In-vivo drug testing is a critical part of drug development. While we can learn a lot about a potential treatment by studying its chemical and physical properties and testing it in cells in a dish or test tube, nothing can truly replicate the complexity of a living organism. Testing drugs in mice, fish, and other animal models of a disease is an important step in demonstrating if a drug appears promising enough – and safe enough – to move into a clinical trial in humans.

To learn more about how in-vivo testing works, how these models are developed and validated, and what new models the ALS Therapy Development Institute (ALS TDI) is currently working on, we spoke to Dr. Theo Hatzipetros, ALS TDI’s Director of Pharmacology.

Why do we test potential drugs in animal models? What's the purpose of that is in our overall science process?

Dr. Hatzipetros: We test drugs in animal models for two main reasons. First, to see if they are efficacious in treating the disease. Number two is to test whether they are safe to use for humans.

Typically, in our studies we would have two cohorts of mice. Half of them will be dosed with an experimental drug and the other half will be the control group. Then, we monitor them over a period of time, looking for differences between the treated and the control groups in terms of their appearance, their weight maintenance, the disease onset and their survival duration. Ultimately, we are trying to identify whether the drug-treated animals fair better that the control animals.

Why do we primarily use the SOD1 G93a mouse specifically at ALS TDI?

The model that we use is the best one available right now, both in terms of modeling ALS and performing drug screens. What I mean by that is that, it resembles ALS in humans very well. The mice look as if they have ALS. They eventually develop paralysis. Internally, they look like humans with ALS. They have more neuron loss, they have macrophage activation. They have all those things that we consider to be ALS.

They’re also conducive to drug screening. Their disease is aggressive, but not too aggressive, so it gives you the chance to intervene pharmacologically. It also doesn't take a long time for these mice to develop and progress through the disease. The life expectancy of these mice is around 120 to 140 days. This means we can screen more drugs in these models than in one that lives to be, say, 250 days.

We also know this model very well. We have used it for over 20 years at ALS TDI. We know how it behaves. If a drug is working, we’re likely to find it in this model.

Are there mouse models of ALS with other genetic mutations? Why doesn’t ALS TDI currently use them?

There are several ALS animal models out there with mutations on many genes, but none of them is as well characterized as the SOD1 model, especially when it comes to drug screening. If you do a PubMed search, you will find very, very few publications where they've screened drugs when in non SOD1 model.

Having said that, ALS TDI has investigated many of them. We’ve characterized many of the developed models. We found that none of them replicate the symptoms of ALS as well as our current model. Even those that do display some symptoms of ALS, it usually takes a long time for those symptoms to appear, and they're not conducive to drug screening. So, there might be other models out there, but we believe that the SOD1 model that we use is still the best model of ALS.

Some of our more recent research programs use transgenic mouse models that involve mutant TDP43, C9orf72 dipeptide repeats, and mutant Profilin-1. These are all mutant genes and proteins associated with different types of ALS. A lot of these studies are carried in the Jim Heller Mouse Investigation Room housed within our Animal Care Facility.

Why don’t we use a model of Sporadic ALS?

An animal model of sporadic ALS model would be great. That's what we all wish we had. It would be the ideal model to use as most cases of ALS are sporadic, but as long as we don't know the exact causes of the disease, we cannot create a sporadic ALS model. That's why we rely on genetic models for the time being.

So, do we know if the findings in the SOD1 model will be useful for people with other forms of ALS?

The downstream effects of the SOD1 Mutation, or any ALS causing mutation are very similar. In humans ALS looks similar no matter what gene mutation you have, or if you have sporadic ALS. So, we hope that the drugs that we develop tackling the downstream consequences of the SOD1 mutation would benefit all types of ALS.

But obviously, the drugs that directly target the SOD1 mutation would only apply to people with the SOD1 mutation. But most of the drugs we develop, or we screen, they target the downstream consequences of that mutation.

What sort of other animal besides mice are used to model ALS?

There are zebrafish models of ALS. There are C. elegans [roundworm] animal models of ALS, specifically engineered around SOD1 mutations and FUS mutations. There’s Drosophila, or fruit fly models of ALS. There are also rat models of ALS, in addition to mice. At ALS TDI we already set up a zebrafish facility and we have a C9orf72 zebrafish model of ALS that we plan to study thoroughly this year. Currently, we do not have roundworm, fruit flies or rat models of ALS at ALS TDI.

What goes into validating a mouse model or any other animal model?

It takes months and months of carefully examining the animals at different levels. At a phenotypic level [how the disease presents and progresses], at a histological level [how it appears in the animal’s tissues], at the molecular level, at a genetic level, and trying to identify. similarities and dissimilarities between those models and ALS in humans. The more similar they are to ALS, the better it is in terms of using them in our research.

Some animal models can be useful for basic ALS research even, though they might not necessarily be appropriate for drug screening. To be used for drug screening, you have to have the additional component of having their disease be aggressive enough to test drugs within a relatively short time period. I would say that any model with a disease duration of anything longer than six months would not be desirable because it will require a lot of time to get your answer and a lot of drug to be used.

In contrast, if a model is too aggressive it might not be biologically relevant and no matter how good your drug is, these mice will not improve. It's one thing developing a model to study the disease and another thing to develop a model to test a drug, it’s a lot harder to have a model that models ALS and at the same time is conducive to drug screening.

So what sorts of models is ALS TDI working on right now?

We will be doing work on the C9orf72 zebrafish model early this year. Also, in the first half of this year, we will bring in mouse models based on mutations on the TDP43, Profilin-1, and C9orf72 ALS causing genes. We’ll spend some time characterizing them, validating them, see how well they replicate ALS, and if we end up discovering that these mice serve our needs, we might use them for drug screening purposes next year or even later this year.


What to do next

  • Learn more about ALS TDI’s research to find new treatments for ALS here.
  • Watch Dr. Hatzipetros’ presentation from last year’s ALS TDI Summit, on the in-vivo testing of a potential new ALS treatment at ALS TDI.