New research reveals the cellular mechanisms by which memory-encoding neuronal networks emerge
Tell me where dwell the thoughts, forgotten till thou call them forth? Tell me where dwell the joys of old, and where the ancient loves, And when will they renew again, and the night of oblivion past, That I might traverse times and spaces far remote, and bring Comforts into a present sorrow and a night of pain? Where goest thou, O thought? To what remote land is thy flight? If thou returnest to the present moment of affliction, Wilt thou bring comforts on thy wings, and dews and honey and balm, Or poison from the desert wilds, from the eyes of the envier?
In his epic poem, Visions of the Daughters of Albion, William Blake wonders about the nature of memory, its ability to mentally transport us to distant times and places, and the powerful emotions, both positive and negative, that our recollections can evoke. The poem contains questions that remain highly pertinent today, such as what happens to our long-lost memories, and how do we retrieve them?
More than two centuries later, the mechanisms of memory storage and retrieval are the most intensively studied phenomena in the brain sciences. It’s widely believed that memory formation involves the strengthening of connections between sparsely distributed networks of neurons in a brain structure called the hippocampus, and that subsequent retrieval involves reactivation of the same neuronal ensembles. And yet, neuroscientists still struggle to answer Blake’s questions definitely.
Now, a team of researchers at the University of Geneva have made another important advance in our understanding of the neural mechanisms underlying memory formation. Using a state-of-the-art method called optogenetics, they show how the neuronal ensembles that encode memories emerge, revealing that ensembles containing too many neurons – or too few – impair memory retrieval.
Optogenetics is an extremely powerful technique that involves introducing algal proteins called channelrhodopsins (ChRs) into neurons. This renders the cells sensitive to light, such that specified groups of them can be switched on or off, using pulses of laser light delivered into the brain via optical fibres, on a timescale of milliseconds.
In recent years, researchers have used optogenetics to label hippocampal neurons that become active during memory formation in the mouse brain, and to manipulate the labelled ensembles in various ways. In this way, they can reactivate the same ensembles to induce memory retrieval; switch fearful memories on or off; convert negative memories into positive ones, or vice versa; and even implant entirely false memories into the brains of mice.
The new research, led by Pablo Mendez and the late Dominique Muller, who tragically died in a gliding accident in April of last year, builds on this earlier work. They created genetically engineered mice expressing ChR in granule cells on one side of the brain, in the dentate region of the hippocampus. Granule cells are the principle neurons in this region of the hippocampus, which are thought to be critical for hippocampal functions such as memory and spatial navigation. They placed the animals into large cages, allowing some of them to explore their new environment. Meanwhile, they optogenetically activated random granule cells in some of the mice, but not others.
When they dissected and examined the animals’ brains 45 minutes later, the researchers found spatial exploration evoked activity in ensembles of hippocampal neurons, as determined by levels of cFos, a so-called ‘immediate early’ gene that is switched on quickly when neurons start to fire. Importantly, mice allowed to explore their cages had higher numbers of cFos-expressing granule cells than those left in their home cages for the duration of the experiment, and those that received optogenetic stimulation during the exploration had significantly higher numbers of cFos–positive neurons than those that did not.
This showed that spatial exploration evokes activity in ensembles of dentate granule cells, and that randomly altering the activity of these networks with optogenetic stimulation increases the size of the ensembles, or the number of cells within them.
But does manipulating the size of the ensembles have any effect on behaviour? To find out, Mendez and his colleagues placed mice expressing ChR in their hippocampi into another cage, and gave them several mild electric shocks. With repetition of this treatment, the mice quickly learn to fear the cage, and quickly freeze up when returned into it, even when they are not given more shocks.
This time, the researchers optogenetically stimulated random granule cells in some of the mice, but not others, during the training, in order to increase the size of the neuronal ensemble that encodes the fearful memory. These mice exhibited less freezing behaviour when returned to the same cage than others who received no stimulation. But the stimulation also created artificial fear memories, such that the animals froze up in other situations, too.
Inhibition of random granule cells had the same effect, suggesting that merely altering the number of neurons in the ensemble interfered with the animals’ ability to recall the fearful memories. These findings are consistent with those of an earlier study, which also showed that inhibiting or stimulating granule cell activity impairs contextual learning.
To understand why this might be, the researchers performed another series of experiments, using microelectrodes to record the activity of neurons in slices of hippocampal tissue. These experiments showed that optogenetic stimulation of granule cells produces a robust response in neighbouring interneurons, which release the inhibitory neurotransmitter GABA.
Thus, the firing of granule cells leads inhibitory interneurons, which dampen adjacent granule cells and prevent them from entering the ensemble. In this way, interneurons appear to stabilize newly-formed memories by regulating the number and distribution of granule cells involved in encoding memories. Activating or silencing random granule cells upsets this process and alters the number of granule cells, which may make the new memories unstable.
“In this study, we used a simple form of memory, the memory of a spatial context, but the challenge is studying how more complex experiences are memorized, and how the brain deals with the storage of multiple experiences,” says Mendez. “Understanding these questions could help us to understand the limits of the brain’s storage capacity.”
Stefanelli, T., et al. (2016). Hippocampal Somatostatin Interneurons Control the Size of Neuronal Memory Ensembles. Neuron, 89: 1-12. DOI: 10.1016/j.neuron.2016.01.024 [Abstract]
guardian.co.uk © Guardian News & Media Limited 2010