CHANGING THE CODE

Can a Faulty Gene Be Saved?

by Margaret Wahl

Ever since the discovery in 1986 of the gene for dystrophin, the protein that’s missing in Duchenne muscular dystrophy (DMD), scientists and physicians have been trying to figure out how to compensate for its loss.

An obvious solution is to insert a new dystrophin gene, a technique usually referred to simply as gene therapy (see “Bridge Over Troubled Waters,” January-February).

But in the years since gene therapy experiments began, other ideas about how to compensate for errors in the dystrophin gene have arisen. These ideas range from repairing the gene, to modifying the way the cell interprets the language of the genetic code, to changing the activity of a gene that can code for a dystrophin substitute.

These techniques, based on the same genetic research that spawned the concept of gene therapy, can be loosely described as genetic modification. The following pages present four MDA-supported investigators who are among those who have begun to make genetic modification strategies a reality. While their focus is on Duchenne MD, if any of these techniques prove successful, it’s possible they could be applied to therapies for other neuromuscular diseases.

To understand the steps these investigators have taken, it helps to have a little knowledge of how the genetic code leads to the manufacture of proteins.

Breaking the Code

Genes are made mostly of DNA, which is composed of nucleic acids attached to phosphate and sugar groups. When nucleic acids are attached to these other chemical groups, they’re called nucleotides.

From Gene to Protein One strategy to restore dystrophin production is gene repair, which aims to correct the dystrophin DNA sequence. Another is exon skipping, which involves changing the cutting and splicing mechanisms of the cell so that errors in the pre-mRNA are spliced out. Stop codon read-through likewise takes place at the RNA level, coaxing the cell to ignore a stop signal. Finally, utrophin upregulation ignores the dystrophin gene and attempts to boost production of the utrophin protein at any step from the utrophin DNA through protein synthesis.

The four nucleic acids in DNA are adenine (A), guanine (G), cytosine (C) and thymine (T), and when they’re arranged in specific sequences, they form a code that will ultimately determine the composition of protein molecules, which make up most cellular structures and carry out almost all cellular functions in the body.

Some nucleotide sequences are instructions for specific amino acids, the building blocks of proteins. Other sequences, known as stop codons, tell the cell’s mechanisms it’s time to stop reading the code.

Still others, splice sites, determine which parts of the genetic code will be reflected in the final protein’s components, and which parts will be cut out. The parts of the code that are destined to be cut out are known as introns, and the parts to be left in are called exons.

DNA is double-stranded, with the nucleic acids — A, G, C and T — stuck together between the strands like rungs of a ladder. The bonds between the strands are specific as well. Adenine is supposed to pair only with thymine, and guanine only with cytosine.

The first step in protein production is the building of RNA from DNA, a process called transcription.

RNA is very similar to DNA but differs in a few ways: It’s single-stranded; it contains the nucleic acid uracil (U) where DNA contains thymine (T); and its sugar groups aren’t exactly the same as DNA’s.

It’s from RNA that the final recipe for protein production will come, but not directly. RNA is first produced as a “rough draft,” known as pre-mRNA, and later edited to a shorter, final draft, known as messenger RNA or mRNA. The mRNA forms the template for final protein manufacture, known as translation.

At any stage in this process, errors can occur. But these stages also offer a possibility for either natural or laboratory-engineered correction.

Errors in the gene for dystrophin are usually one of two types. In deletions, parts of the coding sequence are missing. This leads to gaps in the RNA and then in the protein, and often interferes with the reading of otherwise correct information that follows the gap.

In the second type of error, point mutations, the wrong nucleic acid is inserted in place of the right one. Sometimes, point mutations cause premature stop codons, which tell the cell’s machinery to stop reading the genetic recipe before all the instructions have been read. A premature stop codon can be formed when only one nucleic acid is misplaced.

Fixing the Code

One genetic modification strategy is DNA repair. It’s complete, it’s likely to be permanent, and it takes advantage of a natural cellular repair process. But so far, it’s been hard to get it to work well enough to be meaningful for people with Duchenne MD.

Another idea is coaxing the cell to “run a stop sign.” Chemical compounds that can cause “stop codon read-through” are under intense investigation, with clinical trials anticipated later this year.

	DNA is double-stranded and contains the recipe, or code, for the body’s proteins through varying arrangements of four nucleic acids: adenine (A), guanine (G), cytosine (C) and thymine (T).

Then there’s the possibility of changing the genetic code at the point at which the pre-mRNA has been made, with an error, but the final mRNA hasn’t yet been formed. Causing the cell to skip over the error-containing parts of the pre-mRNA and make a slightly shorter, but error-free, final RNA, can lead to a dystrophin molecule that’s highly functional. Known as exon skipping, this technique is gaining support.

Another tactic makes use of the fact that dystrophin has a near twin, a protein called utrophin that looks and acts very much like it but isn’t located in the same place in muscle cells and isn’t made in very large amounts. Increasing utrophin’s production and changing its location no longer seem far-fetched strategies to molecular biologists.

Repairing DNA

Thomas Rando, M.D., Ph.D.

Affiliation
Stanford University, Stanford, Calif.
Strategy
DNA Mismatch Repair
Status
Animal models

by Margaret Wahl and Erik Misner

Several decades ago, scientists discovered that bacterial cells (which are simpler than animal cells in several ways) had an amazing ability: They could detect and fix errors in their DNA by an efficient and precise process that came to be known as mismatch repair. The term reflects the principle that each of the four nucleic acids in DNA has only one other with which it can correctly “match,” or combine

Mismatches occur when the wrong nucleic acid is placed on one of the rungs of the ladderlike DNA structure. When bacterial repair systems detect a mismatch in any of the rungs, they move in and repair it.

Later, scientists recognized that more complex cells, including human cells, also have the ability to repair DNA mutations. However, it wasn’t until about 10 years ago that investigators began to consider using this mechanism as a therapy.

“Really, the question was, if a cell has the ability to correct mutations that occur during life, can we get the cell to correct inherited mutations?” says Tom Rando, an associate professor in the Department of Neurology and Neurological Sciences at Stanford School of Medicine. “It’s one thing to recognize a normal biological function. It’s quite another to harness that function and get it to do what you want.”

About 1995, he and others began to investigate that question seriously. By that time, Rando had earned doctoral degrees in cell and developmental biology and in medicine, both from Harvard, and had started studying electrophysiology, particularly the electricitylike activity that transmits signals in the nervous system.

He wasn’t particularly interested in muscle diseases until, as a young trainee in neurology at the University of California at San Francisco in the late 1980s, he was introduced to the MDA clinic. One genetic disease in particular — myotonic muscular dystrophy — captivated him.

Myotonic dystrophy involves both myotonia, the inability to relax muscles on command, which results from abnormalities in nerve signals, and dystrophy, involving degeneration of muscle.

Rando’s interest in myotonic dystrophy, he says, “was the transition between being interested in the electrical properties of the nervous system and getting into the muscular dystrophies.”

Mice, Dogs and Chimeras

The UCSF lab was using the mdx mouse, which has a point mutation that causes a premature stop codon in the dystrophin gene, as a model for studying muscular dystrophy. Rando recalls, “We thought maybe we should try and see if we could direct the cell’s own mismatch repair mechanisms to correct the mdx point mutation.”

Rando’s group and a group working with a dog model of Duchenne MD published papers in 2000 showing that gene repair of point mutations was possible in both types of animals, although it was very inefficient.

“This was all proof of principle, rather than looking at therapeutic efficacy,” Rando says. “We were just trying to see how it worked, how efficiently it worked, and what some of the hurdles might be.”

Rando’s original plan was to use molecules made of both DNA and RNA. These were known as chimeric molecules, a chimera being a beast in Greek mythology that combines parts of different animals.

Nowadays, his group uses molecules made solely of DNA, because these are much easier to make and use. He calls this method oligonucleotide-mediated gene repair. (Oligo means few, and there are only a small number of nucleotides in each repair molecule.)

Repairs That Last

Rando believes the gene-repair approach to treating muscular dystrophy “avoids many disadvantages of other forms of gene therapy.” For one thing, the repair would likely be permanent, since it affects the genes in their natural place on the chromosome, while many forms of gene therapy insert a gene that stays outside the cell’s chromosomes and will likely eventually be lost. For another, it requires no viruses, which can have unpredictable effects.

“At the end of the therapy, you have a completely normal gene,” Rando says. “It’s truly a repair.”

Rando says the technique isn’t yet effective enough to be meaningful to patients, but he remains optimistic.

“We have some new generations of oligonucleotides that we’re trying, and what we’re looking for are better ways to deliver them and higher levels of efficiency. The question will be, as with all these therapies, how many of the hurdles can be overcome. None of them are impossible.”

Running a Stop Sign

Lee Sweeney, Ph.D.

Affiliation
University of Pennsylvania, Philadelphia
Strategy
Stop codon read-through
Status
Human trials in DMD expected this year

by Paul Muhlrad

Muscle biology was mainly an intellectual curiosity for Lee Sweeney when he first set up his University of Pennsylvania research laboratory. His perspective changed as he got to know people who had Duchenne MD.

“I started giving talks in front of some of the parents and going to meetings where I actually had some interaction with some of the patients and their families. And, you know, it put a human side on what to me had been just sort of an esoteric disease,” he says. “At that point I decided that I really needed to try to work on therapeutics.”

Sweeney, chairman of the Physiology Department at Penn, is a member of the Scientific Advisory Board of the biotechnology company PTC Therapeutics, which is developing a new drug that he thinks could treat as many as 10 percent to 15 percent of those with DMD. The drug, dubbed PTC124, targets premature stop codons and may also work in other forms of muscular dystrophy, as well as for certain other genetic disorders, such as hemophilia and cystic fibrosis.

For years, molecular biologists have been trying to replace faulty genes with working versions. Unfortunately, says Sweeney, “We can’t replace a gene at this point in time.” (Human trials to do so in DMD are expected to begin next year.)

Same Gene, New Protein

A more pragmatic approach than traditional gene therapy, he says, might be to coax the faulty gene into making a protein that works. That’s what PTC124 does to genes with premature stop codons.

PTC124 sticks to ribosomes — the cells’ protein factories — and prompts them to interpret a premature stop codon as a normal codon. Instead of aborting assembly of the protein, the ribosome inserts a protein building block — an amino acid — and continues making a complete protein chain until it encounters the normal termination codon, which the ribosome correctly interprets as a stop.

Based on results from preclinical studies in animals and cultured cells, Sweeney is optimistic about using PTC124 to treat boys with DMD who have premature stop codons in their dystrophin genes.

When lab mice with a premature stop codon in the dystrophin gene were given the drug, their dystrophin protein levels reached 25 percent of those of healthy mice, and their disease stabilized. And, Sweeney points out, “our mouse model is almost a worst case.”

The dystrophin-deficient mice have a very early premature stop codon, and, the earlier the premature stop codon occurs in a gene, the more difficult it can be to correct. Sweeney speculates that in humans “maybe 20 percent [of normal dystrophin protein levels] would lead to sort of a mild disease, and 50 percent would probably be enough to eradicate the disease.”

Elisabeth Barton and H.Lee Sweeney at the University of Pennsylvania are studying Duchenne MD in mice with hopes of developing treatments. Photo by Addison Geary

In January, the U.S. Food and Drug Administration awarded Orphan Drug designation, a set of financial incentives to encourage pharmaceutical companies to develop drugs for rare diseases, to PTC Therapeutics, for PTC124 development. Phase 1 clinical trials of PTC124 in healthy people, which concluded last year, show promising results. Humans don’t break down the drug nearly as fast as mice, which should make it much easier to administer effective doses.

A new phase 1 trial using multiple dosage levels of PTC124 began this year, and Sweeney anticipates that phase 2 clinical trials in boys with DMD will begin this spring or summer, if regulatory approvals are obtained.

“I’m incredibly hopeful that this is going to work in some of the patients, maybe stabilize or at least slow [DMD progression] in some and maybe even stop the disease in others,” Sweeney says. “And so, I’m just anxious to start treating people. I’m very excited about it.”

Bandaging RNA Errors

Stephen Wilton, Ph.D.

Affiliation
University of Western Australia, Perth
Strategy
Exon Skipping
Status
Animal models

by Margaret Wahl

Steve Wilton likes to call his gene modification strategy a “genetic Band-Aid.” Like Tom Rando, Wilton and colleagues are using oligonucleotides — short pieces of genetic material — to change the way cells read genetic instructions.

But there are some key differences between Rando’s gene-correction strategy and Wilton’s Band-Aid technique. Wilton’s plan involves blocking the part of the “rough draft” genetic blueprint (pre-mRNA) that contains a flaw.

His oligonucleotides are called antisense, because they stick to and block parts of the RNA that the cell would ordinarily read (make sense of).

At this stage, the cell normally cuts out the parts of the RNA — the introns — that won’t be part of the final message, leaving only the parts of the final RNA message — the exons.

But an error in the gene that makes it into the pre-mRNA can cause the cell to cut and splice in the wrong place, so that the final instructions contain material that should have been edited out, or don’t contain material that should have been left in (a splice site error).

In the early 1990s, Wilton had completed his doctoral training at the University of Adelaide in Australia and was working for a small biotechnology company in Perth, making molecular biology compounds that other researchers would use. “I was getting very sick of that, and I wanted to get back to research,” he says, and he began working after hours at the Australian Neuromuscular Research Institute, part of the University of Western Australia.

At the institute, Wilton worked with molecular biologist Nigel Laing, who had been at Duke University in North Carolina with MDA grantee Allen Roses. “That’s how I got involved in muscular dystrophy,” Wilton recalls.

The gene for dystrophin had recently been found, and other genes for neuromuscular conditions were being identified at a rapid pace. While developing genetic testing for DMD and other diseases, Wilton became fascinated by a recently discovered phenomenon known as revertant fibers, muscle cells found in boys with DMD that mysteriously begin making dystrophin despite genetic mutations that should keep them from doing so.

Antisense Makes Sense

Wilton had an idea that the revertant fibers might occur when a glitch in the cell’s gene-reading machinery allowed it to “skip” a genetic error and continue making the protein from instructions on the far side of it.

“I had a limited imagination, I suppose,” he says, “because I couldn’t see any other way that could happen. Gene deletions are a common type of defect in the dystrophin gene, so it seemed logical that a second mutation could overcome the first one, a case of two wrongs making a right.”

Wilton’s imagination was correct. Exon skipping, as the phenomenon came to be called, was the mechanism by which dystrophin-containing fibers sometimes occurred despite mutations in the dystrophin gene.

But it wasn’t until October 1996, while listening to a lecture by Richard Kole of the University of North Carolina at a gene therapy conference in Lake Tahoe, Nev., that Wilton began to think about how to make exon skipping a treatment for DMD.

Kole was talking about using antisense constructions to block splice site mutations in the beta-globin gene, which underlies the blood disease thalassemia. Wilton recalls that, as his thinking strayed to the implications for dystrophin gene alteration, “it was like being hit by a brick.”

If splice site mutations could be blocked by antisense oligonucleotides, he thought, why not try blocking normal splice sites to keep error-containing exons from being included in the final mRNA?

Wilton and Kole struck up a conversation, and Kole agreed to send Wilton some antisense constructs. “A month after that, we had exon skipping working in some cultured cells.”

These days, having obtained equipment to synthesize antisense oligonucleotides quickly and relatively inexpensively in his own lab, Wilton says he’ll try “blocking anything” that looks like it will help someone with a dystrophin mutation.

“We have some exons where the donor splice site — that’s the one at the back of the exon — works really well. We’ve got other targets where it’s the front exon splice site — the acceptor — and sometimes it’s somewhere in the middle. Sometimes we get all three working. You can’t say there’s a best target as far as we have been able to tell.”

In 2003, he and his colleagues showed that a particular antisense oligonucleotide can overcome a premature stop codon mutation in the dystrophin gene in mdx mice and allow the animals to produce normal levels of dystrophin in a large number of muscle fibers.

Wilton now has the ear of a major pharmaceutical company that’s interested in applying exon skipping to DMD. “Honestly,” he says, “the last couple of years have been unbelievable, and with the support of industry and MDA, we’ll soon see if exon skipping is going to be a viable treatment.”

Coaxing a Stand-in

Bernard Jasmin, Ph.D.

Affiliation
University of Ottawa, Canada
Strategy
Utrophin upregulation
Status
Animal models

by Margaret Wahl

"The beauty of a drug-based therapy is that you can affect all the muscle fibers,” says Bernard Jasmin, a professor in the Department of Cellular and Molecular Medicine at the University of Ottawa who’s been working on that type of strategy for Duchenne MD.

“Ideally,” he notes, “somebody takes a pill, and it stimulates expression of [a protein] in all the muscles, which right now is a major hurdle for gene delivery.”

Jasmin says he got into studying muscles and nerves because he was “very much into sports” but not very good at them. He didn’t want to be a physician because his mind wanders and he doesn’t like sticking to a schedule. So, in 1985, Jasmin began doctoral studies in biology at the University of Montreal, concentrating on the biochemistry and physiology of muscles and nerves.

Later, doing postdoctoral research at the Pasteur and Jacques Monod Institutes in Paris, he focused on the neuromuscular junction, the place where a fiber from a nerve cell meets a specialized area of a muscle fiber. That focus led him, not surprisingly, to the emerging study of proteins unique to the junction. One of those proteins, identified in the late 1980s, closely resembled the newly discovered dystrophin, the muscle protein missing in DMD.

Originally dubbed dystrophin-related protein and later renamed utrophin, it was found to come from a gene on chromosome 6. Dystrophin is made from a gene on the X chromosome, so it could be assumed that boys with DMD would have intact utrophin genes.

(Jasmin is quick to point out that, although he had speculated about utrophin’s existence, he had little to do with actually identifying it. The credit for that, he says, goes to Kay Davies at the University of Oxford, Lou Kunkel at Harvard, and Tejvir Khurana, now at the University of Pennsylvania.)

Although utrophin is close to dystrophin in both structure and function, there’s at least one key difference between the two proteins. During fetal development and perhaps a little beyond, utrophin is present all around the muscle fiber, interacting with clusters of proteins stuck in its surrounding membrane. As the animal or person matures, utrophin is replaced almost entirely by dystrophin, with one exception. At the neuromuscular junction, utrophin remains throughout life.

By the mid-1990s, investigators were asking a lot of questions. Could utrophin stand in for dystrophin? Is there a mechanism that shuts off utrophin everywhere except the junction as an organism develops? And, if so, could it be disabled, allowing utrophin to resume the position that it has during fetal life?

Protein Pathways

Jasmin’s and other groups set out to identify specific pathways that underlie the utrophin-to-dystrophin switch and to make these targets for drug discovery.

Jasmin says his goal is to identify molecules that can trigger or enhance the stimulation of pathways to put utrophin all around the muscle fiber, and “to try to have them specific enough so that you’re not going to have side effects that will do something else. This is where the challenge is.”

Early last year, his group showed that when dystrophin-deficient mice were bred with mice producing higher than normal amounts of the protein calcineurin, the utrophin protein appeared all around the fiber, where dystrophin would have been placed, and it reduced fiber damage.

It may be easier to inhibit something that’s putting a brake on utrophin than to directly increase (upregulate) production of the protein, Jasmin notes. And calcineurin, it turns out, is just the kind of brake release Jasmin and colleagues have in mind.

The brake itself, it seems, is another protein, JNK1. Once the JNK1 brake is overcome by calcineurin, more utrophin can be made, and it extends to areas outside the neuromuscular junction.

In December, Jasmin and colleagues, including MDA grantee Lynn Megeney at the University of Ottawa, showed that corticosteroids like prednisone and deflazacort increase calcineurin activity, which in turn stimulates utrophin production, and that this is likely to be at least part of the reason for their beneficial effects in Duchenne dystrophy.

“We did the proof of principle in the mdx [dystrophin-deficient] mouse, showing that if you stimulate the calcineurin pathway, the mice will be better; and now we find an explanation for the beneficial effect of a drug that is actually used in the clinic. So the whole story is pretty tight as far as we’re concerned.”