Meet the Microglia
American Academy of Neurology meeting report
SAN DIEGO—Researchers have been on a mission to understand microglia—the brain’s resident immune cells—since they were discovered more than 100 years ago. These complex, highly plastic, specialized cells combat infection and engulf debris in the central nervous system (CNS), but they can also become destructive. The details of how microglia act in MS remain murky; a better understanding of their roles under both healthy and inflammatory conditions might yield new clues about how to monitor and treat this disease. Two lines of work presented at this year’s annual meeting of the American Academy of Neurology investigated microglia. One produced a long-sought single marker of microglia and identified a unique microglial “signature,” while another study used a radioactive tracer to visualize these cells in MS patients.
Oleg Butovsky, working in collaboration with Howard Weiner at the Center for Neurologic Diseases at Brigham and Women’s Hospital in Boston, Massachusetts, used sophisticated tools to reveal molecular signatures of microglia (which are a type of macrophage) that were distinct from those of peripheral macrophages that regularly infiltrate the brain. “His findings provide an outstanding advance in our understanding,” said Monica Carson, who studies microglia at the University of California, Riverside, and was not involved in the study.
In another important step, Butovsky recently generated an antibody marker that labels the surface of microglia—and only microglia—in both homeostatic and inflammatory states, allowing researchers to distinguish these cells from peripheral macrophages (Butovsky et al., 2012). Although past studies—including one led by Carson (Schmid et al., 2009)—have generated signatures for microglia, they have typically consisted of a panel of expressed genes rather than a single marker. “He’s gone beyond what we and others have done,” she said.
Butovsky’s survey confirms and builds on findings of the last 20 years, Carson said, which have revealed much about the origins of microglia (Lassmann et al., 1993). Unlike peripheral macrophages, which are short-lived and regularly regenerated throughout life from cells from the bone marrow, microglia are long-lived cells that are “born” very early in development. They emerge in the yolk sac during the first trimester, travel to the brain, and then proliferate and differentiate in that unique environment after birth. Microglia can divide under certain conditions in adulthood, but their supply is never replenished from another peripheral source. “That tells you that there’s something fundamentally different about the brain that demands a terminally differentiated, long-lived macrophage,” Carson said. Although peripheral macrophages are present in the inflamed brain, Butovsky said that his work profiling microglia provides further support for the idea that they are distinct from these other myeloid lineages and not replenished from peripheral sources.
Why the push to understand these highly specialized glia? It’s an important cell type: “They’re activated in all neuroinflammatory conditions as part of a CNS response—anything like stroke, Alzheimer’s disease [AD], brain injury, tumor, and so on,” Butovsky said. By having a better handle on microglia’s basic biology, we might learn about how to modulate them. The ability to manipulate destructive microglia into a beneficial state might provide an opportunity for therapeutic interventions. But the first step, he said, is to understand the fundamental properties of these cells and to be able to distinguish them from recruited macrophages. The new antibody, called 4D4, does just that, specifically labeling microglia from mouse brain.
Butovsky and his colleagues first isolated microglia from the brains of mice, separating them from other cells using antibodies to several known microglial proteins. Then the researchers injected these cells into rats, which produced a handful of antibodies specific to microglia, the best of which turned out to be 4D4. Using the markers, “even under neuroinflammatory conditions, we can differentiate the microglia from the recruited cells,” Butovsky said. 4D4 also doesn’t recognize astrocytes or neurons. “It’s literally specific only to resident microglia,” Butovsky said, “but we still don’t know what exactly the antibody recognizes.” That unknown doesn’t detract from 4D4’s value, Carson said. The epitope may not even be a protein, she said, which would be the usual suspect. “Antibodies bind to three-dimensional structures, which can include sugars, cellular modifications, even plastic—whatever is in their environment,” she said.
Butovsky hunted for a microglial signature in adult mouse brain by starting with a gene expression microarray profile and proteomics studies, as other researchers who were studying microglia have done. The team then narrowed its search for specific markers using an MG400 chip, which Butovsky called “quite exciting new technology,” to look for a signature that was unique to microglia. The new tool quantifies cellular RNAs using probes that bind directly to RNAs, bypassing the need to manipulate them in any way. The technique requires a trifling 3000 cells, and it can simultaneously assess the levels of up to 800 RNAs, including messenger RNAs (mRNAs) and microRNAs (miRNAs)—small noncoding RNAs that interact with mRNAs and affect their expression. In collaboration with Jack Antel and Craig Moore at McGill University in Montreal, Canada, the researchers identified mRNA and miRNA profiles in human microglia that were similar to those seen in mouse. (Human microglia were isolated from adult brain tissue from patients undergoing surgery for intractable epilepsy.) In both humans and mice, the cellular signatures were unique to microglia. (In a separate presentation, Moore described the miRNA profile of human microglia and peripherally derived macrophages in more detail, work that further supports a pathogenic role for the miR-155 miRNA in MS.)
Importantly, Butovsky investigated microglia that were freshly isolated from adult rodent and human brains; in contrast, many other studies have used cells from neonatal mouse brain or that had been cultured in a dish. This is significant in part because, despite their distinct lineage, microglia “are highly plastic; they do what they’re told,” Carson said. Butovsky found heterogeneity among the microglial mRNA and miRNA signatures based on the brain region where the cells had resided. Cells maintained in culture—with no neurons, astrocytes, or other brain cues to influence them—likely differentiate into a cell type quite unlike that in the mammalian brain. By studying cells cultured in artificial environments, scientists might have overestimated microglial toxicity, Carson said: “When you appreciate that, under homeostatic circumstances, they’re not toxic, that changes the game.” When microglia do become harmful, that represents a severe dysfunction, she said.
As for microglia’s “activated,” harmful state, Butovsky saw “defining signatures specific to each disease” in microglia from animal models of MS, AD, and amyotrophic lateral sclerosis (ALS). The activation state is not as simple as turning the cells “on,” he said, and might need to be characterized as microglial-ALS, -MS, -AD, and so forth. Carson agreed that activation is “never going to look like the cartoon” typically used to denote the homeostatic and the activated, harmful states. Carson likens the behavior of microglia in different conditions to a spectrum or sphere of possible activation states, with patterns and ratios of molecules holding sway rather than a simple on-off switch. Nevertheless, she said, “the cartoons are useful to give ideals, but we know those rarely exist in biology.”
While the development of the 4D4 antibody and the microglial signatures promise to help scientists pinpoint microglia in pathology studies, another talk given at the AAN meeting described how positron emission tomography (PET) imaging could be used to detect these cells in vivo. To visualize active microglia, Marios Politis of Imperial College London used a radioactive compound called PK11195. This ligand binds to an outer mitochondrial membrane protein termed translocator protein, which is expressed as microglia become activated. In an earlier study, his team imaged the brains of MS patients and found higher levels of the marker—visible on a PET scan—than in healthy controls (Politis et al., 2012). After being treated with the disease-modifying medication natalizumab for a year, the same patients had lower levels of PK11195 labeling in their gray matter, and the extent of labeling seemed to correlate with patients’ disability score. There is some debate about whether PET ligands like PK11195 are specific for microglia or whether they also label reactive astrocytes (Lavisse et al., 2012). Although it would be nice to know specifically which cell types are active, Politis said, “at the end of the day it may not matter,” if the label provides a useful tool to track disease processes in the brain.
Although peripheral macrophages have so far been the immune cells at the center of attention in MS research, microglia are emerging as sophisticated and crucial cells in both health and disease. These new tools for understanding and visualizing microglia will help discern their role in MS, perhaps yielding new therapeutic or diagnostic strategies.
Key open questions
- What are the beneficial, homeostatic roles of microglia in the brain?
- Do these cells become harmful and contribute to disease pathology or disability in MS?
- If so, can such harmful microglia be manipulated into a beneficial state in living brain?
- Might reactive microglia be used to track and stage the disease process in patients?