Introduction

Although the experience of our different sensory modalities is unique, the brain integrates information across modalities to form a coherent picture of the environment, enhancing our existence and survival. Numerous perceptual studies in humans have demonstrated that stimulation in one sensory modality can alter the response to stimulation in another sensory modality (Welch & Warren, 1986). However, noninvasive evoked potential recording is one of the few means available to study the neurophysiology of human multisensory integration.

Original evoked potential work by Walter demonstrated that large regions of the human cortex are responsive to multiple types of sensory stimulation (Walter, 1965). In addition, it was shown that the simultaneous presentation of stimuli in two different sensory modalities produces significantly larger evoked responses than unimodal stimulation, concurrent with a reduced reaction time on behavioral tests (Andreassi & Greco, 1975; Cigánek, 1966; Hershenson, 1962; Morrell, 1968; Nickerson, 1973). Currently, event-related potentials are being used to study features of multisensory integration, confirming and extending the earlier results (Aunon & Keirn, 1990; Costin, Neville, et al., 1991; Fort, Delpuech, et al., 2002a, 2002b; Foxe, Morocz, et al., 2000; Foxe, Wylie, et al., 2002; Giard & Peronnet, 1999; Hansen & Hillyard, 1983; Kenemans, Kok, et al., 1993; Molholm, Ritter, et al., 2002; Naatanen, Paavilainen, et al., 1993). Yet the interpretation of human multisensory-evoked potential studies is limited by a lack of basic understanding about the neurogenesis of these responses. Thus, there is a need to conduct invasive multisensory-evoked potential studies in animals to provide an essential bridge to understanding the physiological basis of similar phenomena recorded noninvasively in the extracranial human response.

Early evoked potential mapping studies in animals began this task by systematically mapping evoked responses from numerous cortical locations and defining putative multisensory cortex, typically corresponding to regions where unimodal responses converged or overlapped (Berman, 1961a, 1961b; Bignall & Imbert, 1969; Bignall & Singer, 1967; Thompson, Johnson, et al., 1963; Thompson, Smith, et al., 1963; Woolsey, 1967; Woolsey & Fairman, 1946). With the introduction of the microelectrode, multisensory electrophysiology became increasingly focused on information obtainable through single-cell and unit recording. Although this research has led to the discovery of individual multisensory neurons in the cortex and other structures, as well as to the establishment of many basic principles governing multisensory integration (for a review, see Stein & Meredith, 1993), the reliance on recording from single cells or small groups of cells has biased our understanding of these phenomena toward single-cell response properties and away from how larger populations of cells may participate in multisensory integration. Thus, multisensory-evoked potential studies in animals not only provide a method for establishing the neurogenesis of similar population responses in humans, they also provide another essential link, that between the multisensory responses of single cells and those of larger cellular aggregates. In this way, evoked potentials permit exploration of the spatiotemporal properties of multisensory integration in wide areas of multisensory cortex and functional delineation of the location and borders of both multisensory and unimodal cortex for more detailed electrophysiological and anatomical investigation.

Although the closest animal model in which to study the electrophysiology of human multisensory integration is the monkey (Schroeder & Foxe, 2002; Schroeder, Lindsley, et al., 2001), there is growing interest in evoked potential analysis of multisensory integration in rat (Barth, Goldberg, et al., 1995; Brett-Green, Walsh, et al., 2000; Di, Brett, et al., 1994; Mirmiran, Brenner, et al., 1986; Ramachandran, Wallace, et al., 1993; Toldi, Fehér, et al., 1986) and cat (Toldi, Fehér, et al., 1984; Toldi, Rojik, et al., 1981) cortex, where the functional anatomy of multisensory-evoked responses may be studied in a more thoroughly understood nervous system and tentatively extrapolated to higher mammals.

In this chapter we describe new methods for high spatial resolution mapping of unimodal and multisensory-evoked potentials in rat cortex. We demonstrate how these new methods can be used to identify cortical regions that are uniquely responsive to multisensory stimulation, and how the locus of putative multisensory-zones may be defined relative to the cortical anatomy revealed with cytochrome oxidase staining. Finally, we discuss preliminary results obtained with multisensory- evoked-potential-guided intracellular recording and anatomical tracing, which has implications for discriminating the functional anatomy of thalamocortical multisensory integration in the rat and perhaps in higher species, including man.