Pulling huntingtin-associated DNA repair proteins out of HD patient cells

Blog post by Dr. Tamara Maiuri

Now that the cross-linking, fractionation, and oxidative stress conditions have been worked out to pull huntingtin and its interactors out of the “easy-to-work-with” HEK293 cells, it’s time to try out the system in HD patient fibroblasts.

These cells come from real HD patients and controls such as a sibling or spouse. The Truant lab is working on developing a panel of cell lines from different patients with different CAG lengths in their huntingtin genes. We have immortalized the cell lines with hTERT, which allows us to grow them indefinitely—a great resource to be shared with the HD research community.

One thing we and others have noticed is that the cells from HD patients grow much faster than the control cells. Both types of fibroblasts grow quite slowly and don’t yield much protein to work with. This means a fair bit of waiting around for cells to grow! Since the HD cells grow faster, there’s always more of them. For this reason, I used HD cells, bearing 43 CAG repeats, to test out the co-immunoprecipitation conditions previously worked out in HEK293 cells.

In this experiment deposited to Zenodo, I found that the conditions work quite well to pull huntingtin and its associated DNA repair proteins out of HD patient fibroblasts. One problem I ran into (other than the sloth-velocity growth rate), is that the mab2166 antibody doesn’t pull down huntingtin very well out of 3NP-treated fibroblasts. This was confirmed in a second experiment, also added to Zenodo.

Luckily, mab2166 is not the only huntingtin-specific antibody available. I compared the ability of another antibody called EPR5526, to pull down huntingtin and its associated DNA repair proteins. I found that even though EPR5526 pulled out slightly less huntingtin protein, there was more of the DNA repair protein APE1 associated with that pool of huntingtin protein. It could be that the EPR5526 antibody better recognizes huntingtin in the conformation it takes upon oxidative stress (previous work from our lab showed that huntingtin changes shape upon oxidation, which may also explain why mab2166 doesn’t recognize it as well from 3NP-treated cells). Whatever the reason, it looks like EPR5526 is the way to go for this application.

So the preliminary work is done and I now have the conditions right to identify the proteins associated with huntingtin upon oxidative stress by mass spectrometry. The slow growth of the fibroblasts is a major limiting factor, however. After speaking with the Sick Kids mass spec facility about how much protein is needed, I project it will take several months of growing up cells, treating them with 3NP, and freezing them down until I can collect enough material. Look for an update on this front in the fall!

In the meantime, I’ll be looking ahead at how we’re going to analyze the huntingtin interactors that we identify. The point of this project is to find proteins important to the DNA repair process in HD. So I need to find a meaningful way to measure differences in DNA repair between normal and HD cells, then test the effects of the interacting proteins on those differences. We’ve already shown there’s more DNA damage in HD fibroblasts compared to control cells, using a “comet assay”. But that system is labour intensive and not the best choice going forward. I’ve been exploring another way to test DNA repair in cells called a “GFP reactivation assay”. I’ll report my progress in forthcoming blog posts.