| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Neurobiology and Anatomy, Wake Forest University School of Medicine, Winston-Salem, North Carolina
Submitted 6 September 2004; accepted in final form 30 December 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Binns and Salt 1996; Frens and Van Opstal 1998; Jiang et al. 2001; King and Palmer 1985; Meredith and Stein 1986a; Peck 1996; Perrault et al. 2003; Populin and Yin 2002; Wallace et al. 1996). The ability to integrate sensory information in this way is of substantial value in facilitating the role of the SC in controlling orientation behaviors (Stein et al. 1989).
Previous work has shown that multisensory interactions in SC neurons (and in multisensory neurons in other brain structures) (see Brett-Green et al. 2003; Calvert et al. 2001; Chudler et al. 1995; Duhamel et al. 1998; Fuster et al. 2000; Wallace et al. 1992, 2004) are critically dependent on the spatial and temporal relationships of the cross-modal stimuli that are presented as well as on their physical characteristics. For example, when presented with cross-modal stimuli in close spatial and temporal proximity, an SC neuron typically exhibits a significantly enhanced response, whereas when those same stimuli are displaced from one another in space and/or time, that neuron either fails to integrate them (i.e., they are dealt with as 2 separate events) or the integration results in a depressed response (Jiang and Stein 2003; Kadunce et al. 1997, 2001; Meredith et al. 1987; Meredith and Stein 1986a, b, 1996). Additionally, the magnitude of the response enhancement has been found to be greatest when the individual components of the multisensory stimulus are weakly effective when presented on their own (Meredith and Stein 1986a). As these stimuli become increasingly more effective, the magnitude of the response enhancement declines.
Collectively, these spatial, temporal, and inverse effectiveness principles of multisensory integration provide an important predictive framework for understanding how stimulus characteristics shape multisensory interactions in SC neurons. Furthermore, these principles have also been helpful in predicting the effects of various multisensory stimulus combinations on overt behaviors (Corneil et al. 2002; Diederich et al. 2003; Frens et al. 1995; Hughes et al. 1994, 1998; Jiang et al. 2002; Stein et al. 1989; Wilkinson et al. 1996). However, despite their utility, these principles are not sufficient to completely explain the wide range of multisensory interactions that can be observed in SC neurons. For example, whereas some neurons exhibit multisensory response enhancements several times greater than would be predicted based on the sum of their individual unisensory responses, others show maximum enhancements that are less than a simple addition of these responses. Such variability suggests that there are additional factors contributing to the multisensory product. One plausible possibility is that there are inherent differences in the capacity of different SC neurons to integrate their cross-modal inputs (Nozawa et al. 1994; Perrault et al. 2003; Quessy et al. 2000).
Given the importance of the magnitude of the unisensory response for the multisensory product (i.e., inverse effectiveness), we posited that one factor that may play an important role in multisensory interactions is a neuron's dynamic range—the range of responses over which it codes stimulus changes. The present study examined the importance of this factor among visual-auditory SC neurons by systematically manipulating the effectiveness of the visual and auditory stimuli over a neuron's complete dynamic range and examining the consequences of such changes on the multisensory product. The data indicate that the underlying computational operation governing a multisensory response differs substantially among neurons with different dynamic ranges. Portions of this work have been published in abstract form (Perrault et al. 2000).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Implantation procedure
Prior to electrophysiological recording, a recording well/head-holding device was implanted in the skull (McHaffie and Stein 1983). Animals were rendered tractable with an intramuscular injection of ketamine HCl (20 mg/kg) and acepromazine maleate (0.2–0.4 mg/kg). Anesthesia was induced and maintained with isofluorane (1–4%) after endotracheal intubation. During surgery, hydration was maintained with intravenous infusion of lactated Ringer solution (4–8 ml/h) via the saphenous vein. This was followed by postsurgical subcutaneous administration of lactated Ringer (30 ml/kg). During surgery, expiratory CO2 was monitored and maintained between 3.5 and 4.5%. Body temperature was monitored with a rectal probe and maintained at 37–38°C by means of a thermostatically regulated heating pad. Sterile artificial tears were placed in the eyes to prevent corneal drying during the duration of the procedure. To ensure adequate levels of anesthesia, the electroencephalogram (EEG) was recorded by means of a silver wire attached to a bone screw overlying frontoparietal cortex. A high-amplitude, synchronous EEG, indicative of a deeply anesthetized state, was maintained for the duration of the surgery. Once anesthetized, the animal was placed in a stereotaxic frame, and a craniotomy exposed the cortex overlying the SC. A stainless steel chamber that provides access to the SC as well as holds the head was affixed to the skull using stainless steel bone screws and orthopedic bone cement. The margin of the wound was trimmed to form a clean interface with the implant. Analgesics (butorphanol 0.1–0.4 mg/kg or ketoprofen 1–2 mg/kg) were given as needed during recovery.
General recording procedure
Electrophysiological recordings began 7–10 days after surgery. The animal was rendered tractable with an intramuscular injection of ketamine (20 mg/kg) and acepromazine (0.4 mg/kg). The animal was intubated and the head-holder was attached to a mounting plate to stabilize the head without wounds or pressure points. A cannula was placed in the saphenous vein for the continuous delivery of anesthetic (ketamine: 4–8 mg · kg–1 · h–1), paralytic (pancuronium bromide: 0.2 mg · kg–1 · h–1), and fluids (lactated Ringer: 4–8 ml/h). Paralysis and artificial respiration are necessary because eye movements can produce significant displacement of visual receptive fields. Maintenance of adequate levels of anesthesia was done by monitoring multiple vital signs, including expiratory CO2, heart rate, and EEG. For purposes of receptive field mapping, the pupils were dilated with 1% atropine sulfate and corrective contact lenses were placed on the anesthetized (0.5% proparacaine hydrochloride ophthalmic solution) corneas to adjust for retinoscopically determined refractive errors. The optic discs were rear-projected and focused onto a 91-cm-diameter translucent hemisphere placed 45 cm from the eyes. Receptive field maps acquired from multiple sessions in the same and different animals could then be registered by aligning the position of the optic disc. After each recording session (1–3/wk), paralysis and anesthesia were reversed, and on return of normal respiration and locomotion, the animal was returned to its home cage. Experiments generally lasted between 8 and 12 h.
Neuronal isolation and recording
Parylene-insulated tungsten microelectrodes (tip diameter: 1–3 µm, impedance: 1–3 M
at 1 kHz) were positioned via an X-Y translational stage and lowered to the dorsal surface of the SC (identified by characteristic visual activity) by means of a manually driven micromanipulator. Once at the SC surface, the electrode was advanced into stratum opticum (the transitional layer between the superficial and deep SC) by means of a hydraulic microdrive. From here, the electrode was advanced in 10-µm steps while presenting visual and auditory search stimuli as in previous studies (Meredith and Stein 1986a; Wallace et al. 1993). Single units were isolated (criterion signal: noise = 3:1) and digitized by means of a window discriminator (FHC). Neural activity was amplified and monitored, and data were collected using a customized suite of software that employs the 1401 Plus data acquisition system (Cambridge Electronic Design). In the current study, only visual-auditory multisensory neurons were examined.
Sensory stimuli
Although stationary light flashes were the primary visual stimulus (because of the fixed temporal nature of the stimulus), moving visual stimuli were used as well. Stationary visual stimuli consisted of the 50- to 100-ms illumination of a light-emitting diode (LED; 660 nm 
) placed within the receptive field (see Receptive field mapping). Moving visual stimuli consisted of slits, bars, or spots of light projected onto the translucent hemisphere, the movement speed, amplitude, and direction of which could be independently controlled. Whereas the intensity of stationary stimuli was computer controlled, the intensity of moving stimuli was controlled using neutral density filters. In both circumstances stimulus intensity ranged from 0.11 to 13.0 cd/m2 with a background luminance of 0.10 cd/m2. Auditory stimuli were delivered in a free-field setting and consisted of 50- to 100-ms duration broadband noise bursts (20 Hz to 10 kHz). These stimuli were digitally synthesized and delivered through speakers that could be positioned at any location in auditory space. Auditory stimulus intensities ranged from 50.6 to 70.0 dB sound pressure level (SPL) against a background SPL of 50.0 dB. For quantitative tests of multisensory integration, visual and auditory stimuli were typically delivered at the most sensitive location within the excitatory receptive field (see following text).
Receptive field mapping
The borders of each visual receptive field were mapped onto the translucent hemisphere by moving the optimum stimulus, projected from a handheld pantoscope, from the periphery inward from all directions until an enclosed responsive area was defined. Auditory receptive fields were mapped using brief (50 ms) broadband noise bursts delivered from a speaker that could be positioned at any location on a hoop that could be freely rotated about the animal's interaural axis. The typical steps in speaker location represented 
15° of auditory angle in both the azimuthal and elevation dimensions. The location of the stimulus was randomly varied, and a positive response (i.e., a location within the receptive field) was one in which the stimulus-evoked response was readily discernible above background activity. For purposes of receptive field mapping, auditory stimuli were 15 dB above the neuron's previously determined threshold. Receptive fields were transposed from the hemisphere and plotted on standardized representations of visual and auditory space.
Dynamic range determination
To determine the intensities to be used in dynamic range testing, threshold and saturating stimulus intensities were first determined. For threshold determination, stimulus intensity was progressively lowered until no response was elicited from the neuron. The minimum intensity value that resulted in an observable response was set as threshold. For saturation determination, stimulus intensity was progressively raised until no observable increase in response was noted. The first intensity value at which the response no longer increased was set as the saturating intensity. For quantitative analysis of dynamic range, these two qualitatively chosen intensity values served as the upper and lower bounds, and a minimum of 10 intervening intensities were chosen to examine the neuron's visual and auditory dynamic ranges. To determine the neuron's multisensory dynamic range, these same 10+ intensities were then index-matched (Fig. 1). For example, a visual stimulus intensity that elicited a threshold response was paired with an auditory stimulus intensity that elicited a threshold response. Post hoc tests were used to determine if the chosen stimuli were adequate for assessing the neuron's dynamic range. Using these tests, threshold was defined as the intensity that elicited a response that was 1 SD above mean spontaneous activity. Saturation was defined as the minimum intensity at which three successive increases in stimulus intensity failed to result in a significant increase in response. Neurons were said to have a dynamic range in a given modality if they contained a minimum of three intensity values that resulted in responses that could be statistically distinguished from one another. Conversely, neurons were determined to lack a dynamic range in a given modality if manipulations of stimulus intensity failed to yield any significant differences in response.
|
Analyses
Two principal methods of analysis were used to examine the integrative characteristics of the sampled neurons. The first method was one that has been used in many prior studies of multisensory integration (Meredith and Stein 1983, 1986a,b; Wallace et al. 1993) and is termed the interactive index. This metric represents the relative change in a multisensory neuron's dominant unisensory response (e.g., visual) induced by the concurrent presentation of a stimulus from the nondominant modality (e.g., auditory). The interactive index was computed using the following formula
 |
The second method of analysis was a calculation of multisensory contrast. This metric serves as a way of evaluating the multisensory response as a function of the response predicted by an addition of the two unisensory responses. Mean multisensory contrast was computed using the following formula
 |
). Histology and euthanasia
The x-y coordinates of each electrode penetration and the depth of each neuron from the dorsal surface of the SC were systematically recorded. During the final few recording sessions, a series of electrolytic lesions were made by injecting small currents (10–20 µA DC for 5–10 s) through the recording electrode. At the end of the final experiment, the animal was killed with an overdose of sodium pentobarbital (100 mg/kg iv) and perfused transcardially with saline followed by formalin (10%). The brain was blocked, and the SC was cut into 40–50 µm coronal frozen sections. Neuron locations were reconstructed from histologically prepared Nissl-stained sections using standard methods (Meredith and Stein 1986a,b).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
A total of 109 visual-auditory neurons were studied in the multisensory (i.e., below stratum opticum) SC layers. The responses from 84 of these neurons were examined in sufficient quantitative detail to allow a determination of their dynamic ranges to visual and auditory stimuli as well as to their multisensory combination. A neuron's dynamic range was operationally defined as the mean difference in the number of spikes evoked at threshold and at saturation (see METHODS for more detail on this procedure).
Multisensory SC neurons were divided into three general dynamic range categories on the basis of their responses to systematic changes in stimulus intensity (Fig. 2). The first category exhibited monotonic changes in response to changes in both visual and auditory stimulus intensity until saturation was achieved [dual-modality dynamic range neurons (DMDR), 29%]. In the second category, changes in stimulus intensity modulated neuronal responsiveness in only one of the two effective modalities. These single-modality dynamic range neurons (SMDR) represented approximately half of the total population with there being a nearly equal incidence of those having a visual (26%) and those having an auditory (23%) dynamic range. Finally, 22% of the neurons were unaffected by changes in stimulus intensity in either modality [no dynamic range neurons (NDR)].
|
|
|
Dynamic ranges were also assessed in response to multisensory stimuli. To facilitate this analysis, visual and auditory stimuli were index matched for intensity (see METHODS and Fig. 1), enabling an examination of the multisensory response profile at comparable points along the neuron's unisensory dynamic ranges. An example of the results of this index matching is shown in Fig. 5 for a representative DMDR neuron. Note that the maximal multisensory responses exceeded the maximal unisensory responses. This general finding was found in each of the three dynamic range categories (NDR, SMDR, DMDR) as is evident in Fig. 6. Intriguingly, for a number of NDR neurons the presentation of multisensory stimuli effectively created a dynamic range where none was evident under conditions of unisensory stimulation.
|
|
Multisensory integration was assessed at different points along each neuron's dynamic range using two complementary metrics. First, the standard metric for assessing multisensory integration, the interactive index, was calculated. This measure, which has been the convention for quantifying multisensory interactions, relates the multisensory response to the larger of the two unisensory responses and illustrates the proportionate gain provided by having a second channel of sensory information. The second metric was multisensory contrast. This measure relates the multisensory response to a model founded on the predicted sum of the two unisensory responses and provides a view as to the integrative operation performed by the neuron on these inputs. A contrast value of zero represents a response that is equivalent to the predicted addition of the two unisensory responses, positive contrast values (i.e., superadditivity) indicate multisensory responses exceeding the additive prediction, and negative contrast values (i.e., subadditivity) indicate responses that are less than the additive prediction. Both the interactive index and multisensory contrast were calculated because they offer somewhat differing views into the integrative process. Perhaps more importantly, no study to date has attempted to relate between these measures, a feature that was intrinsic to the current study design.
Examples of the results of these analyses are presented for four different SC neurons in Fig. 7. For each of these examples, the principal line graph shows the results of the contrast analyses (i.e., mean contrast as a function of stimulus effectiveness for points along the dynamic range). Furthermore, the three inset bar graphs in each plot show the mean visual, auditory, and multisensory responses at three different points along the dynamic range along with the proportionate multisensory gain (i.e., % interaction) at each of these points. For the neuron depicted in the top left quadrant, the contrast analyses reveal that virtually all tested interactions exceeded the additive prediction. Furthermore, the bar graph insets show that when using the interactive index, large proportionate response enhancements are seen regardless of stimulus effectiveness. Note, however, that the magnitude of these enhancements declined from >1,000% at low levels of effectiveness to 320% at high levels of effectiveness. Because contrast values were invariably positive in these neurons (Fig. 8, top left), they were classified as superadditive, and represented 18% of the population (Fig. 7, pie graph).
|
|
Neuronal categorization is independent of stimulus features
To determine whether changes in the nature of the stimuli would alter these neuronal categorizations, parallel data sets were gathered in 14 neurons using two different types of visual stimuli. Whereas the first data set employed the standard stationary stimulus described in the preceding text, the second set used a moving visual stimulus. Although the shape of the dynamic range functions evoked by stationary and moving stimuli were generally different (data not shown), the size of the dynamic range was not dependent on the type of stimulus used (Fig. 9A). Most importantly, when multisensory contrast was plotted as a function of stimulus effectiveness for moving versus stationary stimuli in this neuron, the plots were virtually identical (Fig. 9B). This invariance in the shape of the contrast functions for the two different types of visual stimuli was a general feature of the sampled population (Fig. 9C), suggesting that these categories are robust and general features of the multisensory population.
|
In an effort to extend the current predictive framework for multisensory interactions, we examined the relationship between a neuron's dynamic range and its multisensory contrast category. Indeed, a strong relationship was found between these two (Fig. 10). Most striking was the finding that neurons capable of exhibiting superadditive interactions (i.e., superadditive and superadditive/subadditive neurons) had a strong tendency to have either no or very small dynamic ranges, whereas additive/subadditive and purely subadditive neurons tended to have large dynamic ranges.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Binns and Salt 1996; Jiang et al. 2001; Kadunce et al. 1997; King and Palmer 1985; Meredith and Stein 1983, 1986a; Peck 1995; Peck et al. 1995; Perrault et al. 2003; Stein and Arigbede 1972; Wallace et al. 1993; Zangenehpur and Chaudhuri 2001). Although the current study focused exclusively on visual-auditory neurons, the broad similarities in the integrative characteristics of SC multisensory neurons suggest that these results may be applicable to the entire population of multisensory neurons in this structure. Multisensory SC neurons exhibit significant variability in their sensory responses
The response magnitude of multisensory SC neurons to both uni- and multisensory stimuli was found to vary widely in the population of neurons sampled. Although response variability has been apparent in previous studies, it has not been clear to what extent it is due to variations in stimulus characteristics (e.g., intensity, duration, etc.), the pattern of converging inputs, or inherent characteristics of different SC neurons. Here we show that different SC neurons respond very differently to stimuli with similar physical characteristics and that these response differences are predictable when viewed in the context of a given neuron's dynamic range.
Substantial differences in the responsiveness of different SC neurons were manifested in differences in their dynamic ranges. That these differences in some neurons were likely due to variations in their inputs (e.g., see Huerta and Harting 1984; Jiang and Stein 2003; Jiang et al. 2001, 2002; Meredith 1999; Meredith and Clemo 1989) is suggested by the observation that multisensory neurons often exhibited striking differences in their visual and auditory dynamic ranges with nearly half of the sample having a substantial dynamic range for one modality and no dynamic range for the other modality. Nonetheless, in those neurons with a demonstrable dynamic range in both modalities, there was a strong correlation between the size of the visual and auditory dynamic ranges, a feature that suggests that there are intrinsic determinants of neuronal responsiveness as well.
Another consistent feature of multisensory neurons was that the size of the multisensory dynamic range was larger than the size of the unisensory dynamic range(s). This finding, which is in keeping with previous observations, see (Jiang et al. 2001), suggests that the maximum information-carrying content of these neurons is achieved only in response to multisensory stimuli. That the multisensory response is the most robust response of these neurons also has relevance for the principle of inverse effectiveness (Meredith and Stein 1986a; Stein and Meredith 1993). This feature of multisensory integration, in which declining levels of response enhancement are seen with progressive increases in stimulus effectiveness, could be explained by a ceiling effect. However, because maximal multisensory responses always exceed maximal unisensory responses, this possibility seems unlikely.
Multisensory contrast reveals categorical distinctions in the SC population
In comparison to the continuous measure of multisensory integration provided by the interactive index (which ranged from 32% to >1,000% in the current study), measures of multisensory contrast enabled a differentiation of the multisensory population into discrete categories. This metric, which assesses the multisensory response against a prediction using both unisensory responses, is being used increasingly as an effective tool to assess the integrative operation carried out by multisensory neurons (Nozawa et al. 1994).
The present data lend themselves to a division of the multisensory responses of SC neurons into four contrast categories. In three of these categories (superadditive/subadditive, additive/subadditive and subadditive), contrast values became increasingly negative with increases in stimulus effectiveness, and the similarities in the shapes of their response functions suggest that they employ similar mechanisms in their synthesis of multisensory information. The fourth category was distinct in that it showed superadditive interactions regardless of stimulus intensity. As a population, these neurons were found to have either compressed dynamic ranges or, more commonly, no demonstrable dynamic ranges. The enhancement in these neurons appears to involve a nonlinear amplification of the unisensory inputs to produce a superadditive response. Possible candidate mechanisms for supporting superadditive interactions include but are not limited to the integration of subthreshold inputs, specialized biophysical properties of the multisensory neurons (e.g., complement of ion channels, channel kinetics, etc.), local circuit effects that serve to amplify responses, and/or a gating mechanism imposed from an extrinsic source(s).
In support of an extrinsic gating mechanism is recent data suggesting that two cortical areas are critical for the manifestation of integrated multisensory responses in SC neurons (Jiang and Stein 2003; Jiang et al. 2001, 2002; Wallace et al 1992, 1993). Thus selective and reversible deactivation of the anterior ectosylvian sulcus (AES) and the rostral lateral suprasylvian cortex (rLS) compromises the integrative capacity of their multisensory SC targets. Such deactivation renders the multisensory responses of these neurons indistinguishable from their constituent unisensory responses. Intriguingly, these cortical inputs appear to be essential for multisensory integration regardless of the contrast operation (i.e., superadditive, subadditive, etc.) utilized by the neuron. Thus it appears that along with the role of extrinsic cortical influences in gating the capacity of multisensory neurons to integrate their multiple channels of sensory input, other factors play an integral role in determining the end product of these integrative operations.
One clue as to the identity of these factors may lie in the relationship found to exist between dynamic range and contrast category. Thus neurons capable of exhibiting superadditive interactions (i.e., superadditive and superadditive/subadditive neurons) were very likely to have no or small dynamic ranges, whereas neurons exhibiting only additive or subadditive interactions typically had extensive dynamic ranges. Such a result points to neuronal responsiveness as a major determinant of multisensory operational mode. In further support of this suggestion is the finding that multisensory interactions appear to be dependent on both levels of spontaneous activity and absolute sensory responsiveness (Perrault et al. 2003).
Unresolved questions include potential differences that might exist in the projection patterns of these different multisensory neuronal types, and in a related question, the functional significance of these different operational modes. Although no obvious differences were noted in the location of each of the multisensory neuronal types, the sample size is too small to support any firm conclusions about spatial distributions. Given our observation of higher spontaneous firing rates in the rostral SC and the strong relationship between spontaneous firing rates and neuronal responsiveness (Perrault et al. 2003), one might predict that neurons with larger dynamic ranges and thus less superadditive potential would be found in these rostral locations. Such a result would fit within the framework of rostral SC zones being concerned with fixation-related processes mediated by intrinsic connectivity, and more caudal SC locations being concerned with localization functions that are dependent on SC outputs to premotor circuits (Meredith and Ramoa 1998; Munoz and Guitton 1989; Munoz and Wurtz 1993; Peck and Baro 1997). It is in such extrinsic projections that nonlinear enhancements may serve an important role, possibly by increasing the probability of an action and/or speeding response times.
Together, the results of the current study illustrate that multisensory integration in SC neurons, in addition to being strongly dependent on stimulus factors (i.e., their spatial and temporal relationship, their relative effectiveness), also appears to be governed by neuron-specific factors. Although the relative contributions of these intrinsic and extrinsic factors are likely to differ from neuron to neuron, it is now clear that the stimulus-dependent modulations of multisensory integration in SC neurons take place within intrinsic limitations characteristic of each neuron. In addition to their obvious mechanistic implications, these results should be used to guide more biologically plausible models of multisensory integration, a rapidly emerging field of inquiry.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Current address of T. J. Perrault: Dept. of Neural and Behavioral Sciences, Pennsylvania State University School of Medicine, Hershey, PA 17033.
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: M. T. Wallace, Dept. of Neurobiology and Anatomy, Wake Forest University School of Medicine, Winston-Salem, North Carolina 27157 (E-mail: mwallace{at}wfubmc.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Binns KE and Salt TE. Importance of NMDA receptors for multimodal integration in the deep layers of the cat superior colliculus. J Neurophysiol 75: 920–930, 1996.
Brett-Green B, Fifkova E, Larue DT, Winer JA, and Barth DS. A multisensory zone in rat parietotemporal cortex: intra- and extracellular physiology and thalamocortical connections. J Comp Neurol 460: 223–237, 2003.[CrossRef][ISI][Medline]
Calvert GA, Hansen PC, Iversen SD, and Brammer MJ. Detection of audio-visual integration sites in humans by application of electrophysiological criteria to the bold effect. Neuroimage 14: 427–438, 2001.[CrossRef][ISI][Medline]
Chudler EH, Sugiyama K, and Dong WK. Multisensory convergence and integration in the neostriatum and globus pallidus of the rat. Brain Res 674: 33–45, 1995.[CrossRef][ISI][Medline]
Corneil BD, Van Wanrooij M, Munoz DP, and Van Opstal AJ. Auditory-visual interactions subserving goal-directed saccades in a complex scene. J Neurophysiol 88: 438–454, 2002.
Diederich A, Colonius H, Bockhorst D, and Tabeling S. Visual-tactile spatial interaction in saccade generation. Exp Brain Res 148: 328–337, 2003.[ISI][Medline]
Duhamel JR, Colby CL, and Goldberg ME. Ventral intraparietal area of the macaque: congruent visual and somatic response properties. J Neurophysiol 79: 126–136, 1998.
Frens MA and Van Opstal AJ. Visual-auditory interactions modulate saccade-related activity in monkey superior colliculus. Brain Res Bull 46: 211–224, 1998.[CrossRef][ISI][Medline]
Frens MA, Van Opstal AJ, and Van der Willigen RF. Spatial and temporal factors determine auditory-visual interactions in human saccadic eye movements. Percept Psychophys 57: 802–816, 1995.[ISI][Medline]
Fuster JM, Bodner M, and Kroger JK. Cross-modal and cross-temporal association in neurons of frontal cortex. Nature 405: 347–351, 2000.[CrossRef][Medline]
Huerta MF and Harting JK. The mammalian superior colliculus: studies of its morphology and connections. In: Comparative Neurology of the Optic Tectum, edited by Vanegas H. New York: Plenum, 1984, p. 687–773.
Hughes HC, Nelson MD, and Aronchick DM. Spatial characteristics of visual-auditory summation in human saccades. Vision Res 38: 3955–3963, 1998.[CrossRef][ISI][Medline]
Hughes HC, Reuter-Lorenz PA, Nozawa G, and Fendrich R. Visual-auditory interactions in sensorimotor processing: saccades versus manual responses. J Exp Psychol Hum Percept Perform 20: 131–153, 1994.[CrossRef][ISI][Medline]
Jiang W, Jiang H, and Stein BE. Two corticotectal areas facilitate multisensory orientation behavior. J Cogn Neurosci 14: 1240–1255, 2002.
Jiang W and Stein BE. Cortex controls multisensory depression in superior colliculus. J Neurophysiol 90: 2123–2135, 2003.
Jiang W, Wallace MT, Jiang H, Vaughan JW, and Stein BE. Two cortical areas mediate multisensory integration in superior colliculus neurons. J Neurophysiol 85: 506–522, 2001.
Kadunce DC, Vaughan JW, Wallace MT, Benedek G, and Stein BE. Mechanisms of within- and cross-modality suppression in the superior colliculus. J Neurophysiol 78: 2834–2847, 1997.
Kadunce DC, Vaughan JW, Wallace MT, and Stein BE. The influence of visual and auditory receptive field organization on multisensory integration in the superior colliculus. Exp Brain Res 139: 303–310, 2001.[CrossRef][ISI][Medline]
King AJ and Palmer AR. Integration of visual and auditory information in bimodal neurons in the guinea pig superior colliculus. Exp Brain Res 60: 492–500, 1985.[ISI][Medline]
McHaffie JG and Stein BE. A chronic headholder minimizing facial obstructions. Brain Res Bull 10: 859–860, 1983.[CrossRef][ISI][Medline]
Meredith MA. The frontal eye fields target multisensory neurons in cat superior colliculus. Exp Brain Res 128: 460–470, 1999.[CrossRef][ISI][Medline]
Meredith MA and Clemo HR. Auditory cortical projection from the anterior ectosylvian sulcus (Field AES) to the superior colliculus in the cat: an anatomical and electrophysiological study. J Comp Neurol 289: 687–707, 1989.[CrossRef][ISI][Medline]
Meredith MA, Nemitz JW, and Stein BE. Determinants of multisensory integration in superior colliculus neurons. I. Temporal factors. J Neurosci 7: 3215–3229, 1987.[Abstract]
Meredith MA and Ramoa AS. Intrinsic circuitry of the superior colliculus: pharmacophysiological identification of horizontally oriented inhibitory interneurons. J Neurophysiol 79: 1597–1602, 1998.
Meredith MA and Stein BE. Interactions among converging sensory inputs in the superior colliculus. Science 221: 389–391, 1983.
Meredith MA and Stein BE. Visual, auditory, and somatosensory convergence on cells in superior colliculus results in multisensory integration. J Neurophysiol 56: 640–662, 1986a.
Meredith MA and Stein BE. Spatial factors determine the activity of multisensory neurons in cat superior colliculus. Brain Res 365: 350–354, 1986b.[CrossRef][ISI][Medline]
Meredith MA and Stein BE. Spatial determinants of multisensory integration in cat superior colliculus neurons. J Neurophysiol 75: 1843–1857, 1996.
Munoz DP and Guitton D. Fixation and orientation control by the tecto-reticulo-spinal system in the cat whose head is unrestrained. Rev Neurol (Paris) 145: 567–579, 1989.[Medline]
Munoz DP and Wurtz RH. Fixation cells in monkey superior colliculus. I. Characteristics of cell discharge. J Neurophysiol 70: 559–575, 1993.
Nozawa G, Reuter-Lorenz PA, and Hughes HC. Parallel and serial processes in the human oculomotor system: bimodal integration and express saccades. Biol Cybern 72: 19–34, 1994.[CrossRef][ISI][Medline]
Peck CK. Visual-auditory integration in cat superior colliculus: implications for neuronal control of the orienting response. Prog Brain Res 112: 167–177, 1996.[ISI][Medline]
Peck CK and Baro JA. Discharge patterns of neurons in the rostral superior colliculus of cat: activity related to fixation of visual and auditory targets. Exp Brain Res 113: 291–302, 1997.[CrossRef][ISI][Medline]
Peck CK, Baro JA, and Warder SM. Effects of eye position on saccadic eye movements and on the neuronal responses to auditory and visual stimuli in cat superior colliculus. Exp Brain Res 103: 227–242, 1995.[ISI][Medline]
Perrault TJ, Vaughan JW, Stein BE, and Wallace MT. Dynamic response ranges in multisensory superior colliculus neurons. Soc Neurosci Abstr 26: 1220, 2000.
Perrault TJ Jr, Vaughan JW, Stein BE, and Wallace MT. Neuron-specific response characteristics predict the magnitude of multisensory integration. J Neurophysiol 90: 4022–4026, 2003.
Populin LC and Yin TC. Bimodal interactions in the superior colliculus of the behaving cat. J Neurosci 22: 2826–2834, 2002.
Quessy S, Sweatt AJ, Stein BE, and Stanford TR. The influence of stimulus intensity and timing on the responses of multisensory neurons in the superior colliculus: comparison to a model's prediction. Soc Neurosci Abstr 26: 1221, 2000.
Stein BE, Meredith MA, Huneycutt WS, and McDade L. Behavioral indices of multisensory integration: orientation to visual cues is affected by auditory stimuli. J Cogn Neurosci 1: 12–24, 1989.
Stein BE, and Arigbede MO. Unimodal and multimodal response properties of neurons in the cat's superior colliculus. Exp Neurol 36: 179–196, 1972.[CrossRef][ISI][Medline]
Stein BE and Meredith MA. The Merging of the Senses. Cambridge, MA: MIT Press, 1993.
Wallace MT, Meredith MA, and Stein BE. Integration of multiple sensory modalities in cat cortex. Exp Brain Res 91: 484–488, 1992.[ISI][Medline]
Wallace MT, Meredith MA, and Stein BE. Converging influences from visual, auditory, and somatosensory cortices onto output neurons of the superior colliculus. J Neurophysiol 69: 1797–1809, 1993.
Wallace MT, Ramachandran R, and Stein BE. A revised view of sensory cortical parcellation. Proc Natl Acad Sci USA 101: 2167–2172, 2004.
Wallace MT, Wilkinson LK, and Stein BE. Representation and integration of multiple sensory inputs in primate superior colliculus. J Neurophysiol 76: 1246–1266, 1996.
Wilkinson LK, Meredith MA, and Stein BE. The role of anterior ectosylvian cortex in cross-modality orientation and approach behavior. Exp Brain Res 112: 1–10, 1996.[ISI][Medline]
Zangenehpour S and Chaudhuri A. Neural activity profiles of the neocortex and superior colliculus after bimodal sensory stimulation. Cereb Cortex 11: 924–935, 2001.
This article has been cited by other articles:
|
|
 |
|
  C. Kayser, C. I. Petkov, and N. K. Logothetis Visual Modulation of Neurons in Auditory Cortex Cereb Cortex, July 1, 2008; 18(7): 1560 - 1574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. N. Carriere, D. W. Royal, and M. T. Wallace Spatial Heterogeneity of Cortical Receptive Fields and Its Impact on Multisensory Interactions J Neurophysiol, May 1, 2008; 99(5): 2357 - 2368. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Leo, N. Bolognini, C. Passamonti, B. E. Stein, and E. Ladavas Cross-modal localization in hemianopia: new insights on multisensory integration Brain, March 1, 2008; 131(3): 855 - 865. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Alvarado, T. R. Stanford, J. W. Vaughan, and B. E. Stein Cortex Mediates Multisensory But Not Unisensory Integration in Superior Colliculus J. Neurosci., November 21, 2007; 27(47): 12775 - 12786. [Abstract] [Full Text] [PDF] |
) and visual (
) modalities in SMDR neurons (left) and DMDR neurons (right). Note that there were no statistical differences in the size of the mean dynamic ranges between these categories (see text).
), visual (
) dynamic range plots for this neuron. Note that for all points along these response continua, multisensory responses exceed either of the unisensory responses and that the multisensory response saturates at levels far higher than either of the component responses. Square waves above the histograms depict the onset and offset of the visual (V) and auditory (A) stimuli.
) and moving visual stimulus (