16 November 2011. Microglia have a split personality in neurodegenerative disease—sometimes good, sometimes bad. What transforms a microglial Jekyll into an inflamed Hyde? Maybe fractalkine signaling, or the lack of it. This neuronally produced chemokine soothes microglia and dampens neuroinflammation (see Cardona et al., 2006; Ransohoff et al., 2007), but the pathway’s role in disease has proved maddeningly hard to pin down. Two new papers offer clues as to why. In the November 9 Journal of Neuroscience, researchers led by Carmelina Gemma at the University of South Florida, Tampa, describe fractalkine’s physiological job. Using fractalkine receptor knockout mice, they report its signaling pathway is necessary for synaptic plasticity and learning. In the same issue, scientists led by Cristina Limatola at Italy’s Sapienza University, Rome, report that wild-type and fractalkine knockout mice with ischemia respond in opposite ways to fractalkine signaling. The chemokine protects wild-type mice, but harms knockouts. Together, the two papers emphasize how this pathway depends on context, and that scientists still have much to learn before findings can be translated into therapies.
Neuronal fractalkine (aka CX3CL1) can be either soluble or membrane bound. Its receptor, CX3CR1, occurs only on microglia (see Harrison et al., 1998). Although several studies suggest that fractalkine signaling protects neurons in various neurodegenerative conditions, such as Parkinson’s disease (see Pabon et al., 2011), the field has been plagued by contradictory results. For example, in Alzheimer’s disease models, scientists have found that CX3CR1 knockouts have more tau pathology than controls, suggesting the pathway protects the brain (see ARF related news story on Bhaskar et al., 2010). However, knockouts deposit less amyloid than controls (see Lee et al., 2010) and lose fewer neurons (see ARF related news story on Fuhrmann et al., 2010), an indication that the pathway is harmful.
“That fractalkine can be neuroprotective is pretty clear, but it’s clear also that it depends on the stage of the disease and the microenvironment of the brain,” Gemma told ARF. Absence of the pathway leads to an inflammatory state, which, depending on the role of microglia in the disease, could be good or bad, she said.
To pin down the role of fractalkine in healthy brains, Gemma’s group compared wild-type mice to CX3CR1 knockouts and heterozygotes on behavioral and electrophysiological measures. The researchers had previously reported that adult mice lacking CX3CR1 produce fewer new neurons than do wild-type animals (see Bachstetter et al., 2011). In the current study, first authors Justin Rogers and Josh Morganti extended the findings to behavior, showing that knockouts learned more poorly than controls in the Morris water maze, in fear conditioning paradigms, and on motor tasks. At the cellular level, the authors found weaker long-term potentiation (LTP), a measure of synaptic plasticity, in hippocampal slice cultures from knockout mice than those of wild-type. Rogers and colleagues demonstrated a gene dosage effect, with CX3CR1 heterozygotes showing an intermediate phenotype for neurogenesis, LTP, and hippocampal-dependent learning. The authors noted, however, that both heterozygotes and knockouts struggled similarly with motor learning, a cerebellum-dependent task.
Looking for the downstream mechanism, the authors found that CX3CR1-deficient mice made higher-than-normal levels of the microglial proinflammatory cytokine IL-1β. When they added the IL-1β receptor antagonist IL-1ra to hippocampal slice cultures from knockout mice, LTP bounced back to normal levels. Similarly, infusing IL-1ra into the ventricles of live mice through a guide cannula for four weeks rescued hippocampal-dependent learning, but not motor learning.
“I thought the paper was quite impressive,” said Terrence Town at Cedars Sinai Medical Center, Los Angeles, California, who was not involved in the study. In particular, the IL-1ra rescue experiments cement the link between IL-1 and the effects on synaptic plasticity and behavior, Town said.
Can scientists conclude fractalkine signaling is good in healthy brains? Not so fast. Recent work published by Limatola’s group in October (see Maggi et al., 2011) gave the exact opposite results in a slightly different CX3CR1 knockout. It had better LTP than controls and learned the water maze better. Town said the divergent results do not surprise him, because fractalkine signaling is complex and context dependent. He speculated that there may be a modifier gene that interacts with the fractalkine pathway and varies in different mouse strains. For her part, Gemma pointed out that her study used only male mice, while the experiments by Limatola’s group used only females, and suggested the contradictory results could represent a gender difference. Town agreed that is also possible, since estrogen signaling affects neuropathology in several disorders.
In their Journal of Neuroscience paper, Limatola’s group shone the spotlight on ischemia. This condition seems to be an exception to the general rule that fractalkine protects diseased brains, with previous studies reporting that fractalkine and CX3CR1 knockouts fare better after ischemia than do control mice (see Soriano et al., 2002; Dénes et al., 2008). To take a closer look, first author Raffaela Cipriani blocked carotid arteries to induce ischemia in wild-type mice, CX3CR1 knockouts, and fractalkine knockouts, and then compared the effect of injecting exogenous fractalkine into the brain ventricles. In agreement with earlier studies, the authors found that mice without fractalkine signaling fared better after ischemia than wild-type mice; in addition, giving fractalkine to knockouts worsened brain damage. However, in wild-type mice, and also wild-type rats, the chemokine had the opposite effect. Exogenous fractalkine protected cells, reduced the ischemia area, and improved motor abilities. Looking for what might underlie this difference, the authors compared microglia from knockouts and wild-type mice in an ischemic, i.e., low-oxygen, in-vitro environment. When given fractalkine, wild-type microglia in this environment became less efficient at phagocytosis, while microglia from fractalkine knockouts did not change. On the other hand, after fractalkine stimulation knockout microglia released less TNFα, which appears to be neuroprotective in ischemia (see Lambertsen et al., 2009), but wild-type cells continued to pump out the factor.
“Our idea is that the difference [in microglia] we observe in vitro could explain many of the differences people find in experiments performed on wild-type animals and knockout animals,” Limatola said. She speculated that when neurons and microglia lack normal communication throughout development, the microglia mature differently. She is investigating whether microglia attempt to compensate for the lack of fractalkine by making more receptor, which might cause them to respond unpredictably to exogenous chemokine. Alternatively, microglia that never see fractalkine might turn on a different suite of genes than they otherwise would. Using microarrays, Limatola will compare the gene expression of wild-type and fractalkine knockout microglia, and plans to further characterize microglia in vivo in different mouse strains. In addition, she will probe the role astrocytes play in the neuron-microglial conversation.
Limatola suggested that wild-type animals might make better models for human therapy than knockouts, since most people presumably have normal fractalkine signaling. Gemma agreed, saying the distinction between wild-type and knockout mice supports using the former in preclinical research. Town demurred, however, noting that it is not yet clear which mouse model is more relevant to people. “We have to be careful in terms of the translational side,” Town said. He believes the fractalkine pathway could be a worthwhile therapeutic target, especially in AD, because of its striking effects on pathology when manipulated in mouse models. The new results emphasize that, depending on the model system, context, and the type of pathology under consideration, the fractalkine pathway can play opposing roles, Town said. “We need to figure out why that is. Otherwise, there will be significant roadblocks to translation.”—Madolyn Bowman Rogers.
Rogers JT, Morganti JM, Bachstetter AD, Hudson CE, Peters MM, Grimmig BA, Weeber EJ, Bickford PC, Gemma C. CX3CR1 deficiency leads to impairment of hippocampal cognitive function and synaptic plasticity. J Neurosci. 2011 Nov 9;31(45):16241-50. Abstract
Cipriani R, Villa P, Chece G, Lauro C, Paladini A, Micotti E, Perego C, De Simoni MG, Fredholm BB, Eusebi F, Limatola C. CX3CL1 is neuroprotective in permanent focal cerebral ischemia in rodents. J Neurosci. 2011 Nov 9;31(45):16327-35. Abstract