Visual influences on ferret auditory cortex

Abstract

Multisensory neurons are now known to be widespread in low-level regions of the cortex usually thought of as being responsible for modality-specific processing. The auditory cortex provides a particularly striking example of this, exhibiting responses to both visual and somatosensory stimulation. Single-neuron recording studies in ferrets have shown that each of auditory fields that have been characterized using physiological and anatomical criteria also receives visual inputs, with the incidence of visually-sensitive neurons ranging from 15% to 20% in the primary areas to around 50% or more in higher-level areas. Although some neurons exhibit spiking responses to visual stimulation, these inputs often have subthreshold influences that modulate the responses of the cortical neurons to sound. Insights into the possible role played by the visual inputs can be obtained by examining their sources of origin and the way in which they alter the processing capabilities of neurons in the auditory cortex. These studies suggest that one of the functions of the visual input to auditory cortex is to sharpen the relatively imprecise spatial coding typically found there. Because the extent to which this happens varies between cortical fields, the investigation of multisensory interactions can also help in understanding their relative contributions to auditory perception.

Fig. 1

Organization of ferret auditory cortex. Auditory cortex in the ferret is located on the ectosylvian gyrus (EG), which is conventionally divided into the middle, anterior and posterior (MEG, AEG and PEG) regions. The inset shows the location of each of the auditory fields on the EG superimposed upon their sound frequency organization, as revealed using intrinsic optical imaging in a single, typical animal (from Nelken et al., 2004). A1, primary auditory cortex; AAF, anterior auditory field; PSF, posterior suprasylvian field; PPF, posterior pseudosylvian field; VP, ventral posterior field; ADF, anterior dorsal field; AVF, anterior ventral field; fAES, anterior ectosylvian sulcal field; sss, suprasylvian sulcus; pss, pseudosylvian sulcus.

Fig. 2

Distribution of visual sensitivity in ferret auditory cortex. (A) Bar graph showing the relative numbers of unisensory auditory (white), unisensory visual (black) and bisensory (gray) neurons recorded in each cortical field. The actual number of neurons recorded in each field is given at the top of the columns. (B) Five example penetrations (whose locations are indicated by the numbers in C). Each panel represents a single penetration made with a linear array electrode. Frequency–response areas are plotted for all sites at which units were significantly driven by pure tone stimuli. Sites in which there were significant multisensory or visual activity are outlined in red or green, respectively. (C) Distribution of response types recorded in the auditory cortex of one representative ferret. The borders between the different cortical fields are indicated by the dashed lines and were estimated from the frequency–response properties of the acoustically-responsive neurons. The locations of auditory (blue dots), visual (green triangles) and bisensory (red diamonds) neurons are plotted on the surface of auditory cortex. Note that bisensory neurons include those in which both visual and auditory stimuli produced a significant response, as well as neurons in which one modality of stimulation significantly modulated the response to the other stimulus. Based on Bizley et al. (2007) and Bizley and King (2008).

Fig. 3

Summary of inputs from visual cortex to auditory cortex. The lateral view of the ferret brain shows the main source of visual cortical input for each region of the auditory cortex. Visual areas 17 and 18 innervate the auditory core areas (blue), visual area SSY and the PPc principally innervate the anterior belt and para-belt areas (green), while visual areas 20a and 20b project to the posterior auditory belt areas (red). A summary of the cortical multisensory and visual inputs to each auditory field is provided in the lower panel. There are also very sparse inputs from visual areas 19 and 21 to various auditory areas. PPr, rostral posterior parietal cortex; PPc, caudal posterior parietal cortex; SSY, suprasylvian cortex; 3b, primary somatosensory cortex; S2, secondary somatosensory cortex; S3, tertiary somatosensory cortex; D, dorsal; R, rostral. Other abbreviations as in Fig. 1. Scale bar is 1 mm. Based on Bizley et al. (2007).

Fig. 4

Raster plots and spatial receptive fields (based on spike counts) for three example neurons. (A) A unisensory visual neuron whose response was unaffected by simultaneous auditory stimulation. (B) A neuron whose auditory spatial receptive field was sharpened by simultaneous visual stimulation. (C) A neuron in which the bisensory response was less spatially sensitive than either unisensory response. (D) Scatter plot showing the mutual information (in bits) transmitted about the location of the stimulus for unisensory visual stimulation (red) and unisensory auditory stimulation (blue) against that obtained with bisensory stimulation. In the case of the blue crosses, most of the points lie above the x = y line, indicating that bisensory stimulation increased the spatial information in the response relative to that produced by unisensory auditory stimulation. Based on Bizley and King (2008).

Fig. 5

(A–E) Population azimuth–response plots (based on spike counts) for each of five auditory cortical fields (A1, AAF, PPF, PSF and ADF). The colour scale indicates the normalized spike rate plotted at different azimuthal angles as a function of time (each stimulus came on at time 0 for 100 ms). The first column shows the population response to auditory stimulation, the second to visual stimulation and the third to combined visual–auditory stimulation, while the normalized spike rates for each are combined in the azimuth–response profiles in the last column. (F) Box-plot showing the MI transmitted by neurons in each of these five cortical areas about the location of spatially and temporally coincident bisensory stimulation. The spatial MI values obtained for ADF were significantly higher than all the other cortical areas. (G) The proportion of neurons in each cortical area for which the responses to bisensory stimulation conveyed more information about stimulus location than the response to sound alone. Based on Bizley and King (2008).