From Futura
If a microscopic film crew could make a movie of the cellular landscape of your brain, aside from the blood coursing through the brain’s 650 km of blood vessels, most cells would be largely motionless except microglia. They would be the hyperkinetic dancing stars of this production.
There are three types of glia cells that together make up half of the brain’s cells. Their smallest subtype, microglia owe their name to their tiny cell bodies, but from these diminutive stationary centres they extend an array of elaborate branches that are constantly in motion. “It’s amazing. Every second, you see something different,” says Professor Rosa Paolicelli, a microglia researcher at the University of Lausanne, France.
In a healthy brain, a microglia’s processes constantly bend, probe, extend, and retract. Occasionally, they touch a synapse and attach for five or so minutes, before uncoupling and moving on. Other times, a branch touches a neuron’s cell body, spending, on average, 25 minutes attached.
And this is only one aspect of the dynamic nature of microglia. When a brain is injured, microglia will retract their processes to assume an amoeboid form and migrate to the site of damage. They will also proliferate under certain conditions due to injury or disease.
But then microglia come from a nomadic tribe of cell types. They are immune cells born in the periphery that flood into the central nervous system very early in development to take up lifelong residency there. In humans, they enter already in the second month of gestation – before true neurons even exist.
Later in development, a barricade forms around the central nervous system. The blood-brain barrier is created by a tight coupling between blood vessel cells and prevents bacteria and viruses – as well as circulating immune cells – from entering the brain.
This means that microglia are the brain’s resident immune system. For much of the 101 years that scientists have known microglia, they were viewed as cells that become active only in a crisis – be it upon infection or in response to injury or degenerative illness.
Their amoeboid form was seen as their active state. When assuming a branched form, they were thought to be “resting”. Today, Paolicelli tells her students that “to call microglia resting is an offence – they are anything but resting.”
Equipped with an array of new techniques, microglia research has been expanding rapidly in the 21st century. The aims are twofold: one, to work out exactly what these ceaselessly moving cells do in the healthy brain, and two, to resolve how they contribute to disease. With progress being made on both fronts – but with many mysteries remaining – “the field”, says Paolicelli, “is exploding.”
1919: GLIA’S BIG BANG
Microglia were conclusively described for the first time in 1919 by Spanish scientist Pio del Rio Hortega. However, if we look at Alois Alzheimer’s drawings today, it is clear that, around a decade earlier, Alzheimer had unwittingly seen microglia when describing the pathology of the disease that now bears his name.
In those days, uncovering the brain’s microscopic structure depended critically on developing stains that labelled only select populations of cells. Hortega’s mentor Santiago Ramon y Cajal had won the Nobel Prize in 1906 for showing that brains are composed of networks of discrete neurons connected to one another by synapses.
He could only do so because the co-recipient of that prize, Camillo Golgi, had developed a method for labelling neurons. Cajal went on to characterize astrocytes – a previously known type of glia named for its star-like morphology – in great detail. But he could never satisfactorily label other non-neuronal cell types and referred to what was only glimpsed as the brain’s “third element”.
His mentee Hortega, though, found two new staining techniques and showed that the third element was not one element but two: the cells now called oligodendrocytes and microglia.
In four papers written in Spanish, Hortega in 1919 described the morphology of microglia in healthy, degenerating, and injured brain tissue, showing that microglia vary in different brain regions. He argued that the cells change form, from branched to amoeboid, and become migratory in pathological scenarios. He also described how they consume cellular debris by the process of phagocytosis, and inferred that they are not brain cells per se, but immune cells. Finally, he speculated that as neurons assemble into circuits during development, some neurons die and some neurons’ processes are removed, and that microglia might clear the brain of these remains. Only recently, Professor Helmut Kettenmann of the Max Delbruck Center for Molecular Medicine, Berlin, Germany, helped translate Hortega’s four papers into English, annotating them to highlight their contribution to many issues in microglia research. The translations were published in 2016 under the title “The Big Bang for Modern Glial Biology”.
Kettenmann says that Hortega’s correct assertion that microglia were related to phagocytic cells of the immune system “was kind of intuition”. Debates raged about where exactly the cells came from until their origins were convincingly pinpointed as primitive macrophages in 2010 by Florent Ginhoux, then at Mount Sinai School of Medicine, USA, and colleagues.
For many decades after Hortega’s discoveries, most microglia research – of which there was not much – focused on these cells being activated only by pathological events. Of note, says Kettenmann, is that in the 1960s and 1970s Georg Kreutzberg, then working at the Max Planck Institute of Psychiatry, Munich, Germany, showed that if the facial nerve was injured, microglia digested the synapses associated with this nerve, introducing the concept of microglia as “synaptic strippers”.
In the 1980s and 1990s, researchers found ways to grow microglia in cell culture and started to better understand these enigmatic cells, showing, for example, that they release and respond to many chemical messengers associated with the immune system and inflammation, such as cytokines.
Then, after the millennium, microglia research truly gathered pace and also turned back to the intact brain. This was aided by genetically modified mice. Microglia are the only cells in the brain that make the receptor for the cytokine fraktaline. In a modernday equivalent of Hortega’s staining techniques, researchers elegantly visualized microglia in the brain of living mice by replacing their fraktaline receptor gene with the gene for the green fluorescent protein (GFP).
These mice allowed scientists to make the first movies of microglia in the living brain. Using mice, two independent groups observed how incredibly dynamic microglia are. “Absolutely nobody had suspected that,” says Kettenmann. The techniques allowed researchers to watch in real-time how microglia respond to disturbances – they reach out their branches towards trouble. However, it begged the question of what microglia were doing when they were supposedly resting?
A LIFETIME’S WORK
Paolicelli entered the field as a PhD student in 2007, when a side project suddenly took off. Cornelius Gross, her supervisor at the EMBL in Rome, Italy, had acquired the fraktaline receptor-GFP mice, but instead of viewing them as only a visualization tool, he pondered the actual function of the fraktaline receptor. When he looked at the abundance of fraktaline itself in the developing hippocampus – a brain region critical for memory formation – he saw a spike in fraktaline expression shortly after birth. What’s more, fraktaline was made by neurons, suggesting a way that neurons might talk to microglia.
Because of how the GFP gene was inserted into these mice’s genome, if a mouse had two copies of the GFP gene, it no longer made fraktaline receptors. Paolicelli therefore used the mice to see what happened if this line of neuron to microglia communication was cut. The result? Hippocampal neurons had more synapses. It appeared that microglia prune away excess synapses during normal development. Supporting this conclusion, fragments of synapses were seen engulfed inside microglia.
Another group published work soon after, in 2012, showing that another messenger network known to be used by the immune system – complement signalling – also mediated communication between neurons and microglia and controlled synaptic pruning in the developing visual system.
“At that time, that was something pretty new,” Paolicelli says. “Most of the studies were just in pathology.” Ever since, synaptic pruning by microglia – first observed by Kreutzberg under pathological conditions – has been studied as a vital aspect of normal development and a contributor to brain plasticity in adulthood.
Microglia sensing the fraktaline released by neurons is just one of a number of channels by which neurons and glia message one another. Current research seeks to identify the full range of signals that coordinate pruning, including signals that draw the microglia toward synapses. Once there, a synapse may be tagged with either “eat me” or “spare me” signals.
One signal that microglia are especially attuned to is ATP. ATP is a fundamental part of energy processing in all cells, but it is also released by neurons – when they are either damaged or highly active. And microglia – owing to a receptor called P2Y12 – move towards spillages of ATP.
“Neurons speak their language, and microglia can listen in and translate this to respond accordingly,” says Professor Ukpong Eyo, a microglia researcher at the University of Virginia, USA. He says that a central concept that has emerged from watching microglia behaviour in the brain is that they are surveillance cells. The cells’ motile processes respond to many different signals by constantly and fleetingly contacting neurons and gobbling up unneeded parts of neurons. All this seems to point to the cells’ primary function: endlessly checking that the neurons in their vicinity are okay.
Eyo warns, however, that much remains unknown about what precisely microglia do most of the time. Imaging data, his own included, have revealed much about how microglia move but they have provided few insights into the consequences of these movements. “We can say they’re spending five minutes over here, ten minutes over there,” he says, “but what are they doing during those ten minutes?” It is often unclear if anything of substance has actually happened to a neuron or synapse after contact. What has been shown is that under certain circumstances, such as following a stroke, the processes of the microglia spend more time attached to neurons. Potentially, they respond to physical damage caused by the stroke, but nobody knows for sure.
Eyo also notes that there have been convincing studies showing that microglia – in development and during learning – help induce the formation of synapses, the opposite of pruning. How this occurs will require further study, and Eyo says that the field in general still needs more methods for interfering with specific aspects of microglia activities to unearth their full range of functions.
BACK TO PATHOLOGY
Eyo’s interest in how neural activity signals to glia led him to study epilepsy, in which neurons become hyperactive. He showed that during seizures, excess neural firing attracts microglia via ATP release. Strikingly, when Eyo blocked P2Y12 receptors or removed microglia altogether, seizures became worse, indicating that normally microglia recruitment quells epileptic activity. “What we’re trying to identify now – which is more difficult – is how they’re doing that,” he says.
In Berlin, Kettenmann is looking at how the masses of activated microglia present in gliomas affect the progression of this aggressive form of brain cancer. The microglia appear to help create space for the tumours to grow and also release substances that keep the cancer cells alive.
Epilepsy and glioma are, however, just two of the diseases in which microglia are being investigated today. Others include multiple sclerosis, chronic pain, motor neuron disease, Parkinson’s disease, autism, schizophrenia, depression, and the disease in which microglia were first seen about 115 years ago – Alzheimer’s.
Dr Aleksandra Deczkowska, a postdoc in the laboratory of Professor Ido Amit at the Weizmann Institute of Science, Israel, has been studying a subpopulation of microglia termed disease-associated microglia (DAM) that appear in brains affected by Alzheimer’s. “DAM are a child of the fact that we have a new technology, single-cell sequencing, that lets us see the cells as they are for the first time,” she says.
Taking thousands of microglia from a mouse model of Alzheimer’s and profiling the genes that each individual one expressed allowed the researchers to identify a small group of cells that looked different from the majority. These cells were also present in mouse models of motor neuron disease and multiple sclerosis. And microglia expressing the same tell-tale genes were seen in dissected brains of people who had had Alzheimer’s.
The transition to the DAM state depends on the activation of a particular receptor called TREM2. It is activated by various molecules found in degenerating brains, including amyloid-beta – the molecule that congregates into the plaques that define Alzheimer’s disease. TREM2 activates microglia much like pattern recognition receptors on other immune cells, which respond to generic molecules on infectious agents such as bacteria.
Deciphering what DAMs do, however, is proving challenging. These cells engulf and clear amyloid plaques and may be protective, but new data from other laboratories suggest that these microglia might, in fact, also help seed the plaques.
The most compelling argument for microglia playing a protective role in Alzheimer’s pathology, Deczkowska says, is that a number of gene variants identified as increasing the risk of developing the condition affect genes expressed highly or uniquely by microglia. This suggests that deficits in microglia biology might somehow take an aging brain down a path toward dementia.
Paolicelli has also pivoted to studying neurodegeneration, probing how the elimination of synapses by microglia might contribute to disease progression across a range of conditions – especially if those microglia contain problematic gene variants. Although the hallmark of Alzheimer’s disease is amyloid deposition, amyloid levels actually correlate poorly with dementia severity, suggesting that for certain people its presence is not problematic. A much better predictor of cognitive decline is the amount of synaptic loss.
“For too long these genes have been studied with a biased focus on amyloid removal,” Paolicelli says. But the relationship between microglia, plaques, and synaptic pruning is complex. In a study in which Paolicelli increased the phagocytic capabilities of microglia in mice, she saw that microglia cleared away more amyloid but also pruned more synapses.
To pin down the exact balance of protective and damaging microglia functions, the full diversity of their contributions at different disease stages is needed. Deczkowska is now moving on to study the effects of aging on microglia, hypothesizing that a decline in these cells’ functioning may relate to the fact that Alzheimer’s is a disease of old age. Paolicelli speculates that if certain factors lead microglia to react differently to potential disease-provoking events, it could tip the balance towards bad outcomes. “Specific dysfunction in microglia could be sufficient to originate neurodegeneration,” she says.
This is a challenge for every disease microglia are implicated in – are they starting the pathology or merely responding? And does this matter for targeting them for potential treatments? What has become clear is that microglia activation is neither an all or nothing event nor a simple one size fits all response. Eyo highlights that the way microglia change in epilepsy has essential differences from the classically described activated state. And Kettenmann and many others are looking at microglial changes associated with schizophrenia and autism as potential triggering mechanisms for these conditions in early development.
“The big question is what comes first,” says Kettenmann. “In the old days, people said the microglia respond to injury and the injury came first. Now the game is open.”