Cadherin in Zebrafish

Dear Moe,

We can certainly target a new epitope in the C-terminus. Before we start, could you try a brief experiment. We previously found that cadherins are best labeled with methanol fixation. Because adherens junction is packed into a polymer, PFA is unable to fully exposed the proteins in the matrix. Could you follow this protocol and try the cells:

https://bicellscientific.com/education/article/protocol-for-staining-cadherin/

BiCell Tech Team

Dear BiCell team

Sorry for being late to report our results.

As you suggested we tried methanol fixation. I think it worked! The pre-fixation time by PFA before methanol seemed to be a critical step. However it’s still weak. I’m wondering if we could really detect the endogenous CDH6. I’ll attach my protocols and results in a PDF file (cdh6 validation-2-1.pdf). Please tell us if you think we could improve anything.

We also did western blotting and it also went well. Next we are planning to check by mutant fish samples.

Best regards,
Moe

Hi, Moe,

Please simply dissect out tissue or embryo, rapidly freeze on dry ice and cut cryosection at -20oC. Then directly dip cryosections into cold methanol (precooled to -20oC) and fix for 20min at -20oC. Take out slide and air dry 30min. Then rehydrate with PBS and follow staining protocol.

Very likely this is a great antibody!

BiCell Tech Team

Hi, BiCell

Thank you for your suggestion. I already tried that protocol. But it was difficult for us.

Since early stage embryos are too small, we routinely perform whole mount staining than making cryosections.

So I tried that protocol in whole mount staining. I’ll attach my protocols and results in a PDF file (cdh6 validation-3-1.pdf).

Unfortunately, without PFA fixation, embryos collapsed and we could not even move on to the blocking step.
That is why I examined different PFA pre-fix condition. Sorry I should mention that first.

Now I think it might be better to stop examining at this early stage.

Because the time point of my interest is around 24 hpf which is capable for making cryosection.

Over expressed CDH6-HA might not exist till that time point so, as a positive control, I will check the signal at the kidney which is already reported. I guess it takes a little time to check, because we are not familiar to cryosection, but I think we have to do it.

I hope we are getting closer to the goal!

Best regards,
Moe

Hi, Moe,

Very promising data! I think this antibody will work, and anti-extracellular domain antibodies are always more useful than anti-intracellular domain antibodies. This antibody may also be used to label live cells in flow cytometry or even neutralizing studies. I think our strategy of using short-peptide can better detect protein in its native conformation.

https://bicellscientific.com/about/

Regarding experiments, I would encourage you to try a bit more on whole mount staining since it is very promising. We previously tried ZO-1 on mouse brain slices (100um vibrotome sections) and they worked well. You might want to image with 2-photon instead of a regular confocal since it can give better depths. Methanol fixation doesn’t work so well for whole mount since when sections are thick (100um), methanol cannot fully replace water in cells. Mysteriously, when we do ZO-1, we used 2% PFA fixation instead of methanol. The important factor is to do an overnight antibody incubation in 0.2% Tween containing BLOCK buffer. Antibodies generally only permeate ~100um. We found that when vibratome thickness is over 200um, no binding can be found.

You could also try to cryostat sections of fish embryos at this stage. A simple maneuver is to combine several embryos before putting into OCT, so that you can see where the sample is when sectioning. We also use low-melting agarose to immobilize small samples and cut out agarose pellet and mount into OCT. We previously tried this on organ of corti from inner ear of embryonic mice and surprisingly, it is not hard to spot the tissue when cryostating.

BiCell team

Hi BiCell

Thank you for giving us a lot of suggestions.

I also tried both immunostaining and western blotting for CDH6.

I think your antibody is working very well for immunostaining in both overexpressed samples and normal samples (validation of cdh6 antibody.pdf).

However, we can’t get a positive band in western blotting.
I’ll attach my results.

As I’ve told before, I injected cdh6-HA-P2A-H2B-GFP mRNA to the 1 cell egg as a positive control.
And collected the embryos at 4-6 hpf for immunostaining and western blotting.

I could detect the HA band (blue arrow) but couldn’t detect CDH6 band.

Is it difficult to detect CDH6 by western blotting by this antibody?
Or are there any points to optimize?

Best regards,
Moe

Hi Moe and Keisuke,

This is very beautiful results with methanol fixation! Congrats!

I would recommend to stick to methanol fixation for adherens junction. It’s very mysterious why methanol will work better than PFA. Probably related to how membrane interacts with cadherin proteins in the cell junction.

I am a little confused with the construct design. Is there an IRES sequence separating Cdh6 from H2B-GFP? I figure that it cannot be fused together because H2B (histone 2b) will alter cdh6 protein localization and move it to the nucleus.

It’s hard to say whether the high band in the HA column is real. Are you using the monoclonal antibody for HA from Roche? The pattern doesn’t look like mAb since mAb will give very clean Western.

I am not very familiar with the protocol for fish embryo. The key issue is the amount of protein you can get from the embryo. Our experience with Western is that the more sample to start with, and the more enrichment steps to include, the better results to achieve. I assume you were using whole cell lysate with SDS 1% lysis buffer. Cadherin protein abundance in whole lysate is probably too low. It might be good to pool a few hundreds of embryos, sonicate the membrane and extract membrane proteins with RIPA buffer. Cadherin abundance after these steps will increase very significantly.

BiCell team

Hi BiCell

Yes! I believe we can get acceptable staining data using this antibody! And thank you for your comment.

I inserted P2A sequence between cdh6-HA and H2B-GFP which would be cleaved after translation.
I think this is working. Because from the immunostaining results, HA and CDH6 was detected at the cell membrane while GFP signal was localized in the nucleus.

I think the band size of WB is little bit wired. If we assume the HA band is a real one, it is too large. CDH6-HA is 90 kDa but the band size is 130 kDa which is an appropriate size for un-cleaved premature CDH6-HA-P2A-H2B-GFP. So I think I might failed to collect the membrane protein and only collected the cytoplasmic protein. This time I used a brief method (I’ll attach the reference paper) to extract embryonic protein and did not use RIPA buffer or other lysis buffer. Maybe that was a problem.

And as you pointed I used monoclonal antibody for HA from Roche (3F10). I don’t know why there are so many extra-bands.

Well I will try again by increasing the sample size and using an appropriate protocol for membrane protein.

Best regards,
Moe

You must be logged in to add new comments. Please log in by clicking the link below:
Register or Login