A mouse model of Alzheimer’s disease shows the differences in the thickness of cortex in mice that have (at right) and have not (at left) received the MIT researchers’ noninvasive light flickering treatment.
PHOTO ILLUSTRATIONS: COURTESY OF THE RESEARCHERS
She’d induced brain waves at a rate of 40 oscillations per second in mouse models of early stage Alzheimer’s disease. According to her analysis, the intervention had cut by half a toxic protein associated with the disease.
The reduction was just too dramatic to be true. So she repeated the experiment. Others did, too. The results remained the same. “We’d thought it was a fluke,” says graduate student Anthony Martorell, who worked with co-first author Iaccarino on the study in the lab of MIT neuroscientist Li-Huei Tsai, director of the Picower Institute for Learning and Memory and the Picower Professor of Neuroscience. “It’s been very hard to reduce any Alzheimer’s disease pathology at all in mouse models or in humans, so our results were really surprising.”
As has now been widely reported, the experiments that followed built a body of evidence pointing to the potential therapeutic benefits of shining flickering light into the eyes of patients with Alzheimer’s disease. Advancing silently over decades, ultimately robbing people of memory and identity, the disease and related forms of dementia affect more than 46 million people worldwide, according to the 2015 World Alzheimer Report, and the worldwide cost is expected to rise to $2 trillion annually by 2030. Currently, there are no effective treatments for Alzheimer’s disease.
Because of the urgent need for new ideas addressing Alzheimer’s and dementia, MIT formed the Aging Brain Initiative, with an emphasis on interdisciplinary approaches. Tsai credits the initiative for bringing together senior investigators with complementary talents to dream up creative ways to advance her lab’s research. “I would not have thought to do a lot of what we did without the input of our colleagues. Their thinking was completely out of the box and they brought fresh perspectives,” says Tsai. “We work together because we care about brain aging, but we think differently.”
A gamma gambit
A lot is known about the molecules involved in Alzheimer’s disease and the genes that increase its risk, but less is known about how it affects the brain as a system. Prior to this study, it was known that in later stages of the disease, when sticky amyloid plaques have already built up in the brain, the brain’s gamma waves lose their strength. Gamma waves ripple across the brain about 40 times per second and appear when the brain is doing attentive work, such as forming memories or solving problems.
Tsai’s team wanted to know if gamma waves are also diminished in early stage disease. They chose to study a mouse model with five human genes associated with a risk of Alzheimer’s. The young mice have elevated levels of the toxic amyloid protein, but no plaques. The team also partnered with a fellow member of the Aging Brain Initiative, Ed Boyden ’99, MNG ’99, professor of biological engineering and brain and cognitive sciences at the MIT Media Lab and the McGovern Institute for Brain Research. Boyden develops technology to record brain signals.
Recordings of neural signals in the mice performed by co-first author Annabelle Singer showed weak gamma signals compared to mice without the disease. “The first thing we asked when we saw that was, what happens if you bring gamma back?” says Tsai.
Tsai’s lab had already induced gamma waves in mouse brains using optogenetics back in 2009, so the researchers knew how to do it. Using the same mouse model of early stage Alzheimer’s, they engineered neurons in the hippocampus, the seat of memory in the brain, to respond to laser light. Then they stimulated these cells with light passed through fibers implanted in the brain, testing different rates of flickering. When the light flickered at 40 flashes a second, the stimulated cells responded and induced gamma waves in the brain.
After repeated experiments, the team found that those oscillations reduced toxic amyloid levels in the brain by 50%. Not satisfied, they wanted to understand how gamma was driving these dramatic reductions. “We can be surprised by the phenomenon, but mechanistic insight makes the data much more valuable,” says Tsai.
They confirmed that gamma reduces the production of toxic amyloid, which is governed by one specific process. But to understand changes in the clearance of amyloid, they cast a wide net. For instance, they used RNA sequencing to look at expression profiles of all of the genes in all of the cells in the hippocampus to see how gene expression differs in brains with weak gamma from those with induced strong gamma oscillations.
Such an approach is sometimes called discovery science. It is unbiased and does not rely on a preformed hypothesis. Rather, the idea is to collect vast amounts of information and look for patterns that form new hypotheses. “This is our lab culture,” says Tsai. “We want to look at the big picture because we don’t want to miss anything.”
What they found was that gamma oscillations change the behavior of immune cells called microglia, which are responsible for clearing proteins such as amyloid. These cells increase in number with increased gamma strength and become bigger and more active. “This experiment really paid off,” says Tsai.
The eyes are the window
Shortly after the team had learned that induced gamma oscillations reduce amyloid, they discussed the results with Emery Brown, Edward Hood Tapin Professor of Computational Neuroscience at MIT, who is a member of the Picower Institute and the Aging Brain Initiative. Brown, who is also a physician, was impressed. But he encouraged the team to find a way to induce gamma that wasn’t as invasive as optogenetics, which requires that optical fibers be implanted in the brain.
He even suggested a method: Shine flickering light into the eyes instead of piping it deep into the brain. “Up to that point it never occurred to me to go this one step further to try out noninvasive stimulation,” says Tsai. “We decided to give it a shot.”
Brown’s idea had precedent. Neuroscience experiments done years ago—and considered classics in the field today—had shown that the visual cortex adopts the activity patterns of light signals entering through the eyes. “This has been known for decades,” says Tsai lab postdoc Chinnakkaruppan Adaikkan, who joined the project midway. “But to use it to amplify aberrant oscillations in the Alzheimer’s brain was an interesting idea.”
Boyden’s team used their engineering expertise to fashion a control- lable flickering LED. Tsai’s team studied the effects of the strobe on the visual cortex. “At this point, the project had become very multidisciplinary,” says Tsai.
The intervention not only halved amyloid levels in mice with early stage Alzheimer’s—it also reduced plaques that form in later stages. This finding, which was published along with the team’s other results in Nature in December 2016, makes the intervention potentially relevant for humans. Alzheimer’s symptoms typically do not appear until after plaques have formed. “Most human patients will have plaques in their brains already,” says Tsai.
Looking forward, the researchers have many avenues they’d like to explore. First, though, they’d like to determine how long the effects of the intervention last and whether other modes of sensory stimulation, such as sound or touch, have similar effects on the regions of the brain that process those inputs. Ultimately, the goal is to find multiple ways to noninvasively stimulate the brain so that the induced gamma waves propagate strongly throughout it. “If we can activate gamma in many different brain regions, perhaps we can get a huge area of the brain involved,” says Tsai. “Treating the whole brain will be important for people with Alzheimer’s disease.”