Re: A1

Posted by Stephen Cody on
URL: http://confocal-microscopy-list.275.s1.nabble.com/FW-A1-tp591961p591971.html

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Dear Erik,

Thanks for your reply. It was exactly this kind of informed discussion I
was hoping to stimulate. CLEM to me has always seemed very powerful,
especially as we have conducted a lot of long term time-lapse
experiments with a conventional point scanning confocal. However, CLEM
so far does not seemed to have taken off in the market. I did not want
to sound critical of the technology, quite the opposite. I think it will
be ideal for the types of experiments we have conducted here. I do have
another question though if you don't mind.

Erik writes
> ....and you do not notice the difference between CLEM and non-CLEM
when you look at the images.

This is great news for qualitative imaging. But from the body of your
email you were suggesting it is suitable for more than just qualitative
imaging. Have you compared images analysed with Image Correlation
Spectroscopy techniques such as RICS? Are the CLEM and non-CLEM images
considered equivalent when analysed with such techniques? It is given
that photo-bleaching will presumably be worse in the non-CLEM images.

Cheers
Steve

Stephen H. Cody
Microscopy Manager
Central Resource for Advanced Microscopy
Ludwig Institute for Cancer Research
PO Box 2008 Royal Melbourne Hospital
Victoria,      3050
Australia
Tel: [phone]    Fax: [phone]
email: [hidden email]
www.ludwig.edu.au/labs/confocal.html
www.ludwig.edu.au/confocal

Tip: Learn how to receive reminders about you microscope booking:
www.ludwig.edu.au/confocal/Local/Booking_Hint.htm  

-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On
Behalf Of Erik Manders
Sent: Thursday, 21 February 2008 5:54 AM
To: [hidden email]
Subject: Re: A1

Search the CONFOCAL archive at
http://listserv.acsu.buffalo.edu/cgi-bin/wa?S1=confocal

Dear All,
I would like to add some comments to this discussion.

Guy, you are right, we've had a discussion about CLEM about a year ago
(and
I'm sure, this will not be the last one).

Michael, I think there is misunderstanding about data can be obtained
with
CLEM. For each individual pixel the optimal exposure time is determined.

The integrated fluorescence signal is extrapolated at the end of each
pixel. So, you are right, this extrapolation is a kind of correction
procedure, but this happens internally in the CLEM electronics and you
do
not notice the difference between CLEM and non-CLEM when you look at the

images. So, this means that the extrapolated pixel intensity is
unbiased.
The only difference is that the noise on the signal is increased for
pixels
that allow a shorter exposure time. So, CLEM reduces the exposure time
in
very bright pixels (S/N is too good in non-CLEM) and in dark background
pixels (why would you want to have a good S/N if there is no signal!).
This
reduced S/N in parts of the image where a good s/n is not needed, is the

price for a reduction in photobleaching and phototoxicity. The signal
does
NOT change by CLEM.

This brings me to the point of quantitative imaging with CLEM:
I do really not agree that CLEM should be switched off when you want to
do
quantitative imaging as suggested by Stephen. As I explained: CLEM does
not
change the measured gray values. There is only one reason why people
might
think that CLEM does not give quantitative data and that is
photobleaching.
I will here argue that CLEM does give images that are not less
quantitative
than non-CLEM images.

So, now we come to the point of photobleaching.
Since CLEM uses different exposure times for different pixels, the
sample
is non-uniformly exposed to light. This causes non-uniform
photobleaching.
So, if you scan your sample several times, you cannot use a standard
method
for photobleaching correction. So, Ron Hoebe and I developed a dedicated

correction procedure for CLEM images (which we will make available
later).
This algorithm calculates how much light each pixels in a 3D volume has
received before scanning that pixel and corrects for bleaching. Since we

know of every pixel how much the exposure time was, we can correct. This
is
not difficult, it is only a little bit more complicated. We also
measured
the difference between this dedicated procedure and a standard
correction
methods and found out that the difference is not much (only a few
percents). Ron did a lot of computer simulations to prove that his
program
works, by simulating every single photon for every single pixel (lot of
computation time...).

So, Stephen, you might say: Well when I want to have quantitative
images, I
just turn of the CLEM. Let's see what would happen. When you do turn it
off, your photobleaching will increase again to its non-CLEM value
(about 5
to 10 fold) So you will end-up with a strongly bleached image and a dead

cell! Great! You do a nice quantitative analysis of an image of a dead
cell
that you can hardly see anymore! And you do this only because you are
afraid that the non-uniform bleaching is not corrected optimally?

And what about non-uniform photobleaching in non-CLEM confocal images.
Does
anyone realize that photobleaching is stronger in the middle of an image

than more to the edges? Since we illuminate the sample with a cone of
light, the pixels more to the edges of the volume get a lower light dose
at
the end of a scan. And I have never seen photobleaching correction
procedures that take this into account. So, how quantitative is your
imaging right now (without CLEM)

Finally, we should realize that bleaching is far from linear (especially
in
confocal and TPE microscopy. All bleaching correction procedures assume
linear bleaching (what else can you do??). So, with this in mind: How
quantitative are your images now. I prefer a 7-fold reduction of
photobleaching so that I do not need any correction procedures!!

Aryeh, you are right: CLEM also increases the dynamic range of the
system.
However, your argument "the detector can be operated with a more linear
response to intensity changes" is not fully correct. We do not reduce
the
intensity of laser light, we only reduce the light exposure time. But
since
the exposure time is reduced for very bright pixels, clipping of the
(integrated) signal is history. This makes that you can see details in
the
weakly stained parts while not saturating the brightly stained parts.
-->
larger dynamic range.

Stephan asked: "can someone who has used the CLEM device please
confirm..."
So far, only a hand full of people have used CLEM. Some developers of
Nikon
rested CLEM and Ron and I as the inventors of CLEM have tested it, but
so
far only one person seriously applied CLEM for his biological research.
Winnok de Vos from the group of Patrick Van Oostveld from Ghent, Belgium

visited our lab for months and months to do his live cell imagine. He
spent
many nights and days behind the CLEM microscope. To his experience,
cells
can cope with a certain amount of trouble (ROS, light damage). As long
as
your exposure is under this critical threshold, you are fine. When the
light is too much, you are in trouble. By using CLEM he succeeded to
stay
under this threshold and he just could go on imaging. Without CLEM it
was
about half an hour; with CLEM for more than 24 hours.

So, we should not only focus on the bleaching reduction by CLEM
(although,
where do you find a anti-bleaching reagent that reduces bleaching by a
factor of 5 to 10 ????) but also to the reduction of phototoxicity.
 
At this moment there is only one working CLEM microscope here at the
University of Amsterdam and some prototypes of the Nikon-CLEM. But soon
there will be much more now Nikon implemented CLEM as a standard option
in
the new A1. As soon as I have my Nikon-A1, I will give my comments on my

findings, especially on my first A1-CLEM experiences. I expect a better
CLEM than our own experimental set-up. We'll see...

I think this comment satisfies you for the coming few months....

Kind regards, Erik




---------------
E.M.M. Manders, PhD
Ass. prof. Molecular Cytology
Manager Centre for Advanced Microscopy

Centre for Advanced Microscopy
Swammerdam Institute for Life Sciences
Faculty of Science
University of Amsterdam
E-mail:    [hidden email]
Tel:       [phone]
Fax:       [phone]
.--. .- ....- . --


This communication is intended only for the named recipient and may contain information that is confidential, legally privileged or subject to copyright; the Ludwig Institute for Cancer Research does not waiver any rights if you have received this communication in error.
The views expressed in this communication are those of the sender and do not necessarily reflect the views of the Ludwig Institute for Cancer Research.