SIM redux

[mcammer]

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

We have a practical question about how we know whether a SIM result is showing real structure or not.

We know from EM that the structure we are looking at is densely packed solid containing the protein of interest throughout, not hollow.  However, the SIM result appears hollow.  There is a possibility that the staining doesn't penetrate to the core, but we really don't know.

Has anybody see this sort of problem with SIM where a structure known to be solid is reconstructed with a hollow core?

Illustration at http://www.flickr.com/photos/mcammer/9696746798/ or https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84

We are using OMX Blaze with a 568 (or is it 561) nm laser.

Thank you.

_________________________________________
Michael Cammer, Assistant Research Scientist
Skirball Institute of Biomolecular Medicine
Lab: [phone]  Cell: [phone]

[George McNamara]

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

Hi Michael,

What does SIMcheck tell you about your machine?
http://www2.bioch.ox.ac.uk/microngroup/software/SIMCheck.shtml

If you are using fluorescent antibody detection, do you know the
antibodies penetrate all the way to the core (as you noted). If it
practical to use direct labeled fluorescent Fab instead of full length Abs?

What do GFP (or Venus, which should be brighter) fusion protein(s) show?
Are there any small molecules that find your favorite molecule? ... If
yes, are the small molecules fluorescent or could be conjugated to a
fluorophore? I'm thinking in particular of fluorescent-phalloidins and
fluorescent bungarotoxins as examples. A related method would be to
conjugate fluorophores to your protein of interest and (find a
microinjection setup and) microinject into your cells. Google:  clare
waterman fluorescence speckle microscopy    for where that could go (ex.
http://www.molbiolcell.org/content/22/21/3940.full ).

Does your Blaze also have TIRF/Monet mode?
... if yes, is the structure close enough to the coverglass to use it?
... if no, the OMX could be a great widefield scope to acquire data
(timelapse mode) for 3Bmicroscopy ... http://www.coxphysics.com/3b/  ...
the ImageJ plugin is very slow, even on a small region of interest (so
far I've only played with it on their demo data set).

George


On 9/7/2013 3:58 PM, Cammer, Michael wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> We have a practical question about how we know whether a SIM result is showing real structure or not.
>
> We know from EM that the structure we are looking at is densely packed solid containing the protein of interest throughout, not hollow.  However, the SIM result appears hollow.  There is a possibility that the staining doesn't penetrate to the core, but we really don't know.
>
> Has anybody see this sort of problem with SIM where a structure known to be solid is reconstructed with a hollow core?
>
> Illustration at http://www.flickr.com/photos/mcammer/9696746798/ or https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84
>
> We are using OMX Blaze with a 568 (or is it 561) nm laser.
>
> Thank you.
>
> _________________________________________
> Michael Cammer, Assistant Research Scientist
> Skirball Institute of Biomolecular Medicine
> Lab: [phone]  Cell: [phone]
>
>
>    

--



George McNamara, Ph.D.
Single Cells Analyst
L.J.N. Cooper Lab
University of Texas M.D. Anderson Cancer Center
Houston, TX 77054
http://works.bepress.com/gmcnamara/26/

[Laevsky, Gary S.]

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

Hi Michael,

The first thing I would ask is what does the background look like?

In the SIM reconstruction that I have seen in multiple platforms, there is an artifact that results in a "mottling" type of pattern.  This is regular and uniform throughout the image.  It can also be made to disappear depending on the parameters used for reconstruction.

The explanation I understand is quite basic, in relation to some of the knowledge held by people that monitor this list.  And I'm sure they can expand, but it has to do with the combination of the different phases, during the reconstruction, used to capture the data.

The problem is, that it can be made to disappear by changing the reconstruction parameters.  There is no "set" setting for the reconstruction.  What I used to do is set my parameters so the background was uniform (if you stretch the heck out of the LUT (obviously not the live histogram), you can see the background better).  Then I was confident that I was seeing "real" data.

If you see the pattern in the background, this may be the culprit.  If the background is uniform, then George and Wendy's comments are more relevant.


Best,

Gary



Gary Laevsky, Ph.D.
Confocal Imaging Facility Manager
Dept. of Molecular Biology
Washington Rd.
Princeton University
Princeton, New Jersey, 08544-1014
(O) 609 258 5432
(C) 508 507 1310

On Sep 7, 2013, at 4:58 PM, "Cammer, Michael" <[hidden email]>
 wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> We have a practical question about how we know whether a SIM result is showing real structure or not.
>
> We know from EM that the structure we are looking at is densely packed solid containing the protein of interest throughout, not hollow.  However, the SIM result appears hollow.  There is a possibility that the staining doesn't penetrate to the core, but we really don't know.
>
> Has anybody see this sort of problem with SIM where a structure known to be solid is reconstructed with a hollow core?
>
> Illustration at http://www.flickr.com/photos/mcammer/9696746798/ or https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84
>
> We are using OMX Blaze with a 568 (or is it 561) nm laser.
>
> Thank you.
>
> _________________________________________
> Michael Cammer, Assistant Research Scientist
> Skirball Institute of Biomolecular Medicine
> Lab: [phone]  Cell: [phone]
>

[James Pawley]

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

Hello all,

This important discussion reminds me of what my post-doc tutor, Alan
Boyde used to say "You can only recognize artifacts in microscopy  to
the extent that you can 'view' the same structure in a second (or
third...) way" (where "view" means "gain reliable structural
information about")

This thread started when Michael Cammer told us about a structure
that he had previously studied by (it turns out freeze-fracture,
rotary shadow) TEM that looked different when viewed using SIM. i.e.,
he had followed the rule.

I would really like to hear responses from other listers who who have
1) viewed a structure in one of the enhanced LM methods and 2) looked
at the same structure by another microscopical technique (SEM, TEM
etc). I would be particularly interested to hear from those who
noticed significant differences and also how they interpreted these
differences (might specimen prep play a role.)

Cheers,

JP

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hi Michael,
>
>The first thing I would ask is what does the background look like?
>
>In the SIM reconstruction that I have seen in multiple platforms,
>there is an artifact that results in a "mottling" type of pattern.
>This is regular and uniform throughout the image.  It can also be
>made to disappear depending on the parameters used for
>reconstruction.
>
>The explanation I understand is quite basic, in relation to some of
>the knowledge held by people that monitor this list.  And I'm sure
>they can expand, but it has to do with the combination of the
>different phases, during the reconstruction, used to capture the
>data.
>
>The problem is, that it can be made to disappear by changing the
>reconstruction parameters.  There is no "set" setting for the
>reconstruction.  What I used to do is set my parameters so the
>background was uniform (if you stretch the heck out of the LUT
>(obviously not the live histogram), you can see the background
>better).  Then I was confident that I was seeing "real" data.
>
>If you see the pattern in the background, this may be the culprit.
>If the background is uniform, then George and Wendy's comments are
>more relevant.
>
>
>Best,
>
>Gary
>
>
>
>Gary Laevsky, Ph.D.
>Confocal Imaging Facility Manager
>Dept. of Molecular Biology
>Washington Rd.
>Princeton University
>Princeton, New Jersey, 08544-1014
>(O) 609 258 5432
>(C) 508 507 1310
>
>On Sep 7, 2013, at 4:58 PM, "Cammer, Michael" <[hidden email]>
>  wrote:
>
>>  *****
>>  To join, leave or search the confocal microscopy listserv, go to:
>>  http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>  *****
>>
>>  We have a practical question about how we know whether a SIM
>>result is showing real structure or not.
>>
>>  We know from EM that the structure we are looking at is densely
>>packed solid containing the protein of interest throughout, not
>>hollow.  However, the SIM result appears hollow.  There is a
>>possibility that the staining doesn't penetrate to the core, but we
>>really don't know.
>>
>>  Has anybody see this sort of problem with SIM where a structure
>>known to be solid is reconstructed with a hollow core?
>>
>>  Illustration at http://www.flickr.com/photos/mcammer/9696746798/ 
>>or https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84
>>
>>  We are using OMX Blaze with a 568 (or is it 561) nm laser.
>>
>>  Thank you.
>>
>>  _________________________________________
>>  Michael Cammer, Assistant Research Scientist
>>  Skirball Institute of Biomolecular Medicine
>>  Lab: [phone]  Cell: [phone]
>>

--
James and Christine Pawley, 5446 Burley Place (PO Box 2348), Sechelt,
BC, Canada, V0N3A0,
Phone [phone], email <[hidden email]>
NEW! NEW! AND DIFFERENT Cell (when I remember to turn it on!) 1-604-989-6146

[Guy Cox-2]

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

Jim,

        Check out this reference, from my friend and colleague Filip Braet

Imaging Fluorescently Labeled Complexes by Means of Multidimensional Correlative Light and Transmission Electron Microscopy: Practical Considerations
    Kobayashi, K., Cheng, D., Huynh, M., Ratinac, K.R., Thordardsson, P., Braet, F.
    2012
    Methods in Cell Biology 111 , pp. 1-20
   
                                                Guy

Guy Cox, Honorary Associate Professor
School of Medical Sciences

Australian Centre for Microscopy and Microanalysis,
Madsen, F09, University of Sydney, NSW 2006

-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of James Pawley
Sent: Tuesday, 10 September 2013 5:20 AM
To: [hidden email]
Subject: Re: SIM redux

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

Hello all,

This important discussion reminds me of what my post-doc tutor, Alan Boyde used to say "You can only recognize artifacts in microscopy  to the extent that you can 'view' the same structure in a second (or
third...) way" (where "view" means "gain reliable structural information about")

This thread started when Michael Cammer told us about a structure that he had previously studied by (it turns out freeze-fracture, rotary shadow) TEM that looked different when viewed using SIM. i.e., he had followed the rule.

I would really like to hear responses from other listers who who have
1) viewed a structure in one of the enhanced LM methods and 2) looked at the same structure by another microscopical technique (SEM, TEM etc). I would be particularly interested to hear from those who noticed significant differences and also how they interpreted these differences (might specimen prep play a role.)

Cheers,

JP

>*****
>To join, leave or search the confocal microscopy listserv, go to:
>http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>*****
>
>Hi Michael,
>
>The first thing I would ask is what does the background look like?
>
>In the SIM reconstruction that I have seen in multiple platforms, there
>is an artifact that results in a "mottling" type of pattern.
>This is regular and uniform throughout the image.  It can also be made
>to disappear depending on the parameters used for reconstruction.
>
>The explanation I understand is quite basic, in relation to some of the
>knowledge held by people that monitor this list.  And I'm sure they can
>expand, but it has to do with the combination of the different phases,
>during the reconstruction, used to capture the data.
>
>The problem is, that it can be made to disappear by changing the
>reconstruction parameters.  There is no "set" setting for the
>reconstruction.  What I used to do is set my parameters so the
>background was uniform (if you stretch the heck out of the LUT
>(obviously not the live histogram), you can see the background better).  
>Then I was confident that I was seeing "real" data.
>
>If you see the pattern in the background, this may be the culprit.
>If the background is uniform, then George and Wendy's comments are more
>relevant.
>
>
>Best,
>
>Gary
>
>
>
>Gary Laevsky, Ph.D.
>Confocal Imaging Facility Manager
>Dept. of Molecular Biology
>Washington Rd.
>Princeton University
>Princeton, New Jersey, 08544-1014
>(O) 609 258 5432
>(C) 508 507 1310
>
>On Sep 7, 2013, at 4:58 PM, "Cammer, Michael"
><[hidden email]>
>  wrote:
>
>>  *****
>>  To join, leave or search the confocal microscopy listserv, go to:
>>  http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
>>  *****
>>
>>  We have a practical question about how we know whether a SIM result
>>is showing real structure or not.
>>
>>  We know from EM that the structure we are looking at is densely
>>packed solid containing the protein of interest throughout, not
>>hollow.  However, the SIM result appears hollow.  There is a
>>possibility that the staining doesn't penetrate to the core, but we
>>really don't know.
>>
>>  Has anybody see this sort of problem with SIM where a structure
>>known to be solid is reconstructed with a hollow core?
>>
>>  Illustration at http://www.flickr.com/photos/mcammer/9696746798/
>>or https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84
>>
>>  We are using OMX Blaze with a 568 (or is it 561) nm laser.
>>
>>  Thank you.
>>
>>  _________________________________________
>>  Michael Cammer, Assistant Research Scientist  Skirball Institute of
>> Biomolecular Medicine
>>  Lab: [phone]  Cell: [phone]
>>

--
James and Christine Pawley, 5446 Burley Place (PO Box 2348), Sechelt, BC, Canada, V0N3A0, Phone [phone], email <[hidden email]> NEW! NEW! AND DIFFERENT Cell (when I remember to turn it on!) 1-604-989-6146

[Peng Xi-2]

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

Dear Michael and Jim,
     We have imaged fluorescent nanodiamond with STED and SIM. The FNDs are
single molecule emitters with size of 35 nm diameter. STED has validated
that they are indeed in single particle. These particles are very good dots
for measuring the PSF of the imaging system.
     When we use SIM to image the particles, we found surprisingly that
even with 200nm out of focus, the reconstructed image shows hollow
structure, exactly as what you have seen. The hollow appears to be
insensitive to the parameters of reconstruction process.
     Based on the image you provided, I guess it may be from the fact that
your structure is slightly above the focal plane -- you may want to do a z
series of SIM to verify it.
     You can download the original data from:
http://bme.pku.edu.cn/~xipeng/FND-SIM.rar     Please pay special attention
to the last two frames.

Cheers,
Peng Xi
Ph. D.    Associate Professor
Dept. of Biomedical Engineering, College of Engineering
Peking University, Beijing, China
Tel: [phone]
Email: [hidden email]
http://bme.pku.edu.cn/~xipeng/




On Tue, Sep 10, 2013 at 3:19 AM, James Pawley <[hidden email]> wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/**wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
> *****
>
> Hello all,
>
> This important discussion reminds me of what my post-doc tutor, Alan Boyde
> used to say "You can only recognize artifacts in microscopy  to the extent
> that you can 'view' the same structure in a second (or third...) way"
> (where "view" means "gain reliable structural information about")
>
> This thread started when Michael Cammer told us about a structure that he
> had previously studied by (it turns out freeze-fracture, rotary shadow) TEM
> that looked different when viewed using SIM. i.e., he had followed the rule.
>
> I would really like to hear responses from other listers who who have 1)
> viewed a structure in one of the enhanced LM methods and 2) looked at the
> same structure by another microscopical technique (SEM, TEM etc). I would
> be particularly interested to hear from those who noticed significant
> differences and also how they interpreted these differences (might specimen
> prep play a role.)
>
> Cheers,
>
> JP
>
>
>
>
>  *****
>> To join, leave or search the confocal microscopy listserv, go to:
>> http://lists.umn.edu/cgi-bin/**wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
>> *****
>>
>> Hi Michael,
>>
>> The first thing I would ask is what does the background look like?
>>
>> In the SIM reconstruction that I have seen in multiple platforms, there
>> is an artifact that results in a "mottling" type of pattern. This is
>> regular and uniform throughout the image.  It can also be made to disappear
>> depending on the parameters used for reconstruction.
>>
>> The explanation I understand is quite basic, in relation to some of the
>> knowledge held by people that monitor this list.  And I'm sure they can
>> expand, but it has to do with the combination of the different phases,
>> during the reconstruction, used to capture the data.
>>
>> The problem is, that it can be made to disappear by changing the
>> reconstruction parameters.  There is no "set" setting for the
>> reconstruction.  What I used to do is set my parameters so the background
>> was uniform (if you stretch the heck out of the LUT (obviously not the live
>> histogram), you can see the background better).  Then I was confident that
>> I was seeing "real" data.
>>
>> If you see the pattern in the background, this may be the culprit. If the
>> background is uniform, then George and Wendy's comments are more relevant.
>>
>>
>> Best,
>>
>> Gary
>>
>>
>>
>> Gary Laevsky, Ph.D.
>> Confocal Imaging Facility Manager
>> Dept. of Molecular Biology
>> Washington Rd.
>> Princeton University
>> Princeton, New Jersey, 08544-1014
>> (O) 609 258 5432
>> (C) 508 507 1310
>>
>> On Sep 7, 2013, at 4:58 PM, "Cammer, Michael" <[hidden email]
>> >
>>  wrote:
>>
>>   *****
>>>  To join, leave or search the confocal microscopy listserv, go to:
>>>  http://lists.umn.edu/cgi-bin/**wa?A0=confocalmicroscopy<http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy>
>>>  *****
>>>
>>>  We have a practical question about how we know whether a SIM result is
>>> showing real structure or not.
>>>
>>>  We know from EM that the structure we are looking at is densely packed
>>> solid containing the protein of interest throughout, not hollow.  However,
>>> the SIM result appears hollow.  There is a possibility that the staining
>>> doesn't penetrate to the core, but we really don't know.
>>>
>>>  Has anybody see this sort of problem with SIM where a structure known
>>> to be solid is reconstructed with a hollow core?
>>>
>>>  Illustration at http://www.flickr.com/photos/**mcammer/9696746798/<http://www.flickr.com/photos/mcammer/9696746798/>or
>>> https://nyumc.box.com/s/**gyq3c3cw6p5aofb7gw84<https://nyumc.box.com/s/gyq3c3cw6p5aofb7gw84>
>>>
>>>  We are using OMX Blaze with a 568 (or is it 561) nm laser.
>>>
>>>  Thank you.
>>>
>>>  ______________________________**___________
>>>  Michael Cammer, Assistant Research Scientist
>>>  Skirball Institute of Biomolecular Medicine
>>>  Lab: [phone]  Cell: [phone]
>>>
>>>
>
> --
> James and Christine Pawley, 5446 Burley Place (PO Box 2348), Sechelt, BC,
> Canada, V0N3A0,
> Phone [phone], email <[hidden email]>
> NEW! NEW! AND DIFFERENT Cell (when I remember to turn it on!)
> 1-604-989-6146
>

[Ian Dobbie]

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

Peng Xi <[hidden email]> writes:

>      When we use SIM to image the particles, we found surprisingly that
> even with 200nm out of focus, the reconstructed image shows hollow
> structure, exactly as what you have seen. The hollow appears to be
> insensitive to the parameters of reconstruction process.
>      Based on the image you provided, I guess it may be from the fact that
> your structure is slightly above the focal plane -- you may want to do a z
> series of SIM to verify it.
>      You can download the original data from:
> http://bme.pku.edu.cn/~xipeng/FND-SIM.rar     Please pay special attention
> to the last two frames.

This data is pretty horrible to be honest. Scanning through the stack
there looks to be a squint PSF, or illumination. An orthogonal view
shows that the XZ PSF is bent. Further doing a FFT on the presented data
shows that there is no information beyond about 220 nm, so the image
isn't super resolution at all. I don't think that you can use this data
to show that a single molecule emitter produces a hollow above or below
focus.

Ian

[Peng Xi-2]

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

Ian is correct -- I shouldn't fully trust the commercial system. Honestly I
am pretty surprised at seeing the hollow, which happens to be similar to
what Michael mentioned... I will redo the experiment and keep you posted.

Peng


On Wed, Sep 11, 2013 at 1:22 AM, Ian Dobbie <[hidden email]>wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Peng Xi <[hidden email]> writes:
>
> >      When we use SIM to image the particles, we found surprisingly that
> > even with 200nm out of focus, the reconstructed image shows hollow
> > structure, exactly as what you have seen. The hollow appears to be
> > insensitive to the parameters of reconstruction process.
> >      Based on the image you provided, I guess it may be from the fact
> that
> > your structure is slightly above the focal plane -- you may want to do a
> z
> > series of SIM to verify it.
> >      You can download the original data from:
> > http://bme.pku.edu.cn/~xipeng/FND-SIM.rar     Please pay special
> attention
> > to the last two frames.
>
> This data is pretty horrible to be honest. Scanning through the stack
> there looks to be a squint PSF, or illumination. An orthogonal view
> shows that the XZ PSF is bent. Further doing a FFT on the presented data
> shows that there is no information beyond about 220 nm, so the image
> isn't super resolution at all. I don't think that you can use this data
> to show that a single molecule emitter produces a hollow above or below
> focus.
>
> Ian
>

[Dan Metcalf]

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

Hi Michael,

I suspect in this case, as Ian Dobbie has suggested, that this is a reconstruction
artefact. However, I have seen a labelling artefact that was resolved by SIM and
STORM using instruments built at the National Physical Laboratory but otherwise
not resolvable by conventional imaging. In this case, a large secretory granule
type structure. The immunolabelling was against the content protein but the
antibodies weren't able to penetrate the very centre, which created a hollow core
type structure as the dyes ended up outside.

Regards,

Dan Metcalf

Nikon Instruments UK