The most likely problem is that the concentration is too low, or the cell is dying (and losing dye) before you get to the staining. Your image is low mag, so it is hard to tell what is going on, and paricularly how well labelled the soma is.

Can you provide more details about how you are staining the cells? Are you doing whole-cell patching, or sharp electrodes, in slices or in vivo? What is your electrode solution composition? How are you doing the DAB reaction? Are you following instructions on an ABC kit? How thick are your sections? Have you tried a fluorescent dye instead, and seen substantially different staining?

-Matthew

Hi Matthew! Thank you for your kind help! We use 0.5-1% biocytin, should be enough? We also tried Streptavidin AlexaFluor, and found same problem. I am doing whole cell patch. The slice is 300um. I think probably the dye is leaking. Now I am trying to recover the slice for 30min before fixing it.

When I tried doing something similar years ago, the problem was in pulling off the cell. I was using fluorescent dyes (Alexa 488 hydrazide). Most of the time, when I pulled off, I could see that the cell just faded away, most likely, as you say, because the dye is leaking out from a ruptured cell. I happened to pull off gently enough that one cell stayed intact, which was all I needed at the time, so I didn't really work out the optimal conditions.

A grad student in my lab did this too using neurobiotin (0.2%, see Ngodup et al 2015, PNAS). I seem to remember that to increase his success rate, he used patch pipettes smaller than our usual 1-2 MOhm, like 5 to 10 MOhm. I don't remember if he had better luck when he pulled away very very slowly, or very very fast.

My suggestion would be to refine your approach using fluorescent dyes that you can observe during the experiment, and then switch to biocytin/neurobiotin when you get that working reliably enough.

-Matthew

Thank you for your suggestions! I will try it.