혼자 공부하면서 생각한 것들을 적어가는 공간입니다.
의견 주고받으면서 같이 공부하실 분은 언제든 환영입니다.
Targeted Activation of Hippocampal Place Cells Drives Memory-Guided Spatial Behavior
Nick T M Robinson, Lucie A L Descamps, Lloyd E Russell, Moritz O Buchholz, Brendan A Bicknell, Georgy K Antonov, Joanna Y N Lau, Rebecca Nutbrown, Christoph Schmidt-Hieber, Michael Häusser
PMID: 33159859 PMCID: PMC7754708 DOI: 10.1016/j.cell.2020.09.061
Abstract
The hippocampus is crucial for spatial navigation and episodic memory formation. Hippocampal place cells exhibit spatially selective activity within an environment and have been proposed to form the neural basis of a cognitive map of space that supports these mnemonic functions. However, the direct influence of place cell activity on spatial navigation behavior has not yet been demonstrated. Using an ‘all-optical’ combination of simultaneous two-photon calcium imaging and two-photon optogenetics, we identified and selectively acti- vated place cells that encoded behaviorally relevant locations in a virtual reality environment. Targeted stimu- lation of a small number of place cells was sufficient to bias the behavior of animals during a spatial memory task, providing causal evidence that hippocampal place cells actively support spatial navigation and memory.
특정 공간에서 자신의 위치 정보를 아는 것은 내비게이션 과정에서 매우 중요하며, 동물의 해마(Hippocampus)의 'Place cells'과 내 후각 피질층(Entorhinal cortex)의 'Grid cells'을 중심으로 내비게이션이 이루어진다는 다양한 연구 결과가 있다. 하지만 place cell이 동물의 내비게이션 행동에 직접적으로 어떤 영향을 미치는지에 대해서는 알려진 것이 거의 없다. Hippocampus에 대해 알려진 것들을 요약하면: 1) Memory encoding/retrieval 에 중요함 2) Place cell 존재함, 3) Place cell은 다양한 정보 - object identity, time, valence, retrospective/prospective location - 들을 갖고 있음.
그럼 극단적인 예를 들어, 누군가가 학교 강의실에 앉아 있고 그 때 특정 place cell이 반응할 때, 이 place cell은 현재 그 위치를 코딩하는지, 과거에 그 강의실에서 있었던 일을 기억하는지, 그 자리에서 내가 보고 있는 컴퓨터를 기억하는 것인지, 아니면 내가 일어나서 나갈 것이라는 정보에 반응하는 것인지 (기타 다양한 경우들이 존재할 수 있다!), 정확히 알 수가 없다. 그렇기 때문에 개인적으로 place cell이 동물의 행동에 직접적으로 어떠한 영향을 주는지에 대한 연구는 매우 흥미로운 주제라고 생각한다(아마 hippocampus 관련 논문 중 'Explicit memory creation during sleep demonstrates a causal role of place cells in navigation' 와 가장 비슷한 실험이 아닐까 싶다). 이 논문에서 보여주고자 한 내용은 Do place cells encode spatially relavent behavior information? If so, manipulation of place cells would change animal's behavior. 이라고 생각된다. 이 논문을 처음 읽고 굉장히 신기하다고 생각하였는데, 굉장히 적은 수의 place cell을 인위적으로 자극한 것만으로도 동물의 행동을 바꿀 수 있다는 결과를 보여주고 있기 때문이다.
Figure 1: Experiment setup and brief behavior/neural results
- A: Schematic of the virtual reality (VR) system
- B: An example field of view (FOV) and neurons expressing GCaMP6f and C1V1 (red-shifted channelrhodopsin)
- C: VR track configuration and each zone & Session configuration
- D: Lick rate during the Baseline & No stimulation sessions (Almost no licking behavior outside of the Reward zone)
- E: Moving speed during the Baseline & No stimulation sessions
- F: Five different example neurons (Neurons are stable)
- G: All recorded neurons during baseline block (Place fields cover whole track)
- H: Example response traces during light stimulation (during STIM session). Each neuron received two light stimuli (duration: 100 ms each).
- I: Classification of place cells based on their center-of-mass location. The numbers of place cells from each session/classification (For Start-PC vs. Rew-PC comparisons, a similar number of neurons were used)
Figure 2: Neural response during Reward-zone stimulation sessions
- A: Heat maps from the baseline & stimulation block (epoch). Most neurons maintain their original place field during the stimulation session (Only neurons that have place fields in Reward-zone are stimulated).
- B: Lick raster plots from the baseline & stimulation epochs. The mouse licked more in the stimulation zone when Rw-PCs were stimulated
- C: Averaged PSTH during the baseline epoch and the Rew-PC stimulation epoch. Why did the animals decrease their licking during the Reward zone?
- D, E: Lick rate difference between baseline and stimulation epoch (either Start-PC or Rew-PC). Animals changed their behavior only during the Rew-PC stimulation session.
- F: Control experiments (Non-PC stimulation and no stimulation)
- G: Correlation between the number of stimulated neurons vs. lick rate. (The more Rew-PCs were stimulated, the more licking behavior was observed.)
- H: Simply activating more neurons did not change behavior (Functional-type specific activation is important).
- I:
- J: Lick rate change during the Reward-zone
- K: Summary of behavior changes with two stimulation conditions (Rew-PC stimulation: more licks / Start-PC stimulation: overshoot running)
- L: During the Start-PC stimulation session, when animals ran more (running overshoot), they tended to lick less
Figure 3: Neural response during Start-zone stimulation sessions
- A: Heat maps from the baseline & stimulation epoch (the Start-PC were stimulated).
- B: Example spatial trajectory data (More incorrect trials during Start-PC epoch).
- C, D: Behavior changes induced by Start-PC or Rew-PC stimulation (Opposite changes) Why was occupancy decreased during Rew-PC stimulation?
- E, F: During the Rew-PC stimulation epoch, animals start to decrease their velocity even before the light stimulation.
- G, H, I, J: The declaration was a gradual shift to earlier points. The gradual shift is significant when compared to shuffled data.
Figure 4: Network analysis & Interaction between stimulated place cells and endogenous activity
- A: Factor analysis results and spatially tuned (distributed?) latent variables
- B: Euclidean distance between latent factors. Rew-PC sessions & Start-PC sessions have a pronounced divergence of trajectories (Directly stimulated neurons are included for this analysis).
- C: Example neurons with enhanced activity during stimulation. Probably I didn't understand it clearly. The authors wanted to investigate the effects of non-argeted neurons. But why would they separate neurons with a strongly enhanced or suppressed population?
- D: Suppressed population
- E: Spatial distribution of enhanced cells. Why did 'No Stim' have enhanced cells near the stimulated area? (Shouldn't it be more widely distributed?) In the case of suppressed cells, how come Start-PC (place field in Start-zone) could be suppressed in that area?
- F: Magnitude of enhancement
Figure 5: Stimulation-induced place field remapping
- A: Example of place cells from different sessions
- B: Pre-Post place field correlation. How come Start zone PCs from Non-PC session is not significantly different from No stim session? Why do 'Stimulation zone PCs' have higher values than 'Start zone PCs' and 'Reward zone PCs'? Why do Rew-PCs have higher values? Why does stimulating one category also affect other categories (e.g. stimulating Start-PC also alters Rew-PC)
- C, D: Place field distribution. After the stimulation, place fields shifted toward the stimulation zone.
- E, F: Place field shift (Positive value for Rew-PCs represent moving towards the center of the trace. Opposite for the Start-PCs). What is the interpretation? After the stimulation, place fields shifted toward the center
- G, H: Lick rate change
- I: Correlation between lick rate change and place field shift (mean COM of each Session, since this should have only one value per Session).
Factor analysis: a way to fit a model to multivariate date to estimate interdependence. In the factor analysis model, the measured variables depend on a smaller number of unobserved (latent) factors (MathWorks). Because each factor may affect several variables in common, they are known as "common factors". Each variable is assumed to depend on a linear combination of the common factors, and the coefficients are known as loadings.
- Here, I assume place fields are variables, and 10-fold latent factors were calculated from each FOV.
Latent variable: In statistics, latent variables are variables that are not directly observed but are rather inferred from other variables that are observed (Wikipedia).
