Dear Sylvie,

You may want to have a look at our manuscript on spectral FRET (SpRET: highly sensitive and reliable spectral measurement of absolute FRET efficiency; doi: 10.1017/S1431927610094493), in which we specifically dealt with the issue of autofluorescence and the effect of the brightness of the samples. We found spectral FRET to be well suited for this purpose. Be warned, though, that the emission spectrum of autofluorescence is not always easily characterized, since it may be contributed by several sources, in which case it can differ in nature from area to area. You can try to include several spectra for autofluorescence, assuming that by doing so you cover a significant number of their combinations within the samples you are looking at.
Hope this helps,

Daniel 

Daniel Gitler, Senior Lecturer
Department of Physiology and Cell Biology
Faculty of Health Sciences
Ben-Gurion University of the Negev
Beer-Sheva 84105
Israel

________________________________________
From: Confocal Microscopy List [[log in to unmask]] on behalf of Hoppe, Adam [[log in to unmask]]
Sent: Friday, November 20, 2015 11:38 PM
To: [log in to unmask]
Subject: Re: Autofluorescence in PFRET

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
Post images on http://www.imgur.com and include the link in your posting.
*****

Dear Sylvie and Neil,

Directly applying conventional linear unmixing to FRET data is a doomed
approach as the FRET creates linear combinations of donor excitation(s)
and acceptor emission(s) - which violates the assumptions of a
conventional spectral unmixing -especially for a 3-cube type approach.
There are handful of different approaches that have been specifically
developed to applying linear unmixing methods to spectral FRET while
accounting for the FRET coupling between donor(s) and acceptor(s).

These can certainly be adapted to remove autofluorescence, but in the end
it is a game of signal to noise ratio.  In addition to the approach of Dr.
Periasamy¹s group, our lab and the Neher lab have worked on a refined
linear unmixing approach that works for FRET might be of some use in
solving your problem.

Here¹s the DOI number:
DOI: 10.1371/journal.pone.0064760


Best Regards,
Adam

Adam Hoppe, Ph.D.

BioSNTR Director
Associate Professor, South Dakota State University
Department of Chemistry and Biochemistry


SDSU, SAV 131
Brookings, SD 57007
Tel: 605-688-5315





On 11/20/15, 2:45 PM, "yuansheng sun" <[log in to unmask]> wrote:

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>Post images on http://www.imgur.com and include the link in your posting.
>*****
>
>Dear Sylvie,
>
>I just follow up with a few comments, embedded below.  Thanks.
>
>Sheng
>
>
>
>> 1- Then I have a question to the list regarding this issue: Is unmixing
>>an
>> appropriate technique for experiments where one wants to extract
>> quantitative intensity information from the image? Does anyone have any
>> experience with this?
>
>
>I think so.  When I worked for Peri, we published a few papers using
>spectral FRET, including a 3-color FRET one.  I can send you the papers if
>you are interested.  I also know quite a few other people use spectral
>imaging and unmixing for steady-state and time-resolved (FLIM) FRET.
>
>
>
>
>
>> In spectral unmixing, an algorithm is applied post acquisition to the
>> image to extract the spectrum of each fluorophore from the complex
>>spectrum
>> collected by the detector. I think that the dimmer the signal, the lower
>> the signal to noise ratio and so the higher the risk of errors by the
>> algorithm. So I would be very cautious with calculating FRET efficiency
>>(ie
>> applying a second algorithm) from unmixed images.
>>
>> If your signal is very bright in both fluorophores (assuming you are not
>> using biosensors), you can acquire reference spectra that are quite
>> reliable. Then it is worth trying to unmix then calculate the FRET image
>> but you will need to scrutinize your results for bias in the FRET
>> efficiency in the dim vs bright areas.
>>
>
>You are right - the SNR is a fundamental issue to any quantitative
>measurement.  The PFRET addresses this issue carefully and calculates the
>bleedthrough percentages based on the intensity levels.
>
>
>
>>
>> 2- Concerning the autofluorescence, I assume you have checked that it is
>> really autofluorescence (ie you can see it in the unstained tissue). In
>> that case it might be worth trying to apply Sudan Black before imaging,
>>if,
>> of course it is fixed tissue you are imaging.
>> If instead it is 'unwanted' fluorescence ('sticky' antibody), spectral
>> unmixing will not help since the 'autofluorescence' will have the same
>> spectrum as your signal. Increase the stringency of your labelling
>>instead.
>>
>> All this applies only if you are not the unlucky one who wants to image
>> non biosensor FRET is a live tissue giving strong autofluorescence.
>> In that case I would turn to FLIM.
>>
>>
>Thank you for mentioning FLIM.  I think it is a very good technique for
>FRET.
>
>
>
>> Best of luck :)
>>
>> Med vänlig hälsning / Best regards
>>
>> Sylvie
>>
>> @@@@@@@@@@@@@@@@@@@@@@@@
>> Sylvie Le Guyader, PhD
>> Live Cell Imaging Unit Manager
>> Karolinska Institutet- Bionut Dpt
>> Hälsovägen 7,
>> Novum, G lift, floor 6
>> 14157 Huddinge
>> Sweden
>> mobile: +46 (0) 73 733 5008
>> office: +46 (0) 8 5248 1107
>> LCI website
>>
>>
>> -----Original Message-----
>> From: Confocal Microscopy List [mailto:[log in to unmask]]
>> On Behalf Of Feinstein, Timothy N
>> Sent: den 20 november 2015 15:51
>> To: [log in to unmask]
>> Subject: Re: Autofluorescence in PFRET
>>
>> *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> Post images on http://www.imgur.com and include the link in your
>>posting.
>> *****
>>
>> Spectral mode (confocal?) is a great idea in principle, but you have to
>> keep in mind it is a highly inefficient mode of acquisition.  Many
>>probes
>> and experiments that are routine on a widefield will become quite hard
>>to
>> detect.  If have a bright enough signal that should be no problem.  If
>>the
>> signal is dim then make sure to use large pixels and open the pinhole
>>much
>> wider than usual.
>>
>> Best,
>>
>>
>> Tim
>>
>> Timothy Feinstein, Ph.D.
>> Research Scientist
>> University of Pittsburgh Department of Developmental Biology
>>
>>
>>
>>
>>
>> On 11/19/15, 11:20 PM, "Confocal Microscopy List on behalf of Periasamy,
>> Ammasi (ap3t)" <[log in to unmask] on behalf of
>> [log in to unmask]> wrote:
>>
>> >Hello Neil
>> >As Sheng mentioned you can use the spectral imaging approach. Or you
>> >can use the unlabeled cells to collect the signal to use for background
>> >subtraction. That will take care. The background subtraction should
>> >remove any autofluorescence or detector noise involved in fluorescence
>> >imaging.
>> >Hope this helps.
>> >
>> >Dr. Ammasi Periasamy
>> >Professor & Center Director
>> >http://www.kcci.virginia.edu/people/profile/ap3t
>> >Phone: (434) 243-7602 or 982-4869
>> >Fax: (434) 982-5210
>> >E-mail: [log in to unmask]
>> >Office Location
>> >W.M. Keck Center for Cellular Imaging
>> >Physical and Life Sciences Building, (B 005) At the intersection of
>> >Geldard dr and White head Rd., Mailing or Shipping Address:
>> >W.M. Keck Center for Cellular Imaging (PLSB 005) University of Virginia
>> >Biology, Gilmer Hall, 409 McCormick Rd.
>> >Charlottesville, VA 22904, USA
>> >FRET/FLIM Workshop-March 7-11, 2016:
>> >http://www.kcci.virginia.edu/workshop
>> >
>> >
>> >
>> >-----Original Message-----
>> >From: Confocal Microscopy List
>> >[mailto:[log in to unmask]]
>> >On Behalf Of yuansheng.sun
>> >Sent: Thursday, November 19, 2015 9:27 PM
>> >To: [log in to unmask]
>> >Subject: Re: Autofluorescence in PFRET
>> >
>> >Then, it would be better to use spectral FRET.  Measure the reference
>> >spectrum of the autofluorescence from unlabeled sample.  Use linear
>> >unmixing to remove autofluorescence.
>> >
>> >Sheng
>> >
>> >
>> >Sent from my T-Mobile 4G LTE Device
>> >
>> >
>> >-------- Original message --------
>> >From: "Anthony, Neil" <[log in to unmask]>
>> >Date:11/19/2015  3:06 PM  (GMT-06:00)
>> >To: [log in to unmask]
>> >Cc:
>> >Subject: Autofluorescence in PFRET
>> >
>> >*****
>> >To join, leave or search the confocal microscopy listserv, go to:
>> >http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>> >Post images on http://www.imgur.com and include the link in your
>>posting.
>> >*****
>> >
>> >Hi all, I hope the science treats you well.
>> >
>> >I was wondering if anybody's had any experience with autofluorescence
>> >when using PFRET, or PeriFRET as I like to call it.  I understand that
>> >autofluorescence is not accounted for in the algorithm?
>> >
>> >Peri, I hope all is well with you.  Can I ask for some pointers on how
>> >think about the system if autofluorescence is the mix?
>> >
>> >Thanks in advance for your time.
>> >
>> >Neil
>> >
>> >________________________________
>> >
>> >This e-mail message (including any attachments) is for the sole use of
>> >the intended recipient(s) and may contain confidential and privileged
>> >information. If the reader of this message is not the intended
>> >recipient, you are hereby notified that any dissemination, distribution
>> >or copying of this message (including any attachments) is strictly
>> prohibited.
>> >
>> >If you have received this message in error, please contact the sender
>> >by reply e-mail message and destroy all copies of the original message
>> >(including attachments).
>>