Blocking a toxic fragment: new mouse study highlights the importance of HTT1a in Huntington’s disease

Anew study using precise gene editing in a Huntington’s disease (HD) mouse model provides evidence that a small fragment of the huntingtin (HTT) protein, called HTT1a, is a central driver of HD. By reducing HTT1a levels, researchers delayed signs of HD like protein clumping, restored which genes are switched on or off, and brought biomarkers back in check in mice that model HD. Uncovering the most toxic form of HTT driving disease could have important implications for how we design HTT-lowering drugs.

A quick refresher: CAG repeats and HTT1a

Everyone who has HD has a change in the genetic code of their HTT gene. A repeating run of C-A-G DNA letters expands to have too many repeats. People with 40 or more CAGs in their HTT gene will develop the disease if they live long enough, and longer repeats are generally linked to earlier onset.

The expanded CAG repeat in the HTT gene causes a corresponding expansion in the encoded HTT protein, which is made from these genetic instructions. But the CAG repeat doesn’t just change the protein sequence. It also affects how the intermediate between the DNA and protein – the messenger RNA – is processed.

A copy of the genetic recipe

Messenger RNA is a genetic message molecule. It is made as a faithful copy of the encoded DNA sequence for a given gene, before being used to make the protein this sequence encodes. This is a bit like making a copy of a recipe from a book so you can cook the recipe without having to go back and forth to the original book itself. 

When the CAG repeat is expanded, a hidden signal in the HTT messenger RNA can be activated. This signal is called a “cryptic polyA”, and tells the cellular machinery to stop making the HTT protein too early. This means that instead of making the full-length HTT protein, the cell makes just a small fragment instead, called HTT1a. 

HTT1a: a toxic HTT fragment protein

The HTT1a fragment protein contains the expanded region, encoded by the CAG repeat, which looks like lots of the same amino acid glutamine, repeating over and over. This fragment is highly prone to clumping (a process called aggregation) and has been shown in many animal and cell models of HD to be toxic. 

Importantly, the longer the CAG repeat is, the more of the fragment HTT1a protein that is made. Because the CAG repeat can get longer in some cells of the body in people with HD, a process called somatic expansion, HTT1a is thought to be a key step in how the disease progresses. 

To better understand this, researchers in this study wondered if the increase in the amount of HTT1a that gets made could be the driver linking somatic expansion to nerve cell damage.  

A clever genetic strategy: deleting the fragment causing signals

To test this directly, the team used a mouse model of HD which has almost 200 CAGs in the HTT gene. This is a much higher repeat number than is typically seen in people, but this expedites when symptoms begin to show in the mice, allowing researchers to get answers to their experiments faster. Due to the idiosyncrasies of HD model systems, these are actually called HdhQ150 mice.

Using CRISPR tools, which are like genetic scissors, they deleted a portion of the HTT gene which contains the signal to make the HTT1a fragment form. This should prevent the toxic HTT1a fragment being made in these mice even though the CAG number is high enough to trigger the cells to create the toxic fragment. This new type of HD mouse model was named the HdhQ150ΔI mouse, with the Δ (delta) representing the genetic domain that was removed. 

They also used mice with CAG repeat numbers that do not cause HD to represent the general population without HD. For these mice, they also chopped out the HTT1a-making signal so they had a control for the possible effects of genetic editing, unrelated to the CAG number. This meant they would know if the results they saw were because of the CRISPR editing itself, or truly because HTT1a wasn’t produced. 

What happened to HTT1a levels in these mice? 

As expected, HTT1a levels dropped off in the HdhQ150ΔI mice. The HTT1a RNA message molecule was completely absent in these mice and protein levels were reduced, but still detectable. 

But why would some HTT1a still be found? The researchers suggest that this deletion might have led the genetic machinery to “readthrough” the gene differently, in a way that still produces HTT1a protein, just in very low amounts.

Importantly, levels of full-length HTT, both unexpanded (regular) and expanded (HD form) were unchanged. This is important for downstream experiments so the team could figure out what effects were specifically due to HTT1a reduction, not total HTT levels. 

A major delay in aggregation

A striking result from this study was that toxic HTT clumps, or aggregates, were found in the brain much later for the HdhQ150ΔI mice compared to HdhQ150 mice. 

In standard HdhQ150 mice that model HD, aggregates were spotted in different regions of the brain in mice at 6 months old and continued to increase as mice aged. You may be thinking, “Whoa! 6 months is really young!” And you’d be right. The quick timeline of HD disease features in these mice is caused by the very high CAG repeat number and is intentional so that results can be achieved faster.

However, in HdhQ150ΔI mice that didn’t make the toxic HTT1a fragment, these clumps were delayed by several months, and in some parts of the brain, not evident until the mice were much older. These clumps still contained HTT1a, reinforcing the idea that this HTT fragment is important to kick off formation of these clumps. 

Effects on gene expression

Next the researchers looked at which genes were switched on and off in these different mouse models. This is a well-documented change that happens early on in different HD models, with a specific signature of gene switch changes. 

In the HdhQ150 HD mice, the researchers found that 1,200+ gene switches were altered at the 6 month mark in the striatum, a deep region of the brain most impacted in HD. At 12 months, this reached 2,800+ genes. 

In the mice with less HTT1a, some of these gene switches were restored, with a 25% improvement at 6 months and a 40% improvement at 12 months. When they looked in another part of the brain called the hippocampus, they saw partial improvements there too. 

Interestingly, even the low residual HTT1a levels in HdhQ150ΔI mice were sufficient to start to fix this hallmark of HD. 

Biomarkers: a striking effect

Perhaps the most exciting result in this study came from biomarker levels that the researchers looked at in plasma and spinal fluid that bathes the brain, also called CSF. The team measured NfL, a general biomarker of brain health, as well as BRP39, the mouse equivalent of a human biomarker called YKL-40 that tallies with inflammation. 

In the HdhQ150 HD mice, both biomarkers were increased in CSF by 12 months, indicating that the nervous systems of these mice were sick. However, in the HdhQ150ΔI mice, NfL and BRP39 remained at levels of the regular non-HD mice at 12 and 17 months in CSF, and were also reduced in plasma. 

Interestingly, this return to regular levels of CSF biomarkers happened even though the gene switch signature changes were only partially restored. Because NfL and YKL-40 can be used in human clinical trials to assess brain health, this finding suggests that reducing HTT1a could meaningfully alter HD biology.

Connecting the dots from somatic expansion → HTT1a → aggregation

Somatic CAG expansion is now viewed by many HD researchers as a driver of HD. As CAG repeats expand in neurons most affected in HD, HTT1a levels will rise. The scientists in this study propose a model where: 

1. Somatic expansion increases the CAG repeat and HTT1a production.
2. Increased HTT1a leads to more toxic clumps in these cells.
3. The clumps drive more changes in the gene switches.
4. This then causes the cells to get sick, leading to changes in biomarker levels. 

They propose that this could be slowed or halted by reducing HTT1a levels to interrupt this process.

What does this mean for therapies?

Several HTT-lowering approaches are in development or clinical testing. Many lower only the full-length form HTT. These include the antisense oligonucleotide (ASO) tominersen from Roche, and small-molecule splicing modulators like votoplam and SKY-0515 from Novartis and SkyHawk respectively. 

Others target exon 1 so can reduce levels of both full-length HTT and HTT1a, including the gene therapy AMT-130 from uniQure and the siRNA ALN-HTT02 from Alnylam Pharmaceuticals. 

This study suggests that lowering only the full-length form of HTT may not be as beneficial and reducing HTT1a might be important too. Because the regular HTT protein has important functions in cells, strategies that specifically target HTT1a might be a safer and more efficient approach. 

Important caveats

All that said, there are some important limitations to note of this study. All of this data comes from mice that model HD and have a very long CAG repeat (~195). Mice are a great model for scientists to explore these kinds of ideas, but they don’t live anywhere near as long as people and most people with HD have ~42 CAGs on average, way less than in this model system. 

Still, the data from this study will certainly be a source of new ideas for other teams to follow up on. Stay tuned to HDBuzz for more research in this space. 

Summary

• Researchers made a mouse model of HD that didn’t make as much of the HTT1a fragment, thought to be toxic in HD. 
• By reducing levels of the HTT1a message, the researchers saw delays in the formation of toxic protein clumps made by HTT1a. 
• Which genes were switched on or off more closely mimicked mice without the HD gene in these special HTT1a deficient mice, and many key biomarkers returned to levels closer to normal.
• These findings provide evidence that HTT1a is not just a byproduct of the HD expansion, but a driver of HD biology.