General consensus among Alzheimer’s researchers has it that the disease’s main culprit, a protein called amyloid beta, is an unfortunate waste product that is not known to play any useful role in the body—and one that can have devastating consequences. When not properly cleared from the brain it builds up into plaques that destroy synapses, the junctions between nerve cells, resulting in cognitive decline and memory loss. The protein has thus become a major drug target in the search for a cure to Alzheimer’s.

Now a team of researchers at Harvard Medical School and Massachusetts General Hospital are proposing a very different story. In a study published this week in Science Translational Medicine, neurologists Rudolph Tanzi and Robert Moir report evidence that amyloid beta serves a crucial purpose: protecting the brain from invading microbes.

“The original idea goes back to 2010 or so when Rob had a few too many Coronas,” Tanzi jokes. Moir had come across surprising similarities between amyloid beta and LL37, a protein that acts as a foot soldier in the brain’s innate immune system, killing potentially harmful bugs and alerting other cells to their presence. “These types of proteins, although small, are very sophisticated in what they do,” Moir says. “And they’re very ancient, going back to the dawn of multicellular life.”

Just as amyloid beta is known to do, these antimicrobial proteins can build up and form fibrils that, when not properly regulated, have harmful effects. Yet unlike amyloid beta, their role in the immune system has been widely accepted: They prevent microbes from adhering to host cells, and ultimately trap them. So five years ago the pair of researchers set out to determine whether amyloid beta could also act as a natural antibiotic. They had previously explored their hypothesis in vitro, but in this new study they used worms and mice. For these animal models they compared what happened when amyloid beta was overexpressed to when it was not produced, and found that the former led to increased resistance to infection and longer survival rates.

The team first infected cultured human and hamster cells with a type of fungus called Candida albicans and found that high expression of amyloid beta had a protective effect, doubling the number of cells that were not infected. The researchers then moved on to roundworms, or nematodes, which usually do not survive for more than two to three days once the fungus takes hold. The nematodes with overexpressed amyloid beta, however, were still going strong five to six days after they were infected. Finally, the researchers infected the brains of mice with a strain of salmonella to cause meningitis. Mice that were genetically altered to overproduce human amyloid beta survived nearly twice as long as mice that did not have the protein at all.

Most shocking of all, according to Tanzi and Moir, was that when they injected bacteria into the brains of Alzheimer’s mouse models, amyloid plaques—the hallmark of the disease—formed within 48 hours. “We didn’t know this was even possible,” Tanzi says, “that amyloid plaques would form rapidly overnight.” And in the middle of each plaque was one Salmonella bacterium, supporting the theory that the amyloid deposition had formed around the microbe as an entrapment mechanism—just like LL37 and other established antimicrobial proteins.

“These results are particularly intriguing,” says Anna Palamara, a microbiologist at the Sapienza University of Rome who was not involved in the study. “Previous [research] shows that several infectious agents, including viruses, trigger amyloid beta production and accumulation.” Herpes and influenza are just two of the infections that have been tested by other research teams.

The Harvard team’s new findings provide further evidence that Alzheimer’s could be inadvertently spurred by an infection that causes the formation of too much amyloid. As people get older the immune system and blood-brain barrier become increasingly compromised, making it easier for microbes to sneak into the brain. It wouldn’t take many of these pathogens, according to Tanzi, to cause amyloid buildup. “And that could rapidly start the cascade toward the disease,” he says, “causing tangles and inflammation. You’ve got all three pillars of Alzheimer’s right there.”

“It is possible to speculate that during a mild infection the production of amyloid beta may help,” Palamara says. “But in the presence of persistent or repeated infections, amyloid beta levels may accumulate, exceeding a threshold. In this case its protective role might change to the well-known neurotoxic one.”

The idea that amyloid beta has a positive function in the body could potentially change how scientists approach potential treatments. Instead of attempting to completely eliminate the protein, “we might want to think about just dialing it down,” Tanzi says. Moreover, Moir adds, the drugs in trials now are for the most part designed to reduce inflammation by targeting pathways in the adaptive immune system. But if amyloid production and deposition are innate immune responses, then targeting pathways of innate immunity or the microbes themselves may be the way to go.

They do not expect convincing the scientific community of this to be easy. “This is really going to cause a lot of unrest in the field,” Tanzi says. “Any new revolutionary discovery is first ridiculed, then violently opposed, and finally taken to be self-evident. We’re ready for the ridicule and the violent opposition, and we think we have enough data so that we can look forward to self-evident.”

But the pair has a long way to go. They are now moving forward with a plan to systematically characterize the microbes found in the aging brain. From there they hope to identify the pathogens that may be involved in the onset of Alzheimer’s—as well as those that potentially play a role in other amyloid diseases, such as diabetes.

“We’re at the top of a mountain with a freshly formed layer of snow,” Moir says. “Where you go is where you choose. There’s so much to explore.”