Alzheimer’s Protein May Be a Natural Immune Protein


by Jim Schnabel

May 19, 2010

The protein amyloid beta (A-beta) is best known for its tendency to form sticky, insoluble amyloid fibrils in the brains of people with Alzheimer’s disease. Scientists have long studied the possible ways in which it could drive the disease but have paid much less attention to what its natural, evolved function might be. In fact, many believed that A-beta was a mere waste product, left over from the processing of a larger protein.

Now a team of researchers from a prominent Alzheimer’s laboratory has published preliminary evidence that A-beta does have a natural function: as an antimicrobial molecule to help the brain ward off infections. If proved true, this would affect how researchers develop drugs to stop Alzheimer’s.

“If A-beta is essential, then I think we’d have to think about targeting it in Alzheimer’s the same way we target cholesterol in heart disease—not whacking it with a sledgehammer but dialing it down to safer levels,” says Rudolph Tanzi, one of the study’s senior authors and a member of the Dana Alliance for Brain Initiatives. He is a co-director of the Massachussetts General Hospital’s Institute for Neurodegenerative Disease (MIND), which is affiliated with Harvard Medical School.

For three decades, Tanzi’s research has focused on the genes and their mutations that make people more susceptible to Alzheimer’s. Recently he and a younger investigator at MIND, Robert Moir, began to suspect that A-beta had a relationship to the so-called innate immune system, an evolutionarily ancient system that includes small proteins (“peptides”) that have broad-spectrum antimicrobial effects.

In genetic studies, Tanzi saw hints that irregularities in the functions of innate immunity genes were somehow linked to Alzheimer’s risk. Separately, Moir noticed that A-beta has striking similarities, in its structure and in other properties, to an innate immunity protein, LL-37, a known anti-microbial peptide.

In scientific meetings last year, and more recently in the March issue of the open-access journal PLoS One, Tanzi and Moir and their colleagues presented evidence suggesting that A-beta is indeed an antimicrobial cousin of LL-37. In lab tests against a dozen common microbes, the two short proteins show a similar killing effect, with A-beta more potent than LL-37 against some microbes.

The researchers saw the same effect in tests using brain tissues removed at autopsy:  The A-beta-filled samples from people with Alzheimer’s, compared to samples from age-matched people who didn’t have the disease, could kill microbes more readily, and in proportion to the samples’ A-beta content. The researchers were able to block this effect by blocking A-beta with antibodies, thus showing that A-beta played a causal role.

Circumstantial evidence

“I don’t doubt that those data are correct,” says Charles Glabe, a neurobiologist and A-beta expert at the University of California, Irvine. “We’ve done some of those experiments and can reproduce them.”

Tanzi points to other evidence from prior Alzheimer’s research that seems broadly consistent with the new theory. Traumatic brain injuries, including head injuries and strokes, are associated with increased A-beta production and predict increased Alzheimer’s risk. Such injuries would be expected to activate inflammatory and immune system components in the brain. And the brains of people with Alzheimer’s generally do exhibit signs of increased inflammation in disease-affected areas.

Cardiovascular problems also are associated with increased Alzheimer’s risk; in fact a particular variant of a cholesterol-transport protein, apolipoprotein-E, is linked both to higher Alzheimer’s risk and to higher cardiovascular disease risk. Tanzi suspects that vascular incidents leading to the cutoff of oxygen to brain cells could cause damage and thus trigger increased A-beta production.

“The idea is that any type of injury to the brain will trigger the production of A-beta as part of the innate immune system,” he says.

Moreover, as Tanzi and his colleagues search the genomes of families that have a predisposition for Alzheimer’s, he says, they are finding links to genes relating to stroke and cardiovascular disease risk, as well as genes that regulate the innate immune system.

“It’s starting to paint a picture in which A-beta is still central to the disease, but the upstream events may include increasing insults to the brain as we age, and overreaction to that insult by the innate immune system,” he says.

To accept that A-beta truly is a natural anti-microbial peptide (AMP), Glabe would like to see harder evidence, for example in animal models, that A-beta’s production by brain cells increases in response to infection, the way LL-37 and other AMPs do. “Is it regulated in the same fashion and in the same cells? Does it act like it’s a member of this [AMP] family? That’s an obvious thing to test,” he says.

No such test has yet been published. But Tanzi and his colleagues are currently searching for evidence of infectious agents that may be hidden triggers of Alzheimer’s, including the bacterium chlamydia pneumoniae, a common cause of pneumonia in the elderly. He suggests that if there were a common microbial pathogen that triggers amyloid as we age, then we might be able to find a vaccine to prevent it. Other labs in the past have proposed such a hypothesis, but were met with general skepticism. To Tanzi, the links his group has turned up between A-beta and innate immunity mean that the possibility of an infectious trigger needs to be considered thoroughly, “even if it sounds crazy now.”

Is A-beta’s toxicity good, too?

Another unresolved question is whether A-beta kills microbes in the same way that it seems to kill healthy neurons. Glabe suspects that in both cases A-beta “oligomers,” relatively small clumps of A-beta that are still soluble, weaken key membranes in bacteria and neurons, causing them to leak and die. “It could be a matter of balance, so that at some concentration [A-beta’s toxicity] is good, but if you get too much of it, it’s bad,” says Glabe.

Tanzi suspects that this membrane-weakening effect might work against microbes, but may be only a secondary cause of damage to neurons in Alzheimer’s disease. He thinks it more likely that A-beta’s most direct effect on cognition is to shut down synapses on neurons, as other researchers have suggested is the case. He also sees this as an extension of a natural function. “It makes sense that when you have an injured part of the brain, it isn’t going to work so well, and you would want to quell activity there,” Tanzi says.

If Tanzi’s suspicions are correct, then removing A-beta entirely could interrupt normal brain function and leave it exposed to infectious agents. “Innate immunity is your first line of defense,” he says. “If you don’t have LL-37, for example, you’ll die very young of infection.”

For the moment, arguments like these haven’t made Tanzi and Moir popular with some Alzheimer’s researchers who have invested their efforts in A-beta-eradicating therapies. “But Rob Moir just got a big NIH grant to work on this,” Tanzi says happily. “So it looks like the system is working.”