Learning new things is not enough in this ever-changing world we live in. Erasing outdated knowledge and replacing them with updated versions is just as critical. This ability is what neuroscientists call “cognitive flexibility.”
The lack of cognitive flexibility causes us to rely erroneously on outdated memories and make wrong choices, preventing us from adapting to change. Neurological disorders such as autism, schizophrenia, and early-stage dementia are characterized by a decrease in cognitive flexibility.
A new study shows star-shaped cells called astrocytes in the horn-shaped hippocampus are important for memory processing, and simultaneously regulate and integrate synaptic plasticity in nearby synapses to render memories flexible.
The study led by C. Justin Lee, PhD, and his team at the Center for Cognition and Sociality in the Institute for Basic Science (IBS) in Daejeon, South Korea, is published in an article in the journal Biological Psychiatry titled, “Astrocytes Render Memory Flexible by Releasing D-Serine and Regulating NMDA Receptor Tone in the Hippocampus.”
The underlying cause of reduced cognitive flexibility is a decrease in the function of a neuronal receptor and ion channel called NMDAR (N-methyl-D-aspartate receptors), responsible for synaptic plasticity.
However, how a decrease in NMDAR function translates into impaired cognitive flexibility is not clear. It is also unclear whether hippocampal astrocytes regulate basal NMDAR function and cognitive flexibility. Moreover, while many small molecules (agonists and co-agonists) trigger the physiological response of NMDARs, the source of one NMDAR co-agonist called D-serine has been unclear.
In this study, the researchers showed astrocytes can synthesize D-serine and release it through a calcium-activated channel called Best1. Earlier studies have shown astrocytes release another messenger molecular called glutamate through Best1. Both glutamate and D-serine trigger NMDARs. That astrocytes can release both glutamate and D-serine through the same channel indicates these cells are ideal regulators of NMDAR activity and synaptic plasticity.
“We found that hippocampal astrocytes regulate NMDAR tone via BEST1-mediated co-release of D-serine and glutamate,” the authors noted.
Pivotal for cognitive flexibility is a neurophysiological phenomenon called heterosynaptic long-term depression (hetLTD) where inactive synapses weaken when nearby synapses activate. In this study, the researchers show hetLTD is mediated by astrocytes in the hippocampus.
Lee said, “Previous studies have mostly focused on changes in specific synapses to stimuli. The discovery of this phenomenon where changes in one synapse can induce changes in nearby synapses during learning shows that finding out what happens to the other synapses is important for understanding the mechanism of learning and memory formation.”
Wuhyun Koh, PhD, a scientist at the IBS and the first author of the paper said, “Since each astrocyte is in contact with over 100,000 synapses, astrocytes can control numerous synapses and integrate synaptic plasticity simultaneously.”
The researchers showed mice lacking Best1 exhibit reduced NMDAR tone and consequently impaired hetLTD. In the Morris water maze—a popular experimental paradigm used to measure cognitive flexibility in mice—mice lacking Best1 scored just as well as normal (wildtype) mice in the initial learning session where the mice had to locate a platform hidden under water. However, when the platform was moved to a different location and the mice had to re-learn the new location of the hidden platform, Best1 knockout mice continued to swim to the old location, exhibiting impaired cognitive flexibility, in contrast to the normal mice that were able to re-learn the new location.
To rescue the impairment, when the researchers injected mice lacking Best1 with D-serine during the initial learning phase, it improved NMDAR tone and cognitive flexibility. This indicates memory flexibility is determined at the time of initial learning, which refutes the previously proposed notion that synaptic plasticity only occurs after memory formation, when memories are modified.
Investigating the mechanism further, the researchers show norepinephrine can activate astrocytes, causing them to release both D-serine and glutamate. This implies that the flexibility of memory can be determined by concentration and arousal during learning.
“The study uses two new research protocols: heterosynaptic recording and utilization of GRABNE, norepinephrine sensor, developed in Yulong Li’s lab. The GRABNE sensor was surprisingly optimal for measuring norepinephrine,” said Koh. “It was not easy to measure the heterosynaptic plasticity that occurs when homosynaptic plasticity appears, but using theta glass electrodes, two independent synaptic measurements were obtained and plasticity was quantified.”
In future studies, Koh said, the team intends to explore further along three lines of investigation: mechanisms underlying norepinephrine mediated plasticity, how hippocampal NMDAR tone is modulated in real time, and NMDAR tone and function in astrocytes present in brain regions beyond the hippocampus.
