As you read this sentence, an army of cells patrols your brain. These soldiers slip around neurons, using their gangly appendages to search for threats. If one of them detects a pathogen or injury, it springs into action. Swelling up and descending in a voracious attack, it releases chemicals that signal for its comrades to join the fight.
Known as microglia, these specialised immune cells are our brains’ premier defenders. They protect us from invaders, clear away debris and maintain connections between neurons to ensure the brain remains in peak condition.
Advertisement
Yet, despite their vigilance, microglia can sometimes engage in friendly fire, with a growing body of evidence suggesting they may be the engineers behind some of the brain’s most intractable conditions, such as Alzheimer’s disease and depression. If that is the case, targeting our wayward defenders – or even replacing them with rejuvenated troops – may lead to exciting new therapies.
Microglia were discovered in 1919 by . While experimenting with novel ways of staining brain tissue, he stumbled across these new cells and named them after the ancient Greek words for “small” and “glue”.
This turned out to be an ill-fitting description. Other than their splotchy appearance, microglia have few glue-like qualities. Instead, they are some of the most dynamic cells in the body, roaming the brain with spindly, tentacle-like projections that expand and retract in response to changes in their environment.
How microglia function
Microglia are a form of macrophage, a type of immune cell whose roles include regulating inflammation, controlling infection and gobbling up debris. However, unlike other macrophages, microglia have abilities uniquely adapted to the brain, such as shaping connections between neurons. These junctions, known as synapses, allow the cells to communicate with one another. Infants develop more synapses than they will need later in life, so they are , creating more efficient neural networks – and microglia have a central role in this essential developmental process.
has been at the forefront of this research. In the early 2000s, she was completing a postdoctoral fellowship at Stanford University, California, where she became captivated by the microglia’s relentless poking and prodding in the brain. “No other cell in the brain has that ability to be so dynamic,” she says. “To see that in real time led to all kinds of questions. What are they sensing? What are they doing when they touch a synapse or neuron?”
Her , when she and her colleagues discovered that neurons deposit a protein called C1q on synapses that are later removed. C1q belongs to the complement system, a group of proteins that tag debris, pathogens and other cells for macrophages to remove, kind of like an “eat me” signal. The researchers showed that mice genetically engineered to lack this protein were unable to eliminate synapses, causing nerves in their eyes to develop incorrectly.
at the Broad Institute in Massachusetts used different imaging techniques to further characterise the interactions between microglia and these nerve cells in developing mice. This confirmed that microglia respond to complement proteins by engulfing the synapses they tag.
Pruning appears to occur across many different brain regions. at the European Molecular Biology Laboratory in Italy and his colleagues, for instance, have demonstrated that of adolescent mice, an area that is critical for memory.
When it works well, this synaptic pruning prevents our brains from becoming a tangled mess of redundant connections. Unfortunately, a growing body of research suggests that our microglia can sometimes go rogue – and their malfunctioning may lead to various conditions affecting the brain.
People with that encodes for the complement protein C4, which also triggers synaptic pruning, have an increased risk of developing schizophrenia, for example. Overexcited microglia may also be implicated in depression. Evidence is emerging that microglia in two brain areas critical for regulating emotion – the prefrontal cortex and the anterior cingulate cortex – become active during severe depressive episodes, and activation in the latter region correlates with severity. It may be that these microglia are struggling to regulate inflammation in the brain – a known risk factor for depression. This could explain the mechanism behind certain antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), which trigger microglia to release anti-inflammatory molecules.
The most damning evidence concerns microglia’s potential role in neurodegeneration. Consider Huntington’s disease, a genetic condition in which nerve cells in the brain progressively deteriorate, leading to cognitive, mobility and speech issues. In 2023, Stevens’s lab that were genetically engineered to show symptoms of the illness. They found that complement proteins spur microglia to prune synapses between the cortex and striatum, two brain areas that deteriorate in Huntington’s. By blocking these proteins, the researchers were able to slow synapse loss and prevent cognitive decline in the animals.
We don’t yet have tools to measure complement proteins on synapses in living humans, says Stevens – but you can find signs of them in other tissues. As part of the same study, Stevens and her team analysed spinal fluid from 63 people with Huntington’s disease. On average, those at later stages of the condition had 50 per cent more complement protein in their samples than those with early-stage Huntington’s.
Overreactive microglia may also be implicated in the development of Alzheimer’s. , synapse loss is actually the strongest correlate of cognitive decline in the condition.
A by Stevens and her colleagues showed that C1q accumulates in the brains of mice that have been engineered to develop a rodent equivalent of Alzheimer’s disease, even before the condition’s characteristic plaques form. On average, the animals had three times the amount of C1q in their hippocampus and nine times the amount in their frontal cortex compared with those in a control group. The animals later experienced significant synapse loss in these regions, which are also vulnerable areas in people with Alzheimer’s. Crucially, the team found that an experimental drug inhibiting C1q prevents microglia from excessively engulfing the animals’ synapses.
Our genes may mediate this process. For example, TREM2 is strongly linked to Alzheimer’s disease risk, and that certain versions of this gene may protect against Alzheimer’s by preventing microglia from pruning synapses in response to complement proteins.
Such findings don’t exonerate amyloid plaques. These abnormal proteins tend to accumulate between neurons in regions crucial for memory and other cognitive functions – and microglia show heightened activity around them. “They really aggressively attack those plaques. They swell up. They look different. They proliferate. There are huge numbers of them,” says at the University of California, Irvine.
Microglia’s role in dementia
There is some evidence that microglia can first play a protective role by surrounding the plaques to control their spread. At some point in the condition’s progression, however, they stop being effective. Tellingly, the , which is a well-established risk factor for developing Alzheimer’s disease, appears to impair microglia’s ability to neutralise the threat of these plaques in the brain. For instance, animal studies show APOE4 disrupts communication between microglia and other immune cells in the brain, preventing the cells from clearing plaques.
“You’ve got this situation where perhaps an initial response of the microglia is actually helping the brain, but then, maybe later on, this response drives other parts of [the condition],” says Green. “And so everything becomes a lot more complicated.”
Green and his colleagues’ investigations into this process began in 2014, when he discovered a way of knocking out a brain’s microglia. It turns out that the survival of these cells depends on a receptor called colony stimulating factor one (CSF1). When he and his colleagues treated mice with an experimental drug that inhibits CSF1, nearly all their microglia died. This offered an ideal way to determine what these cells do in the brain. “If you want to know something’s important, you get rid of it and then see how a physiological process continues in its absence,” he says.
In , Green and his colleagues used a CSF1 inhibitor to deplete microglia in mice that had been engineered to develop Alzheimer’s. They found that amyloid plaques failed to form in the animals except in areas where microglia survived. It isn’t clear why that is, but the researchers postulate it could be that the rogue microglia secrete chemicals that spur amyloid proteins circulating in the brain to form plaques.
If microglia are contributing to Alzheimer’s disease and other forms of neurodegeneration, then new treatments might be able to target that wayward behaviour.
Simply removing the aberrant cells – as Green had done in mice – is unlikely to be viable. “Microglia are your protection against pathogens,” he says. “I think that, if we were to get rid of them, it wouldn’t be good over the long term.”
It may, however, be possible to deplete microglia and allow them to grow back in a revitalised form. Green and his team have found that when they stopped treating mice with CSF1 inhibitors, the roughly 1 per cent of microglia that survived . Within three days of removing the drug, about half of the animals’ microglia had returned. By week two, levels were restored to baseline.
“I don’t know of any other cell in the body that can proliferate as quickly and robustly as these remaining microglia,” says Green. “I remember looking down the microscope and it’s like, this can’t be real. It’s one of those things I think you only see once in a career.”
Even more surprising was that the new cells appeared rejuvenated and behaved more like those seen in the brains of young mice. “The idea is that you can basically replace your old, aged, injured microglial tissue with a new one,” says Green.
Researchers at Xiamen University in China and the Mayo Clinic in Florida have . In 2023, they showed that depleting microglia in mice with Alzheimer’s-like symptoms and then allowing them to repopulate reversed cognitive deficits. On average, the animals performed as well on certain memory tests as those without signs of Alzheimer’s, suggesting microglial repopulation could be a viable strategy for treating the condition.
But Green is sceptical, since he suspects the microglia would return to their dysfunctional state over time. “The plaques would just activate the new microglia again, inducing them to take on the same inflammatory state as the old ones,” says Green.
Controlling microglia activation
Nevertheless, it might treat conditions like stroke or brain injury where the initial threat is no longer present. For instance, Green and his colleagues have shown that depleting microglia in mice with damaged neurons in their hippocampus . They performed better on memory tests, had lower levels of inflammatory molecules in their brains and had more synapses between neurons than untreated rodents did.
The ideal treatment would be one that could modulate microglia at different stages in a disease, switching on their inflammatory state when beneficial and turning it off once it becomes detrimental. This may not be as far-fetched as it seems.
While completing a postdoctoral fellowship at Stanford University, had what he calls a silly idea. A mouse in the lab had been genetically engineered to lack microglia. He decided to transplant stem cells – which have the ability to transform into other cell types – from a healthy mouse into this sick one.
The animal, which normally would have died in two weeks, survived for four months. When Bennett looked at its brain, he saw it was filled with stem cells that had become macrophages closely resembling microglia. “To me, this was a little bit of magic,” he says.
The finding raises a tantalising possibility. If a stem cell transplant could replace microglia, we may be able to genetically engineer these cells to fix mutations or to have certain disease-fighting properties, such as secreting substances to protect neurons. “You could programme the cell so that if it senses one thing in the environment, it turns on a gene, but if it senses another thing, it turns off that gene,” says Bennett.
The first challenge in making this dream a reality is finding a way to deplete microglia without killing off the transplanted stem cells. After starting his own lab at the University of Pennsylvania, Bennett and his colleagues to resist drugs that inhibit CSF1. They then transplanted these cells into mice treated with the inhibitors and found that they transformed into macrophages in the animals’ brains, replacing the missing microglia.
They have since repeated this process in mice engineered to develop a rare genetic condition called globoid cell leukodystrophy, which is believed to be driven by aberrant microglia and results in severe neurodegeneration. The results of the study haven’t been published yet, but Bennett says they found the transplant not only reduced neurodegeneration in the mice but doubled the animals’ survival.
Treating Alzheimer’s
He and his team are now using this platform to develop new ways of treating Alzheimer’s disease. Already, they are replacing microglia in mice with stem cells modified to have special receptors for identifying and clearing amyloid plaques. Whether this affects symptoms remains to be seen, says Bennett.
Both Green and Stevens believe microglia replacement holds significant promise for treating various neurological conditions. However, they both also acknowledge that it is still a long way away. “We have a number of obstacles to overcome before it becomes a reality,” says Green.
For example, it isn’t clear if stem cells can fully assume microglia’s functions. Part of the concern is that unlike other macrophages, microglia originate from a structure called the yolk sac, which develops in the uterus during pregnancy. “We want to know what that means and when and how it might matter,” says Bennett.
For any of these treatments to work, we will need a much more detailed portrait of the brain’s guardians. “We need to understand what these microglia are doing in the brain, and then we need to work out how we can modulate them in the right way,” says Green. Only then can we become the commanders of our brain’s defence system so that its army of soldiers never turn against us again.
