Sample mounting

Revision as of 13:54, 13 November 2014

How to mount samples for high resolution microscopy

To get high resolution images takes some care in sample preparation, so the microscope really can see all the tiny details it should.

We know that you cant do super resolution microscopy until you are already a good microscopist who can do normal resolution.

Here are some guidelines to help you get high resolution.

To be able to get the best resolution possible from a microscope, you need to take great care with the sample preparation:

• Correct coverslips (thickness)! Only high precision 0.17 mm exactly coverslips for high resolution lenses
  • e.g. product 1014 at http://www.hecht-assistent.de/uploads/tx_rmproducts/d54_16.pdf
  • ask for them specifically in MPI-CBG stores.
  • the nornal cheap #1.5 coverglasses are NOT suitable for high resolution microscpy as they vary in thickness from 0.16 to 0.19 mm.
  • This is the wrong place to save money. Use the proper ones - they are not expensive.
  • number 1 (#1) coverslips are only for thick (many tens of microns) histological sections, and not for our typical thinner samples.
  • Dont use #1 coverglasses unless you really have the right thick sample - ask if you are unsure.

• Known mounting medium with known refractive index and antifade agent.
  • Remove all excess water/buffer/salts - or even soak/wash in the mounting medium before final mounting
  • Prolong gold is good (as is soft setting vectashied).
  • Mowiol and hard setting vectashield can cause problems with sample smashing artifacts and refractive index inhomogeneity.

• To get high resolution, you need to use a high quality objective lens with a high numerical aperture.
  • High resolution has NOTHING to do with high magnification - magnification is easy - the image on the auditorium screen is really big - but is it high resolution? No.
  • Abbe told us that the resolution is related to the angle of light that can get into the lens, the NA. Bigger NA is better resolution, and you get more light too.

• Correct for spherical aberrations using the right immersion oil / objective lens correction collar setting.

Dead Fixed Samples

Different samples require different ways of sample mounting, with different mounting medium and coverslips of different thickness.

If you get this wrong, your images will suffer from poor z resolution and loss of intensity due to spherical aberration caused by (A) refractive index mismatch between the sample and the lens immersion medium and (B) incorrect sample+coverslip thickness for what the lens is corrected.

There is no substitute for getting the optical properties of the sample matched to the lens you are using. If you do that, the green things will look rounder in z... but they will never "look" isotropic (= same width in z and xy), always a bit "stretched" in z ... thats just how the optics work.

Remember, the image of a point is not a point, its a blob called the Point Spread Function (PSF), which is larger in z and in x and y. See PSF at SVI homepage.

Q) What does that mean for you?

Use coverslips that are exactly always the same thickness (in your case, usually those nice high precision #1.5 from stores which are exactly 0.17 mm).

(The total thickness of the sample and the coverslip should be about 0.17 mm - the lenses are corrected for that thickness) for a very thin sample, like a cell monolayer, only a few microns thick, then use a #1.5 coverslip which is 0.17 mm, but for a very thick histological section that is 200 microns, use a #1 coverslip that is 0.15 mm. So than section plus coverslip = again 0.17 mm.)

If your tissue sections or samples are about 20 microns ...... which coverslip should you use .......? Probably the number #1.5 (0.17mm), because that gives about 0.172-0.175 mm thickness. If you use #1 coverslides, then that 0.15 mm plus 0.002-0.005 mm = about 0.155 mm .... which is very far from 0.17 mm. And in that case the images will be more blurred in z than they should be (if using an oil lens corrected for 0.17 mm, or if the water immersion lens correction collar is set to 0.17 mm !!!)

Watery samples, eg. living samples

For a watery sample, you might be able to embed in 2% low melt agarose in buffer or 50:50 glycerol : buffer.... use a water objective. The correction collar in the fancy C-Apochromat 40x, 60x or 63x 1.2 NA allows you to correct for the coverslide thickness, from 0.15 to 0.19 mm. So if you use that properly you can use #1 or #1.5 cover slides. But you need to get that right ..... or else you get? Yes! Blurry images in z direction!

For living samples there are high quality glass bottomed sample chambers and dishes to fit on inverted microscopes, be careful to get only the ones with high precision 0.17 glass or the equivalent quartz on them:

• Mattek
• IWAKI seems to have vanished
• IBIDI have correct optical path length plastic bottom dishes.

Ibidi is nice in a sense that they send you free examples of their dishes for testing. Just have a look. They do not have glass bottom but some kind of plastic of the correct optical path length like 170 micron glass. Not sure how precise or reproducible they are.

Other companies making chambers:

Warner http://www.warneronline.com/product_info.cfm?name=Find%20Chambers%20by%20Application&id=1390

Pecon http://www.pecon.biz/?page_id=1504

or Live Imaging Services http://www.lis.ch/thechamber.html

This one actually looks quite promising. Since this is a rather small company, they might have customized chambers as well or might make them (depending whether at some point Hartmut will be cheaper again...).

dead dehydrated samples

Use an oilimmersion lens with high NA. Dont use a hard setting mounting medium like Mowiol as it smashes the sample when it set and gives bobly artifacts. You can use prolong gold or fluoromount G for example. Aim for a sample plus coverslip thickness of close to 0.17 mm, since the objective are expecting that to be the case.