In the ensuing Q&A, another member of the class piped up to ask a question about APP, Rice recalls. They asked, “Well, what does this protein do?” Rice was stumped—and not because she hadn’t prepped. Considering the question, she realized that “there really wasn’t a sufficient answer” as to what APP does beyond getting broken down. It triggered quite the classroom discussion, she says. “No one could believe that the field didn’t have a good idea of this.”
Rice, now head of her own lab at the University of Oklahoma Health Sciences Center, isn’t the only researcher to have noted this lack of attention. While APP has for decades been recognized as the source of Alzheimer’s-associated plaques, far less is known about its normal physiological function in the brain, let alone how that function might itself contribute to disease. Scientists do know that the protein is encoded by the highly conserved APP gene, which is expressed in multiple tissues starting in early embryonic development—not just in old age when Alzheimer’s disease typically arises. And a wave of genetic knockout studies in mice and fruit flies in the 1990s and early 2000s revealed possible roles for the protein in stem cell differentiation and cellular signaling. But the details remained murky, and a lack of Alzheimer’s-like symptoms in the knockout animals led many people to assume that these functions were irrelevant to neurodegeneration, says Bassem Hassan, a neuroscientist who studies APP at the Paris Brain Institute. Instead, pathologists and pharmaceutical companies have been far more focused on amyloid-β plaques, which seem to clog up the brain and cause the physiological and cognitive damage that makes the condition so devastating.
“The overriding interest in the community is to understand how Alzheimer’s disease arises and how it can be combatted,” Hassan says. This led to a mentality, he continues, that “if the physiological function of APP is irrelevant to [disease], then why bother [studying it]?”
It’s crazy to think about [these peptides] in disease and not even know what they’re normally doing.
—Heather Rice, University of Oklahoma Health Sciences Center
Hassan is one of a number of researchers working to shift that thinking—and not just for APP and Alzheimer’s. He, Rice, and others are digging into the biology of various proteins that have been relatively understudied outside the context of the neurodegenerative diseases they’re associated with. “Studying the normal function of the protein [makes] people feel uncomfortable,” says Robert Edwards, a neuroscientist at the University of California, San Francisco (UCSF) who works on α-synuclein, which misfolds and forms aggregations in the brains of people with Parkinson’s disease, among other conditions. “They say, ‘Well, what does that have to do with the disease?’ But my feeling is that whatever you learn about the normal function [provides] a foundation on which to build.”
Not only are such studies pointing to roles for some of these proteins in neural communication and brain development, they’re highlighting how these overlooked functions may help researchers understand neurodegenerative disease after all—long before classical signs such as protein aggregations emerge. By focusing on these functions, “the clinically relevant problems become enlightened,” says Hassan. While this attitude has yet to win over everyone, Hassan hopes that it may one day lead to a better idea of how to prevent neurodegeneration “or at least slow down the progression,” he says, “because we understand the fundamentals of what’s going on.”
Building the backstory for proteins involved in neurodegeneration
Sandrine Humbert has been studying the mechanisms underlying Huntington’s disease for more than 20 years. Unlike Alzheimer’s and Parkinson’s—which have identified genetic risk factors in only a portion of cases and are thought to be significantly influenced by environmental factors—Huntington’s is always explained by a mutation in the huntingtin gene. But it’s like Alzheimer’s and Parkinson’s in that the disease is partly characterized by protein aggregations. In people who have the condition, which typically starts causing motor and cognitive symptoms in middle age, mutant huntingtin protein (HTT) clumps inside neurons and invades their nuclei—processes that are linked to brain damage.
Humbert, who heads up a lab at the Grenoble Institute of Neurosciences in France, tells The Scientist that for a long time after the disease-causing mutation in huntingtin was identified in 1993, scientists focused intently on understanding and preventing the toxic effects of mutant HTT in the adult brain. “People figured . . . it would be kind of easy,” what with the causative mutation being known, she says. While some groups got to work characterizing how deformed versions of HTT wreak havoc on neurons, others developed experimental therapeutics designed to lower the expression of the gene or production of the protein; several of those therapeutics are now in clinical trials.
Yet Humbert argues that there’s more to HTT than meets the eye. For example, while a couple of labs have reported that reducing levels of the protein in adult rodent brains has little to no harmful effect, studies by her group and others have suggested that knocking out the regular version of the gene in adult animals causes neuronal dysfunction in some brain regions. Such studies have helped spark a conversation about whether HTT might do something rather important in a healthy brain, Humbert says. “People are now looking more and more at the function of the normal protein.”
In a review of this research published a few years ago, Humbert and Grenoble colleague Frédéric Saudou summarized possible jobs for normal HTT based on dozens of studies carried out primarily in vitro and in lab animals. Purported roles include the trafficking of vesicles and endosomes within neurons, widespread transcriptional regulation, and the control of programmed cell death. (See illustration.) Mutations in huntingtin, the authors noted in their paper, could not only endow HTT with new, cell-destroying characteristics but also compromise the protein’s ability to perform its everyday roles.
Deep dives into other proteins involved in neurodegenerative disease are revealing that many have similarly complex backstories, separate from the aggregation traditionally associated with pathology. For example, independent work by several research groups has revealed a connection between APP and the Wnt family of proteins, which has long been recognized as central to multiple cell-signaling and developmental pathways. Hassan, in collaboration with Rice and other colleagues, recently published experiments in fruit flies and in cultured mouse neurons suggesting that APP is itself a Wnt receptor. Speaking to The Scientist shortly after those results were published, Hassan described APP as a “calibrator” for Wnt signaling, computing ratios of different Wnt proteins to drive decisions about whether a neuron should grow to be long and spindly or short and bushy.
Rice’s own lab reported a couple of years ago in Science that part of APP also binds to a receptor for the inhibitory neurotransmitter GABA on presynaptic neurons, and seems to dampen those cells’ communication with other nerve cells in vitro and in rodents. There have been several reports of APP interacting with other neurotransmitter receptors, she adds. In relation to Parkinson’s disease, meanwhile, Edwards’s lab at UCSF has described in multiple papers over the last decade how α-synuclein could help regulate the release of neurotransmitters and other cargo in cultured neurons and in animal brains.
Together, these and other studies are providing a new level of functional insight into proteins typically studied in the context of rampant brain destruction—revelations that are important for a number of reasons beyond illuminating human biology, say researchers who spoke with The Scientist. For example, determining the normal function of these proteins could point to possible side effects of therapeutics that lower the production of disease-linked peptides, says Edwards. Additionally, it could reveal new approaches for targeting disease-associated proteins, says Rice, who is studying whether it’s possible to exploit some of the other products of APP’s breakdown to counteract the toxic effects of amyloid-β and who is a coinventor on a patent related to APP’s interaction with GABA.
However, several scientists say that they’re most excited by another research idea, one that places normal protein function at the center of how neurodegenerative disease arises. While emphasizing that individual diseases such as Alzheimer’s and Parkinson’s encompass a diverse range of conditions with different symptoms and outcomes, these researchers argue that disruption to disease-associated proteins’ normal roles could itself be a contributor to neurodegeneration. Hassan, for one, says that he fully expects this to be the case, and argues that it offers a much more plausible explanation for disease pathology than the famed aggregation of “toxic” proteins.
Linking normal protein function to neurodegeneration
Despite the persistence of the idea in the pharmaceutical industry and in popular imagination that neurodegeneration is caused by clumps of malformed or misfolded peptides, the hypothesis has long had its critics in the research community. Part of scientists’ skepticism stems from the continued failure of protein-reducing therapies to prevent or slow disease in clinical trials following apparent successes in animal models designed to recapitulate some elements of the pathology. There are also brain imaging studies from the past couple of decades that have frayed the connection between aggregation and disease symptoms. In Alzheimer’s, for example, human studies suggest it’s possible to have amyloid deposits without obvious health problems.
Another challenge to the toxic clump model comes from the study of genetic risk factors for neurodegenerative disease. For example, several rare genetic mutations in SNCA—the gene coding for α-synuclein—are associated with elevated Parkinson’s risk, but the relationship isn’t straightforward. At least in vitro, “mutations in α-synuclein that cause Parkinson’s disease do not clearly correlate with their propensity for aggregation,” Edwards explains. “There are some that reduce their propensity for aggregation, others increase it. So there’s something we’re missing.”
In light of these and other observations, many researchers working on neurodegenerative disease have come to see protein aggregation as a byproduct, rather than the primary cause, of neurological damage, says David Sulzer, a neuroscientist and Parkinson’s expert at Columbia University who has collaborated with Edwards. Still, settling on an alternative explanation for how neurodegenerative diseases arise isn’t easy. Several current hypotheses revolve around inflammation caused by viral infection or defects in cellular waste disposal. “My hypothesis is that there’s a problem in the normal α-synuclein degradation, which leads to the [aggregation] but it’s also leading to cellular damage that eventually ends up causing Parkinson’s,” says Sulzer. He adds that similar waste-disposal hypotheses have been proposed to explain some of the pathology of Alzheimer’s.
Other researchers see a different possibility: that problems in neurodegenerative disease start with disruption to the normal functions of the aggregating proteins themselves.
Investigating this idea in Parkinson’s, Edwards and his colleagues have conducted animal studies that replicate some of the mutations associated with elevated Parkinson’s risk to see how they affect α-synuclein’s activity. One such experiment built on the team’s finding that α-synuclein affects how vesicles fuse with the cell membrane, speeding up the release of their cargo. Using mice and cultured rat neurons, the researchers discovered that SNCA mutations associated with Parkinson’s disease seem to specifically disrupt this function. This means that neurons with these mutations might have lower-than-normal neurotransmitter release, something that could ultimately contribute to neuronal death, Edwards speculates. Of course, only a fraction of people who get Parkinson’s have mutated copies of SNCA. But the team’s work hints that genetic or environmental disruptions to a pathway involving α-synuclein may contribute to disease pathology.
Studying the normal function of the protein makes people feel uncomfortable.
—Rob Edwards, University of California, San Francisco
Hassan has been exploring similar ideas as they relate to APP and Alzheimer’s. For example, he and others have noted that certain rare mutations in APP that are associated with some early-onset forms of the disease may alter APP’s regular activity. Using fruit flies, which have a homologous version of APP called APPL, Hassan’s lab showed that deleting the underlying gene or blocking production of the protein was associated with increased neuronal death, particularly in young flies, as well as problems in intracellular trafficking and other important cellular processes. While losing APP entirely via genetic knockout isn’t the same as having a mutated version of the protein, the findings point to potential neurological consequences of reduced or altered APP function, Hassan argues.
Some groups have also focused on these proteins’ role in mammalian brain development. Humbert, for example, says she’s interested in whether changes in HTT’s activity might set the brain on a developmental path that predisposes it to damage, even if clinical symptoms of Huntington’s don’t show up until a person’s 30s or 40s. She points to evidence from other labs suggesting that animals engineered to express lower-than-usual levels of HTT protein as embryos but normal levels as adults show abnormal brain development as well as later-life neuronal degeneration.
In the last couple of years, her own group has reported that mouse embryos expressing mutated versions of huntingtin show disruptions in cell growth and intracellular dynamics in neurons, leading to differences in neuronal proliferation and, ultimately, a difference in cortex structure between mutant mice and controls. A recent study the lab carried out using small sections of neural tissue from human fetuses hinted at a similar pattern in people: compared with controls, carriers of mutant HTT had a differently developed cortex at the end of the first trimester. The findings may help explain some brain imaging studies that suggest people with Huntington’s have altered cortical structures, as well as behavioral changes, even before the classical cognitive and motor symptoms of the disease emerge, Humbert speculates. “We are showing that abnormal development may contribute to preclinical symptoms.”
Humbert notes that she and Hassan have recently started discussing parallels in their research, despite studying different neurodegenerative diseases. Hassan’s group is now working on understanding APP’s role in brain development and how that role may vary across the animal kingdom—early observations with cultured human cells and rodents, for example, hint that the protein may be more important for neurogenesis in people than it is in mice, Hassan says. Such research is contributing to the existing, although still peripheral, idea that neurodegenerative diseases may have a strong neurodevelopmental component, Hassan says, especially when it comes to inherited versions with a known genetic cause. According to that view, at least in some cases, “damage begins from the beginning,” he adds. It just “manifests itself at different times in your lifetime.”
Continued debate about what really causes neurodegeneration
For now, the evidence for a causal connection between disruption in normal protein function and the onset of neurodegenerative disease is still thin, and researchers working in the field are the first to acknowledge the challenges of closing the gap. Animal studies are only so informative about the human brain—particularly when mice and other lab models don’t develop anything like human neurodegenerative disease. Moreover, in vitro studies of cultured human neurons may not capture what really happens during in utero development, several researchers note. And even if the brain does develop or function differently in people with genetic forms of neurodegenerative disease, these changes may be largely or entirely compensated for by other neural pathways, perhaps reducing their relevance to the eventual onset of symptoms later in life.
A number of scientists also acknowledge that there’s likely more than one route to neurodegeneration. Speaking about Alzheimer’s, for example, Rice says that while she thinks changes in normal protein function might be important, she’s also of the mind that the aggregation of amyloid-β likely plays a substantial role in brain damage. And researchers studying the myriad other genes associated with risk of neurodegenerative diseases, such as PINK1 in Parkinson’s and ApoE in Alzheimer’s, highlight that there may be multiple overlapping mechanisms that lead to the same sort of pathology. “Everybody wonders, ‘When do these paths all converge? Is there a common pathway?’” says Icahn School of Medicine at Mount Sinai’s Deanna Benson, who has found that mutations in LRRK2, a protein kinase that interacts with α-synuclein and is linked to Parkinson’s risk, seem to alter neuronal connections in the developing brains of baby mice.
Despite the outstanding questions, researchers tell The Scientist that they are pleased to see a greater focus on normal protein function starting to percolate through the neuroscience community, with several saying that their ideas are now getting a warmer reception than they did several years ago. Rice, who has spent more than a decade digging into APP, says she hopes the momentum will keep building. In fact, she maintains much the same opinion now about research on APP and proteins like it as she did when she gave her talk back in undergrad: “It’s crazy to think about [these peptides] in disease and not even know what they’re normally doing.”