Re: FRET Acceptor Photo-Bleaching vs Sensitized Emission

[Alessandro Esposito]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Kees,
    I am not surprized at all about differences in values measured by
acceptor photobleaching or sensitized emission FRET. Both methods are
quantitative and, gererally speaking, work well. However, they suffer of
different biases and experimental issues and, therefore, different
results are to be expected.

Indeed, I would advice not to over-interpret FRET efficiency as an
absolute number, at least in the average experiment in biology.

apbFRET is the simplest method, but photoconversion, donor
photobleaching and incomplete acceptor photobleaching will all concur
to bias your results. You can standardize as much as possible your
protocols and compensate for most issues and apbFRET will work.
However, if you have the possibility to use seFRET, perhaps it is a
better way if you experience photoconversion of YFP.

There are many ways to do seFRET. If your stoichiometry is 1:1 you can
even just to ratiometric imaging and pick differences among different
biological samples. Otherwise, you can do seFRET with all controls
needed, but I would advise to use also a construct exhibiting known
FRET efficiency. Also, do several control images on a per session basis.

Even in this case, I would recomand to not really over interpret the FRET
efficiency value you get. If your "biological controls" make sense, then
you can start to tease out differences among different samples.

One line of caution about intensity based methods and modern confocal
microscopes. Sometimes, modern confocal microscopes do not exhibit an
ideal linearity between photon arrived to the detector and signal
acquired. Refrain yourself to use all the available dynamic range,
perhaps sacrifying the top 20% of it (for instance, with 1024 gray levels,
use up to 800, do not be worried only if you measure 1024).

Obviously, there are other methodologies around, but apbFRET or
seFRET should be perfectly ok after you found standardized imaging
conditions and sample preparation that provide reproducible results.

Kind regards,

Alessandro
www.mrc-ccu.cam.ac.uk

[Tim Feinstein-2]

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi all,

IMO most people do not appreciate how hard/complicated it is to make sense out of a single FRET efficiency number drawn from a single (repeated) experiment and derived with a plugin/simple equation.  Too many variables and unknowns.  FRET experiments should have internal controls to calibrate each number to, for example, transfer with a non-interacting negative control as well as a known positive control.  If both proteins are membrane bound, for example, a YFP with the right membrane topology and one transmembrane domain or myristilated/palmitoylated and expressed at comparable levels makes a useful and necessary basis for comparison.  A classic obligate dimer like GABAb1/2 expressed at comparable levels can serve as a positive control.  

Further, any FRET pair in the same compartment will transfer some energy, with the efficiency depending on their collision rate.  All FRET between two molecules should thus be controlled for concentration by graphing transfer efficiency against acceptor expression level (e.g., fluorescence intensity).  Here is a classic example of this control:

http://www.nature.com/nchembio/journal/v4/n2/full/nchembio.64.html

A hyperbolic relationship between FRET and [acceptor] indicates specificity, and it offers a very rough of interaction affinity.  A straight line suggests that your molecules are just bumping into each other.  

Still better, try to measure stimulated emission a live cell or dish while the cell transfers from a known state (e.g., proteins not associated) to an experimental state or vice versa.  If you detect a FRET change while not seeing a comparable change when you express a control acceptor at comparable levels, you can be pretty confident that you see FRET.  

The best of all is to make a biosensor with both FRET partners on a single protein, in which case donor/acceptor expression levels become much less of a concern.  If only we could do that more often.  

All the best,


TF

Timothy Feinstein, PhD
Visiting Research Associate
Laboratory for GPCR Biology
Dept. of Pharmacology & Chemical Biology
University of Pittsburgh, School of Medicine
BST W1301, 200 Lothrop St.
Pittsburgh, PA  15261

On Apr 9, 2013, at 6:57 AM, Alessandro Esposito wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear Kees,
>    I am not surprized at all about differences in values measured by
> acceptor photobleaching or sensitized emission FRET. Both methods are
> quantitative and, gererally speaking, work well. However, they suffer of
> different biases and experimental issues and, therefore, different
> results are to be expected.
>
> Indeed, I would advice not to over-interpret FRET efficiency as an
> absolute number, at least in the average experiment in biology.
>
> apbFRET is the simplest method, but photoconversion, donor
> photobleaching and incomplete acceptor photobleaching will all concur
> to bias your results. You can standardize as much as possible your
> protocols and compensate for most issues and apbFRET will work.
> However, if you have the possibility to use seFRET, perhaps it is a
> better way if you experience photoconversion of YFP.
>
> There are many ways to do seFRET. If your stoichiometry is 1:1 you can
> even just to ratiometric imaging and pick differences among different
> biological samples. Otherwise, you can do seFRET with all controls
> needed, but I would advise to use also a construct exhibiting known
> FRET efficiency. Also, do several control images on a per session basis.
>
> Even in this case, I would recomand to not really over interpret the FRET
> efficiency value you get. If your "biological controls" make sense, then
> you can start to tease out differences among different samples.
>
> One line of caution about intensity based methods and modern confocal
> microscopes. Sometimes, modern confocal microscopes do not exhibit an
> ideal linearity between photon arrived to the detector and signal
> acquired. Refrain yourself to use all the available dynamic range,
> perhaps sacrifying the top 20% of it (for instance, with 1024 gray levels,
> use up to 800, do not be worried only if you measure 1024).
>
> Obviously, there are other methodologies around, but apbFRET or
> seFRET should be perfectly ok after you found standardized imaging
> conditions and sample preparation that provide reproducible results.
>
> Kind regards,
>
> Alessandro
> www.mrc-ccu.cam.ac.uk