detection
Donor channel
 donor absorption wavelength
 donor emission range
 FRET channel
 donor absorption wavelength
 acceptor emission range
 Acceptor channel
 acceptor absorption wavelength
 acceptor emission range

I have written a Matlab application, rFRET, which evaluates such an experiment using different approaches:

  • pixel-by-pixel
  • using summed or mean intensities
  • regression-based approaches
  • maximum likelihood estimation (MLE, see the last paper in the table above)
Syntax: rfret
Registration: a machine specific code is shown when running the program for the first time which you have to send to [email protected]. I will send you a machine-specific unlock code free of charge.
Installation: download rfret.zip, unzip it into a folder on your machine and type rfret at the Matlab command prompt after changing the current folder to the one you installed rFRET into. Alternatively, add the folder to the search path of Matlab.
Help: the Help file also available from within the application.
Sample images: a set of images is available for practicing using the program. The ZIP file contains four folders:
  • "A546-pertuzumab" with four images recorded of a sample labeled with AlexaFluor546-pertuzumab :
  1. A546-pertuzumab_D.tif: donor channel
  2. A546-pertuzumab_T.tif: FRET (“transfer”) channel
  3. A546-pertuzumab_A.tif: acceptor channel
  4. A546-pertuzumab_M.tif: membrane mask generated by manually seeded watershed segmentation

  • "A647-trastuzumab_DOL 1.0" with four images recorded of a sample labeled with AlexaFluor647-trastuzumab (degree of labeling of the antibody was 1):
  1. A647-trastuzumab_D.tif: donor channel
  2. A647-trastuzumab_T.tif: FRET channel
  3. A647-trastuzumab_A.tif: acceptor channel
  4. A647-trastuzumab_M.tif: membrane mask generated by manually seeded watershed segmentation


  • "A546-trastuzumab_DOL 1.1" with four images recorded of a sample labeled with AlexaFluor546-trastuzumab (degree of labeling of the antibody was 1.1):

  1. A546-trastuzumab_D.tif: donor channel
  2. A546-trastuzumab_T.tif: FRET channel
  3. A546-trastuzumab_A.tif: acceptor channel
  4. A546-trastuzumab_M.tif: membrane mask generated by manually seeded watershed segmentation


  • "A546-pertuzumab+A647-trastuzumab" with four images recorded of a sample labeled with AlexaFluor546-pertuzumab and AlexaFluor647-trastuzumab:

  1. A546-pertuzumab+A647-trastuzumab_D.tif: donor channel
  2. A546-pertuzumab+A647-trastuzumab_T.tif: FRET channel
  3. A546-pertuzumab+A647-trastuzumab_A.tif: acceptor channel
  4. A546-pertuzumab+A647-trastuzumab_M.tif: membrane mask generated by manually seeded watershed segmentation.


Updating: the program will automatically check for and download upgrades.

The effect of fluorophore saturation on intensity-based FRET calculations

Fluorescence emitted by fluorophores does not increase linearly with excitation intensity. This phenomenon is called fluorophore saturation. The figure below shows the effect of fluorophore saturation on the emitted fluorescence. Since the photon fluxes shown on the horizontal scale are achieved in regular confocal microscopy, fluorophore saturation takes place under conditions commonly used in research.

While fluorophore saturation itself has been recognized for quite some time, its effect on intensity-based FRET calculations is overlooked. The figure below shows the effect of donor saturation on the FRET efficiency calculated from donor quenching. It can be seen that the apparent FRET efficiency declines significantly as donor saturation kicks in.

The consequences of fluorophore saturation on intensity-based FRET calculations are taken into consideration in the rFRET Matlab applications (see above). In addition, an Excel file is also made available in which most of the formulas have been implemented.


r0ForFret: calculating the overlap integral and R0 for a donor-acceptor pair

The Matlab application, r0ForFret, calculates the overlap integral between the emission spectrum of the donor and the absorption spectrum of the acceptor as well as the R0 value, which is the distance at which the FRET efficiency is 50% between a certain donor-acceptor pair.
Program can be used in two different modes:
  • GUI mode:
Syntax: just type "r0ForFret" at the Matlab command prompt.
Help: the Help file also available from the application.
  • Command prompt mode:
Syntax: [r0,j]=r0ForFret(options)
Help:
The following options are available:

  • 'fd' followed by a variable (required): a variable containing the emission spectrum of the donor in a 1-D array. The spectrum does not have to be normalized, because the program performs normalization.
  • 'absa' followed by a variable (required): a variable containing the absorption spectrum of the acceptor in a 1-D array. If the spectrum contains normalized absorption values, the parameters 'normalized' and 'epsa' must be present in the argument list.
  • 'wavelengthd' followed by a variable (required): a variable containing the wavelength scale of the donor emission spectrum. The unit of the wavelength must be nm.
  • 'wavelengtha' followed by a variable (required): a variable containing the wavelength scale of the acceptor absortpion spectrum. The unit of the wavelength must be nm.
  • 'n' followed by a number (required): the index of refraction of the medium.
  • 'qd' followed by a number (required): the quantum yield of the donor in the absence of the acceptor.
  • 'kappa2' followed by a number (required): the orientation factor.
  • 'normalized': indicates that the absorption spectrum of the acceptor does not contain molar absorption coefficients, but normalized absorption values.
  • 'epsa' followed by two numbers: the first is the wavelength at which the molar absorption coefficient of the acceptor is given by the second number.

E.g.:
Normalized absorption spectrum of the donor:
 [r0,j]=r0ForFret('n',1.4,'kappa2',2/3,'qd',0.8,'normalized','epsa',556,112000,'absa',abs,'fd',emiss,'wavelengthd',wld,'wavelengtha',wla);

Molar absorption coefficients are available in the absorption spectrum of the donor:
[r0,j]=r0ForFret('n',1.4,'kappa2',2/3,'qd',0.8,'absa',abs,'fd',emiss,'wavelengthd',wld,'wavelengtha',wla);

Installation: download the r0ForFret.zip file, unzip it to a folder on your computer. This program is also available from the Matlab File Exchange.