Introduction

Fragile X Syndrome (FXS) is a leading inheritable cause of mental impairment, affecting approximately 1∶4000 males and 1∶8000 females in the United States [1]. FXS arises from a loss-of-function in the FMR1 gene that encodes the Fragile X Mental Retardation protein (fmrp), an RNA binding protein [2], [3]. Although the symptoms of FXS patients vary in severity and expression, characteristic phenotypes include reduced intellectual abilities, hyperactivity, increased seizure susceptibility, and impaired visuo-spatial processing [4]. Hypersensitivity to sensory stimuli, including tactile defensiveness, are also principal symptoms of FXS [5], [6]. At present, there is no known cure for FXS or treatment than reverses the collective pathology.

This Fmr1 mutant mouse exhibits several phenotypes similar to Fragile X syndrome including increased seizure susceptibility and hyperactivity, as well as deficits in spatial and motor learning [7]–[11]. Fmr1 KO mice display abnormal sensory gating during prepulse inhibition as well as cortical hyperexcitability [12]–[14]. However, whether Fmr1 KO mice exhibit differences in somatosensory responsiveness akin to the characteristic tactile hypersensitivity to sensory stimuli in FXS is unknown. In addition, whether Fmr1 KO possess normal tactile-dependent learning is unclear.

The gap crossing (GC) learning paradigm is a prime example of a distance detection/object localization task [15]–[17]. In this task, mice are placed on an elevated starting (‘home’) platform in a light-tight enclosure. Mice rely on their whiskers to explore the dark environment and locate a ‘target’ platform. A computer-controlled robotic system adjusts the distances between the two platforms. At short distances, mice perform the task by detecting the target platform with their whiskers and nose, activating whiskers as well as touch receptors in the skin around the nose. At longer distances mice depend exclusively on their whiskers for tactile information [18]. Successful task acquisition requires intact somatosensory ‘barrel’ cortex. Mice improve their performance on this task with experience; this learning yields a greater percentage of successful crossings of a given distance in successive sessions of trials.

Here we examined the cortical responses to whisker stimulation and the rate of tactile learning for Fmr1 mutants and wild-type (WT) mice. Fmr1 KO mice displayed greater cortical responses to whisker stimulation, normal initial performance on the gap cross task, but significantly lower improvement with experience at whisker-dependent distances. Thus, we conclude that disruption of Fmr1 gene function alters cortical responses to sensory stimuli and perturbs tactile learning.

Materials and Methods

Results

To explore the cortical representations of tactile stimuli by Fmr1 KO and WT mice, we employed optical imaging of intrinsic signals (OIS) to measure cortical responses to whisker stimulation. OIS is a non-invasive measure of tissue reflectance correlated with neural activity [21]–[23]. To measure neural responses to whisker deflection in somatosensory barrel cortex, we adapted a method for measuring intrinsic signals developed to study visual cortex [19] (Fig. 1A, B). We examined the cortical response to stimulation of the C2 whisker for both Fmr1 mutants and WT mice (Fig. 1C). Fmr1 KO mice displayed significantly greater size of response (pixels) relative to WT mice across magnitude of response thresholds (Fig. 1D) (Genotype × OIS magnitude threshold: F(1,54) = 6.851, p = .011, KO n = 10, WT n = 10; two-way ANOVA). The size of the region of response in WT mice was approximately the size of barrels observed in slice preparation of mouse cerebral cortex [24]. Thus, Fmr1 mutant mice exhibit abnormally large responses to whisker stimulation in barrel cortex despite a normal cytoarchitecture [25].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. Fmr1 KO mice exhibit increased evoked activity in primary somatosensory cortex during whisker stimulation.

(A) Schematic showing experimental set up of intrinsic optical imaging over primary somatosensory cortex (black circle) during periodic whisker stimulation. (B) Pictures of the thin skull preparation and example images collected from a wild-type (WT) mouse (left) and Fmr1 KO mouse (right) mouse. Scale bar = 0.4 mm. Rostral (R), Caudal (C), Lateral (L) and Medial (M) coordinates are shown. (C) Representative examples of data collected during a typical imaging session. Above, a time series of pixel values for the cortical location indicated by the asterisk in the Fmr1 KO in panel B. Below, a fast-fourier transform (FFT) of the raw trace extracts the magnitude of the change in reflectance (ΔR/R) corresponding to the frequency of whisker stimulation (red square). (D) The number of pixels within the region of response with ΔR/R magnitudes greater than the threshold indicated on the abscissa for WT (n = 10) and Fmr1 KO (n = 10) mice. The response to whisker stimulation is elevated in Fmr1 KO mice (WT vs. KO, p = .011; 2-way ANOVA).

https://doi.org/10.1371/journal.pone.0109116.g001

To determine if Fmr1 KO mice display associated deficits in whisker-dependent learning, we examined the performance of Fmr1 KO and WT mice on a validated test of tactile learning, the gap cross task (Figure 2). These were separate cohorts of mice from those examined in OIS experiments (Figure 1). Successful performance on this task requires both tactile stimulation of the whiskers and intact somatosensory cortex [16]–[18]. Following one day of habituation training, all mice were tested on the gap cross at a range of distances spanning 3.0 to 6.0 cm.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. The Gap Cross task is a whisker-dependent sensory learning paradigm.

(A) Schematic of the gap cross learning task. Motion sensors positioned at four points along the 2 platforms (labeled #1–4) track the mouse as it moves from the starting platform across a given gap distance to the target platform. (B) Activation of each sensor (grey box) indicates the position of the mouse. (C) Successful crosses are defined as the movement of the mouse from the starting platform to the target platform (green circles). Failures are defined as trials in which the mouse approaches the edge of the home or target platform and returns to the back of the home platform (red crosses).

https://doi.org/10.1371/journal.pone.0109116.g002

To assess improvement in performance with experience, we compared the percent of successful crossing between the first six sessions (1–6) and the subsequent six sessions (7–12) at all gap distances tested for WT and Fmr1 KO mice (Figure 3). Both genotypes displayed a similar high percentage of successful crosses at distances less than 4.5 cm (Fig. 3A, B). At these shorter ‘nose-distances’, mice are able to detect the target platform by touching it with their nose as well as whiskers. WT and Fmr1 KO mice displayed similar improvement with experience at these ‘nose distances’ (Fig. 3A, B, D). The percent of successful crosses increased significantly in the second half of sessions (sessions 1–6 vs. sessions 7–12) (s1–6 vs. s7–12 X ‘nose’ gap distance, WT, F(1,15) = 9.575, p = .007; n = 6; KO, F(1,24) = 23.41, p<.0001, n = 9; WT; two-way RM-ANOVA) (Fig. 3A, B). Both WT and Fmr1 KO mice improved an average of more than 15% at these ‘nose’ distances (Fig. 3D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. Fmr1 KO mice display normal learning on the gap cross assay at shorter gap distances but impaired learning at longer whisker-dependent distances.

(A) The percent successful crosses averaged across the first six sessions and subsequent six sessions across gap distances ranging from 3.0 cm to 6.0 cm for both wild-type mice (black lines, n = 6) and Fmr1 KO mice (blue lines, n = 9). For each distance, the line marker on the left is the average success rate of the first six sessions and the connected line marker on the right is the average success rate of the subsequent six sessions. Error bars represent standard error of the mean. (B) At shorter ‘nose’ distances, both wild-type (WT) and Fmr1 KO mice (KO) improve to a greater percentage of successful crosses between the average of the first six sessions (WT, grey, KO light blue) and the last six sessions (WT, black, KO dark blue). This improvement is statistically significant (WT, p = .007; n = 6; KO, p<.001, n = 9; WT; two-way ANOVA) (C) At whisker-dependent distances, wild-type (WT) improve between early sessions (grey line) and subsequent sessions (black line) despite the lower overall success rate at increasing gap distances. However, KO mice do not display significant improvement between early sessions (light blue line) and later sessions (dark blue line) (WT, p = .002, n = 6, KO, p = .14; n = 9, two-way ANOVA). (D) Average improvement for WT and KO mice at shorter ‘nose’ distances and longer ‘whisker’ distances. WT mice display significantly greater improvement at whisker-dependent distances that KO mice (p = .02, two-tailed t-test).

https://doi.org/10.1371/journal.pone.0109116.g003

In contrast to shorter distances, at longer distances (5.0, 5.5, 6.0) mice rely exclusively on information from their whiskers to detect the target platform. Overall, performance declines with increasing gap distance (Fig. 3A). At these longer distances WT mice improved significantly with experience (s1–6 vs. s7–12 X ‘whisker’ gap distance, WT, F(1,15) = 13.60, p = .002, n = 6, two-way RM-ANOVA) (Fig. 3C) and the percentage improvement was similar in magnitude to that at ‘nose distances’ (WT, p>.9; n = 6, two-tailed t-test). Interestingly, Fmr1 KO mice did not display significant improvement at these whisker-dependent gap distances (s1–6 vs. s7–12 X ‘whisker’ gap distance, KO, F(1,24) = 2.338, p = .14; n = 9, two-way RM-ANOVA) (Fig. 3C). The improvement of Fmr1 KO mice at ‘whisker’ distances was significantly less than that of WT controls (p = .02, two-tailed t-test) (Fig. 3D). Similarities between the two groups in the total number of trials as well as the number of successful crossings for all distances suggest that these deficits were not due to differences in mobility, exploration or motivation on the task (total number of successful crossings Fmr1 KO = 195, SEM = +/−7, WT = 183 SEM = +/−9, total number of trials Fmr1 KO = 366, SEM = +/−8, WT = 369 SEM = +/−24). Thus, Fmr1 KO mice exhibit a deficit in tactile learning that correlates with abnormal cortical sensory representation of whiskers in barrel cortex.

Discussion

FXS is characterized by cognitive impairment, anxiety, developmental delay, increased seizure susceptibility, and behavioral hyper-excitability. Altered sensory responses and reduced sensory integration are also symptoms of FXS. In particular, disrupted somatosensory processing, which is often described as tactile defensiveness, is well documented in patients with FXS [5], [26], [27].

FMR1 KO mice manifest several phenotypes similar to symptoms of FXS [7], [28] These mice have proven to be a useful preclinical model for studying both the molecular functions of FMRP and the morphological and functional disruptions associated with its loss [29]. Fmr1 KO mice display alterations in synaptic function and plasticity [4], [30]–[36], as well as heightened excitability in barrel cortex that may result from dysfunction within inhibitory circuitry [37]–[39]. In addition, they exhibit increased duration and incidence of network excitation, including increased synchrony of network activity and increased interconnectivity between layer V pyramidal neurons [40]–[42]. Despite both the clinical relevance of sensory dysfunction in Fragile X patients, as well as an increased understanding of the altered network state in FMR1 KO mice, it remains unclear whether cortical responses to tactile stimulation are aberrant in the somatosensory cortex of Fmr1 KO mice.

We examined tactile learning by comparing the performance of Fmr1 KO and WT mice on a whisker-dependent sensory learning paradigm, the gap cross task. Fmr1 KO mice displayed initial performance on this task indistinguishable from WT mice and normal learning at shorter ‘nose’ gap distances where sensory information regarding the position of the target platform was presumably more definitive. However, Fmr1 KO mice exhibited significantly less improvement in performance at whisker-dependent distances. Thus, aberrant cortical responses to whisker stimulation we observed in Fmr1 KO correlate with a deficit in a whisker-dependent and barrel cortex-dependent tactile learning task. However, as OIS is an indirect measure of neural activity, we cannot exclude the possibility that either differences in coupling of neural activity to factors mediating the change in tissue reflectance or different effects of isoflurane on WT and Fmr1 KO mice could contribute to the differences in cortical responses to whisker stimulation.

However, this perceptual learning task is not exclusively reliant on cerebral cortex or the somatosensory system. Distance detection and object localization integrates motor and sensory activity of both subcortical and cortical circuitry [43], [44]. Thus, whether the exaggerated whisker representations in barrel cortex contribute to the deficits in tactile learning in Fmr1 KO mice or whether this learning impairment results from aberrant neural circuitry elsewhere in brain is unclear at present. For example, this learning deficit could result from a dysfunction in motor cortex. We considered whether the increased anxiety-like behaviors exhibited by Fmr1 KO mice might contribute to their deficits in the gap cross assay [9], [45]. However, as the initial performance and improvement of Fmr1 KO mice at shorter ‘nose’ distances was similar to WT mice, as was their initial performance at longer ‘whisker’ distances, we propose that the anxiety-like behaviors of Fmr1 KO mice are unlikely to be a major contributor to the learning deficit specific to whisker-dependent gap distances we observe. Future studies will be required both to discriminate between these possibilities as well as ascertain if performance by Fmr1 KO mice on the gap cross assay is improved by administration of mGluR5 antagonists or GABA_B agonists reported to rescue some aspects of the phenotype Fmr1 KO mice [46], [47].