mechanism of gold fiducials

[Paul Paroutis]

*****
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.
*****

I have a few questions regarding the use of gold nano particles as fiducial
markers for STORM imaging that I was hoping the listserv community could help
with:

1) What is the mechanism by which they are visible under standard excitation
wavelengths? It seems that they're based on surface plasmon resonance, and
the amount of scatter across various wavelengths varies according to the
properties of the particle. Is that correct?

2) If so, are there gold fiducials that anyone has used with success across
multiple wavelengths? In other words, using a single fiducial that is visible
across all channels. According to the OD curves, it seems that some are great in
the red, but not so good in blue, etc. Or is this mitigated by the fact that even if
there is a little bit of scatter, at the powers typically used for STORM it will be
visible regardless?

3) Finally, does the size of the fiducial matter? Why would someone use 50 nm
fiducials instead of 100 or 200 nm?

Any advice on all this would be appreciated.

Thanks,
Paul

[WHEELER Ann]

*****
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 Paul, I can have a go:

1) What is the mechanism by which they are visible under standard excitation
wavelengths? It seems that they're based on surface plasmon resonance, and
the amount of scatter across various wavelengths varies according to the
properties of the particle. Is that correct?

Small gold particles are fluorescent, the mechanism is mostly likely the way that other nanoparticles emit light by electrons moving between a homo and lumo state. I would use a 561nm laser and a long pass filter to find them, to confirm this phase contrast can be used, others may have different strategies. However as you state the size of them does determine the emission wavelength. The advantages of the gold fiducials is they never blink. I would use 100nm particles from BBI gold, these can be used to correct drift. However in our hand they do emit weakly.



2) If so, are there gold fiducials that anyone has used with success across
multiple wavelengths? In other words, using a single fiducial that is visible
across all channels. According to the OD curves, it seems that some are great in
the red, but not so good in blue, etc. Or is this mitigated by the fact that even if
there is a little bit of scatter, at the powers typically used for STORM it will be
visible regardless?

We haven't managed to visualise 100nm gold fiducials in blue. However we have used other nanoparticles with more success (which we did report at FOM 2014). Commercial Q-dots can be used for this - although they are variable in size our EM work shows they are 20 - 50nm, it depends on how much resolution is required whether these would be appropriate.

3) Finally, does the size of the fiducial matter? Why would someone use 50 nm
fiducials instead of 100 or 200 nm?

Simply because of the required resolution, I would say. With the kind of experiments I have done in the past the best I could expect was 80nm (because spinning disks eat light). However if I was to use TIRF based methods more could be expected. We've recently been using Q dots which we sized by TEM since we are trying to improve resolution. The major problem is if drift can only be corrected to 100nm this will determine the localisation accuracy of the experiment.

Hope this helps, good luck and please do get in touch if you require more info!

Dr Ann Wheeler
Head of Imaging at IGMM, University of Edinburgh
Edinburgh Super-Resolution Consortium manager:- IGMM



--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.

[Zdenek Svindrych-2]

*****
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 there,
the problem with fiducials in STORM imaging is that they have to be pretty
weak so they don't interfere with the blinking of your sample. Fluorescent
beads are usually too bright, but you may try to find some that don't get
excited too much at your laser wavelength.

You may try nanoparticles, quantum dots, etc, even pure elastic scatterers
could serve that purpose, if your emission filters are not good enough and
some very small part of excitation light leaks through... The size does not
matter, as you usually can localise fiducials much better than single
molecules. Blinking is also not an issue, as you usually don't need high
temporal resolution (the drift should be slow).

Incidentally, I've discovered very weak fiducials in one of EPFL
localisation challenge datasets (I think it was Tubulins - long sequence
here: http://bigwww.epfl.ch/palm/datasets/index.html?p=real-ls ). If you
perform (preferably with imageJ) grouped minimum z-projection with, let's
say, 150 group size, you'll see two fiducial drifting slowly...


There is infinitely (though countably, I suppose) many ways to do STORM,
just try few of them and stick with those that fit your sample-
nstrumentation-processing combination.


Btw, we mostly did cross(auto)correlation drift correction with some
success.

Best, zdenek, www.kcci.virginia.edu






---------- Původní zpráva ----------
Od: WHEELER Ann <[hidden email]>
Komu: [hidden email]
Datum: 5. 12. 2014 17:14:08
Předmět: Re: mechanism of gold fiducials

"*****
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 Paul, I can have a go:

1) What is the mechanism by which they are visible under standard excitation
wavelengths? It seems that they're based on surface plasmon resonance, and
the amount of scatter across various wavelengths varies according to the
properties of the particle. Is that correct?

Small gold particles are fluorescent, the mechanism is mostly likely the way
that other nanoparticles emit light by electrons moving between a homo and
lumo state. I would use a 561nm laser and a long pass filter to find them,
to confirm this phase contrast can be used, others may have different
strategies. However as you state the size of them does determine the
emission wavelength. The advantages of the gold fiducials is they never
blink. I would use 100nm particles from BBI gold, these can be used to
correct drift. However in our hand they do emit weakly.



2) If so, are there gold fiducials that anyone has used with success across
multiple wavelengths? In other words, using a single fiducial that is
visible
across all channels. According to the OD curves, it seems that some are
great in
the red, but not so good in blue, etc. Or is this mitigated by the fact that
even if
there is a little bit of scatter, at the powers typically used for STORM it
will be
visible regardless?

We haven't managed to visualise 100nm gold fiducials in blue. However we
have used other nanoparticles with more success (which we did report at FOM
2014). Commercial Q-dots can be used for this - although they are variable
in size our EM work shows they are 20 - 50nm, it depends on how much
resolution is required whether these would be appropriate.

3) Finally, does the size of the fiducial matter? Why would someone use 50
nm
fiducials instead of 100 or 200 nm?

Simply because of the required resolution, I would say. With the kind of
experiments I have done in the past the best I could expect was 80nm
(because spinning disks eat light). However if I was to use TIRF based
methods more could be expected. We've recently been using Q dots which we
sized by TEM since we are trying to improve resolution. The major problem is
if drift can only be corrected to 100nm this will determine the localisation
accuracy of the experiment.

Hope this helps, good luck and please do get in touch if you require more
info!

Dr Ann Wheeler
Head of Imaging at IGMM, University of Edinburgh
Edinburgh Super-Resolution Consortium manager:- IGMM



--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336."