Figure 2.

The permutation-based logic in the basolateral amygdala (BLA) cell assemblies for processing food experiences. (A) Examples of pyramidal cells with specific to general responsiveness. The top row shows a cell selectively responded to milk; A representative two-event cell responding to rodent diet biscuit (RD) and milk (the second row); A three-event cell responding to biscuit, sugar and milk (the third row); A pyramidal cell responding to all four types of foods (bottom row). Peri-event raster (upper subgraph) and peri-event histograms (lower subgraph) are presented. *P < 0.05, **P < 0.01, ***P < 0.001, Wilcoxon signed-rank test. (B) BLA pyramidal cells followed the power-of-two-based permutation rule to process appetitive information (n = 9 mice, 221 responsive pyramidal cells). As i was increased from 1, 2, 3, to 4, hierarchical clustering plots show the emergence of distinct neural cliques from 1, 3, 7, to 15. The Y-axis lists the number of responsive putative pyramidal cells. Color scale bars indicate the logarithm-transformed responsiveness of putative pyramidal units (10-s post-stimulation) averaged over seven trials. It typically took 5~10 s for mice to consume each pellet or milk. (C) The specific-to-general assembly logic was also present in the simultaneously recorded BLA pyramidal cells from a single mouse (50 out of 110 pyramidal cells were responsive to foods) in the very first recording session. In this mouse, the food order was presented as follows: seven rodent biscuits, seven sugar pellets, seven rice, followed by seven milk droplets. (D) Z-score plots show population responses of each of the 15 distinct neural cliques in the BLA to various food experiences. The number of pyramidal cells belonging to each clique is listed in each subplot. Color lines (blue, orange, green, purple) indicate a given food type.