Hi Leena,

I have some questions and suggestions:

Did you fragment your chromatin prior to Western-blot sample preparation? This helps a lot to release the proteins from crosslinked complexes. Use samples at the step prior to IP (Lysed, dounced, sonicated).

Did you run a positive control for your protein-of-interstest (lysates from same cells that havn't been crosslinked)?

Don't remove stacking gel to check for aggregated protein which might be stuck at the bottom of gel pockets.

Use fresh sample buffer. Incubate 10min at 96°C.

You could also try to only use 1%formaldehyd (freshly prepared from PFA) 5-10min, RT instead of DSP. This worked well in my hands.

What kind of antibody did you use? Poly- or monoclonal?

In my experience polyclonal Ab worked better, since the not all binding epitopes will be masked by crosslinking or other proteins in complex, which will lead to better enrichment during ChIP procedure.

Best,

Kai

Hi,

Thanks for the quick answer :-). I'm making sure we are talking about the same thing.. Is chromatin fragmentation helpful for protein-protein immunoprecipitation or are you talking about chromatin IP?

I did run a control and proteins looked fine there. I attached a pdf, hopefully you can see results from there. I have cut out lanes where I had IP, but results were such that control IP worked fine and nothing/very little was in IP from cross-linked extract.

I didn't check total protein by Coomassie blue staining or other similar protocol, so that's one thing I could do to double check there is same amount of protein in extract from crosslinked cells as in control. At least cell pellet was similar size and I'm assuming cell lysis works for crosslinked sample like it works for control cells. But maybe it's possible cell lysis doesn't work as well as it works for control..?

Did you use 1% formaldehyde for detecting protein-protein interactions? Is there an inactivation step?

Antibodies that I used for WB detection were mostly monoclonal. I had one polyclonal ab but result was similar (very little protein detected in cross-linked extract).

For IP, I used GFP-Trap beads that pull down GFP-tagged proteins (with these beads no antibody conjugation to beads is necessary).

Hi,

sorry, I got a bit confused, since I read ChIP instead of Co-IP. This might be due to the fact that I cannot recall a publication were an IP was done with cross-linked extracts.

If I get you right, you transfected your cells with a GFP-Fusion protein and did an IP on control and crosslinked extract to detect a protein-protein interaction between two proteins, right? When doing so, you really have to make sure that reversal of the crosslink works before performing a WB.

But, if you are only able to detect a positive IP after crosslinking, what does it tell you?

I don't know, if you really want to hear this, but reviewers will probably ask you to do an IP without crosslink, with endogenous proteins, with antibodies directed against your target protein...... And most probably they will ask you to test for direct direct interaction with recombinant proteins, if the interaction you want to investigate is a crucial point in your story.

Best Kai

Hi,

I have a cell line that is stably expressing a GFP-fusion.

I'm trying to co-IP a kinase and its substrate protein. There is substantial evidence that there are multiple phosphorylation events between these two but not much evidence when it comes to co-IP. I found at least one article where an interaction was shown by co-IP. Both proteins were overexpressed.

I have thought about the fact that recombinant proteins may have to be used. This thing I'm trying to analyze is a bit complicated and I'm not sure if this event I'm looking at would occur with purified proteins because co-factors may be involved and I may have to look at this in cells. I have also done co-IP with endogenous proteins using ab against the target protein trying to see the kinase there but interaction is so transient that I cannot see it. That's why I thought I'd try the crosslinker.

At this point I would like to get ideas what to do next. Is the problem reversal of the crosslink, problem in antibody detection (proteins modified by DSP and not recognized anymore), problem in cell lysis or something else.

What do you suggest is the best way to check that reversal of the crosslink works?

Hi,

What I do to control proper crosslinking and its reversal is to take aliquots of total protein samples before and after crosslinking. Aliquots taken after crosslinking are added with protein sample buffer including and excluding reducing agents (DTT or mercaptoethanol, compounds that disrupt DSP). If the crosslinking works fine and you blot a protein known to interact or form complexes, you should see a shifted migration in the lane corresponding to the crosslinked sample prepared with no reducing agent. In the sample prepared with reducing agent the target protein should show normal migration, the same as in the pre-crosslinking sample.

I also see some differences in Coomassie stained gels (crosslinked samples give somehow messy bands in the upper part of the gel), but I believe blotting a complex-forming protein is more conclusive.

I hope this helps somehow.

Regards.

Hi Leena,

My experiment is conjugate antibody EGRF with amine groups on particle surface (Polydopamine coating silica) by DTSSP. I also use NHS for the process. However, it  was  aggregated after conjugating and could not combine. Would you like sharing your experiences about it ?

 Hi Leena,

I have worked with DTSSP, the "impermeable" variation to DSP and using a protein tagged with FLAG epitope, this protein started to dissapear in presence of the crosslinker, I found that this FLAG epitope has 2 lysines and it is used as test control of the crosslinking reaction by the entreprise producing crosslinker. So my advise is check the epitope of your antibody and yes, DSP can prevent the detection .

Best, 

Heidy

I've used Flag-tag to capture proteins after DSP crosslinking. It's possible that the crosslink interferes with the GFP-trap, several of my colleagues have had issues with GFP-trap being really dirty in IP so they all switched back to using Flag-tag. I make sure that the 2-mercaptoethanol in my SDS sample buffer is fresh and boil a little longer (15 minutes) to make sure the disulfide bonds in DSP get reduced prior to SDS-PAGE.

Hi Leena, were you able to troubleshoot the problem? I am having similar problem with DSP cross-linker, my protein concentration is greatly reduced in cross linked sample compared to the control. I use 1mM final DSP(dissolved in DSMO and diluted in cross-linking buffer) to do intracellular cross-linking of my protein of interest in HUH7 cells and the control gets only DMSO in cross linking buffer. I also observe cell clumping after adding the cross linker, and when spun down the cross linked cell pellet looks larger than the non cross linked sample. Would appreciate your input!

Hi Taslim, how many cells are you using? I think that the ratio of DSP/cells (and crosslinking volume) is more critical than just the final DSP concentration. I use 1.25mM DSP on a fairly large number of cells (5x10^8 - 1x10^9 cells in 20ml PBS). I would expect that over-crosslinking might negatively effect protein extraction (and potentially disrupt antibody binding sites). You may want to do a titration of DSP with a set cell concentration and volume. My crosslink protocol is to incubate the reaction for 30 minutes at room temp in PBS and then quench by adding 20mM Tris. Since the sample is crosslinked, you can extract using pretty harsh conditions (SDS etc) so long as the anti-body will still work under those conditions.

Marc Alard Morgan I think you are very right about DSP/cells ratio and crosslinking volume. Every time I added DSP to 8x10^6 cells in 1ml PBS. In an effort to troubleshoot yesterday I repeated the experiment with 1mM DSP in 4ml cell suspension(8x10^6 cells in PBS) and observed much better protein concentration in my DSP treated lysate, although not same as in control(I did both western blot and commassie stain). May be further dilution of cell density would do better. I would definitely try your suggestions of DSP titration with different cell concentration and volume.

Thank you so much for your suggestions!

Taslim Khan another thing you could do to test the extent of crosslinking is run western blots against your target using both reducing and non-reducing conditions. I would expect to see a high molecular weight smear form under non-reducing conditions if there is crosslinking. You probably want to find the ideal ratio of DSP/cells that gives you sufficient crosslinking of your target but also allows for efficient protein extraction.