Superresolution microscopy

[Knecht, David]

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

I am exploring the versatility of super-resolution commercial systems.  Are there any commercial systems that can realistically do super-resolution microscopy with live cells?  Any that can do fixed cells with both fluorescent proteins and antibody based dyes?  Thanks- Dave

David Knecht, Ph.D.
Professor and Head of Microscopy Facility
Department of Molecular and Cell Biology
U-3125
91 N. Eagleville Rd.
University of Connecticut
Storrs, CT 06269
860-486-2200
860-486-4331 (fax)

[Kurt Thorn]

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

On 10/14/2013 7:00 AM, Knecht, David wrote:
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> I am exploring the versatility of super-resolution commercial systems.  Are there any commercial systems that can realistically do super-resolution microscopy with live cells?  Any that can do fixed cells with both fluorescent proteins and antibody based dyes?  Thanks- Dave

This depends a lot on your requirements.  The Nikon N-SIM system
(structured illumination) is supposed to be fast enough for live cell
imaging, but has minimum exposure times of around a second for a single
image, so it's only good for slow processes or cases where you need a
single plane.  There are a number of papers published using STED for
live cell imaging, but fields of view tend to be small (a few microns on
a side).  There are also a number of non-commercial approaches that are
much faster - Hari Shroff's new instant SIM comes to mind.

In our experience most systems will do fixed cells with FPs or small
molecule dyes and are pretty much agnostic on the source of the
fluorophore.  We have done both FPs and dyes on our Nikon N-STORM and
N-SIM systems, for instance.

Best,
Kurt

>
> David Knecht, Ph.D.
> Professor and Head of Microscopy Facility
> Department of Molecular and Cell Biology
> U-3125
> 91 N. Eagleville Rd.
> University of Connecticut
> Storrs, CT 06269
> 860-486-2200
> 860-486-4331 (fax)
>
>

[Unruh, Jay]

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

We have successfully done live cell 3D SIM on the OMX from API (no commercial connection).  I am fairly confident that this is almost solely due so the sensitivity and high S/N ratio of the sCMOS cameras.  Our exposure times are routinely under 200 ms.  The overall acquisition was still around 15 sec for an 8 slice stack so I use "live" here with a grain of salt.  More often we PFA fix and then do SIM.  So far, we have either done FP's or antibodies but not both together.  We work in yeast, so copy numbers are low.  This is our main limitation so far.  For example, nuclear pore complexes (<32 copies per npc) are too dim for this technique.  Highly expressed proteins like cdc42 look quite nice.

In principle 2D reconstruction could increase the speed dramatically as long as you have a thin sample.  Unfortunately, the OMX doesn't have that option for reconstruction right now.

Jay

-----Original Message-----
From: Confocal Microscopy List [mailto:[hidden email]] On Behalf Of Kurt Thorn
Sent: Monday, October 14, 2013 11:22 AM
To: [hidden email]
Subject: Re: Superresolution microscopy

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

On 10/14/2013 7:00 AM, Knecht, David wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> I am exploring the versatility of super-resolution commercial systems.  
> Are there any commercial systems that can realistically do
> super-resolution microscopy with live cells?  Any that can do fixed
> cells with both fluorescent proteins and antibody based dyes?  Thanks-
> Dave

This depends a lot on your requirements.  The Nikon N-SIM system (structured illumination) is supposed to be fast enough for live cell imaging, but has minimum exposure times of around a second for a single image, so it's only good for slow processes or cases where you need a single plane.  There are a number of papers published using STED for live cell imaging, but fields of view tend to be small (a few microns on a side).  There are also a number of non-commercial approaches that are much faster - Hari Shroff's new instant SIM comes to mind.

In our experience most systems will do fixed cells with FPs or small molecule dyes and are pretty much agnostic on the source of the fluorophore.  We have done both FPs and dyes on our Nikon N-STORM and N-SIM systems, for instance.

Best,
Kurt

>
> David Knecht, Ph.D.
> Professor and Head of Microscopy Facility Department of Molecular and
> Cell Biology
> U-3125
> 91 N. Eagleville Rd.
> University of Connecticut
> Storrs, CT 06269
> 860-486-2200
> 860-486-4331 (fax)
>
>

[Andrew York]

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

 My apologies for the plug, and also because it's not a commercial system,
but our paper on instant SIM <dx.doi.org/doi:10.1038/nmeth.2687> is finally
out. It works pretty well on live cells:

http://www.nature.com/nmeth/journal/vaop/ncurrent/fig_tab/nmeth.2687_SV10.html

http://www.nature.com/nmeth/journal/vaop/ncurrent/fig_tab/nmeth.2687_SV11.html
(downloaded version looks better than online version)

 It's not as gentle as spinning disk, mostly because the effective pixel
size is a lot smaller. Other than that, it's like a spinning disk with
twice the resolution. 2D timelapses work especially well, since you don't
have to wait 10 ms for each z-stage motion.

 I don't think there's going to be a commercial version available this
year, but I don't mind helping to build a few of them. It's about a week of
work if you have all the parts. If the government ever reopens, you're
welcome to come try out the one at the NIH.


On Mon, Oct 14, 2013 at 2:01 PM, Unruh, Jay <[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 successfully done live cell 3D SIM on the OMX from API (no
commercial connection).  I am fairly confident that this is almost solely
due so the sensitivity and high S/N ratio of the sCMOS cameras.  Our
exposure times are routinely under 200 ms.  The overall acquisition was
still around 15 sec for an 8 slice stack so I use "live" here with a grain
of salt.  More often we PFA fix and then do SIM.  So far, we have either
done FP's or antibodies but not both together.  We work in yeast, so copy
numbers are low.  This is our main limitation so far.  For example, nuclear
pore complexes (<32 copies per npc) are too dim for this technique.  Highly
expressed proteins like cdc42 look quite nice.
>
> In principle 2D reconstruction could increase the speed dramatically as
long as you have a thin sample.  Unfortunately, the OMX doesn't have that
option for reconstruction right now.
>
> Jay
>
> -----Original Message-----
> From: Confocal Microscopy List [mailto:[hidden email]]
On Behalf Of Kurt Thorn

> Sent: Monday, October 14, 2013 11:22 AM
> To: [hidden email]
> Subject: Re: Superresolution microscopy
>
> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> On 10/14/2013 7:00 AM, Knecht, David wrote:
> > *****
> > To join, leave or search the confocal microscopy listserv, go to:
> > http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> > *****
> >
> > I am exploring the versatility of super-resolution commercial systems.
> > Are there any commercial systems that can realistically do
> > super-resolution microscopy with live cells?  Any that can do fixed
> > cells with both fluorescent proteins and antibody based dyes?  Thanks-
> > Dave
>
> This depends a lot on your requirements.  The Nikon N-SIM system

(structured illumination) is supposed to be fast enough for live cell
imaging, but has minimum exposure times of around a second for a single
image, so it's only good for slow processes or cases where you need a
single plane.  There are a number of papers published using STED for live
cell imaging, but fields of view tend to be small (a few microns on a
side).  There are also a number of non-commercial approaches that are much
faster - Hari Shroff's new instant SIM comes to mind.
>
> In our experience most systems will do fixed cells with FPs or small
molecule dyes and are pretty much agnostic on the source of the
fluorophore.  We have done both FPs and dyes on our Nikon N-STORM and N-SIM
systems, for instance.

>
> Best,
> Kurt
>
>
> >
> > David Knecht, Ph.D.
> > Professor and Head of Microscopy Facility Department of Molecular and
> > Cell Biology
> > U-3125
> > 91 N. Eagleville Rd.
> > University of Connecticut
> > Storrs, CT 06269
> > 860-486-2200
> > 860-486-4331 (fax)
> >
> >

[Reto Fiolka]

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

Andrew, I would be very interested how your instant SIM compares to a
Gustaffson-style 3D SIM (using same NA), in terms of resolution and bleaching.
If NIH is closed, there is a custom built dual color SIM at Janelia that you might
try out if you want to do that.

Your system is drastically simpler than the live 3D SIM systems we built and
probably much more cost effective too.

The single and multicolor SIM system that were built under Mats Gustafsson
were live capable:

http://www.nature.com/nmeth/journal/v8/n12/abs/nmeth.1734.html

http://www.pnas.org/content/109/14/5311.short

I could do about 1 Hz for a slice of 1 micron thickness (Stack of 8 slices),
however this needed very short exposure times and thus bright fluorophores.

The OMX blaze from API achieves a similar volumetric sample rate.

With the Gustafsson style 3D fourier space processing I do not see a possibility
to go 2D in a 3D SIM setup, i.e. one always needs a volume data set to start
with.

Best,
Reto

[Tim Grotjohann]

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

You might be also interested in doing RESOLFT microscopy (see for example
Grotjohann et al., Nature 2011; Grotjohann et. al., eLife 2012; Testa et
al., Neuron 2012). It is inherently live-cell and works with reversibly
switchable fluorescent proteins. RESOLFT uses low light intensities
(kW/cm²-range) and, in comparison to all other superresolution methods (with
theoretically no resolution limit), exposes the lowest light dose to the
sample. Hence phototoxic effects are minimized. Routine resolutions are
50-90nm. For the single beam instrument which is commercially available from
Abberior Instruments, pixel dwell times are typically <400us, your total
image acquisition time will then depend on your field of view and pixel size.
The data aquisition in RESOLFT can be parallelized to achieve much bigger
fields of view (FOV) in short times - see Chmyrov et al., Nature Methods
(10) 2013. Here the FOV and acquisition time are not longer coupled. The FOV
is just limited by the camera size. The resolution was about 80nm with an
acquisition time of ~3s for a 120x100um field of view. A pdf is available
here: http://tinyurl.com/lqc86ha.

Best,
Tim