Cortical Cholinergic Input Is Required for Normal Auditory Perception and Experience-Dependent Plasticity in Adult Ferrets

Effect of cortical ACh depletion on auditory localization performance. A, Approach-to-target localization responses at each stimulus duration. Correct responses are shown in light gray, and incorrect responses are further subdivided into front–back errors (dark gray), left–right errors (black), and unclassified errors (mid-gray). B, Polar plot illustrating the proportion of correct responses (radial axis) at each peripheral speaker position for the two experimental groups when tested with 40 ms duration stimuli. Control animals are shown in black and animals with ME20.4-SAP cholinergic lesions in purple. C, Polar plot showing the mean error magnitude (radial axis, degrees) for each speaker position for a stimulus duration of 40 ms. D, Head orienting responses obtained in the same trials. Mean final head bearing is plotted against stimulus direction for each experimental group. The shaded region represents 1 SD about the mean for the control animals. The broadly sigmoidal dependence of head bearing on stimulus location was conserved across the two experimental groups.

Stimulus duration-dependent effects of cholinergic modulation on localization performance. A, B, Mean ± SD proportion of correct responses and front–back errors, respectively, across stimulus duration for control animals (black) and animals with bilateral cholinergic deafferentation (purple). C, Response time between stimulus onset and the animal licking a peripheral reward spout, fitted using an ex-Gaussian distribution and bootstrapped 100 times, shown relative to stimulus duration for each experimental group. Solid lines represent the response times across all trials and dashed lines for correct trials only. The parameter μ represents the mean ± SD of the normally distributed response times. D, Latency of the initial head movement after sound onset (reaction time) for each experimental group across stimulus duration. The distribution of these reaction times was also described with a high degree of accuracy using an ex-Gaussian distribution, and the value plotted, μ, represents the mean ± SD of the normally distributed head movement latencies.

A, Normalized histogram of response times in the approach-to-target localization task for ferrets with immunotoxin lesions of the NB. These values, which are binned at 100 ms, indicate the time between initiating a trial and licking a peripheral reward spout. An ex-Gaussian distribution was used to model these data in MATLAB using the DISTRIB toolbox available from http://darwin.psy.ulaval.ca/∼yves/distrib.html. The purple line represents the best-fit, ex-Gaussian model. The mean (μ) and SD (σ) of the Gaussian component of this function are represented by the purple circle and bounding lines, respectively. Mean response time (red circle) and SD (red line) further illustrate the non-normality of this distribution. B, Ex-Gaussian best fit of response times measured for a stimulus duration of 2000 ms in the approach-to-target localization task for ferrets with immunotoxin lesions of the NB (purple) and control animals (black), highlighting the delayed peripheral response times for lesioned ferrets at this stimulus duration.

Effect of cortical ACh depletion on behavioral adaptation to a unilateral earplug. A, The proportion of correct responses, averaged over all speaker positions, is plotted for individual animals before earplug insertion (Pre-plug), during training with the earplug in place (Days 1–10), and after earplug removal (Post-plug). Each animal is represented by a different symbol. The mean values for each experimental group are indicated by the solid lines, and the shaded regions show 1 SD about the mean. B, C, Scatter plots showing the calculated linear regressions between the proportion of correct responses and day of wearing an earplug to model the rate of adaptation (slope) to an earplug for the controls (B) and animals with bilateral cholinergic lesions (C). Symbols represent the data from individual animals, solid lines are the best linear fits, and dashed lines show the 95% confidence intervals of these fits.

Changes in response bias and gain during adaptation to a unilateral earplug. A, Confusion matrices of approach-to-target responses for animals with bilateral cholinergic deafferentation and controls. Stimulus and response locations have been transformed in azimuthal space so that the posterior and anterior portions of the circular response space are “folded” onto each other, effectively turning all errors into lateral errors. Regression lines indicate the best linear fit of the stimulus–response function before plug insertion (Pre-plug), at two time points during training with the earplug in place (Plug: Day 1–2, Plug: Day 9–10), and after plug removal (Post-plug), for animals with bilateral cholinergic lesions (purple) and controls (black). Perfect performance equates to a slope of one and a y-intercept of zero. Deviations from these values indicate impaired localization performance. Note that the responses become biased toward the non-occluded ear when the earplug is first inserted. B, Plots illustrating changes in response bias (y-intercept) and response gain (slope), calculated from the stimulus–response linear regressions, for individual animals in both groups. Shaded areas indicate where the left ear was plugged.

Effects of a unilateral earplug on the sound-evoked head orienting errors (stimulus location minus final head bearing) made by control ferrets (left column) and animals with bilateral cholinergic lesions of the NB (right column). Each panel shows data from individual trials together with the fitted quadratic polynomial curves calculated using a robust least-squares method (bisquare weights). Shaded regions indicate the 95% confidence intervals in each case. A, B, Distribution of the errors before insertion of the earplug (Pre-plug) and after its removal (Post-plug) for both groups. C, D, Head orienting errors made during the first 2 d of wearing an earplug (Initial plug). The pre-plug curve fit and 95% confidence intervals are also shown for comparison. Note the greater variability in the size of the errors in the animals with cholinergic lesions (D) than in the controls (C). E, F, Head orienting errors made during the last 2 d (9 and10) of wearing an earplug. The pre-plug curve fit and 95% confidence intervals are again shown for comparison.

Cholinergic inputs from the NB to the auditory cortex. A, B, Coronal sections taken at the level of the NB immunostained to visualize p75^NTR cells in a control animal (A) and an animal in which bilateral ME20.4-SAP injections had been made in the NB (B). Insets show a higher-magnification view of the most medial region of the NB, in which the cells are most densely packed. C, D, Photographs taken at low magnification from coronal sections in the same animals stained to visualize AChE fibers in the ectosylvian gyrus, in which the auditory cortex is located. The regions indicated by the white rectangles in each of these panels are shown at higher magnification in E and F, respectively, and illustrate the density of AChE fibers within the supragranular layers of the auditory cortex. All four pictures were taken with exactly the same light and filter conditions at the microscope, and no additional computer manipulations of contrast or brightness were performed. Scale bars: A, B, 200 μm; insets, 50 μm; C, D, 1 mm; E, F, 100 μm. Cl, Claustrum; D, dorsal; EG, ectosylvian gyrus; ic, internal capsule; L, lateral; LV, lateral ventricle; pss, pseudosylvian sulcus; sss, suprasylvian sulcus.

Distribution and morphology of neurons in the ferret NB that express p75^NTR (left column) or PV (right column). A, B, Pictures taken from coronal sections at the level of the NB in the right hemisphere showing the distribution of p75^NTR (A) and PV (B) IR neurons. C, D, Drawings of the same sections depicting the limits of the NB and the location of p75^NTR (C) and PV (D) IR neurons. E, F, Higher-magnification images of p75^NTR (E) and PV (F) positive neurons, with the squares in A and B representing the respective regions where these pictures were taken. Cl, Claustrum; D, dorsal; ic, internal capsule; M, medial; PEG, posterior ectosylvian gyrus. Scale bars: A–D, 0.5 mm; E, F, 50 μm.

Bilateral ME20.4-SAP injections in the NB result in a local loss of p75^NTR-IR cells in that nucleus without affecting other cholinergic groups in the basal forebrain. These pictures were taken from coronal sections at the level of the NB (A, B) and medial septum (C, D). A and C are from a control animal, and B and D are from an animal in which ME20.4-SAP injections had been made in the NB. Scale bars, 0.5 mm. Cl, Claustrum; ic, internal capsule; D, dorsal; M, medial; PEG, posterior ectosylvian gyrus.

Effect of degree of cholinergic deafferentation on auditory localization performance. Proportion of correct responses (averaged over all loudspeaker locations) at five different stimulus durations for all animals with bilateral ME20.4-SAP immunotoxin injections in the NB, ordered by the percentage of loss of p75^NTR-IR neurons in the NB relative to control animals (see Table 2). Increasing cholinergic deafferentation led to progressively greater performance deficits, particularly at shorter stimulus durations. The degree of performance impairment positively correlated with the extent of cholinergic cell loss in the NB. Values plotted are means ± SD.