Your body's immune system plays a critical role in health and disease. In the last issue of Quest, we introduced the major parts of the immune system, and saw how autoimmune disease develops.

In this article, we turn to the barrier to successful gene therapy posed by the immune system.

It's little wonder that gene therapy is so hard: Over thousands of years of evolution, your body's immune system has been exquisitely tuned to combat exactly the sort of cellular invasion that gene therapy requires.

Almost all current gene therapy strategies use viruses to deliver genes to your body's cells. Viruses are very good at this: They've perfected their ability to find, enter and infect your cells with their genes.

In this regard, they represent highly evolved gene delivery vehicles (vectors), seemingly ideal for the very task required of them by gene therapy.

But if viruses have perfected their art of attack, your immune system has perfected equally well its art of self-defense: Once found, infected cells are killed to prevent the spread of infection.

While this thwarts a cold virus attack on your respiratory system, for instance, it also thwarts researchers' best efforts to cure neuromuscular disease by introducing therapeutic genes into muscle.

WEAPON OF CHOICE

One of the earliest, and still the most widely used, gene vectors is the adenovirus.

"It's been used in vaccine programs, and it's known to be pretty safe to work with," says MDA-funded researcher Dr. Paula Clemens of the University of Pittsburgh. "Also, it's relatively easy to manipulate its genes, and we can grow a lot of it fairly easily. And adenovirus, unlike some other vectors, will infect cells that aren't dividing."

This is especially important for muscle work, since muscle cells lose the ability to divide as they mature. Adenovirus is also a rather large virus, with more room for therapeutic genes than most other vectors. This is critical for treatment of Duchenne muscular dystrophy, since the gene for dystrophin, the protein missing in this disease, is the largest gene known.

Unfortunately, despite their many advantages, adenoviral vectors provoke a very strong response from the immune system.

While scientists didn't expect the immune system would ignore adenoviral vectors, not many researchers focused on it in the early days of gene therapy research.

"When people began thinking about gene therapy, it was clear this was going to be an issue," Clemens says. "It wasn't addressed experimentally early on, because first you had to do a proof of principle for gene delivery, to find out if you could even make gene delivery work. The immunology issues weren't addressed early, but they weren't a total surprise."

As it turns out, though, these issues currently represent one of the major challenges to successful gene therapy for virtually every disease being addressed.

OVERCOMING THE IMMUNE DEFENSE

MDA grantee Dr. George Karpati of McGill University in Montreal explains that there are three potential sources of immunologic reaction.

"First, there is the covering, or capsid, of the virus, which is made of protein. This is what the virus uses to latch onto the cells, to get inside. A multitude of viruses are taken into the cell, where these capsids come off. So this is a huge foreign protein load, right into the cell."

The invaded cell uses its MHC proteins (molecules that distinguish each person's immune system as unique) to show the immune system fragments of these capsids. In essence, says Karpati, "it says to the immune system, "I've been inundated with a foreign protein! Get rid of me!'"

How can this reaction be prevented?

"It's very difficult," Karpati says. "One way to do it is to suppress the immune system of the host. The big question is for how long. Can we use just a few weeks of immunosuppression to prevent the host from reacting to the capsid, or will we need to go longer? The answers are still out on this."

Karpati's group is experimenting with short-term administration of a new immunosuppressant drug, FK506, also called tacrolimus. FK506 blocks the action of T cells (the ranking members of the immune system that search for foreign proteins to destroy). So far, his experiments show good results with only a few weeks of treatment.

"With this treatment, our animals maintain a nice high level of gene expression (production of a protein by a gene) for two months. After that, though, the expression starts dropping," Karpati says.

The likely reason for this drop goes to the second target of immune attack: viral proteins made by the host cell after infection.

Anticipating this problem, researchers have used vectors specifically modified to prevent expression of viral proteins.

"When we remove what's called the E1 region of the viral genome," Karpati says, "the vector should no longer be able to express its own genes (make its own proteins). And to a large extent, this is what happens. But you still get so-called "leaky expression,' even after E1 removal, and these proteins are also immunogenic (producing an immune response)."

In contrast to the enormous but short-term antigen load (a protein is called an antigen when it's the subject of immune surveillance) created by the capsids, leaky expression is a low-level, sustained problem, since the viral proteins are made continuously after infection. Even after the capsid crisis has passed, the immune system will respond to these, killing off the cells that express them.

The MHC molecule presents peptides (protein fragments) for T cells to examine for possible destruction. Will they present viral peptides, or peptides from transgenes?

FLANKING MANEUVERS

A number of strategies have been designed to reduce leaky expression. Removing another viral region, called E4, has helped. MDA-funded researcher Jeffrey S. Chamberlain of the University of Michigan and a group including Clemens went one step further: They removed all of the viral genes from their vector, leaving only the capsid and some snippets of structural DNA to sandwich their dystrophin gene.

"There were two goals for our large-capacity vector," Clemens says. "The first was to simply accommodate the dystrophin gene." She points out that earlier vectors didn't have enough room to fit the entire dystrophin gene. Instead, they used a "minigene," whose ability to substitute for dystrophin wasn't fully established.

"The second reason was to prevent the virus from expressing its own proteins." With no viral genes, there can be no leaky expression. With no leaky expression, there should be no second-level immune attack.

Or so it would seem on first glance.

However, if the vector has no genes of its own, it can't make more of itself in the cell cultures used to grow it in the first place. To solve this problem, the vector must be incubated with a full-fledged adenovirus, called a "helper virus," which supplies the crucial machinery for growth. The vector is then separated from the helper before using it for gene transfer experiments. Unfortunately, the separation isn't perfect, so the solution containing the vector still has some helper in it.

These viruses infect the cells along with the vector, and again the result is leaky expression. While better separation techniques may help, it's not clear yet whether the helper contamination can be reduced enough to avoid setting off an immune attack.

Clemens also points out that each gene delivery system comes with its own unique set of problems.

Most gene therapy experiments include a marker gene as part of the therapeutic package. The gene codes for a protein called beta-galactosidase, or beta-gal, for short. When specially treated, beta-gal will make the infected cells turn blue, allowing researchers to find them easily. Clemens has shown that the beta-gal delivered by adenovirus can be immunogenic, and so her group is currently developing a vector without it.

"However," Clemens notes, "some research has shown that when you deliver beta-gal with a different virus, called adeno-associated virus (AAV), you don't get much of an immune response. So you have to look at each system independently."

Even with most of its genes intact, AAV doesn't seem to provoke much immune reaction. Unfortunately, it's much too small for the full dystrophin gene.

For now, the problems of helper contamination and leaky expression remain, continuing to provide targets for the immune system.

What other problems are there?

THE ENEMY WITHIN

"The third potential source of immune reaction is the transgene protein (the protein made from the transferred gene) itself," Karpati says.

During development, your immune system learns "who you are" when it's exposed to all of your own proteins in the thymus (a small gland in your chest). But if you've never made dystrophin because your gene for it is defective, your immune system may think dystrophin is a foreign protein, and may react to it when it sees it.

It's a chilling prospect for treating disease if the immune system will attack the cure.

"However," Karpati notes, "in a boy with Duchenne, dystrophin doesn't have to be an entirely foreign protein, because of the revertant fibers found in the muscles of most boys with DMD."

Revertant fibers are muscle cells that have reverted to expressing dystrophin, probably because their dystrophin gene has re-mutated to a useful form. There aren't enough of them to prevent muscle weakness, but there should be enough to teach the immune system that dystrophin is a self-protein.

"If that's the case, then the immune system of a boy with Duchenne should have been educated properly, so that it would not recognize dystrophin as a foreign molecule," Karpati explains.

While this would seem to put to rest the third problem, the situation is yet more complicated. In fact, Karpati says, "we know that antibodies to dystrophin still form a great deal, when we introduce the vector with the dystrophin gene."

At first glance, this appears to be an insurmountable problem for gene therapy -- after all, antibodies are one of the immune system's most potent and long-lasting weapons against foreign proteins.

"However," Karpati points out, "these antibodies do not seem to be particularly harmful. Why they form and why they don't cause trouble is still unknown."

In the face of all this, it's probably fair to say that most researchers take a cautiously optimistic view of dystrophin's potential to provoke an immune response. Nonetheless, its involvement remains a possibility that must be tested in further experiments.

If dystrophin reaction does turn out to be a problem, the gene therapists may turn to new strategies being developed by researchers in organ transplantation.

SIMILAR STRATEGIES

While the particulars are different, gene therapy and transplantation share some similar immune obstacles. Currently, long-term acceptance of a transplanted organ requires long-term suppression of the host's T cells with drugs such as FK506 and cyclosporine. Nonetheless, these T cells remain ever ready to rise up again, and often do if the drugs are stopped.

To find alternatives to this situation, transplant researchers are now taking advantage of new findings in immunology about how T cells respond.

As Dr. Daniel B. Drachman of Johns Hopkins University in Baltimore pointed out in the last issue of Quest, T cells need two signals to mount an immune attack. Cyclosporine and FK506 interfere with the first of these signals. A new class of experimental drugs targets the second signal.

When a T cell has its second signal blocked while its first is being activated, it becomes unresponsive, and may die off. As a result, the immune system loses the cells that are responding most strongly to the new antigens. The result is immunologic tolerance for the transplanted organ. If dystrophin does turn out to be immunogenic, inducing tolerance to it may be the only way to stop the immune reaction.

What about solving all the problems of immune reaction by inducing tolerance to the capsids and other viral proteins?

Researchers are hesitant to pursue this for now. Even if it's possible, it would probably increase the risk of infection from other, natural adenoviruses. No longer seen as the enemy, they would be more likely to cause a runaway infection in the host.

Because there are many kinds of adenoviruses, it may be possible to find one, or engineer one, that is different enough from the most infectious types to allow safe toleration. At the moment, however, this possibility remains only speculation.

So if leaky expression can be stopped by some modification of the vector, if the initial capsid reaction can be suppressed and if reaction to dystrophin is not a significant problem, what other barriers remain for gene therapy?

Researchers will still have to grapple with getting the vector to all the cells that need it, and only to those cells. This may require targeting strategies that are still in development.

And, Karpati says, "there are probably things we don't know about yet because we've never reached the point where they would show up. For instance, might the cell nucleus itself have a way of recognizing this foreign DNA, and eliminating or silencing it? It doesn't look like it, but in order to be sure, we have to get past those initial points where the cell is wiped out by the immune system."

Where does all this leave gene therapy for Duchenne? Clemens points out that in its own way, the field of gene therapy is having to work out many of the same issues seen by transplant researchers in the early years of that field. "The trajectory for gene therapy is really similar to what we saw in organ transplants, except we aren't as far along," she says.

New ideas, experiments and strategies are needed to make progress at each step.

"For the immune system problems," Clemens says, "there are immunosuppressant cytokines, costimulatory blockers, soluble receptors: a whole laundry list of things. And like everything else in science, it will probably turn out to be more complicated than we expect. There probably won't be some single magic bullet to solve this problem; it's going to require finding the right combination."