The last five years have seen marked advances in understanding the muscle fiber membrane, a thin sheath that surrounds each muscle fiber (long muscle cell) and keeps some substances inside the fiber and others out.

Illustration (click for enlarged picture)

By the late 1980s, it was clear that dystrophin, the first protein to be linked to a muscular dystrophy (MD), was closely associated with the membrane. But it wasn't until the mid-1990s that MDA-funded research teams, such as that of MDA grantee Kevin Campbell at the University of Iowa, helped researchers see the larger picture by identifying a cluster of proteins now known as the dystrophin-glycoprotein complex (DGC) (see illustration). When all or part of the DGC is missing, the theory goes, the muscle membrane gets the brunt of the forces of contraction and is much more likely to tear.

The DGC, sandwiched between two layers of the membrane, almost certainly plays a part in protecting that delicate structure from the forces generated by muscle contractions. Some investigators believe it may also have other functions, such as transmitting signals between the inside and outside of the fiber, and nearly all believe in its mechanical protective function.

When all or part of the DGC is missing, the theory goes, the muscle membrane gets the brunt of the forces of contraction and is much more likely to tear. Such tears are a very serious problem; they allow substances to leak in and out of the muscle fiber randomly. This kind of damage is likely to occur when the DGC isn't intact and, to worsen matters, muscles without intact DGCs are also missing the part of the mechanism needed to repair a torn muscle fiber -- an intact gene that would normally participate in the repair work.

Large tears, left unrepaired, eventually lead to the death of a muscle fiber due to uncontrolled biochemical traffic in and out of the fiber and probably to "wounded cell" signals that activate biological waste management mechanisms designed to clear away dead and dying cells.

The question facing MD researchers today is, can the muscle membrane be saved? The new science of gene therapy has provided unprecedented hope that indeed it can -- if the relevant genes can be inserted into the muscle and produce ("express") their proteins at the membrane.

Six genes are now attracting the attention of gene therapists and muscle specialists; each, by its absence, has been associated with a form of MD.

DYSTROPHIN, SARCOGLYCANS, LAMININ-2 LINKED TO MDs

In 1986, MDA researchers made a landmark discovery by finding the gene that, when flawed, leads to Duchenne muscular dystrophy (DMD), a relatively common and devastating disorder linked to the X chromosome and affecting boys almost exclusively.

The gene responsible for DMD carries instructions for a protein named dystrophin (see illustration), a very large, rodlike molecule that has recently been shown to be part of the DGC and to link it to the inside of the muscle fiber.

When dystrophin is absent, several other proteins in the DGC are absent or diminished, presumably because they lack their usual attachment sites. Without dystrophin, proteins known as the sarcoglycans, dystroglycans and syntrophins are also lost. When dystrophin is merely less than sufficient, the DGC is, not unexpectedly, less severely affected, and a muscle disorder that's less severe than Duchenne -- known as Becker muscular dystrophy (BMD) -- results.

The '90s have brought the identification of several other proteins in the DGC, their corresponding genes and the disorders associated with their complete or partial absence.

Of particular interest today are the proteins known as sarcoglycans and laminin-2. Several studies, many funded by MDA, have shown that each of the four known sarcoglycan proteins -- alpha, beta, gamma and delta -- can, when absent or diminished, cause a muscular dystrophy ranging in severity from a disorder that looks much like DMD to one that looks more like BMD. These muscular dystrophies have been classified as different forms of limb-girdle dystrophy (LGMD). (This classification is a little misleading and has been challenged by some researchers, because LGMD was until recently used only to describe a relatively late-onset and benign form of MD, while sarcoglycan deficiencies can cause disorders with a much broader range of severity.)

Mutations in any of the four sarcoglycan genes cause partial to complete loss of a sarcoglycan protein and affect the entire group of sarcoglycans, which apparently depend on each other to form a stable cluster. One missing sarcoglycan generally causes the loss of the other three.

Laminin-2 is a protein with three chains, one of which has been linked to a form of MD. The laminin alpha-2 chain, when absent or diminished, causes congenital muscular dystrophy, which varies in severity. (Here, too, terminology has become a problem. The term "congenital," which means "onset at birth" and has traditionally been used to describe a severe, early-onset disorder, has become inadequate to describe what is now known to be a disorder that can begin later in life and show lessened severity if laminin-2 is partially preserved.)

The absence of the laminin alpha-2 chain affects the whole laminin-2 protein. It doesn't seem to affect the rest of the DGC, but the loss of laminin-2 is devastating because this protein forms a key linkage between the fiber membrane and structures outside the fiber.

CAN GENE THERAPY SAVE THE DGC?

Animal experiments indicate that replacing a missing membrane protein can probably restore the entire DGC. (A note of caution, however, is that many animal experiments designed to explore this concept were done in so-called transgenic animals, which carry the new gene from before birth; in humans, therapeutic genes will be inserted at a later stage of development, which may affect the outcome.)

A notably successful postnatal experiment in animals is a recent one in which a sarcoglycan deficiency was corrected in hamsters using a technique that can be used in humans. The inserted sarcoglycan gene restored the entire DGC and fully restored membrane function. Some recent gene insertion experiments with utrophin, a close chemical cousin of dystrophin, have shown similar restoration of the DGC in mice, leading scientists to hope for such results in human patients.

The highest hurdle in muscle gene therapy is "a delivery problem" -- how to deliver the gene to the muscle fiber, make it last and express its protein, and prevent destruction of the new protein or the fiber itself by the immune system.

The highest hurdle in muscle gene therapy is what researchers loosely term a delivery problem -- how to deliver the gene to the muscle fiber, make it last and express its protein, and prevent destruction of the new protein or the fiber itself by the immune system.

Viruses are frequently used to insert genes into cells because evolution has provided them with highly efficient mechanisms to invade cellular territory and deposit genetic material there. Viral "vectors," as viruses are called when they're used to transport therapeutic genes, are the targets of major research efforts. (Non-viral vectors are also in development, but they haven't proved as effective as viral vectors so far.)

Several MDA-supported researchers were among those reporting tangible technical progress at the first meeting of the American Society of Gene Therapy (ASGT), which took place May 28-31 in Seattle.

ADENOVIRUSES CAN CARRY LARGE GENES

Both the dystrophin gene and the laminin alpha-2 gene are very large. If viral vectors will be used to transport these genes, the large adenovirus seems best suited. However, the adenovirus has been shown to cause the immune system to ravage it, the therapeutic protein produced in animal gene therapy experiments and probably even the cells that harbor the new genes. The challenge is to sneak the viral vector carrying the therapeutic genes into muscle without attracting the attention of the body's immune surveillance systems.

Some genetic regions in the adenovirus, such as one called E3, seem to dampen the immune response, rather than enhance it, noted one ASGT conference report from Marshall Horwitz of Albert Einstein College of Medicine in New York. Horwitz implied that perhaps this genetic region should be included in gene therapy vectors.

At the same meeting, MDA grantees Paula Clemens of the University of Pittsburgh and Jeffrey Chamberlain of the University of Michigan reported progress in the design of "gutted" (empty) adenoviral vectors that are nothing but viral shells used to insert therapeutic genes into cells. These gutted vectors lack all their own genes but have a viral outer covering, or capsid, that allows the vector entry into the cell.

Even these vectors can trigger an immune response, the lecturers noted, which may result in part from immunogenic proteins in the viral capsid. For this reason, immunosuppressant drugs may be necessary with adenovirus-based gene therapy, at least for a short time after the gene and its viral vector are injected.

The researchers, undaunted, continue to tweak the system. Clemens has found that a commonly used "marker" protein, called beta-galactosidase, which turns cells carrying new genes blue for easy identification, is itself a stimulator of the immune system. Beta-gal, as it's known to the scientific community, is a handy tool for experimenters but isn't necessary to a vector's effectiveness. It will be eliminated in human trials of gene therapy.

Chamberlain's group has also tinkered with the gutted adenovirus to make it less likely to alert the immune system. The gutted viruses need another virus for their production in the laboratory; these other viruses are termed "helper viruses."

Although gene therapists such as Chamberlain and Clemens do their best to keep the helper virus out of the vector that's to be used in experiments, traces of it often remain and are themselves a red flag to the immune system. Chamberlain's group is using a new type of helper virus designed to be less "immunogenic." His group has also made changes to the gutted vector itself.

"We think we have a promising system," Chamberlain said at the May meeting, noting that adult mice with competent immune systems have tolerated the new gutted vectors for three months so far, even though they received vectors made with the old helper virus. He says he thinks vectors made with the new helper virus will be even better.

The sarcoglycans are smaller proteins and arise from smaller genes than do dystrophin or laminin alpha-2, and it's possible they'll be carried by vectors other than adenoviruses in the future (see "AAVs Evade Immune System," below).

However, in an experiment done in Campbell's lab, a gene for delta-sarcoglycan carried by a modified adenovirus was successfully inserted into the delta-sarcoglycan-deficient muscles of hamsters, where it restored the DGC without creating an immune response. The findings, published in the May issue of Molecular Cell, were reported at the ASGT meeting by Campbell's colleague Kathleen Holt. The reasons for the lack of an immune response with delta-sarcoglycan when a notable immune response has been measured with dystrophin delivered by an adenovirus aren't entirely clear; for now, they're a nice surprise.

Campbell's lab is working with mice to develop gene therapy using laminin alpha-2 for congenital muscular dystrophy. The Iowa team plans to use a gutted adenoviral vector developed by Chamberlain's group to deliver the laminin alpha-2 gene.

AAVs EVADE IMMUNE SYSTEM, CAN CARRY SMALL GENES

Much of the ASGT meeting was devoted to extolling the significant virtues of a relative latecomer to the gene therapy scene -- the adeno-associated virus as vector.

The adeno-associated virus (AAV) is a tiny virus, able to carry only very small genes, but it seems to slip past the immune system virtually unnoticed, a definite plus for long-lasting gene therapy. Another secret to the longevity of genes transferred with this virus is that the AAV inserts itself into one of the cell's already existing chromosomes and therefore seems to become a permanent part of the cell's makeup. By contrast, the adenovirus stays apart from the cell's chromosomes and might eventually be lost. (Some researchers say that's a safety check and that integration into a chromosome may be a safety hazard; such concerns will be examined by the Food and Drug Administration before any human trials can begin.)

Sarcoglycan genes are small enough to fit inside an AAV. At least two MDA-supported groups, both at the University of Pennsylvania, are developing gene therapy for limb-girdle dystrophy using sarcoglycans inside AAV vectors.

H. Lee Sweeney is heading one team and Hansell Stedman another; both are working with MDA grantee James Wilson at their institution. Wilson heads the University of Pennsylvania's prestigious Institute for Human Gene Therapy, which has the capacity to develop and test vectors quickly and on a large scale. Campbell's group may also pursue gene therapy strategies to treat LGMD.

Olivier Danos of Genethon in France, who spoke at the May meeting, is also working with AAVs as gene therapy vectors and considers the sarcoglycan genes "very good candidates" for AAV delivery systems.

HUMAN TRIALS NOW IN FDA'S HANDS

"Human trials are closely scrutinized by the FDA," says Bill Moore, who is trained in neuroscience and coordinates MDA's research program. Moore and several MDA-supported researchers have held several meetings with FDA representatives to plan upcoming human trials in DMD and to consider such trials in LGMD.

"We're being cautious, because we want to anticipate and avoid any errors or requests to repeat studies because of non-compliance with FDA regulations or some oversight in study design," Moore says, acknowledging the frustrations of families awaiting news of these trials. "Compliance with FDA regulations from the very beginning saves a lot of time, effort and money contributed by the public to support research."

One outcome of the preliminary meetings has been a decision to test the immunosuppressant medication known as FK506 in boys with DMD before they're given a gene therapy construct. The original plans for a gene therapy trial called for using FK506 to dampen a possible immune response to the gene therapy but didn't include a separate test of FK506.

"This separate test of the drug will determine the effects of FK506 when administered by itself," Moore says. "Then, when the drug therapy is given with the gene transfer, the researchers will be able to separate their effects." MDA grantee Jerry Mendell at Ohio State University plans to oversee this trial .

Moore says there is cautious optimism about upcoming trials for both DMD and LGMD in the near future and that, even if things aren't perfect, it's time to move to human testing. "With all the different techniques and technologies under development, the probability for success is good. We need to keep in mind, though, that research is in a constant state of change, always evolving better techniques and technologies. What we're about to test in these initial trials will probably be greatly modified later, based on new research results."

The first human trials of muscle-directed gene therapy are likely to begin later this year or early next year, depending on the FDA's response to MDA-supported proposals. Their primary goal will be to test the safety of the transferred genes and their viral vectors and to give researchers concrete ideas to take back to the lab. Changes in the vectors and other parts of the delivery systems will almost certainly be made, but the ultimate goal of treating MD by restoring the muscle fiber membrane has never seemed so near.