Scalable time representation of CA1 neurons in the temporal bisection task. (A) Schematic of the task (task 1). The rats ran on the treadmill for long or short intervals and then selected the left or right arm associated with water reward for long or short intervals, respectively. Sets of intervals were 10 s (long) and 5 s (short) in blocks 1 and 3 and 20 s (long) and 10 s (short) in block 2. (B, a to c) Firing patterns of three representative time cells in blocks 1 (left), 2 (center), and 3 (right). Top: Raster plots of long- and short-interval trials. Bottom: PETHs of long (solid)– and short (dotted line)–interval trials. (C) Firing patterns of time cells in blocks 1 (left), 2 (center), and 3 (right) (n = 454 units from seven rats). Each row represents PETHs of single neurons during long-interval trials in each block. The color scale represents the firing rate of each neuron. The neurons were ordered according to the peak time in block 1. (D) Population vector analyses of the data (C). Left: Autocorrelation of the population vector matrix of block 1. Center and right: Cross-correlations of blocks 1 and 2 and blocks 1 and 3, respectively. The color scale represents Pearson’s correlation coefficient. (E) Distributions of the scaling factors of time cells between blocks 1 and 2 (blue) and blocks 1 and 3 (green). Credit: Science Advances, doi:10.1126/sciadv.abd7013


Scalable representation of time in the hippocampus


Leave A Reply