Ever feel groggy first thing in the morning? Imagine how animals must feel after months in deep hibernation. Thanks to cold-shock response proteins that help replace synapses lost during extended slumber, hibernators manage to awaken with their cognitive faculties intact. Those same proteins may falter in neurodegenerative disease, according to a study in the January 14 Nature. Researchers led by Giovanna Mallucci of the University of Cambridge in England reported that artificially boosting one of them, RBM3, protected mouse models of Alzheimer's and prion diseases. Whether this protein plays a role in synaptic repair in humans, and whether it could point the way to a therapy to slow neurodegeneration, are still open questions.
During hibernation, the body dials down functions to a minimum. The brain, being the most energetically costly organ, takes a big metabolic cut, and does so by dismantling synapses. In what Mallucci referred to as the most intense example of synaptic plasticity in nature, these withered synapses reemerge again as the animals awaken (see Popov and Bocharova, 1992; Magariños et al., 2006). “It’s an extreme form of what happens in our brains all the time, which is breaking down and reforming synapses,” Mallucci said.
Mallucci, whose lab primarily focuses on synaptic loss during prion diseases, wanted to tease apart the pathways that tip synapses toward destruction and regeneration. She opted to use a model of hibernation, the starkest example of synaptic recovery.
To simulate hibernation, first author Diego Peretti of the University of Leicester in England injected mice with 5’-adenosine monophosphate (5’-AMP), a metabolite that disrupts oxidative phosphorylation and prevents animals from generating heat. The AMP rapidly reduced core body temperature and put mice into a deep torpor before they even had a chance to shiver up some warmth, Mallucci said. The researchers then transferred the animals into a fridge kept at a chilly 16-18 degrees Celsius, a temperature other small mammals typically experience when hibernating. Electron microscopy revealed that hippocampal neurons lost about a quarter of their synapses after the mice spent 45 minutes “hibernating,” but rose back to normal after the animals were gradually rewarmed.
How would neurodegenerative disease affect this striking display of structural plasticity? The researchers performed the same cooling/warming experiment on 5xFAD mice (a model of AD) and on Tg37+/- mice infected with prions. Tg37 mice overexpress the prion protein PrP, so injecting them with infectious prions leads to a rapid loss of neurons and eventually death. Synapse loss is an early symptom of disease in both models, though it occurs at around seven weeks post-infection in the prion mice, and at 4 months of age in the 5xFAD mice. When the researchers cooled and rewarmed the mice prior to these time points (at four to five weeks post-infection in the prion mice or at 2 months of age in the 5xFAD mice), synapses recovered completely. However, when the researchers cooled and rewarmed the mice just before impending synapse loss (at six weeks post-infection in the prion mice or 3 months in the 5xFAD mice), synapses did not recover, suggesting that something about the neurodegenerative process prevented this structural plasticity.
To determine why, the researchers first turned to cold-shock proteins. Previous studies had reported that RBM3 switches on in the mouse hippocampus in response to hypothermia, and that it protects cultured neurons from cell death when the temperature drops (see Tong et al., 2013; Chip et al., 2011). In neurons, the protein localizes to dendrites, where it promotes protein synthesis (see Dresios et al., 2005; Smart et al., 2007). The researchers found that RBM3 levels rose in the hippocampi following cooling in normal mice or in young 5xFAD and prion-infected animals. However, by 3 months of age in 5xFAD mice or six weeks post-infection in the prion mice, RBM3 levels stayed put during hibernation.
Could this sluggish RBM3 response thwart synaptic recovery? The researchers addressed this by tweaking RBM3 expression in several ways. First, they used cooling to turn on RBM3 in wild-type mice or prion-infected mice very early in disease. When they monitored RBM3 expression in these scenarios, they were surprised to find that the protein remained elevated for several weeks following episodes of cooling and rewarming. This early cooling treatment protected the prion mice from the subsequent loss of synapses that normally occurs as the disease progresses. The hibernation stint also preserved synaptic transmission (as measured by excitatory post-synaptic currents), and reduced neuronal cell death in the hippocampus. The mice also burrowed and recognized novel objects just like wild-type animals. While the prion-infected TG37 mice died young despite their early cooling treatment, they survived an average of six days longer than PrP mice that were always cozy in their cages. Importantly, all of the benefits of early cooling were nixed when the researchers treated the mice with shRNAs that turned down RBM3 expression, or when the mice were cooled and rewarmed later in the disease.
In the grand finale set of experiments, the researchers bypassed cooling and injected lentiviruses expressing RBM3 directly into the hippocampi of mice early in prion disease. They achieved a threefold overexpression of RBM3 that afforded the mice the same synaptic, behavioral, and survival benefits that early cooling had, and rescued flagging protein synthesis observed in neurons nine weeks post-infection. Boosting RBM3 expression also allowed synapses in both neurodegenerative disease models to recover completely after cooling. Conversely, knocking down expression of RBM3 accelerated disease progression in the prion model, and hastened synapse loss in both models. Synapses and memory even took a hit in normal mice deprived of RBM3 expression, as they lost synapses and did not recognize novel objects as well as control mice. This suggested that the cold-shock protein may play an important physiological role in normal synapse upkeep.
Mallucci hypothesized that RBM3 promotes synaptic plasticity and staves off neurodegeneration by raising levels of protein synthesis in dendrites. “Synapses are so dependent on their key synaptic proteins for assembly and function,” said Mallucci. Because synapses often reside distant from the cell body, local translation at the dendrite is important to ensure a ready supply of such proteins and thus facilitate synaptic recovery, she added.
Thomas Arendt of the University of Leipzig, Germany, commented that the study was interesting and extended previous work in the field, but wondered how RBM3 tied in with some of the hallmarks of AD neuropathology, namely tau phosphorylation. He, and others, had previously reported that hibernation triggered hyperphosphorylation of tau in a variety of hibernators, including ground squirrels, hamsters, and black bears, and that, much as the synapses recover, tau becomes dephosphorylated again upon waking (see Arendt et al., 2014, and full comment below). Mallucci told Alzforum that they did measure changes in tau and as with previous reports, phosphorylation increased during cooling and decreased following warming in normal mice and in their two models of neurodegenerative disease. She suspects that problems with synaptic recovery in these models occur independently of tau phosphorylation. Mallucci also reported that levels of other key disease-associated culprits—Aβ oligomers in the 5xFAD mice and misfolded prion proteins in prion-infected mice—were not altered by cooling or by changes in RBM3 expression.
Lest anyone opt for extended plunges in ice baths to boost RBM3 and avoid neurodegeneration, Mallucci warns, “Please don’t try this at home”—it would cause hypothermia and bring on a host of other unsavory problems. While hypothermia has been used therapeutically to prevent cellular damage during surgery or brain damage in many contexts, including stroke, in an uncontrolled setting it is extremely dangerous, she said. Instead, Mallucci plans to dig deeper into the RMB3 pathway to understand why this molecule shuts down in the context of AD and prion disease, and whether boosting its expression might help slow neurodegeneration. As Mallucci and other researchers pointed out, the role of RBM3 in the human brain remains unknown.—Jessica Shugart