Adaptive Whisking in Mice

Abstract

Rodents generate rhythmic whisking movements to explore their environment. Whisking trajectories, for one, appear as a fixed pattern of whisk cycles at 5–10 Hz driven by a brain stem central pattern generator. In contrast, whisking behavior is thought to be versatile and adaptable to behavioral goals. To begin to systematically investigate such behavioral adaptation, we established a whisking task, in which mice altered the trajectories of whisking in a goal-oriented fashion to gain rewards. Mice were trained to set the whisker to a defined starting position and generate a protraction movement across a virtual target (no touch-related tactile feedback). By ramping up target distance based on reward history, we observed that mice are able to generate highly specific whisking patterns suited to keep reward probability constant. On a sensorimotor level, the behavioral adaptation was realized by adjusting whisker kinematics: more distant locations were targeted using higher velocities (i.e., pointing to longer force generation), rather than by generating higher acceleration (i.e., pointing to stronger forces). We tested the suitability of the paradigm of tracking subtle alteration in whisking motor commands using small lesions in the rhythmic whisking subfield (RW) of the whisking-related primary motor cortex. Small contralateral RW lesions generated the deterioration of whisking kinematics with a latency of 12 days post-lesion, a change that was readily discriminated from changes in the behavioral adaptation by the paradigm.

Activity in Barrel Cortex Related to Trace Eyeblink Conditioning

“…The second approach dropped all assumptions about the shape of the learning curve and estimated change points of learning scores in similar ways as done previously ( Gallistel et al, 2004 ). Figure 2 C shows the CUSUM ( http://www.variation.com/cpa/tech/changepoint.html ; Durstewitz et al, 2010 ; Chakrabarti et al, 2022 ; see above, Materials and Methods), which is the progression of the cumulative distance between the series of learning scores and their mean. The negative peaks between two points at which the CUSUM curve departed from the bootstrapped population allowed us to identify significant change points.…”

“…To this end we calculated the cumulative sum chart (CUSUM) curve, which is the cumulative difference between a series of data points and their mean [ with the time series, , indices of time series, and the mean of the time series; Fig. 2 C ; http://www.variation.com/cpa/tech/changepoint.html ; Durstewitz et al, 2010 ; Chakrabarti et al, 2022 ]. The negative peaks of CUSUM curves ( Fig.…”

Discerning state estimation and sensory gating, two presumptive predictive signals in mouse barrel cortex

“…3A-C). A brain-machine interface was established based on a real-time state engine to monitor 7N activity with respect to baseline and threshold of neuronal multi-unit activity at a temporal resolution of 0.1 ms. To enable a trial, the mice had to keep 7N activity below a baseline for a minimum interval of 500 ms, upon which the light would go off to signal to the mice that the trial start is now enabled 22 . After the animals initiated the motor command, the trial started and a sensory consequence would be presented with a pre-defined delay with respect to its onset.…”

“…The experiments of the present study are based on an established whisker reach task in head-fixed mice 22 . Before behavioral training, the mice were implanted with a single recording electrode in facial nucleus (7N) at a site where microstimulation evoked whisker movements, and either a 2x2 mobile wire-electrode-array (7 mice), or a 16-electrode linear silicon-electrode-array (5 mice), in the C2 barrel column.…”

“…After complete recovery from all surgical interventions, and habituation to head-fixation 40 , the mice were trained on a whisker reach task with the same equipment and exactly following the strategy as reported before 22 . After the behavioral training, mice consistently and precisely performed reaching movements (no touch) of their left whisker, starting from a predefined baseline location or behind (located just rostral to their whisker resting position), and surpassing a threshold location more rostral.…”