First comes a tricky virus. Then a blow to the head. The result: an autoimmune assault on the brain, and debilitating epilepsy.
The ancient Greeks called epilepsy the sacred disease; they mistook its seizures for the actions of a powerful and mysterious god. Two thousand years later, neurologists know that seizures occur because nerve cells in the brain fire too rapidly, but the cause of that hyperexcitability is still unknown. Although some of the more than 40 forms of epilepsy can be managed by drugs, others are stubbornly resistant. No drug, for example, can relieve the devastating seizures that occur in patients with Rasmussen’s encephalitis, an extremely rare form of epilepsy that strikes children under the age of ten. The seizures, which tend to occur in only one side of the brain, can be so relentless that in some cases the offending hemisphere must be excised.
A recent study has penetrated the mystery of Rasmussen’s encephalitis. It suggests that patients with Rasmussen’s suffer terrible seizures because their body mistakenly produces antibodies against the receptors that bind glutamate, the most common chemical messenger in the brain. Disturbed immune reactions play a role in diseases as diverse as rheumatoid arthritis, diabetes, and lupus; they have also been implicated in disorders of the central nervous system, such as multiple sclerosis and myasthenia gravis. The new study extends their reach even further. It provides tantalizing support for the idea that autoimmune reactions--in which the immune system essentially attacks parts of the body--may underlie not only some kinds of epilepsy but other neurological diseases that have no known cause.
A breakthrough in epilepsy research was scarcely what neurobiologists Scott Rogers and Lorise Gahring at the Veteran Affairs Medical Center in Salt Lake City intended. Their stated task was humbler: to measure the distribution of glutamate receptors in the brain. Receptors are proteins on the surface of cells that act as gates; receptors on nerve cells allow the cells to exchange chemical messengers with one another. For each type of messenger, there is a particular type of receptor that binds to the messenger much as a lock fits a key. By seeing where glutamate binds to glutamate receptors in the brain, Rogers and Gahring hoped to understand its role in neurological diseases.
The researchers decided to hunt the receptors using antibodies: molecules that recognize foreign substances in the body and tag them for destruction. Since the proteins that make up the glutamate receptors are normally found only in brain tissue and not in the bloodstream, the immune cells that patrol the bloodstream would naturally perceive them as foreign. By injecting those proteins into the blood, the researchers reasoned, they could provoke the immune cells into producing antibodies, which they could then collect and use to seek out glutamate receptors in the brain.
Rogers’s group did the experiment with rabbits: they injected the animals with their own glutamate receptor proteins and waited for the rabbits’ immune system to create antibodies. At that point the experiment took a dramatic turn. After receiving a couple of injections of one protein, two of the three rabbits began twitching. They seemed to be in the throes of an epileptic seizure.
The result was so provocative that Rogers enlisted the help of James McNamara and Ian Andrews, epilepsy specialists at Duke University Medical Center. When McNamara and Andrews examined the rabbits’ brain tissue, they saw a familiar inflammatory pattern. There were clumps of immune cells around the blood vessels, says McNamara. It looked just like what I see in patients with Rasmussen’s encephalitis. Such clusters, he adds, would never be found in a healthy brain.
In a healthy brain, the blood capillaries are tightly encased in a barely permeable layer of cells known as the blood-brain barrier. This cellular seal protects the brain from various substances in the blood that would interfere with the electrical and chemical activity of nerve cells. It permits only tiny, essential molecules, such as oxygen, glucose, and hormones, to squeeze through--but not antibodies or other large molecules, to say nothing of whole immune cells. (The brain has its own immune cells, the microglia.) In some cases, however, the blood-brain barrier can be breached--for instance, by a head injury. What McNamara saw in the rabbit brain tissue suggested that antibodies were somehow penetrating the blood- brain barrier and calling in the immune artillery to assault the glutamate receptor proteins.
Perhaps, he and Rogers decided, that type of assault was also kindling the periodic seizures that characterize Rasmussen’s encephalitis. If so, then patients with Rasmussen’s should have antibodies to the glutamate receptor proteins in their blood. McNamara examined the blood of Rasmussen’s patients and that of healthy controls. He found antibodies to a particular receptor protein--the same one that had triggered seizures in the rabbits--only in the blood of the patients with Rasmussen’s. What’s more, when he filtered the antibodies from the blood, the patients’ seizures subsided.
Researchers have long speculated that Rasmussen’s encephalitis is either a chronic infectious disease or an autoimmune disease. But there has never been much evidence for either hypothesis. The only clue has been that most children get Rasmussen’s after a head injury, measles, or an upper respiratory tract infection. Now Rogers, McNamara, and their colleagues propose that the seizures are triggered in part by antibodies to a glutamate receptor protein. But why does the body start making antibodies to one of its own brain proteins? And how do those antibodies get through the blood-brain barrier?
The scenario McNamara envisions involves a double whammy. We think people get antibodies because they get infected with a microorganism like a bacterium or virus that has a protein that is structurally similar to this glutamate receptor, he says. So say I get infected and my body mounts an immune response. If simultaneously I get a little bump on the head, then those antibodies can gain access to the glutamate receptors in the brain. That can cause focal inflammation that causes seizures that further open the blood-brain barrier.
The result is a vicious cycle: antibodies enter a break in the blood-brain barrier, they cause inflammation, the inflammation causes seizures, the seizures cause more breaks in the blood-brain barrier, which provoke more seizures. The seizures all occur in the region where the blood-brain barrier was initially breached by a blow to the head, says McNamara, which would explain why they are confined to one hemisphere. (One problem with this theory is that the rabbits developed seizures without ever being bumped on the head; it may be that rabbit brains are just not as well insulated as human brains.)
Although Rasmussen’s is exceedingly rare--there were just 48 cases in the United States between 1957 and 1987--epilepsy is not. It affects 2.5 million Americans. McNamara thinks his double-whammy scenario-- a blow to the head followed by an autoimmune attack on the brain--may account for other forms of epilepsy as well. A promising candidate, he says, is a form of epilepsy that affects between 12,000 and 20,000 Americans: those patients have inflamed brain tissue in the temporal lobe that strongly resembles the tissue seen in Rasmussen’s patients.
The most common cause of epilepsy in adults in the U.S. is trauma, says McNamara. People get whacked on the head and then six months or a year later develop an epileptic condition. I’m not saying that an immune mechanism is responsible. But I am saying that we need to think about this differently now.