Layo,

All arteries have a distinct layer, the internal elastic membrane, just beneath the endothelial layer. Veins either lack this structure or it is very thin and indistinct. In tissue fixed by standard histological methods, this layer is easily identifiable when stained with H&E or Verheoff’s stain. If your methodology precludes using standard tissue staining or requires antibody localization, an anti-elastin antibody should stain this layer heavily allowing you to easily distinguish arteries from veins.

Good luck,

RW

Thank you for this help Robert. How would I go about labeling all vessels within my tissues, so that I could then isolate between the arteries (via H&E stain) and the veins.

To answer that, I need to know what you are studying, what is your goal? Is your goal to enumerate the numbers, locations, and sizes of arteries and veins in brain taken from mice treated in different ways? Are you doing immunostaining for other antigens? Are you planning to do frozen sections or chemically fixed tissue sections? Will this be a light microscope or EM study or will you use some other methodology for quantitating vessels? With some further information, perhaps I or others can suggest the simplest approach.

RW

The goal is to assess vascular organization throughout the cortical layers in mice, so we need to label the vessels for analysis (which would include checking the density of vessels are different areas of the cortex). And a light sheet microscope will be used, on the full hemisphere sample. So having a method that would work with tissue samples as large as a mouse hemisphere (12 x 7 mm) would be ideal. No other immunostaining, or frozen sections will be used. Thank you for your input.

Sounds like an interesting project. I assume that the tissue be chemically fixed as with formalin or paraformaldehyde, dehydrated and then cleared with xylene. If so, then you should be able to use eosin Y as the stain for internal elastic membrane. Eosin is alcohol soluble so it will penetrate thicker tissue samples, though I’m not sure how thick. I usually use it on 10-20 um thick tissue sections, but it should penetrate much deeper than that in thicker tissue. It is also highly fluorescent, if you intend to use a fluorescence filter system, but it stains connective tissue a bright orange-red color in plain white light.

You should check a histological staining manual (Humason) for eosin staining methods. Also, take a look at a histology atlas (DiFiore or Sabotta) or textbook (Fawcett) to see what the internal elastic membrane looks like with eosin staining. It’s quite distinctive.

Hope this helps,

RW

Thank you for this Robert, your comments are very helpful. And your assumptions are correct regarding my project.

How would I go about using eosin to label just the elastic membrane, when it seems to be stain for cytoplasm? Maybe I misunderstood something..

As well, would there be an effective way to tag a fluorophore to the eosin Y, such that the vessels could be visualized using a light sheet? This would be very useful.

And thank you for mentioning those books, I've been reading them and they are exceptionally useful.

Best,

Layo

Yes, eosin stains cytoplasm, collagen, elastin, and a range of other tissue components. However, it intensely stains internal elastic membrane and IEM has a distinct structure so it ought to be easily recognizable. Unfortunately, RBCs will stain heavily with eosin. This may interfere with vessel identification. If you are able to perfuse the vascular system in the brain with buffer or fixative to remove most blood cells prior to fixation, then this problem would be avoided.

Also, histological staining is a bit of an art in that the intensity of staining can be adjusted by the concentration of the dye, the staining time, and the amount of destaining done with alcohol. The staining will never be as specific as immunostaining with an antibody or DNA staining with DAPI, but it should be sufficient to allow you to identify the types of blood vessels. Using histo dyes is a lot like adjusting brightness, contrast, saturation in photographs on the computer. The result can be optimized as needed.

As I mentioned, Eosin Y is itself a fluorophore. It can be visualized using FITC or RITC filters. You ought to just do a little pilot study to fiddle with the technique. I have not used eosin in the way you intend, so I can’t guarantee it will give you the results you want, but it seems a likely choice.

RW