Periodic Reporting for period 4 - HIRESMEMMANIP (Spiking network mechanisms underlying short term memory)
Reporting period: 2021-04-01 to 2022-09-30
Spiking patterns including persistent activity and multi-neuronal sequences are regularly observed during short term memory (STM). Yet it is unknown whether the patterns are required for memory maintenance, an outcome of it, or an epiphenomenon.
Understanding the spiking network mechanisms underlying STM will solve a long-standing question, and may allow interfacing the neuronal system at its own spatiotemporal resolution, supplementing lost functions.
Project objectives thus included (1) development of STM tasks that are suitable for intact and unrestrained lab animals (freely-moving mice). (2) determination of spike timing that is required for sustaining STM. (3) supplementing spontaneous activity during memory maintenance by synthetic spike patterns.
Understanding the spiking network mechanisms underlying STM will solve a long-standing question, and may allow interfacing the neuronal system at its own spatiotemporal resolution, supplementing lost functions.
Project objectives thus included (1) development of STM tasks that are suitable for intact and unrestrained lab animals (freely-moving mice). (2) determination of spike timing that is required for sustaining STM. (3) supplementing spontaneous activity during memory maintenance by synthetic spike patterns.
We trained freely-moving mice on several novel STM tasks that employ auditory, whisking, visual, or intra-cortical cues. A necessary condition for success was a previously undescribed hybrid strain of mice. Distinct tasks were developed for working memory and for priming, and the hybrid mice achieved stable performance on all tasks. For the first time, a clear priming effect was observed in mice.
We developed new hardware for multi-neuronal bidirectional control: optics (0.5 g dual-color laser assembly) and CMOS chips (2 x 0.7 mm 32-channel current sources). Using the technology, we generated arbitrary multi-neuronal spike patterns deep in the brain of freely-moving mice.
We discovered a new putative phase code for spatial position of the animal termed theta phase rolling.
We discovered that information transmission between a group of PYR and a single postsynaptic INT increases information and improves precision. Transmission increases reliability timescale approximately tenfold, converting a rate code into a temporal code.
Finally, we developed a paradigm to modulate reference memory, associating injected spiking patterns with behavior. The mice learn to associate a new rule with behavioral choice in a single session.
We developed new hardware for multi-neuronal bidirectional control: optics (0.5 g dual-color laser assembly) and CMOS chips (2 x 0.7 mm 32-channel current sources). Using the technology, we generated arbitrary multi-neuronal spike patterns deep in the brain of freely-moving mice.
We discovered a new putative phase code for spatial position of the animal termed theta phase rolling.
We discovered that information transmission between a group of PYR and a single postsynaptic INT increases information and improves precision. Transmission increases reliability timescale approximately tenfold, converting a rate code into a temporal code.
Finally, we developed a paradigm to modulate reference memory, associating injected spiking patterns with behavior. The mice learn to associate a new rule with behavioral choice in a single session.
The novel behavioral, biological, electronic, mathematical, optical, and signal processing tools developed in the project will facilitate research of multiple questions in systems neuroscience.
Some of the discoveries made have implications beyond understanding the spiking mechanisms underlying STM. Phase rolling may encode various types of information, and error correction and temporal coding during synaptic transmission are likely to be general properties of converging neuronal networks.
Some of the discoveries made have implications beyond understanding the spiking mechanisms underlying STM. Phase rolling may encode various types of information, and error correction and temporal coding during synaptic transmission are likely to be general properties of converging neuronal networks.