Robert Moir
© Aram Boghosian for STAT
Robert Moir prepares a microbial broth in his lab at Massachusetts General Hospital's Genetics and Aging Research Unit.
Robert Moir was damned if he did and damned if he didn't. The Massachusetts General Hospital neurobiologist had applied for government funding for his Alzheimer's disease research and received wildly disparate comments from the scientists tapped to assess his proposal's merits.

It was an "unorthodox hypothesis" that might "fill flagrant knowledge gaps," wrote one reviewer, but another said the planned work might add little "to what is currently known." A third complained that although Moir wanted to study whether microbes might be involved in causing Alzheimer's, no one had proved that was the case.

As if scientists are supposed to study only what's already known, an exasperated Moir thought when he read the reviews two years ago.

He'd just had a paper published in a leading journal, providing strong data for his idea that beta-amyloid, a hallmark of Alzheimer's disease, might be a response to microbes in the brain. If true, the finding would open up vastly different possibilities for therapy than the types of compounds virtually everyone else was pursuing.

But the inconsistent evaluations doomed Moir's chances of winning the $250,000 a year for five years that he was requesting from the National Institutes of Health. While two reviewers rated his application highly, the third gave him scores in the cellar. Funding rejected.

Complaints about being denied NIH funding are as common among biomedical researchers as spilled test tubes after a Saturday night lab kegger. The budgets of NIH institutes that fund Alzheimer's research at universities and medical centers cover only the top 18 percent or so of applications. There are more worthy studies than money.

Moir's experience is notable, however, because it shows that, even as one potential Alzheimer's drug after another has failed for the last 15 years (the last such drug, Namenda, was approved in 2003), researchers with fresh approaches - and sound data to back them up - have struggled to get funded and to get studies published in top journals. Many scientists in the NIH "study sections" that evaluate grant applications, and those who vet submitted papers for journals, have so bought into the prevailing view of what causes Alzheimer's that they resist alternative explanations, critics say.

"They were the most prominent people in the field, and really good at selling their ideas," said George Perry of the University of Texas at San Antonio and editor-in-chief of the Journal of Alzheimer's Disease. "Salesmanship carried the day."

Dating to the 1980s, the amyloid hypothesis holds that the disease is caused by sticky agglomerations, or plaques, of the peptide beta-amyloid, which destroy synapses and trigger the formation of neuron-killing "tau tangles." Eliminating plaques was supposed to reverse the disease, or at least keep it from getting inexorably worse. It hasn't. The reason, more and more scientists suspect, is that "a lot of the old paradigms, from the most cited papers in the field going back decades, are wrong," said MGH's Rudolph Tanzi, a leading expert on the genetics of Alzheimer's.

Even with the failure of amyloid orthodoxy to produce effective drugs, scientists who had other ideas saw their funding requests repeatedly denied and their papers frequently rejected. Moir is one of them.

For years in the 1990s, Moir, too, researched beta-amyloid, especially its penchant for gunking up into plaques and "a whole bunch of things all viewed as abnormal and causing disease," he said. "The traditional view is that amyloid-beta is a freak, that it has a propensity to form fibrils that are toxic to the brain - that it's irredeemably bad. In the 1980s, that was a reasonable assumption."

But something had long bothered him about the "evil amyloid" dogma. The peptide is made by all vertebrates, including frogs and lizards and snakes and fish. In most species, it's identical to humans', suggesting that beta-amyloid evolved at least 400 million years ago. "Anything so extensively conserved over that immense span of time must play an important physiological role," Moir said.

What, he wondered, could that be?


Moir, a native Australian, isn't sure where he gets his anti-establishment streak, but during his undergraduate days down under he took a microbiology course from Nobel laureate-to-be Barry Marshall, who bucked orthodoxy for years by believing that a bacterial infection (H. pylori) causes ulcers. Marshall even infected himself to prove the point. "Everyone thought he was crazy," Moir recalled. "He was crazy, but he was also right."

As a Ph.D. student at the University of Melbourne in the early 1990s, he was the lead author of an Alzheimer's paper in the Journal of Neurochemistry and co-author of many more in such leading journals as Neuron and Cell. They mostly stuck to the amyloid dogma, but even then, "Rob was always a dogged researcher, meaning like a dog with its bone," said Colin Masters, the distinguished Alzheimer's researcher at Melbourne under whom Moir, now 57, studied. "He never gives up and has never been one to follow the standard line."

In 1994, Moir changed hemispheres to work as a postdoctoral fellow with Tanzi. They'd hit it off over beers at a science meeting in Amsterdam. Moir liked that Tanzi's lab was filled with energetic young scientists - and that in cosmopolitan Boston, he could play the hyper-kinetic (and bone-crunching) sport of Australian rules football. Tanzi liked that Moir was the only person in the world who could purify large quantities of the molecule from which the brain makes amyloid.

Moir initially focused on genes that affect the risk of Alzheimer's - Tanzi's specialty. But Moir's intellectual proclivities were clear even then. His mind is constantly noodling scientific puzzles, colleagues say, even during down time. Moir took a vacation in the White Mountains a decade ago with his then-6-year-old son and a family friend, an antimicrobial expert; in between hikes, Moir explained a scientific roadblock he'd hit, and the friend explained a workaround.

Moir's inclination toward unconventional thinking took flight in 2007. He was (and still is) in the habit of spending a couple of hours Friday afternoons on what he calls "PubMed walkabouts," casually perusing that database of biomedical papers. One summer day, a Corona in hand, he came across a paper on something called LL37. It was described as an "antimicrobial peptide" that kills viruses, fungi, and bacteria, including - maybe especially - in the brain.

What caught his eye was that LL37's size and structure and other characteristics were so similar to beta-amyloid, the two might be twins.

Moir hightailed it to Tanzi's office next door. Serendipitously, Tanzi (also Corona-fueled) had just received new data from his study of genes that increase the risk of Alzheimer's disease. Many of the genes, he saw, are involved in innate immunity, the body's first line of defense against germs. If immune genetics affect Alzheimer's, and if the chief suspect in Alzheimer's (beta-amyloid) is a virtual twin of an antimicrobial peptide, maybe beta-amyloid is also an antimicrobial, Moir told Tanzi.

If so, then the plaques it forms might be the brain's last-ditch effort to protect itself from microbes, a sort of Spider-Man silk that binds up pathogens to keep them from damaging the brain. Maybe they save the brain from pathogens in the short term only to themselves prove toxic over the long term.

Tanzi encouraged Moir to pursue that idea. "Rob was trained [by Marshall] to think out of the box," Tanzi said. "He thinks so far out of the box he hasn't found the box yet."

Moir spent the next three years testing whether beta-amyloid can kill pathogens. He started simple, in test tubes and glass dishes. Those are relatively cheap, and Tanzi had enough funding to cover what Moir was doing: growing little microbial gardens in lab dishes and then trying to kill them.

Day after day, Moir and his junior colleagues played horticulturalists. They added staph and strep, the yeast candida, and the bacteria pseudomonas, enterococcus, and listeria to lab dishes filled with the nutrient medium agar. Once the microbes formed a thin layer on top, they squirted beta-amyloid onto it and hoped for an Alexander Fleming discovery-of-penicillin moment.

They got it in 2009. The nearly invisible layers of bugs had turned into visible floating clumps, the result (Moir determined through microscopy) of beta-amyloid molecules cross-linking into a tangled skein - Tanzi calls it a "nano-net" - that trapped and killed them.

Bugs in dishes are one thing, and pretty distant from brains. Dozens of patients and families have donated Alzheimer's brains to MGH, however, and they were at Moir's disposal. He took tiny bits of tissue from 32 of them (12 to 24 hours after death) and also from 13 healthy brains. After grinding up each sample in a sterile mortar-and-pestle ("low tech, but effective," Moir said), he shot a broth of candida into each.

In the mush from Alzheimer's patients' hippocampus, which is ravaged by the disease and contains lots of beta-amyloid, candida growth plummeted compared to its riot of reproduction in both healthy brains and beta-amyloid-free Alzheimer's cerebellum. But when Moir added antibodies that sideline beta-amyloid, the candida grew in the hippocampus like the yeast in a good sourdough.

That, Moir thought, was persuasive evidence that beta-amyloid fights microbes. The first journal he submitted his results to, Science, "looked like they would take it," Moir recalled. But after what editors told him was "a consultation with our Alzheimer's disease experts," it rejected the paper. So did three other journals.

It was Moir's first glimpse at what he was up against. It wouldn't be his last.

He and his colleagues finally published their discovery in March 2010, receiving what he calls a "mixed reception." "The old idea that amyloid-beta is intrinsically pathological had been around for 30 years," Moir said. "Some test tube experiments weren't going to change that."

He'd have to do more work, this time in animals. Fortunately, he had just received a five-year NIH grant, his first, based in part on the data in that 2010 paper. The $400,000 a year was just enough to cover his salary and a postdoc - and to buy $400 mice and other research necessities.

Moir and his colleagues started with "Alzheimer's mice," genetically engineered to express human beta-amyloid and develop amyloid plaques by about five months (30-something in human years). They injected salmonella into the brains of month-old, plaque-free animals.

The bacteria acted like a hunk of cow dropped into a tank of piranha: Amyloid fibrils emerged practically instantly, and plaques formed within 48 hours, ensnaring the bacteria and preventing brain infection. It was more evidence that beta-amyloid is not a neuro-toxic mistake but a defense against microbes - which enter the brain more frequently as people age and the blood-brain barrier becomes leaky.

They wrote up that result in 2014 and excitedly sent it off for publication, to first one journal and then another five. Most didn't even send it out for peer review, instead rejecting it with the standard email brushoff that "this paper doesn't fit our needs." "What they meant," Moir said, "was f- off."

The good news was that as they were amassing rejections, they were collecting more and more evidence - eventually, in three kinds of cells, including human neurons growing in lab dishes, and four kinds of animals - that beta-amyloid fights microbes.

Still, there were no takers for their paper, and Moir was publishing little else - a potential death spiral in a publish-or-perish era. He could easily have gotten the paper into a lesser journal, said Alzheimer's scientist Ashley Bush of Melbourne, but "the ideas in it were too important. I have to hand it to him: He took the risk because of his belief in the data."

In late 2015, an editor at Science Translational Medicine, which had rejected the paper, heard a talk Tanzi gave about his research with Moir. Orla Smith (who wasn't aware of the rejection) approached Tanzi afterward and encouraged him to submit the study to her journal.

It was accepted and published in 2016, two years after the scientists had first tried. The reception was better than Moir expected. The journal's editors called the findings a "rehabilitation of an amyloid bad boy" which identified "inflammatory pathways as potential new drug targets for treating" Alzheimer's. It was named one of the top five neurology advances for 2016. A key Alzheimer's expert wrote that the findings "deserve to be followed up right now."

"Right now," of course, was already years after the discovery of how beta-amyloid protects against microbes could have seen the light of day. And by then, Moir's NIH funding had run out.

Moir was hopeful, with his paper finally published, that he could get a new grant, but he would soon discover just how entrenched were the amyloid true believers. Though dozens of drugs targeting beta-amyloid had failed during the decade he had been pursuing his unorthodox ideas, the bulk of NIH Alzheimer's funding still backed conventional projects.

He was trying to keep his head above water. After his first NIH grant ended in 2014, Moir had received a $500,000 research grant from the Boston-based Cure Alzheimer's Fund, which was founded by venture capitalists. It's receptive to studies that don't necessarily adhere to the amyloid hypothesis, said CEO Tim Armour, because it has "a higher tolerance for risk" than NIH does. Moir's work, Armour said, "addressed a nagging scientific question: Why would evolution preserve amyloid plaques if they are simply toxic byproducts?" Answering that "would have a dramatic impact on the way the field understands the disease."

NIH disagreed. In 2016, one of its study sections evaluated Moir's proposal to see whether herpes simplex virus 1 (HSV-1), which causes cold sores and can reach the brain, might promote both amyloid plaques and tau tangles. He wanted to test the role of HSV-1 in both mice and a 3D network of human neurons that Tanzi and his colleagues had created. Because it readily formed plaques and tangles, they called it "Alzheimer's in a dish."

The NIH study section had other ideas, according to copies of the evaluations Moir provided STAT. Two reviewers praised the proposal, giving it scores of 2 (out of 9, where 1 is the best) on criteria such as significance and innovation. If the third had been comparable, Moir would have been funded.

Instead, the third review was a death sentence, rating Moir's proposal 6 or 4 on key criteria. It complained that Alzheimer's "develops in only a fraction of people infected with HSV-1," as if that ruled out a connection - after all, fewer than 1 in 10 smokers develop lung cancer. It dinged Moir for being a mere assistant professor and for having had only one previous NIH grant. And it complained that the main novelty of his proposal was studying HSV-1, "whose relationship to Alzheimer's remains tenuous." Moir thought that turning hunches into facts was why scientists do science.

He probably had the bad luck to be assigned a reviewer "who's either completely unqualified, or simply says, 'I don't believe it' [that microbes have anything to do with Alzheimer's], or who's a dyed-in-the-wool believer in the old amyloid hypothesis and doesn't even read the scientific literature anymore," Tanzi speculated. "This is why the most boring, incremental stuff gets funded. If it's a Hail Mary, they figure nah, no one will ever catch that" - so risky research is rejected.

When a grant proposal receives crazily divergent reviews, as Moir's did, the study section is supposed to discuss it to see whether the outlier might have misunderstood something. That didn't happen.

STAT asked study section chairman Dr. J. Paul Taylor, a neurogeneticist at St. Jude Children's Research Hospital, why. He said that although "widely divergent scores would typically lead to discussion," on the key criteria these review are not "widely divergent." The proposal likely fell "in the very fat middle of the pack and on the bubble for getting discussed."

The verdict stood.

Dr. Richard Hodes, director of the National Institute on Aging, said it "does its best" to make sure there are appropriate experts on study sections (which are not institute-specific). Asked whether scientists who stick to old orthodoxies can torpedo innovative research, he called it "a great question," and one NIA officials worry about. NIA "feels a strong responsibility" to not "be too narrow in our focus" when it comes to Alzheimer's research, he added. It supports a clinical trial of the antiviral drug valacyclovir in Alzheimer's patients who test positive for herpes simplex, for instance.

Using CureAlz funding, Moir did the experiments that NIH didn't think important enough to support: testing whether beta-amyloid can protect against HSV-1 and two other herpes viruses, called HHV-6A and HHV-6B.

The pathogens he'd used before, although useful for establishing the general principle, weren't thought to be terribly relevant to Alzheimer's. HSV-1 was: More than 100 papers had linked it to Alzheimer's, such as by showing that its genes are found within amyloid plaques. None, however, had shown it triggered beta-amyloid formation and then plaque deposition. "That was what we wanted to address," Moir said. "Can herpes accelerate amyloid-beta deposition?"

To find out, he injected herpes viruses into the Alzheimer's-in-a-dish. Within 48 hours, beta-amyloid encased the viruses in nano-nets and prevented them from infecting neurons. But it also triggered the formation of plaques between neurons, which in turn sparked formation of tau tangles inside them. Mice genetically engineered to produce human beta-amyloid did the same, quickly developing virus-trapping beta-amyloid plaques.

Virus-trapping plaques also set off raging inflammation, as immune responses tend to do. That raised the possibility that a chronic herpes presence in the brain, which becomes more likely with age, might trigger sustained activation of the beta-amyloid immune response and therefore inflammation. That inflammation, and not beta-amyloid, might be what's killing synapses and neurons.

This research, too, had a tortured path to publication, racking up rejections in 2017. Early this year a similar paper began making the rounds, reporting that Alzheimer's brains are, compared to healthy ones, awash in HHV-6A and HHV-7, especially in regions most ravaged by the disease. That paper was published in Neuron this past July. So was Moir's.

Together, the complementary papers made headlines around the world, with outside experts saying viruses had to be considered possible causes of Alzheimer's.

Armed with the new data, Moir earlier this year tried one more time for an NIH grant, to study in greater detail how herpes viruses might spur Alzheimer's and, if so, what therapies might work.

If he and other scientists are right that beta-amyloid is an antimicrobial, that the brain goes on an amyloid-making immune rampage in response to pathogens, and that the rampage ignites neuron-killing inflammation, it suggests very different therapeutic approaches than the 30-year pursuit of amyloid destroyers.

One hint of what those approaches should be comes from a 2018 study in Taiwan. It found that people with a herpes virus infection are at 2.5-fold higher risk for dementia than similar people without that infection - and that those treated with anti-herpes drugs were 92 percent less likely to develop dementia than those whose infections were left untreated.

"It used to be thought that stopping the plaques early was 'primary prevention,'" Tanzi said. "I think primary prevention is stopping the microbes." Treatment would mean leaving amyloid mostly alone (since it protects the brain from herpes and other viruses) but targeting inflammation, a biological fire that "kills 10 neurons for every one killed by amyloid and tau directly," he said. "Neuroinflammation is where we're going to find [Alzheimer's] drugs."

Tanzi chairs the scientific advisory board of AZTherapies, a Boston-based biotech that is recruiting patients with early Alzheimer's for a clinical trial testing whether inhaling a powdered form of the immune-targeting drug cromolyn reduces dementia. (The nose connects to the brain.)

One reviewer of Moir's latest NIH grant proposal called it "fascinating and innovative," with "the potential for answering the vital question" of what causes late-onset Alzheimer's. (The early-onset form is genetic.) Moir, the reviewer said, "has consistently produced creative, thought-provoking work."

Another, however, slammed the proposed research, questioning whether it "will be relevant to understand AD pathogenesis" since, in his view, amyloid formation might not matter in Alzheimer's. (It's the disease's defining trait.)

The proposal was turned down in June. "One bad review can sink a new and innovative idea just because it's new and innovative," Moir said matter-of-factly. He can't work up much animosity anymore.

This month, however, he got an unheard-of email from NIH: The agency had found some extra money lying around in its budget. Would he please respond to the reviewers and resubmit his proposal? An over-the-moon Moir did. He expects to hear back in a few weeks.