Skip to main content

Long-term Investigation of Functional Excitatory Synapses: Linking Plasticity, Network Wiring and Memory Storage

Periodic Reporting for period 3 - LIFE synapses (Long-term Investigation of Functional Excitatory Synapses: Linking Plasticity, Network Wiring andMemory Storage)

Reporting period: 2020-04-01 to 2021-09-30

The nature of the physical substrate of memory – or engram – is probably one of the longest studied mysteries in neuroscience, and yet it still remains elusive. In recent years, the search for the engram has gained new momentum due to the possibilities of selective activation and silencing of specific neurons in the brain. Recent studies suggest that the engram could be defined as the subset of neurons that is necessary and sufficient to cause recall of a specific memory when activated. But where is the engram when the neurons are not active? Most likely, ‘lasting alterations’ during memory formation are encoded in the connections between nerve cells, termed synapses. The connection pattern and strength of these synapses determine how information is flowing between nerve cells and thereby underlie memory recall. This raises the possibility that the engram could be encoded in a pattern of altered synapses, not a pattern of cell bodies. Long-lasting potentiation or depression of synaptic efficacy is thought to underlie learning and memory formation suggesting that the engram could be stored in the strength of synapses. Yet, most excitatory synapses in the brain are highly plastic and show pronounced morphological dynamics. It is therefore not clear to what extent engrams can be stored in a network of synapses and how functional and structural changes of individual synapses contribute to the engram.

The central aim of this project is to identify synapses participating in the engram and to study their morphological stability and functional properties, but also to identify general rules that determine the lifetime of synapses and their participation in information processing. We develop novel approaches to precisely control synaptic circuits and neuronal networks with light or targeted pharmacological interventions termed "optogenetics" and "chemogenetics", respectively. We use these techniques in combination with molecular markers of synaptic and neuronal activity to investigate functional synapses and neuronal populations in their native circuit over the time scale of weeks, relevant for memory encoding and long-term storage. Thus, by connecting functional long-term analysis of single synapses and neuronal circuits with morphological observations the project will fill a wide gap in our understanding of how synapses contribute to formation and storage of memories.
During the first half of the project, we worked on the improvement of existing and development of novel optogenetic tools for neuronal activity manipulation. We engineered novel anion-conducting channelrhopsins termed Phobos and Aurora, which enable inhibition of neuronal activity at a wide optical spectrum and at different temporal scales. We demonstrated their superiority over other, commonly used optogenetic silencing methods that are based on ion pumps, for silencing of neuronal circuits in Drosophila larvae. In addition, we contributed electrophysiological measurements in neurons for collaborative projects working on different types of light-sensitive proteins. One study described the crystal structure of the red shifted, excitatory channelrhodopsin Chrimson, leading to the development of a new, even further red-shifted version of Chrimson termed ChrimsonSA. A second study described MerMAIDs, a novel family of naturally-occurring anion-conducting channelrhodopsins, assembled from the Tara Oceans metagenomes.

In parallel, we worked on improved optical methods to investigate synaptic transmission. Using two-photon imaging at single Schaffer collateral synapses together with patch-clamp electrophysiology and biophysical modeling, we measured novel, genetically encoded sensors of fast glutamatergic transmission. In another collaborative effort, we developed a new, genetically encoded optical sensor termed SynTagMA that enables identification of thousands of active synapses at the same time. We demonstrated its applicability in vivo as tool to map active synapses at hippocampal interneurons and to track populations of active hippocampal neurons associated with a particular behavioral state.

Moreover, we investigated the long-term consequences of synaptic plasticity on the stability of single Schaffer collateral synapses during the following days. This work demonstrates that strengthening synapses reduces their risk for elimination from the circuit. We further showed that sequential synaptic potentiation and depression interact over days to influence synapse stability, with the last plasticity event fully determining synaptic survival probability. Moreover, we addressed whether synaptic potentiation is maintained at synapses, showing that individual synapses revert to their original strength already 24 hours after the plasticity-inducing event. Long-lasting circuit modifications are therefore more likely governed by removal or stabilization of synapses. Thus, functional plasticity at the single synapse serves to set its lifetime and thereby determines whether it is retained in the pathway or not. The long-term consequence is that initial functional adaptation s of synapses, seen as potentiation or depression of synaptic pathways are transformed into changes in circuit wiring, which, from a macroscopic point-of-view, have the same effect on synaptic pathways.

During the third funding period, we published a novel optogenetic tool for bidirectional control of neuronal activity. The brain is a complex organ composed of many different types of nerve cells that are connected to each other with highly specific patterns. To better understand brain function and dysfunction, including neurological disorders, one must be able to manipulate nerve cells with high precision and specificity in the living organism. In principle, optogenetic manipulations allow the activation and inhibition of the same population of neurons to test their sufficiency and necessity for a particular brain function. However, existing optogenetic tools do not easily allow such bidirectional manipulations with light. Moreover, there are still some additional remaining challenges in the field of optogenetics that demand refinement and further development of optogenetic actuators with new biophysical properties.
We developed a novel optogenetic tool, termed BiPOLES (Bidirectional Pair of Opsins for Light-induced Excitation and Silencing), that expands the possibilities for optical manipulation of neuronal networks. BiPOLES is comprised of an inhibitory, blue-light-sensitive anion-conducting channelrhodopsin fused to an excitatory, red-light-sensitive cation-conducting channelrhodopsin in a single protein. BiPOLES allows multiple new applications including (1) potent excitation and inhibition of the same neurons with red and blue light, (2) exclusive red-light activation of a neuronal subpopulation in multicolor experiments, thus enabling mutually exclusive dual-color excitation of two distinct populations when used together with a second blue-light sensitive channelrhodopsin, and (3) optical tuning of the membrane voltage.
Due to its utility for a wide range of research questions, its versatile functionality and its applicability in numerous model systems - as demonstrated in this study for worms, flies, mice and ferrets - BiPOLES fills an important gap in the optogenetic toolbox and might become the tool of choice to address a number of yet inaccessible problems in neuroscience.

Finally, we show how different anesthetics alter spine dynamics, hippocampal activity and memory consolidation. Understanding how different anesthetics affect the brain, particularly the hippocampus, is important for both clinicians with human patients and experimental scientists who work with animals. We recorded brain activity from the hippocampus while mice were anesthetized using one of three common combinations of general anesthetics: isoflurane, ketamine/xylazine (Keta/Xyl), and medetomidine/midazolam/fentanyl (MMF). Brain activity was recorded electrically and by calcium imaging. Both recording methods showed that each drug changed brain activity in the hippocampus compared to wakefulness or natural sleep. We found a number of differences in how the specific anesthetics affected the brain. For example, Keta/Xyl strongly reduced overall calcium activity, while MMF affected its rate much more than the duration. Furthermore, all anesthetics affected the stability of synaptic connections between brain cells in the hippocampus. Keta/Xyl disturbed synaptic stability most drastically, reflecting its strong disturbances of neuronal calcium activity. Recovery time also differed; brain activity returned to normal in about 45 minutes after isoflurane anesthesia, but it took close to 6 hours for the other two drugs. Similarly, the mice showed signs of retrograde amnesia after both Keta/Xyl and MMF anesthesia. But after isoflurane anesthesia—the condition, which showed the mildest disturbances compared to natural sleep—they could still remember what they had learned before the surgery. Knowing these varying effects on the hippocampus and memory formation should be useful for doctors or experimenters when considering which method to use.
With our ability to chronically follow individual synapses with a known history both functionally and structurally over many days, we were able to investigate the structure-function relationship at single synapses over time scales that were not accessible until now. Having established an all-optical way to control synaptic plasticity allowed us to directly test the effects of plasticity on synapse survival. Usually, investigations of synaptic plasticity with single-synapse resolution are limited to only a few hours due to the invasive nature of commonly used recording techniques, such as whole-cell patch-clamp recordings or NMI-glutamate photolysis. Imaging methods that permit long-term investigations of synapses usually are limited to structural/morphological aspects, thus ignoring the functional state of the synapse. Our work showed that modifications of synaptic strength are not maintained over long time periods, but rather are converted to a "survival signal" instructing synapses whether to remain part of the circuit or to get removed.

A second highlight that emerged during the first funding phase is the development of a new optogenetic sensor to label thousands of active synapses at once, termed SynTagMA. Due to their small size and vast number, synapses are challenging to investigate in large numbers at once. This new probe overcomes limitations of commonly used technologies, such as acute two-photon calcium imaging, expanding the number of synapses accessible for analysis by orders of magnitude. Upon binding of calcium and simultaneous violet-light exposure, SynTagMA turns from green to red, allowing "snapshots" of active synaptic populations at defined time points. In addition, SynTagMA facilitates repeated labeling of active neuronal populations due to concomitant expression in the cellular nucleus, dramatically improving the time resolution compared to commonly used techniques relying on immediate-early gene expression such as cFos or Arc. In the future, SynTagMA may allow investigating the role of large synaptic populations in memory formation and storage with single-synapse resolution at high temporal precision.

The development of BiPOLES will enable many scientists to perform experiments that were not possible before. Inhibiting and activating the same neuronal populations enables them to test their sufficiency and necessity for a particular brain function. Aside from this, other manipulations become possible, such as all-optical voltage and excitability tuning of neurons either with single- or two-photon light.
Nerve cell expressing an optogenetic silencer (shown in yellow) and a morphology marker (cyan)