For nearly two decades, academic and industry researchers working to find ways to slow the progression of Alzheimer’s disease have focused chiefly on the amyloid-β plaques that accumulate among neurons. Dozens of clinical trials have tested drugs designed to remove or reduce these plaques, but successes have been few. Aducanumab, Biogen’s amyloid-attacking antibody drug (brand name Aduhelm) that was approved earlier this year following a long drought in new treatments for Alzheimer’s disease (AD), has been mired in controversy after scientists raised questions about the drug’s efficacy.
This lack of progress has prompted many research groups to look instead at non-neuronal cells in the brain, and in particular, at immune cells known as microglia. Vital in both developing and mature brains, these cells help shape neurons, control how they communicate, keep an eye out for pathogenic intruders, and mediate neuroinflammation. This last role has emerged as particularly important as researchers uncover evidence that inflammation is linked to many neurological diseases—including AD—as well as to other conditions associated with aging.
Many scientists have been waiting for the pharmaceutical industry to take notice of this link. “We knew all this ten years before, the rest of the world just didn’t pay attention to it,” says Jean Harry, a neurotoxicologist at the National Institute of Environmental Health Sciences in Durham, North Carolina. Key players in driving change have been recent genome-wide association studies (GWAS), which have pointed to AD–associated mutations in genes that are highly expressed in microglia, strengthening the evidence for links between these cells and the disease. “You can’t ignore it anymore,” says Bobbi Fleiss, a microglial neurobiologist at RMIT University in Melbourne, Australia.
Now, despite a handful of concerns about the overall benefits and safety of targeting microglia, there is indeed commercial interest in the approach for AD and other neurodegenerative diseases. Several companies have microglia-targeting therapeutics in preclinical testing or even early clinical trials, and investments are coming in for smaller companies looking to get into the space. Late last year, for example, Vigil Neuro, a startup based in Cambridge, Massachusetts, secured $50 million in Series A funding to develop new microglia-targeting drugs that aim to slow down the progression of AD. Chief Scientific Officer Richard Fisher says that the evidence from GWAS, coupled with ongoing stem-cell and animal studies, have “pointed in the direction that, in neurodegenerative disease pathobiology, microglia are critical.”
Microglia-targeted therapeutics developed for Alzheimer’s disease
Microglia are not your ordinary immune cells. Their activity affects the formation of synapses and memories, and the overall health of neurons. They are important for repairing damaged brain tissue and for clearing dead cells and molecular debris, such as incorrectly folded proteins. Microglia also communicate with other non-neuronal cells through signaling proteins called cytokines to regulate various neural and inflammatory functions.
In a healthy brain, microglia are usually in a homeostatic state, surveilling the environment for disturbances. But they can also become activated by cues such as neuronal death, causing them to go into damage-response mode, increasing their phagocytic activity and sending out pro-inflammatory cytokines. A bit of activation is usually beneficial, but genetic mutations or faulty cellular signaling can make microglial responses go awry—for example, by keeping the cells permanently “on,” leading to or worsening a host of diseases. Recent studies in mice and humans suggest that microglia, through their different activation states and responses, are partly culpable in Parkinson’s disease, multiple sclerosis, stroke, and even Down syndrome.
In a healthy brain, microglia are usually in a homeostatic state, surveilling the environment for disturbances.
The evidence is particularly strong for microglial involvement in AD. Several mouse studies have found that microglia located near amyloid-β plaques look and behave differently from those located away from the plaques: the cells nearby enter what researchers call a disease-associated state, and are smaller and express more inflammatory and phagocytic genes. These microglia become a major source of neuroinflammation, which, if left unchecked, can worsen AD. Mutations in genes expressed in microglia can also contribute to Alzheimer’s risk. For example, certain receptors present on microglia recognize amyloid-β and another AD–associated protein called tau, prompting the cells to clear these deposits and prevent their diffusion to nearby neurons. People who carry mutant versions of these receptors are at a much higher risk of AD.
Companies are tackling this biology in a couple of ways: “either push cells back towards homeostasis or reduce neuroinflammation,” says Tarek Samad, senior vice president and head of research at Lundbeck, a Denmark-headquartered pharma company that recently launched a research cluster focused on therapeutic targets for modulating neuroinflammation. Vigil Neuro, for example, wants to push microglial cells in AD patients toward a disease-associated state, which some research suggests could increase the clearance rate of amyloid-β deposits and eventually reduce neuroinflammation. The company plans to use an antibody to activate an innate immune receptor called TREM2, which sits on the surface of disease-associated microglia and encourages surrounding microglia to convert to a disease-associated state. Research over the past few years has “just made a very compelling story that TREM2, and innate immunity in general, are a big part of the story for all neurodegenerative diseases,” says Fisher. Vigil’s antibody, VGL101, is currently in the preclinical phase, with initial tests in monkeys showing promising results, he adds.
More-established companies such as Alector and Denali Therapeutics, both based in San Francisco, are also developing TREM2 agonists; Alector has partnered with pharma giant AbbVie to test their TREM2-targeting molecule in Phase 2 clinical trials, while Denali and Takeda Pharmaceutical Company are testing their own drug in early human studies. TREM2 is also a key focus of Muna Therapeutics, a new company based in Denmark and Belgium that recently secured $73 million in Series A funding to develop small-molecule drugs for neurodegenerative diseases in partnership with Axxam, a research organization based in Milan, Italy. “The time is right to pursue these therapeutics,” says Rita Balice-Gordon, Muna’s CEO.
Other companies are trialing drugs for additional targets (see table below). For example, aside from overactivation, high levels of microglial proliferation also seem to worsen AD. This process is regulated by the receptor CSF1R, and studies have shown that knocking out the gene that codes for the receptor in mice stops microglia from multiplying. Inhibiting CSF1R pharmacologically seems to have a similar effect: Janssen Pharmaceutical’s small-molecule inhibitor reduces microglia proliferation and alleviates neurodegeneration in mouse disease models, Simon Lovestone, Janssen’s vice president and disease area stronghold lead for neurodegeneration, tells The Scientist in an email. The company is currently running a Phase 1 trial that partly aims to test how well the molecule blocks CSF1R in vivo.
Therapeutic candidates for other diseases
While AD has attracted the most attention, other neurodegenerative diseases are on the radar as well. Frontotemporal dementia (FTD), a condition marked by shrinkage of the frontal and temporal lobes of the brain, is often misdiagnosed as AD, as patients have similar cognitive issues.
One of the key genetic drivers of FTD is a mutation in the granulin (GRN) gene, which normally encodes a protein called progranulin. In healthy individuals, progranulin regulates the inflammatory activity of microglia; in people lacking functional progranulin, the cells produce excessive amounts of cytokines and accumulate cellular waste. To offset the effects of defective or insufficient progranulin, Alector is currently conducting Phase 3 trials of a monoclonal antibody that elevates progranulin levels, and recently announced a two-year partnership with GlaxoSmithKline to refine the antibodies further. Denali Therapeutics, with Takeda, also has a recombinant progranulin protein in development that aims to make up for the deficit in FTD patients’ brains.
Parkinson’s disease (PD) has been another major focus of researchers developing microglia-targeting therapeutics. PD features protein deposits in the brain—usually of α-synuclein—that can lead to its characteristic tremors and cognitive problems. There are several genetic risk factors for PD, one of which involves mutations in LRRK2, a gene coding for a protein kinase. Overactive mutants of this enzyme can increase α-synuclein production, which can activate microglia and trigger neuroinflammation. To combat this, Denali, in partnership with Biogen, has developed an LRRK2 inhibitor, which is currently in a Phase 1 clinical trial. Biogen is also testing an antisense oligonucleotide drug, developed in collaboration with Ionis Pharmaceuticals, against LRRK2 mRNA.
Other therapeutic directions may be on the horizon. Microglia seem to be involved in diseases such as amyotrophic lateral sclerosis and multiple sclerosis through a signaling pathway known as TREM2-APOE. In microglia that are in a disease-associated state, this pathway promotes phagocytosis, worsening neurodegeneration. A recent study in mice showed that knocking out TREM2 could tone down the signal and return the microglia to a harmless homeostatic state.
Several companies are developing drug candidates to manipulate the activity of microglia, both in Alzheimer’s disease and in other brain disorders. A selection of experimental therapeutics undergoing clinical or late-stage preclinical research is below.
Stage of development
TREM2: Membrane receptor that activates microglia in disease
Alector, with AbbVie
Denali, with Takeda
CSF1R: Membrane receptor that regulates microglial proliferation
Janssen, with the University of Oxford
SIGLEC-3/CD33: Genetic risk factor and membrane receptor that regulates microglial activation
Alector, with AbbVie
MS4A4A: Genetic risk factor and membrane protein that regulates TREM2 activity
RIPK1: Membrane and cytosolic protein that regulates inflammation and cell death
Denali, with Sanofi
IL-1β: Secreted extracellular and cytosolic protein that regulates inflammation
Progranulin: Secreted extracellular protein that regulates inflammation
Denali, with Takeda
LRRK2: Genetic risk factor and membrane protein that regulates cell signaling
Denali, with Biogen
Paused after Phase 1
Biogen, with Ionis
Obstacles to targeting microglia in neurodegenerative disease
The path to commercial success for many of these therapeutics contains a number of hurdles. For one thing, microglia are very closely related to macrophages, immune cells found everywhere in the body. This means that several of the genes expressed by microglia, such as TREM2, can also be expressed by macrophages. Diego Gomez-Nicola, a neuro-immunologist at the University of Southampton in the UK, says he is particularly worried about how manipulating these genes or the proteins they produce will affect the systemic immune response. As a case in point, researchers reported a couple of years ago that mice treated with a CSF1R blocker were more susceptible to West Nile virus. “Industry is usually risk averse, and if you put a risk in front of them, they will try not to look at that risk until they have to,” says Gomez-Nicola.
Fisher and Balice-Gordon contend that any drastic effects on the immune system will show up in the Phase 1 safety and efficacy studies. “To date, there is no compelling data that agonizing TREM2 in the periphery in Alzheimer’s disease leads to detrimental consequences on macrophage function,” says Balice-Gordon. Fisher adds that in Vigil’s preclinical trials of its TREM2 agonist on monkeys, the team has seen no peripheral activation of macrophages, but there is a boost in biomarkers of microglial activation.
Different regions of the brain show different types and levels of microglial activity.
The dynamic nature of microglia also complicates the picture: at any given point in time, different regions of the brain show different types and levels of microglial activity—especially in disease, when the brain has both activated and resting microglia (see “Hidden Variation” below). But most therapeutic agents will indiscriminately reach all microglial cells, not just the ones responsible for inflammation at a particular moment. To allow microglia to carry out their normal, beneficial functions in the brain with minimal interruption, any microglia-targeting therapeutic “should be something that you could quickly get cleared out of the system,” says Harry.
She also expresses skepticism about therapies that target single genes and cell types, pointing out that in the case of microglia, drug developers often ignore the fact that these cells work in coordination with other cells such as astrocytes. Lundbeck’s Samad counters that, in cases where both microglia and other cells are implicated in disease, companies could turn to combinatorial therapies, as the immuno-oncology field has been doing for a while.
On a larger scale, the roles of microglia in disease are viewed through a blurry lens, says Richard Ransohoff, a partner at Third Rock Ventures with three decades of experience as a practicing neurologist. He identifies certain “conceptual problems” with microglial research. “[One] is the thought error of believing that microglia are primarily about host defense and injury repair and clearance of debris, and that it’s loss of those functions that creates disease. It’s simply not the case,” as they also regulate neuronal function. This second aspect of microglia biology is often overlooked in disease research, he adds.
While Ransohoff acknowledges that these cells are compelling therapeutic targets, he says it’s possible that the current animal disease models, on which most drug development studies rely, are “worthless.” He points out that a lot of the cognitive and behavioral tests used in mice do not have true correlates in humans. The transcriptional profiles of microglia also differ between mice and humans, possibly limiting the utility of these models for drug development.
Aside from these concerns about the basic biology of microglia, Gomez-Nicola says he’s curious as to whether the recent FDA approval of aducanumab will spur a shift from targeting microglia back to targeting amyloid-β accumulation. Samad says he remains optimistic that interest in microglia and neuroinflammation will persist, noting that there isn’t yet any drug—aducanumab included—that can significantly slow disease progression in AD patients. “So that leaves ample room for innovation in therapies,” he says. “If anything, folks and stakeholders will double down to really continue the innovation journey.”
Although microglia were first identified in 1919, scientists only recently started describing different subpopulations of these cells. Since the early 2000s, microglia, because of their close relation to macrophages, have been classified as either M1 (pro-inflammatory) or M2 (anti-inflammatory). Over the last few years, though, studies using single-cell transcriptomics have revealed a baffling array of microglial subtypes in the brain. Cells seem to be so specialized that even neighboring microglia sometimes react differently to the same stimulus. The M1/M2 paradigm breaks down in the face of such variation (which peaks during early development but persists to a considerable degree into adulthood), and many researchers are now calling for a better system to classify microglial states.
Differences between microglia are amplified in disease. For example, a cell activated in an Alzheimer’s disease patient might engulf only amyloid-β plaques, while one activated in a stroke patient can eat up entire neurons. Such differences can arise two kinds of variation: inherent differences among the microglia, which usually result from variation in gene expression patterns or epigenetic modifications, and differences in activation state during disease, which depend on specific pathologies and the location of the microglia. For example, in some diseases, only cells near the site of damage turn on phagocytosis, while those away from the site remain inactive.
Diego Gomez-Nicola, a neuroimmunologist at the University of Southampton in the UK, says that acknowledging these variations in the microglial responses is vital for moving forward with therapeutics, “because all of this information mostly comes from rodent studies. And we don’t yet . . . have that degree of granularity in humans.” In particular, microglial variation makes it tricky for companies to target only certain subtypes of the cells, notes Rita Balice-Gordon, CEO of Muna Therapeutics, which recently secured funding to develop small molecules to target microglia for neuro-degenerative diseases. Single-cell and spatial transcriptomics will allow researchers to narrow down on the genetic signatures of microglia that are involved in disease. Nevertheless, Balice-Gordon says, “I think understanding the range of microglial behaviors, those that support good inflammation, those that support bad inflammation, and how to modulate that . . . still remains a challenge.”