Microscopy Setups - What works best?

SVSUNEURO

Interested in hearing about microscopy/imaging solutions that you are all using. Please share your imaging setup and pros/cons of each. We are interested in brain tissue imaging at levels from systems/cells to synapses using multiple fluorophores (IH)? However, any general hints and tips would be appreciated.

What technologies work best for you folks?

marzena

We use Zeiss spinning disk setup. It works fine although we have available only 10x objective, thus resolution is not the best. It is suitable for imaging up to 2 mm (mouse brain slice). It is very useful as it allows to make tiles and z-stack at the same time.

LauraLynch

We had a demo of a Zeiss light sheet system at a shared facility last month. Getting half a mouse brain into the chamber required some ingenuity. However, acquisition is very fast. The resolution isn't as good as 2-photon microscopy.

SVSUNEURO

After quite a bit of research and conversation with several labs and vendors that have done work with some of the larger labs that were on the cutting edge of this type of imaging, I built my NSF-MRI around the Zeiss Lightsheet Z.1 I was hoping to hear from the original authors / developers of CLARITY, but through the grapevine it was suggested that they are using a technology that is not yet publicly available. Spinning disk was another option, but filesize issues for very large samples was a big drawback.

SVSUNEURO

I think that though Zeiss did not identify the lightsheet for this purpose they have embraced it as a major driver for this instrument. They now have a specialized chamber and objective set for the lightsheet, not cheap...but what in science is?

LauraLynch

Yeah, we had to be creative in order to get half a mouse brain in. The commercial Zeiss setup is meant for a 1 ml syringe-wide agar plug with samples (fish larvae, fly larvae) embedded in an agar plug extruded from the syringe. We took a plastic transfer pipet, cut off the narrow end, and made an agar plug with the Focus-Cleared brain embedded into the agar at one end. It was far from ideal: we had to use tape to secure the pipet, the brain curved sideways (bad mounting), and the sample repeatedly hit the chamber's edge when it moved through the light field. Nonetheless, we did get some good images.

Another drawback is that the imaging chamber should be filled (20 or 40 ml, I don't remember which) with FocusClear. That was too expensive for a demo, so we used 40% glycerol instead. We could see the FocusClear dripping off the brain into the Glycerol. That was painful to watch.

SVSUNEURO

For several reasons, we started with CLARITY, but switched to developing SeeDB.

The costs associated with the reagents for CLARITY and the challenges associated with getting consistent results (sometimes it worked, sometimes not, even with no changes to methods).

The biggest issue with SeeDB was antibody penetration for IHC (My model is in Rat CNS), Riken could only get about 100um penetration, we are now able to get over 1mm penetration. Our issue is now the confocal we use (Olympus FV10i). It has a very limited working distance and only two choices of objectives (10x or 60x) with digital magnification and it is very slow so 3mmx3mmx500um image in three channels takes almost three full days to scan.

So, still not there yet, but getting close. Some of our images/movies are posted here: www.brl-svsu.org/SeeDB

SVSUNEURO

@LauraLynch It would be great if you could/would be willing to post images/movies of the mouse brain from the lightsheet if you could. The Zeiss rep mentioned that there were some labs working on this but didn't have examples.

LauraLynch

Here ya go:

Light sheet, Thy1EYFP mouse injected at cerebellum HVI with GCaMP6f GFP, 16 days passive clearing, imaged in Focus Clear/40% glycerol:

https://drive.google.com/file/d/0B5H6exz8a5XAZWlRMzA3bFBQeHc/edit?usp=sharing

https://drive.google.com/file/d/0B5H6exz8a5XAV0djZ0lUNmdMZ0U/edit?usp=sharing

The same sample, in FocusClear, on a 2-photon setup:

https://drive.google.com/file/d/0B5H6exz8a5XARmQtR25QMWFKNUk/edit?usp=sharing

JustDave

@SVSUNEURO Can you provide details as to the modifications you made to the SeeDB protocol to enhance IHC staining efficiency?

Varf

Hi everyone. I don't work with brain tissue however, I have the same problem with very low a/body penetration =( . At the same time with DAPI I had very good results: 200um samples was completely stained and there was no problem to make a z-stuck of images.
@SVSUNEURO I saw pictures at www.brl-svsu.org/SeeDB - they're awesome. What was the improvement that you've mentioned?

SVSUNEURO

More than happy to, just email me from your personal account (Jeff Smith : jsmith12 [AT] svsu.edu I would rather not hijack the CLARITY boards to discuss SeeDB.

We are running another couple of brains with additional refinements. But, I will say that this article was very helpful and might be helpful to others that are working with IHC with thick samples: "A Method for 3D Immunostaining and Optical Imaging of the Mouse Brain Demonstrated in Neural Progenitor Cells" Plos One 8, Published: August 06, 2013 DOI:10.1371/journal.pone.0072039 Jacqueline A. Gleave, Jason P. Lerch, R. Mark Henkelman, Brian J. Nieman

Again, my tissue is fixed adult rat brain, so some adjustment might be needed for your applications. We have been working pretty diligently on the refinement of these methods for the last six months or so. So I would expect any minor change in tissue sample may lead to very significant refinements in the procedures for your application. The other major difference is that all IHC is done BEFORE SeeDB, whereas the CLARITY protocol has IHC steps after clearing.

Varf

Thank a lot @SVSUNEURO for your reply!

I looked at the article and realized that the low penetration of a/b may be due to absence of antigen retrieval step and Proteinase K digestion (however, I always do it for 5 um paraffin IHC). Also, I was quite surprised with 100% Methanol freeze/thaw step - never heard about it before.
Guess, that's how you've got deeper staining!
Will use methanol and Proteinase K for Clarity cleaned samples

What I'm doing now to improve the quality of staining is vacuum exposure for 20min on ice for each a/body (1ry, 2ry, and Straptavidin).

Will let you know if it improves anything.

KeithSiew

@varf

I'd be very interested to know the outcome of your steps to increase antibody penetration. Are you intending to apply the methanol and Proteinase K steps before or after CLARITY has been applied?

Regards,
Keith

Varf

Hi, yes, gonna try it this week. Will let you know in 10-14 days if it works any better after Proteinase and Methanol.

Varf

@SVSUNEURO , in the article you've mentioned ( "A Method for 3D Immunostaining and Optical Imaging of the Mouse Brain Demonstrated in Neural Progenitor Cells") they use 10 mg/ml of Proteinase K. It's very very high concentration! For IHC on paraffin we are used to work with 5 µg/ml! I wish there's no mistake in the article...

SVSUNEURO

Folks, in my hands there was no need for ProK, if you look at their data the greatest improvement came after the addition of the freeze/thaw steps. In our examples mentioned above, all we used was that single additional step.

pensare

interesting SVSNEURO, but do you perfuse your tissue with 1 % PFA solution or 4 % PFA solution?

Varf

I'm trying two ProtK concentrations: 10µg/ml and 1mg/ml for 10 min at 37 degrees. Didn't take a risk with 10mg/ml - seems to be too high. And yes, before I did Methanol freeze/thaw cycles. Now it's with 1ry a/b. Will tell you if there's any improvement in couple of weeks.

Matasha_D

@SVSUNEURO‌ Hi we are trying to develop a CLARITY+SeeDB method. Have you tried combining the two? So far, we tried SeeDB alone without success (mouse hemibrain) and then combined method in which we performed CLARITY steps upto ETC followed with SeeDB with some success (brain quite clear but very swollen). Are you performing SeeDB alone, if so are you using their standard protocol or are you combining any steps from CLARITY with SeeDB?
Thanks