Arresting ALS

Every day I receive emails from ALS patients and their loved ones, which is both gratifying and heart breaking. The time from onset of the disease to death can be just a few years, and patients are desperate for some hope. The possibility that my work might help these people is a large part of what has kept me going for over 20 years, despite all the setbacks and frustrations.

My ALS journey actually started with an interest in oxidative stress. I was studying the oxidant peroxynitrite, which mediates tyrosine nitration – a process you can find in stroke, diabetes, heart disease, neurodegenerative disorders and many other conditions. One of the major antioxidant defenses that prevents the formation of peroxynitrite and protects the body from oxidative stress is the protein superoxide dismutase (SOD). But SOD also has a dark side – we discovered that it can catalyze tyrosine nitration, speeding up oxidative damage.

In 1993, it was found that some inherited cases of ALS are caused by mutations in the SOD-producing gene SOD1 (1). Our research group hypothesized that in these patients, SOD would catalyze tyrosine nitration and make the disease worse – a toxic gain-of-function. Soon after, we discovered that mutant SOD did exactly that (2), and ever since we’ve been focused on better understanding the role of SOD in ALS, and ultimately finding a therapy.

A cloud with a copper lining?

In many diseases – even intractable enigmas like Alzheimer’s – therapies work in mouse models, but cannot be translated into humans. In ALS – a disease whose pathology has been cited since 1824 and named since 1874 (3) – there had never even been a single functionally effective therapy in a mouse model. To this day, only one drug – riluzole – has been approved for use in the treatment of ALS, and that came onto the market 20 years ago, and increases life expectancy by only 2–3 months (4).

Over the years, we made many discoveries that kept pushing our knowledge a little bit further, but nothing that we could pursue as a possible therapy. The SOD1 mutant mice still died at around 130 days. Eventually, we decided to shift focus – if we couldn’t slow the disease process, could we speed it up? If we know how to break something, we reasoned, we might get a better idea of how to fix it.

Copper and zinc are essential for the maturation of SOD, and in ALS mouse models a lack of copper binding causes accumulation of mutant, immature SOD. A paper published in 2007 revealed that if you overexpress the copper chaperone for SOD (a metalloprotein named CCS) in wild-type mice, the mice are perfectly fine. But if you overexpress CCS in an ALS mouse model, the mice start dying six or seven times faster (6). That was an interesting discovery, because all human ALS patients have comparatively high CCS relative to SOD.

A former student of mine – Blaine Roberts – visited our laboratory and talked about a compound being studied by the Florey Institute in Melbourne (where he is now head of metalloproteomics), called copper-ATSM (CuATSM). The compound is traditionally used in PET imaging but Peter Crouch’s lab, alongside Roberts, had shown that it can improve locomotor function in ALS model mice (7). CuATSM delivers copper to the brain, and we suspected that it would counteract the copper deficiency seen in ALS mouse models.

A breakthrough

We acquired mice from the CCS study (6) – which developed ALS symptoms relatively slowly – and cross-bred them with the standard (SOD1-G93A) ALS mouse model, resulting in mice that died in 8–14 days. The first hurdle we faced in testing the drug with this model was how to get the CuATSM into tiny four or five day old mice, who were already runts. Our solution came when we found that CuATSM is soluble in dimethyl 
sulfoxide (DMSO).

As we pipetted the CuATSM/DMSO solution onto the backs of the baby mice, it was absorbed through the skin within minutes, turning it bright red. Soon after, we watched the color fade as the solution was distributed into the subcutaneous fat, then subsequently to the brain and other organs via the bloodstream. Now the question became – would CuATSM affect disease progression?

Straight away, the mice started to improve markedly. They gained weight, and a few days later, far from being at death’s door, they showed no signs of disease. They began to develop quite normally, and to our surprise they passed the crucial 130-day mark, then 150 days, then 200. It was around day 230 when the first mouse became sick, but the others made it beyond their first birthday – unheard of in ALS research!

But our first response wasn’t elation, instead it was “something has to be wrong”. Our first thought was that the transgenes could have become inactivated, and only once we had ruled that out did we allow ourselves to get a little excited. Even then, we wanted to be really sure that the findings were airtight, so we immediately embarked on a series of very carefully controlled trials. It was difficult to bite our tongues, because we saw these results just as the ice bucket challenge was becoming a global phenomenon and everyone was talking about ALS (see "Putting ALS on Ice").

We spent a lot of time considering sample sizes, consulting with statisticians about the setup and animal specialists about breeding the mice. We also blinded the study since the CuATSM/DMSO could stain the fur. We did that all to maximize the amount of good data we could obtain. Initially, we were worried about CuATSM toxicity but it soon became clear that there was no serious toxic effect on the mice, so we were able to increase the dose considerably. Ultimately, we settled on treating pups from the age of 5 days, with 30 mg/kg/dose of CuATSM twice daily, and the results surpassed all our expectations. We extended the lives of the standard SOD1 mutant ALS mouse model by 25 percent, and the ALS mouse model with both SOD1 mutation and overexpressed CCS by an amazing 500 percent (8). What’s more, cessation of CuATSM caused the mice to develop ALS symptoms, and restarting therapy rescued them.

ALS Genes

Genotypes of familial ALS and their associated chromosomes

Trials and tribulations

We carried out a lot more experiments to reproduce certain elements of our findings, and compared our results with those of the drug’s developers in Melbourne, so by the time the paper came out we felt confident in our results. Despite this, we were met with skepticism from some of the peer reviewers, who were concerned that the mouse model may not translate into human patients. Only 2–7 percent of all ALS patients have a SOD1 mutation, and some consider this familial form a separate disease to sporadic ALS. Although it makes up the vast majority of cases, very little is known about the etiology of sporadic ALS – which is why most ALS researchers study familial forms. I believe there is a definite possibility that the drug might work in sporadic ALS patients – after all, the SOD1 mutation amplifies traits of the wildtype protein. But even if the drug does only work in the SOD1 mutation patients, it would be a much needed breakthrough for the disease.

The Melbourne group have taken the lead in developing the compound for Parkinson’s disease and ALS. Their license was granted to a newly formed company called Procypra, based in the USA, which is developing CuATSM for treatment of Parkinson’s and ALS.

Where to next?

The identification of new genes is driving the field forward – we need to find out what is making things go wrong before we can start to think about how to fix it. There are now over 20 different types of genetic mutations linked with ALS, each one opening up new avenues of research (see "ALS Genes" and "Research Roundup"). However, it’s unfortunate that the SOD1 gene has been neglected by researchers who feel that, after decades of study with very little progress, it’s a dead end. I’m hopeful that our results will change that perception and encourage other researchers to start using this model again.

Right now, the team and I are working really hard on finding other variants of the CuATSM drug that will deliver enhanced results; the drug that we have right now works much better than the first 20 versions we tried, so there’s definitely hope that we can keep improving. We’re also advancing the mass spectrometry methods that we use to measure what’s happening inside the motor neurons, to give us an even clearer view into the mechanisms of disease. Thanks to those efforts, we already have a pretty good idea about how SOD causes motor neurons to die, but we want to keep honing those ideas until we can address all the potential criticisms.

I’m really excited about the field right now. The ice bucket challenge put ALS in the spotlight and injected much-needed funding. Now we must continue to drive research forward until we have an effective treatment in the clinic. We understand so much more about the mechanisms of the disease, that it no longer seems like a hopeless task. Research has reduced the disease to something we can attack, and hopefully defeat.

Joe Beckman is the Principle Investigator, and Burgess and Elizabeth Jamieson Chair, in Healthspan Research at the Linus Pauling Institute, Oregon State University, OR, USA.