Fast-Track Pharmacy

New technology could propel rapid drug discovery for DMD and SMA

by Dan Stimson on July 31, 2001 - 5:00pm

In the mid-1980s, when MDA-supported researchers began searching for the defective gene underlying Duchenne muscular dystrophy (DMD), the scientific community at large was pessimistic about the outcome.

MDA grantee Christian Lorson's research on SMA has helped to lead to a high-tech screen for drugs to treat the disorder.

In contrast to previous gene hunts, this one was launched with no clue about the gene's identity. Lacking the option to select a handful of candidate genes, the researchers chose to conduct a broad sweep for the DMD gene using markers — genetic signposts that were inherited along with DMD in different families. To many scientists, the effort was a typical "fishing expedition" — casting out a line with little chance of catching anything.

But by 1986, a team of researchers had reeled in the DMD gene, an accomplishment that paved the way for research into gene therapy and stem cell therapy for the disease.

Today, MDA-supported researchers are continuing to fine-tune those therapies, but they're also launching a second wave of fishing expeditions — this time, to uncover pharmaceutical therapies for DMD and other neuromuscular disorders, including spinal muscular atrophy (SMA) (see "Fishing for Drugs").

Rather than use traditional pharmacological approaches to evaluate just one or a few drugs at a time, they're using new technology that rapidly sifts through thousands of drugs and drug targets with potential clinical value.

This approach — called high-throughput screening in the research world — is again drawing skepticism, but even the skeptics have to be impressed with its possible rewards. These high-tech searches have the potential to significantly add to the limited drug treatments currently in use or in clinical trials for DMD (see "DMD Drugs in the Pipeline") — all with the same remarkable speed of the DMD gene hunt.

Making up for lost dystrophin

The DMD gene encodes dystrophin, a protein normally found beneath the surface (membrane) of muscle cells. As illustrated at right, dystrophin appears to provide an essential link between proteins within the membrane and proteins that make up the cell's skeleton (or cytoskeleton).

Many researchers believe that dystrophin and its associated chain of membrane proteins, together called the dystrophin glycoprotein complex (DGC), form a scaffold that protects the muscle cell from the mechanical forces of contraction. In any case, genetic defects that cause a loss of dystrophin — or essential pieces of it — destabilize the DGC and lead to the characteristic muscle wasting of DMD.

The obvious aim of therapy for DMD is to compensate for dystrophin deficiency, and researchers have envisioned several pharmaceutical treatments that might accomplish that feat. One strategy is to search for drugs that could stimulate intact dystrophinlike proteins to take dystrophin's place in the DGC.

A related idea is to custom-design a drug that could directly take over some of dystrophin's function or perhaps rearrange other proteins in the DGC so they're not as dependent on dystrophin.

And a final approach is to single out entirely new targets for drug intervention, by identifying unknown genes and proteins that contribute to the progression of DMD.

Delivering a boost to dystrophinlike proteins

At least two proteins have a function similar to that of dystrophin, and efforts are under way to identify drugs that could boost those proteins in dystrophin-deficient muscle.

Utrophin is a small protein that looks a lot like dystrophin; it actually stands in for dystrophin in fetal muscle, but gets largely replaced by dystrophin and ends up in small patches in mature muscle. Alpha-7-beta-1 integrin (shortened here to "integrin") is a protein found in a unique complex that's functionally related to the DGC.

MDA-supported researchers have shown that when dystrophin-deficient mice are genetically engineered to overproduce utrophin or integrin, they're protected against DMD. Kay Davies, at the University of Oxford in England, did the work on utrophin, while Stephen Kaufman, at the University of Illinois in Urbana, did the work on integrin.

Since there's no obvious drug that can boost utrophin or integrin, Davies and Kaufman are independently preparing to screen thousands of chemicals that might be up to the task. Because these high-throughput screens require massive resources, including a comprehensive stock of chemicals, both Davies and Kaufman are working with biotechnology companies.

For his screen, Kaufman is collaborating with Aurora Biosciences, based in San Diego. Brian Pollok, vice president of Discovery Biology at Aurora, says the company has a chemical library that "tries to cover as much chemical space as possible." Some chemicals in the library are traditional druglike molecules and others are less conventional, derived through a process known as combinatorial chemistry, he says.

Developing an efficient way to screen for an integrin-boosting chemical will take several months, says Pollok, but once a screen is set up, it's possible to test as many as half a million chemicals in just one week.

With help from the Long Island, N.Y.-based biotechnology company OSI Pharmaceuticals, Davies has already done a pilot screen to identify chemicals that stimulate the "A-promoter" — an on-off switch in the utrophin gene. She's screened hundreds of thousands of chemicals, and found some promising candidates that are now undergoing further testing. Recently, she discovered that the utrophin gene has several other promoters, so she's in the process of revamping the old screen.

Marius Sudol uses three-dimensional computer models of dystrophin and beta-dystroglycan to help him understand how the two proteins interact.

"We're hopeful that if we can design a screen to cover all of the promoters, then we'll have a multiple-target screen that is more likely to work," she says.

Though drug screens for DMD are still at an early stage, Aurora Biosciences and MDA-supported researchers have already seen glimmers of success with a drug screen for SMA (see "Fishing for Drugs"). Since drug screens for DMD will probably follow a similar design, that's good news not just for SMA treatment, but for DMD treatment as well.

Custom designing drugs to repair the DGC

While Davies and Kaufman pursue drugs to boost utrophin and integrin, MDA grantee Marius Sudol is taking a closer look at dystrophin and other proteins of the DGC. Sudol, a biochemist at Mount Sinai School of Medicine in New York, believes that examining the physical interactions between dystrophin and its partners in the DGC could enable the design of drugs that fit into the DGC and compensate for missing dystrophin.

For years, Sudol has focused on a tiny region of dystrophin that acts as a critical attachment point between dystrophin and another DGC protein called beta-dystroglycan (see illustration). Studying that tiny region, called the WW domain, "gives researchers a new tool to probe the [DGC] and might lead to the discovery of small molecules that can regulate [DGC] assembly," says Sudol.

The WW domain is present in many proteins besides dystrophin, including caveolin-3, the protein that's defective in one type of limb-girdle muscular dystrophy (LGMD). For all proteins that have it, the WW domain appears to be an important site for interaction with other proteins.

To gain a better understanding of the WW domain's function, Sudol is collaborating with AxCell Biosciences Corp., a biotechnology company in Newtown, Pa., that specializes in charting protein-protein interactions. Using a patented technique, AxCell recently completed a high-throughput screen and identified more than 69,000 protein interactions of the WW domain. At least a subset of those could hold clues to drugs that might compensate for the loss of dystrophin and its WW domain, Sudol says.

In his own laboratory, Sudol's research on the WW domain has recently led to progress toward drugs for Alzheimer's disease. In recent experiments, he designed small pieces of protein (peptides) to inhibit a WW domain-containing protein involved in Alzheimer's, and found that the peptides had beneficial effects in a cell culture model of the disease.

A designer drug that mimics or enhances the function of the WW domain might make an appropriate treatment for some cases of DMD, Sudol says.

Sudol predicts that large-scale screens for other protein domains that regulate formation of the DGC will yield further insights into drug treatment for DMD and other types of muscular dystrophy, like LGMD.

"I believe that protein-protein interaction screens will have a tremendous impact on muscular dystrophy research," he says.

New drug targets: an array of possibilities

Eric Hoffman, a molecular biologist at the Research Center for Genetic Medicine in Washington, is at the forefront of high-throughput screens to identify new drug targets for DMD treatment.

Hoffman was part of the team that discovered the dystrophin gene. Now, he and others are realizing that a deficiency of dystrophin sets off a chain reaction of harmful effects on other genes needed in muscle. Ultimately, the muscle's overall pattern of gene activity or expression runs amok, with some genes getting inappropriately turned on and others inappropriately turned off.

Identifying those incorrectly expressed genes — and finding drugs to correct their expression — might provide an effective way to treat the disease, Hoffman says.

Fishing for genes

"What's critical is to tease apart all of these effects, and figure out which ones are important," says Hoffman. "Which ones are directly caused by problems with dystrophin? We need to build pathways that go from gene to gene. Then, we can target these pathways with drugs."

To use the gene microarray as a snapshot of gene expression in muscles affected by DMD, Hoffman starts with a muscle biopsy from a person or animal with the disease. In a series of chemical procedures, the biopsy is stripped down to yield a "soup" of active genes, and those genes are then tagged with a fluorescent dye.

When the soup is poured over a gene microarray, an active gene in the soup sticks to a matching gene fragment on the microarray, creating a fluorescent spot. An inactive gene (not present in the soup) leaves a blank spot on the microarray.

As late as 1995, putting together this genetic jigsaw puzzle would have been nearly impossible; back then, scientists could only look at the expression of one or a few genes at a time. But thanks to a recent invention called a gene microarray — a tiny chip of glass neatly arrayed with thousands of gene fragments — it's now possible to get a panoramic snapshot of up to 10,000 genes at once.

Using a set of gene microarrays, Hoffman recently scanned through some 6,000 genes, and found about 150 that stand out with highly abnormal expression patterns in DMD.

"What we're doing now," Hoffman says, "is looking at all those genes, figuring out what they do, and building the pathways that connect them together." Hoffman says this process is the key to identifying what he calls pathway-directed drugs — drugs that can compensate for abnormal pathways of gene expression.

Sometimes, he says, there's an obvious fit between an abnormally expressed gene and a pathway-directed drug.

One of the genes that lit up on Hoffman's microarrays turned out to be a signpost for dendritic cells, immune cells that promote inflammation and probably contribute to muscle damage in DMD. Oxatomide, a drug that inhibits dendritic cell activity, is already being tested in clinical trials for DMD.

High-throughput screens designed to find drug treatments for spinal muscular atrophy are starting to yield some promising "hits," says Brian Pollok of Aurora Biosciences, the biotechnology company running the screens.

In SMA, the death of muscle-controlling nerve cells (motor neurons) in the spinal cord leads to muscle weakness and wasting, causing death during infancy in the most severe cases. Nearly all cases of SMA are caused by defects in the SMN1 gene, which encodes an essential protein called survival motor neuron (SMN). Although everyone has a backup SMN gene called SMN2, it normally doesn't produce enough protein to fully substitute for the missing SMN1 product.

Fortunately, experiments on mice have shown that a genetically engineered boost of SMN2 can compensate for a deficiency of SMN1. Inspired by those results, Aurora's ongoing screens are designed to identify drugs that can stimulate SMN2.

Christian Lorson helped design a system that allows rapid screening through candidate SMA drugs by looking for a simple change in treated cells.

MDA grantees Christian Lorson of Arizona State University in Tempe and Elliot Androphy of Tufts University School of Medicine in Boston are helping design those screens. For the screens to work, "the most important thing is developing a 'read-out' method that will allow you to quickly scan the effects of a large number of chemicals," says Lorson.

Toward that goal, Lorson and Androphy attached the SMN2 gene to a gene that encodes green fluorescent protein (GFP), a green-glowing protein naturally found in certain jellyfish. When the SMN2-GFP gene is inserted into cells in culture, the cells (unlike the jellyfish) normally remain unassuming and dim, barely visible under a microscope. But if the cells suddenly crank up their levels of SMN2-GFP protein, they glow bright green.

This simple visual distinction — black vs. green — allows a rapid search for drugs that can jump-start SMN2.

"You hope for several preliminary hits, knowing that the end product will be whittled down to a handful of useful drugs. It's a hit-and-miss strategy," says Lorson.

So far, the strategy is paying off. Using a similar method, Aurora has already completed a preliminary screen for SMA drugs.

"We've screened half a million compounds [on cultured cells], and there are some that show an effect," says Pollok. "Once we single out a really promising compound, it will take between 18 months and five years to make it a drug and evaluate it for safety," he says.

No votes yet
MDA cannot respond to questions asked in the comments field. For help with questions, contact your local MDA office or clinic or email publications@mdausa.org. See comment policy