The ALS Association: Research Archive November 27, 2007

November 27, 2007

Drug Discovery in ALS: Animal Models are Problematic but Crucial

By Richard Robinson
Science Writer

Second in a series of three articles

In late September 2007, The ALS Association brought together 30 scientists, clinicians, and industry researchers for a three-day conference on “Drug Discovery, Biomarkers, and Clinical Trials in ALS.” This is the second of three articles on that conference.

Successful clinical trials for ALS drugs depend on a long preclinical development process. Key aspects of that process were a major focus of The ALS Association’s September conference.

The SOD1 Mouse: How good a model is it?

A major challenge in ALS is that the cause of the disease in most cases is not known. The exception is the gene SOD1 (Cu/Zn superoxide dismutase 1). Mutations in this gene account for about 20% of cases of familial ALS, and 2% of all cases of ALS. Several different SOD1 mutations have been introduced into mice; these mice develop muscle weakness and die prematurely. The G93A SOD1 mouse is so far the main animal model for ALS, both for understanding the disease and for discovering potential therapies.

Researcher Peter Lansbury, Ph.D. (Link Medicine Corporation), is studying the effect of mutations on the protein the gene makes. Normally, two SOD1 proteins link together like two identical Lego� pieces. Some of the mutations prevent this linkage. How this leads to ALS is not known, but Dr. Lansbury is working to develop drugs that allow the two proteins to stay linked together, expecting this may have a beneficial effect.

Unfortunately, like almost every animal model of every disease, the SOD1 mouse model is by no means perfect. It cannot mimic every feature of human SOD1 ALS, simply because mice and humans are different in important ways. One crucial difference is the length of their motor neurons, the nerve cells that control muscle that die off in ALS. Human motor neurons are many times longer than their mouse counterparts. This provides a challenge when relying on transport systems along the axons to deliver treatments to the cell body as is being considered in gene therapy approaches. The human neuron may have transport problems that the mouse never encounters, simply because of that extra length.

Equally important, the mouse model may not be a good mimic of other, non-genetic forms of ALS, which affect the large majority of patients. These difficulties have led some researchers to question the utility of the mouse model. “I’ll defend the mouse models, but we have to use them correctly,” said Robert Brown, M.D. (Harvard Medical School). Researchers must constantly keep in mind that the mouse can only be used to ask certain questions, and the validity of the answers it gives must then be proved in humans with ALS.

Nonetheless, the hope is that the SOD1 mouse may be able to provide some clues to other forms of ALS. Researchers think that while there may be several different kinds of initiating events—a gene mutation, an environmental toxin, or some combination—most forms of ALS converge in a “final common pathway” to produce motor neuron death and ultimately the clinical disease. They hope that while the SOD1 mutation is only one starting point, it leads “downstream” to events that are shared with most other forms of the disease. Thus, studying these downstream events in the mouse will provide insights that apply to other forms of ALS. “The great promise is not only to develop an SOD1-targeted therapy, but to develop tools and identify ways to move forward in the other forms of ALS,” said Tim Miller, M.D., Ph.D. (Washington University).

One such downstream event may be a dysfunction in another kind of cell in the brain and spinal cord, called astrocytes. Astrocytes don’t do the communicating and controlling that motor neurons do. Instead, they help maintain neurons, by feeding them, processing their waste, and performing other chores. “Astrocytes are dysfunctional in ALS,” according to Jeffrey Rothstein, M.D., Ph.D. (Johns Hopkins University), who has studied these cells extensively. It may be that by helping astrocytes, motor neurons can be helped as well.

Another potential clue to the ALS disease process comes from the lab of Michael Strong, M.D. (London Health Sciences Centre, Ontario), who presented new results showing that in ALS, a protein called TDP43 (TAR DNA binding protein 43) gets stuck in clumps of cellular debris in affected motor neurons. TDP43 normally helps transport genetic messengers out of the cell’s nucleus, and if it cannot do this, the cell may suffer. Dr. Strong is continuing to research how these events may influence the development of ALS.

Discovering Targets and Developing Drugs

Each of these theories of disease provides insights which may be used to discover potential drug targets. A target is something in the disease process that can be influenced by a drug. A target might be a harmful molecule that can be blocked, or a process that can be altered, or a deficiency that can be overcome. In ALS, potential targets are discovered through basic research on disease mechanisms. Once a potential target is identified, it must be validated—shown to actually influence the disease process—in animal models. Target discovery and validation in animal models are currently the most challenging aspect of drug development for ALS.

“I think the discussion of the challenges of translating animal model results into the clinic was an extremely important aspect of this conference,” said Lucie Bruijn, Ph.D., science director for The ALS Association. “There is not often an interaction between clinicians running trials and scientists testing compounds in the animal model. I think this was a unique component of the workshop, and represents a trend that is emerging in the field.”

Once a target has been validated, there are many well developed tools available for finding drugs that can “hit” the target, potentially altering the course of disease. “Drug discovery is an iterative process,” according to Robert Pacifici, Ph.D. of CHDI (Cure Huntington’s Disease Initiative). The process usually begins with a rapid screen of hundreds or even thousands of compounds, looking for a small effect on the target. This is carried out in individual cells, or even in a test tube, rather than in a whole animal. Such “high-throughput” screening is often best done by a lab specializing in this approach, serving many different research groups in the field. Once a “lead compound” is found that has some effect, the chemists go to work making variations of the molecule, trying to optimize its ability to interact with and influence the target.

Many factors must be considered in drug development. If a drug is ever to make it through clinical trials, it must be able to be absorbed, distributed, metabolized, and excreted safely within the body, and drug developers keep all these in mind as they manipulate the compound. And if the drug is ever going to make it to market, it must be patentable, or no drug company could afford to invest the research and development costs in the first place. “It is extremely important, right from the start, to keep the patent in mind,” Dr. Lansbury noted.

One way to move quickly from target to clinical proof-of-concept is with drugs already approved by regulatory agencies. One strategy is to combine approved drugs, since combinations may create synergistic effects not seen or even anticipated with individual drugs acting alone. “By systematically screening through millions of pairwise combinations, the CombinatoRx drug discovery technology has the potential to provide a renewable and untapped source of new therapeutics,” according to Jane Staunton, Ph.D., from CombinatoRx, Inc, a biotech firm pioneering this approach. With over two thousand FDA-approved compounds, the number of potential drug-drug combinations is over two million. Such combinations have the potential to amplify a small effect of one drug, or allow a lower dose to have a stronger therapeutic effect than it would otherwise have thereby decreasing or limiting toxicities associated with a larger dose.

Once a candidate drug (or combination of drugs) is identified, its activity is validated through secondary assays and preclinical testing in animal models, both for safety and efficacy. From there the product candidates can typically move directly into phase 2a proof-of-concept clinical testing to further validate the activity seen in screening and animal models. Once proof-of-concept testing is complete the company will decide whether to forward integrate itself or seek partners for later stage development.

In the final installment of this series, we will look closely at clinical trial design and use of biomarkers in ALS trials, two important factors for increasing the speed and decreasing the size of trials, both vital for quickening the pace of drug trials for ALS.