The N6-furfuryladenine Hypothesis in HD

The summer of 2018 marks the third manuscript to come out of the lab defining the role of huntingtin in DNA damage repair, and a defective signaling pathway in HD caused by the proximity of expanded glutamine tracts to the kinase substrate site in huntingtin N17, the first 17 amino acids of huntingtin.

Since 2011, we have defined this site as hypo-phosphorylated in HD.  Since 2011, academic and industrial labs have reported that huntingtin is under-phosphorylated at many sites, and in all model and samples taken from humans.

(Data from Hung et al., 2018, showing hypo-phosphorylation in human HD fibroblasts from clinic). Here’s Claudia:

 

With the help of a donor gift of a Nikon A1 confocal microscope, the CFI/OIT Leaders Opportunity Fund for a super-resolution SIM device, Laura Bowie set on to her PhD thesis work with CIHR doctoral Scholarship to set up an unbiased screening protocol in which a robotic stage took random shots around a  96 well plate well, with a library of natural compounds applied. This was supposed to be a pilot screen of only 130, to test out the microscopy acquisition and then use non-supervised machine sorting to sort the images, so from front-to-back, there is no possible investigator bias.

Here’s Laura:

However, we got hits, most of them were anti-oxidants, which lead to our 2016 discovery that huntingtin was a ROS sensor via a single methionine at position eight, and this oxidation proceeded phosphorylation. The other hits affected the NFKb/ IKK kinase pathways, pathways we discovered in 2011. But we got one lone hit, distant and distinct in 3D prinicipal component analysis space: N6-furfuryladenine.

The rest, is now recent history.

The net results: YAC128 mice get better, and brain levels of huntingtin drop. Huntingtin hypo-phosphorylation is now restored to normal, with sub-micromolar levels required.

This was all outstanding mouse work from Melanie Alpaugh and Simonetta Sipione at the University of Alberta.

But what is N6-FFA? 

An astute eye will find an adenosine core, the same core of nature’s energy molecule, ATP, but count the positions around the rings from nitrogen 1, and at nitrogen 6 you will find a furfuryl ring -the 5 -sided ring with double carbon bonds and an oxygen. In DNA, this is what happens when you oxidize DNA, and this isn’t good. Instead of pairing with thymine (A-T), it can pair with guanine (A-G) -that’s not supposed to happen. So, the DNA gets fixed. In dividing cells, it get fixed by mismatch repair, which sees the N6-FFA:G mismatch and removes it. But in brains, neurons don’t divide, so they rely on Base Excision Repair or Nucleotide Excision Repair, in which a single nucleotide is removed and replaced, or a small patch of DNA. This is how oxidized guanines are removed and throw out of the body as garbage. We can even detect N6FFA in human pee.

N6FFA -Reduce, Reuse, Recycle

Neurons are weird cells. They don’t work like most other cells. They don’t divide, they are huge and spindly and they are energy hogs. Over half the energy in your body is used in the brain at rest, and it doesn’t really slow down or take a break. This means the brain takes huge amounts of fuel, and burns a lot of ATP. Neurons also rely heavily of recycling nucleotides, or salvaging, because at times, they run out of fuel and ATP is just not around. So the brain makes energy like other cells by oxidative phosphorylation and glycolysis (blow dust off the old biochemistry textbook). What happens if we don’t have the enzymes to salvage nucleotides? We get severe diseases, usually fatal in children. All this burning means pollution, in the form of reactive oxygen species, or ROS. Neurons need to get rid of ROS, or ROS will go nuts and react with everything around it, especially DNA.

Enter Nick Hertz

This is Nick.

In 2013, the Kevan Shokat lab at UCSF, with student Nick Hertz, did a screen to find molecules that could be used by a mutant form of Pink1 kinase, a form found in families with familial Parkinson’s disease.  (digression, Nick Hertz is the great Grandson of Gustav Hertz, who won the 1926 Nobel Prize with James Franck in Physics). What Nick discovered was that N6FFA can be salvaged to form weird triphosphate, with that adenosine core, and this triphosphate can be used by mutant Pink1 to restore it’s activity lost in Parkinson disease. We know this is also true to HD because a version of N6FFA that cannot be salvaged doesn’t work to fix mutant huntingtin hypo-phosphorylation. so, N6FFA is not the active molecule, it is a pro-drug, a compound that gets converted to an active form. This is a “neo-substrate” for a kinase, not ATP or GTP. This is an aspect of drug discovery called pharmacokinetics, or what the body does to the drug. Pharmacodynamics is what the drug does to the body. Both are essential to understand to make a lead like N6FFA into a drug. With David Litchfield, (another uncool biochemist) at  Western University, he could show that the huntingtin kinase, CK2, can also use this “neo-substrate”, while the Shokat lab thinks there are only two kinases that can do this: CK2 and Pink1. (digression #2, Litchfield was a graduate of McMaster Biochemistry Department).

“But why not just use ATP?!?” – several reviewers and pharma executives

Somehow in the last 40 years, biochemistry became uncool. We don’t teach it in the detail of the past because science is just moving too damn fast and we have more and more stuff to teach. It’s too bad, because biochemistry is what led to drugs that worked 40 years ago, and still work today.  It’s also complex, with steps and pathways and feedback loops and big wall charts no one looks at. The human neuron can undergo energy crisis, times where all the ATP is burned up, under periods of stress. Stress like high ROS levels due to human aging, which only increase as the brain gets older. DNA damage triggers an even called PARylation – PAR is poly-ADP-ribose, these are chains that grow out of sites of DNA damage that act like nets catching PAR-binding proteins, and when they form, they need to get removed or they will inhibit energy metabolism by draining NAD+, halting glycolysis, and ATP release from the energy plants of the neuron, the mitochondria. So, sorry “experts” but ATP is not universal and always abundant, and energy deficits is a long-standing observation in HD, which makes the comment from an HD expert rather bizarre.

There is defective DNA repair in HD (2017)

There is defective DNA repair in HD (2018)

The point is, there is defective DNA repair in HD.

To read more about PAR and PARylation, see the work in real-time with Dr. Maiuri.

What does this say about the Amyloid hypothesis in HD?

The amyloid-like hypothesis in HD is fraught with problems. There is no explanation as to why aggregates of huntingtin can accumulate for decades, when neuronal protein half-lives don’t exceed 20 days.

Plus, the whole AD drug thing.

Regions with aggregates don’t map to regions of pathology in HD, and while there are phenotypes of disease in human cell from HD patients, we can’t find aggregates unless we force them to happen in tiny fragments overexpressed.  This is explained  by much hand -waving in the research community for 25 years, typically based around synthetic hyper-allele models of disease, that usually involves cutting off 97% of the protein.

The effect of N6FFA, and HD Genome-Wide Association Studies, suggests that aggregates are an effect, not cause in HD. DNA damage and energy crisis proceed protein misfolding. Misfolding is just a symptom of a sick neuron, particularly a sick ER. When we restored this signaling, mutant huntingtin inclusions disappear.

So, by unbiased microscopy screening, we hit the same answers as by unbiased GWAS in HD 2015. and again in 2017. Maybe hypothesis-driven research was not the best approach for HD.

 

So N6FFA is a drug? 

No. N6FFA is a lead. As a compound, it has very poor pharmacological properties for dosing in humans, but, derivatives are looking very promising to overcome these hurdles already, we just need to make sure we don’t gain any toxicity (N6FFA is a natural human metabolite). This is dull, pedantic work that needs to be done and tested in more animal models before we plan to trial in humans.

We also present the N6FFA hypothesis in a Youtube video abstract: