Is huntingtin a PAR-binding protein?

Blog post by Dr. Tamara Maiuri

Previously I explained our rationale for hypothesizing that huntingtin may bind poly ADP-ribose (PAR). If so, this could be the way it gets to damaged DNA, and this might be dysregulated in HD. Since we know DNA repair genes impact whether HD patients get sick early or late in life, this is a good place to look for problems—then we can look for drugs that fix the problems.

But first things first: does huntingtin bind PAR? There are many “domains” or regions of proteins that are capable of binding PAR: PAR-binding motifs, macrodomains, and WWE domains to name a few. A quick scan of the huntingtin sequence revealed four potential PAR-binding motifs. We can look at the recently solved huntingtin structure to get some more clues:

This is pure speculation at this point—many regions of the protein were left out of this structure so it’s too early to know for sure—but it’s fun to guess: the barrel of the huntingtin solenoid is the right size to accommodate DNA, as it is the same size as the DNA binding regions of other DNA repair proteins such as MSH2, MSH6, and PCNA (pictured above). Similar to MSH2, three of the potential PAR-binding motifs within huntingtin are exposed to the outer surface of the DNA binding region, suggesting a mechanism by which huntingtin is recruited by PAR, followed by direct binding to DNA. (Direct binding of purified full-length huntingtin to DNA has been shown by Dr. Rachel Harding and deposited to Zenodo: https://zenodo.org/record/801606#.Wr4eG5PwZ-V).

One way to get a clue about this is to immobilize huntingtin protein on a membrane, then overlay it with purified PAR polymer. If huntingtin binds PAR, it too will be stuck to the membrane. After washing away unbound PAR, you can detect whatever is left with an anti-PAR antibody. This is called a PAR overlay assay.

The first few experiments (deposited to Zenodo) look promising. Purified full length huntingtin (produced by Dr. Harding) reproducibly bound PAR in several experiments, as did a fragment made up of amino acids 78-426 (which conveniently contains one of the potential PAR binding motifs mentioned above). I also tried two preparations of expanded huntingtin (Q46 and Q54) in one experiment, with confusing results: huntingtin Q54 bound PAR while huntingtin Q46 did not. This could be because the Q46 prep was from an older stock—I’m currently repeating the experiment to find out what’s going on.

These are very early results, and I need to make mutations in the potential PAR-binding motifs to see whether they are specifically mediating the PAR interaction, or whether this is nonspecific binding. But we may be on to something here… in the next blog post I’ll share more preliminary data implicating PAR in huntingtin chromatin recruitment in cells.