THE JOURNAL OF COMPARATIVE NEUROLOGY 464:415– 425 (2003)

Somatosensation in the Superior Colliculus of the Star-Nosed Mole SAMUEL D. CRISH,1 CHRISTOPHER M. COMER,1 PAUL D. MARASCO,2 2 AND KENNETH C. CATANIA * 1 Laboratory of Integrative Neuroscience, University of Illinois at Chicago, Chicago, Illinois 60607 2 Department of Biological Sciences, Vanderbilt University, Nashville, Tennessee 37235

ABSTRACT The superior colliculus (or optic tectum in nonmammals) plays a critical role in the visual system and is essential for integrating sensory inputs to guide eye and head movements. However, what is the role of the superior colliculus (SC) in species that depend almost exclusively on touch? In this study we examined the SC of the star-nosed mole, a subterranean mammal that, instead of using vision, explores its environment using its tactile star. The star acts like a mechanosensory eye with a central tactile fovea that is constantly shifted in a saccadic manner. Multiunit microelectrode recordings were used to determine the topography and receptive field organization of somatosensory inputs to the SC and to test for visual and auditory responses. Here we report an SC dominated by somatosensory inputs in which neurons in all layers responded to mechanosensory stimulation, forming a topographic representation of contralateral body dominated by the mechanosensory star. Receptive fields were large, and appendage representations overlapped, suggesting that the SC may use a distributed, population code to guide the saccadic movements of the mole’s touch fovea. No auditory or visual responses were recorded from the SC, although neurons in the neighboring inferior colliculus responded to auditory stimuli. Layers IVb–VII were identified, and a layer superficial to IVb contained neurons that responded to somatosensory stimulation, suggesting that there are unique patterns of afferents in the star-nosed mole’s SC. J. Comp. Neurol. 464:415– 425, 2003. © 2003 Wiley-Liss, Inc. Indexing terms: somatosensory; topography; mechanosensory; evolution; optic tectum

The superior colliculus (SC) has primarily been studied for its role as part of the vertebrate visual system. There is good reason for this emphasis, as vision is the dominant sensory modality of many vertebrates and this is particularly true for primates and carnivores, from which much of our information about the superior colliculus is derived. In barn owls— experts at auditory localization—vision can be said literally to dominate tectal organization by guiding the structure of auditory receptive fields during critical periods of development (Knudsen and Brainard, 1991; Knudsen, 1999). A similar instructive role for visual inputs has been suggested in mammals, as somatosensory maps in the SC have disproportionately large representations of areas in spatial register with the overlying representation of central vision (Stein et al., 1975; Dra¨ger and Hubel, 1976). Given the important role the SC plays in visual orientation and the dominant influence visual inputs play in functional organization during collicular development, we wondered how the SC would be organized in an animal © 2003 WILEY-LISS, INC.

with rudimentary eyesight that instead depends on somatosensation to explore its environment. Star-nosed moles (Fig. 1A) are good candidates for this investigation because their somatosensory system is organized in many ways like a visual system. There is a tactile fovea on their mechanosensory star (Fig. 1B) that is constantly shifted in a saccadic manner when moles explore their surroundings (Catania and Kaas, 1997; Catania, 1999). Thus, despite

Grant sponsor: Fundamental Neuroscience Training; Grant number: T32 MH64913 (P.M.); Grant sponsor: National Institutes of Health; Grant number: MH-58909 (K.C.); Grant sponsor: the Searle Scholars Program. *Correspondence to: Kenneth C. Catania, Vanderbilt University, Department of Biological Sciences, VU Station B 351634, Nashville, TN 37235. E-mail: [email protected] Received 20 February 2003; Revised 24 April 2003; Accepted 25 April 2003 DOI 10.1002/cne.10791 Published online the week of August 11, 2003 in Wiley InterScience (www.interscience.wiley.com).

416

Fig. 1. The star-nosed mole’s sensory specializations and superior colliculus. A: The face and forelimbs of a star-nosed mole showing the unusual tactile snout ringed by 22 fleshy appendages. B: Schematized body and nose showing the approximate sizes of body parts and the numbering system for the nasal appendages. The 11th appendage (on each side) acts as the tactile fovea. C: The ventral surface of the mole’s brain showing the large trigeminal nerves carrying information from the star and the tiny optic nerve. D: A transverse section of the midbrain processed for the metabolic enzyme cytochrome oxidase (CO) showing the thin superior colliculus (SC) and the central gray

S.D. CRISH ET AL.

(CG). The arrow marks a series of rostrocaudally oriented fiber bundles that correspond to layer IVb in mice (see Discussion and Fig. 7). E: An illustration of our flattening technique for the superior colliculus, adapted from similar procedures used for cerebral cortex. F: A flattened superior colliculus sectioned parallel to the dorsal surface to show its areal extent (SC) and its relationship to the inferior colliculus (IC) and pretectum (PT). The arrow marks part of a series of rostrocaudally oriented fiber bundles that correspond to layer IVb in mice (see Fig. 7).

STAR-NOSED MOLE SUPERIOR COLLICULUS poorly developed vision, star-nosed moles make frequent and precise orienting movements to novel stimuli— behaviors that in other mammals are mediated by the SC (see Stein, 1998, for review). A number of interesting questions about the mammalian SC may be addressed in star-nosed moles. First, how important is the SC in species with rudimentary eyesight? Does the SC guide orienting behavior in the absence of vision, or has it degenerated in star-nosed moles? Have some collicular layers been lost, as has been suggested for blind mole-rats (Cooper et al., 1993a,b), or could layers have changed their usual functional roles? How mutable is the SC in the course of evolution? More specifically, differences between collicular and cortical somatosensory receptive fields have been attributed to the role the colliculus plays in integrating visual and somatosensory information (Stein and Meredith, 1993). How then are somatosensory receptive fields in the colliculus structured in a species that presumably does not need to keep somatosensation in register with vision? Here we begin to address these questions by investigating the organization of the star-nosed mole’s SC with multiunit microelectrode mapping of visual, auditory, and somatosensory responses.

MATERIALS AND METHODS Microelectrode recordings were made from the SC of 6 moles (Condylura cristata) collected under PA permit number COL00087. Moles were anesthetized with an IP injection of 0.6 mg/kg urethane (15% solution) followed by a supplement of 0.1 mg/kg after 10 minutes. This was followed by two to five IP injections of 0.10 ml ketamine (10 mg/ml, diluted in sterile saline) to induce a surgical plane of anesthesia. Body temperature was maintained with a hot water bottle, the mole was placed in a headholder, and a craniotomy was made over the caudal cortical hemisphere. Caudal cortex was removed by aspiration to expose the SC. A photograph of the collicular surface was made to mark microelectrode penetrations. Single tungsten microelectrodes were lowered into the colliculus (1–1.5 M⍀ at 1,000 Hz) to record multiunit activity. Neuronal responses were amplified, viewed on an oscilloscope, and monitored on a speaker. Selected penetrations were marked with microlesions. The photograph of the colliculus was used to identify the locations of electrode penetrations relative to lesions made in the colliculus. In cases in which it was difficult to view the colliculus, stereotaxic coordinates were used to locate electrode penetrations relative to lesions. Tactile receptive fields were defined by stimulating the skin, deflecting the vibrissae, or bending hairs on the body with a small probe or fine brush. Probe tip diameters ranged from 200 ␮m to 2 mm. Visual stimuli were presented with an opthalmoscope using moving bars of light, circles, and full-field flashes. Neurons at 14 electrode penetrations were tested for auditory responses. Tests for auditory responses were made at 100-, 200-, and 300-␮m depths. For two penetrations neurons were also tested at 400 ␮m. Auditory stimuli consisted of clicks, taps, and white noise moved to different locations around the contralateral side of the animal’s body at a distance that was varied from approximately 12 to 24 inches. The click stimulus was 1.3 ms in duration and 64 dB SPL intensity. Spectral energy ranged from 5 to 15 kHz, peaking at 8 kHz

417 with a 10-dB octave fall-off. The broad-band stimulus was 2 seconds in duration and 62 dB SPL intensity. The magnitude spectrum was flat from 0.2 to 9.5 kHz, with a 5-DB/octave fall-off. Auditory stimuli were characterized using a 1/4-inch field microphone (ACO Pacific, Belmont, CA) coupled to a computer-controlled calibration system (Tucker-Davis Technologies) (Alachua, FL) in a sound attenuating chamber (Industrial Acoustics). Although the stimuli used did not elicit neuronal responses from the SC, neurons in the inferior colliculus responded vigorously to these stimuli. After recording sessions, moles were given a lethal dose of sodium pentobarbital (100 mg/kg) and perfused with phosphate-buffered saline (PBS; 0.9%) (pH 7.4) followed by 4% paraformaldehyde in PBS. The brain was postfixed for 30 minutes and cryoprotected in 30% sucrose. The SC was dissected free from the brainstem, and a wedge of tissue was cut from below the SC so that both hemispheres could be flattened out on the freezing microtome. This was accomplished by positioning the dorsal surface of the SC against the ice and applying pressure to the inverted block of tissue (Fig. 1E). The flattened colliculus was cut at 50 ␮m and processed for cytochrome oxidase (Wong-Riley and Carroll, 1984). Sections were photographed with a digital camera (Nikon D1x) attached to a Wild M400 Photomakroskop and subsequently imported into Adobe (San Jose, CA) Photoshop. Contrast and brightness were adjusted digitally, and images were then placed into Adobe Illustrator 9.0 files in order to add labels and scale bars to the final figures. All procedures were reviewed and approved by the Vanderbilt University Animal Care Committee and conform to NIH guidelines.

RESULTS Before discussing the responses of neurons in the mole’s SC, it is useful to make some basic observations about the external appearance of the brain. A ventral view (Fig. 1C) allows us to compare cranial nerves II and V to assess the star-nosed mole’s dependence on vision versus touch from an anatomical perspective. Consistent with the star-nosed mole’s fossorial habitat (Hamilton, 1931) and the central role of the star in exploratory behaviors (Catania and Kaas, 1997), the massive trigeminal nerves— carrying mechanosensory information from the star— dwarf the optic nerves carrying information from the small eyes. Moving centrally, it is immediately apparent from transverse sections (Fig. 1D) that the mole’s SC is proportionally thinner than that of more visual mammals such as carnivores and primates (Sze´ kely, 1973; Kanaseki and Sprague, 1974; Huerta and Harting, 1984) or rodents (Wiener, 1986; Paxinos and Watson, 1986). It is approximately 600 – 800 ␮m in thickness from the surface to the central gray in formalin-fixed material, or about 60% of the thickness (1.2 mm) of mouse SC (Franklin and Paxinos, 1997). The deepest layers of the star-nosed mole SC are well formed and similar to those observed in other species (see Discussion), whereas the superficial layers are clearly reduced (Fig. 1D). To investigate the responses of neurons in the starnosed mole SC, we collected data from 183 electrode penetrations in six moles. In two cases both left and right superior colliculi were investigated, providing eight colliculi from which our summary of responses was derived. For ease of interpretation across cases, data are presented

418 from the perspective of the left SC. For 33 electrode penetrations distributed across the colliculus, we provided visual stimulation to the contralateral eye with an opthalmoscope using moving bars of light, circles, and full-field flashes while monitoring activity from neurons between the collicular surface and a depth of 300 ␮m. No units or multiunit clusters were found that responded to visual stimuli. Neurons at 14 electrode penetrations were tested for auditory responses between the surface and a depth of 300 ␮m, using clicks, taps, and white noise moved to different locations around the contralateral side of the animal’s body. As was the case for visual stimuli, we obtained no responses to auditory stimuli in the SC. Auditory responses were readily obtained from neurons at penetrations made in the inferior colliculus, and this served as a useful method for determining the caudal extent of the SC during our recordings. In contrast to visual and auditory stimuli, tactile stimuli to the mole’s body produced strong responses across most of the extent of the SC at a wide range of depths. Our main goal was to map the topographic arrangement of body parts and to determine the size and shape of the corresponding receptive fields across the extent of the SC. For this purpose we developed a simple method of flattening the SC and sectioning it parallel to the surface (Fig. 1E,F) in order to relate microlesions accurately to the locations of electrode penetrations marked on photographs of the SC (Figs. 2– 4). Tactile receptive fields were defined by using small hand-held probes to stimulate the star, deflect vibrissae, or move hairs on the body. The strongest responses were obtained at depths from 100 to 300 ␮m, but responses were clear upon first entering the most superficial cell layers of the SC and continued as deep as 700 ␮m. Neurons responded well to stroking or depressing the skin surface and to deflection of hairs or vibrissae. Responses were rapidly adapting such that neurons responded poorly or not at all to sustained pressure. However, little habituation was observed for transiently applied stimuli. A large proportion of the rostrolateral SC responded to tactile stimulation of the appendages from the contralateral star. This is illustrated in Figure 2A-C, which shows a row of closely spaced electrode penetrations progressing from rostral to caudal in the lateral SC. Neurons at all the penetrations responded to stimulation of the star, and the row was bounded by microlesions (Fig. 2B) that allowed the extent of the star representation in the rostrocaudal dimension to be appreciated in a histological section. As the row of electrode penetrations proceeded from rostral to caudal, there was an orderly progression of receptive fields on the ventral aspects of the star, moving from the medially situated 11th appendage (the somatosensory fovea) to the peripheral appendages. Note that the receptive fields were not only restricted to the ventral half of the star but were also relatively large (Fig. 2C). The representation of the 11th appendage was clear but not particularly magnified. Rows of electrode penetrations made further medial in the SC corresponded to more dorsal portions of the star at the rostral extent of the SC, whereas caudomedial parts of the SC represented parts of the mole’s body (Fig. 2D–F). For example, the lesion at point U in Figure 2D and E marks the location of neurons that responded to stimulation of the forelimb, whereas all the neurons at penetrations rostral and lateral to this location responded to stim-

S.D. CRISH ET AL. ulation of the star. The progression of receptive fields on the body was consistent with the progression observed on the star, in that neurons at more caudal locations responded to stimulation of progressively more caudal body parts. The tail was represented at the caudalmost extreme of the SC, and, conversely, the 1st and 11th appendages—at the rostral extreme of the mole’s body— were represented at the rostral extreme of the SC. Sampling a larger area of the SC with more widely distributed electrode penetrations revealed both the rostrocaudal and mediolateral topography of the body representation. For example, neurons at penetrations C and J in case 9 (Fig. 3) were located at approximately the same rostrocaudal position but were at extremes of the mediolateral axis of the SC. Both responded to stimulation of the distal face, but at opposite extremes; neurons at penetration C responded to stimulation of the hairs on the chin (ventral body), whereas neurons at penetration J responded to hairs on the dorsal snout (and to parts of appendages 1–3 of the star). Note also the lesions at D and A (Fig. 3): neurons responded to stimulation of dorsal and ventral parts of appendage number 11, respectively. In general, the representations of different appendages overlapped one another (Fig. 4). A compilation of all the somatosensory responses from the eight colliculi investigated provides a consistent and clear topography of the body representation in the starnosed mole’s SC (Fig. 5). The body has an upright orientation with head (and star) located rostrally and tail located caudally. The most obvious feature of the body representation is the enormous representation of the star (Fig. 6). Within the representation of the star, the topography of the sensory surfaces was consistent and orderly across cases, mirroring the representation of the body (Fig. 5B). That is, the star was upright in the colliculus such that appendage 1 was represented just medial to appendage 11 at the rostral border of the SC, with the remaining appendages 2–10 forming a pinwheel pattern in the clockwise direction. Appendage 11, the somatosensory fovea, was not obviously magnified in the collicular representation. Although the receptive fields on the star were of a small absolute size for collicular neurons (often only a few square millimeters in area), they were nevertheless much larger than those found for neurons in the somatosensory cortex of the star-nosed mole (Fig. 5D). Most receptive fields included multiple appendages, which is rarely observed at the cortical level. Corresponding to the large receptive field size, there was considerable overlap in the representations of neighboring appendages. Figure 5C illustrates the extent of this overlap from all eight colliculi examined in this study, showing only the odd-numbered appendages for clarity. Because this is a compilation of results from six different animals superimposed on a generic colliculus, the actual degree of overlap for any individual animal is probably somewhat less than illustrated. The overlap of receptive fields for selected individual cases is illustrated in Figure 4.

DISCUSSION The overall topographic arrangement of body parts in the star-nosed mole’s SC is consistent with that found in other species and other sensory systems (Kruger, 1970;

STAR-NOSED MOLE SUPERIOR COLLICULUS

419

Fig. 2. The results of microelectrode recordings from the superior colliculus (SC) of the star-nosed mole. A: The locations of electrode penetrations relative to the SC borders in case 7. IC, inferior colliculus. In this case a row was made from rostral to caudal. Stars mark the locations of microlesions. B: A section of the flattened SC processed for the metabolic enzyme cytochrome oxidase. Note the locations of microlesions corresponding to the row of penetrations illustrated in A. C: Receptive fields on the star for the electrode penetrations A–K. In this case all the receptive fields were located on the star, which clearly takes up a large proportion of the SC. As electrode penetrations progressed caudally, receptive fields on the star progressed from the midline (appendage number 11) to the more lateral appendages. D: The locations of electrode penetrations relative

to the SC borders in case 6. Stars mark the locations of microlesions. E: A section of the flattened SC processed for the metabolic enzyme cytochrome oxidase. Note the locations of microlesions. The star at “G” indicates the position of a microlesion visible in a deeper section. F: Receptive fields on the star and body for the electrode penetrations A–Q. As penetrations progressed caudally in the SC, the receptive fields moved from the midline of the star (appendage 1) to lateral appendages (4 and 5), and then to the body. Penetrations A–P cover the entire rostrocaudal extent of the colliculus and illustrate the orderly topography of the body representation. Just caudal to the SC, neurons in the inferior colliculus responded to auditory stimulation (Aud). Rostral is left and medial is up.

Cynader and Berman, 1972; Stein, 1981; Chalupa, 1984; Stein and Meredith, 1993). The representation of the body in the deeper somatosensory layers (IV–VII) of sighted mammals is generally in register with the overlying visual representation such that the body has an upright position (dorsal body facing the midline) with the head and face represented rostrally and the tail and hindlimb repre-

sented caudally (Stein et al., 1975,1976; Stein, 1984; Dra¨ ger and Hubel, 1976; Wallace and Stein, 1996). The representation of the star-nosed mole’s body also corresponds to a common pattern for the representation of sensory space in the optic tectum of reptiles (Hartline et al., 1978; Gaither and Stein, 1979; Stein and Gaither, 1983). The representation of appendage 11, the somato-

420

S.D. CRISH ET AL.

Fig. 3. The results of microelectrode recordings from the superior colliculus (SC) of case 9. A: The locations of electrode penetrations relative to the SC borders. Stars mark the locations of microlesions. B: A section of the flattened superior colliculus (SC) processed for the metabolic enzyme cytochrome oxidase. Note the locations of microle-

sions corresponding to the penetrations in A. The star at “K” indicates the position of a microlesion visible in a deeper section. IC, inferior colliculus. C: Receptive fields on the star and body for the electrode penetrations A–L. Rostral is left and medial is up.

sensory fovea of the star, is in approximately the same relative location as the representation of the retinal fovea, or area centralis, in superficial layers of more visual mammals. However, it is clear that vision does not dictate the organization of sensory space in star-nosed moles, as has been suggested for the somatosensory maps of other species. The results raise a number of issues for discussion, such as how information is differently represented in the SC and neocortex, what parallels there are between visual and somatosensory representations of SC information, and how the star-nosed mole’s SC differs anatomically from that of sighted mammals.

tion of the star, which is relevant to our current findings in the colliculus because heavy corticotectal connections have been reported to originate from nonprimary somatosensory areas (Stein et al., 1983), particularly when projecting to deep layers of the SC (Huerta and Harting, 1984). Although such connections have yet to be investigated in the star-nosed mole, the unusually large nonprimary representations in star-nosed moles could play an important role in this circuit. The star representation takes up approximately 40% of the body representation in all of the cortical somatosensory areas combined (27). Similarly, the star takes up roughly half of the body representation in the SC (Figs 5, 6). In both cases this enlargement is a testament to the behavioral importance of the star in exploratory behaviors. However, there are some clues to the way information is represented differently in the cortex and the colliculus that seem consistent with two fundamentally different roles in processing sensory information. There were three main differences between the cortical and collicular representations that were evident from our data. First, the relative degree of magnification for the somatosensory fovea (appendage 11) differed between the SC and cortex. In the cortex, appendage 11 is greatly magnified, taking up at least 25% of the S1 star representation and a similar proportion of S2 (Catania and Kaas,

Comparison of cortical and collicular representations In star-nosed moles, both the cortical and collicular somatosensory areas consist of topographic representations of the contralateral body surface. However, there are a number of differences in the way sensory information is represented in the two structures. The most obvious difference is that somatosensory cortex contains two distinct representations of the mole’s body and three representations of the star (Catania, 2000). These include primary somatosensory cortex (S1), secondary somatosensory cortex (S2), and a third representation (S3). Star-nosed moles have a particularly large and distinctive S2 representa-

STAR-NOSED MOLE SUPERIOR COLLICULUS

Fig. 4. Overlap of the representations of different nasal appendages shown for recordings from the star representation in different cases. The numbers correspond to the appendages on the star represented in each region. In contrast to the segregated and discrete representations of the appendages in neocortex (Catania and Kaas, 1995), the representations of the appendages in the SC showed considerable overlap. Scale bar ⫽ 1 mm.

1995, 1997; Catania, 2000). We were therefore expecting to find a large representation of the 11th appendage in the colliculus, but this was not the case, Rather, the representation of the 11th appendage was comparable in size to the other appendages (Fig. 5). This may reflect the different role of the fovea in the visual versus the somatosensory system. In the visual system, for example, there may be many small corrective or exploratory saccades that would presumably require a correspondingly large amount of processing territory in the colliculus, whereas this does not appear to be the case for foveation with the tactile star (Catania and Kaas, 1997). A second difference between the cortex and the colliculus was the overlap of appendage representations for the star. In the mole’s cortex, the representation of each ray of the star is remarkably discrete, such that a single isomorphic module of the cortex represents each separate appendage. In contrast, the representations of the rays in the SC overlapped each other extensively (Figs. 4, 5C). This is similar to the rodent whisker system, in which cortical barrels are distinct representations of each whis-

421 ker (Woolsey and Van der Loos, 1970), but overlapping representations have been reported for the whiskers in the SC (Dra¨ ger and Hubel, 1976). Little overlap was observed between the star as a whole and the representations of other body parts. Finally, the sizes of the receptive fields in the SC were much larger than in the cortex (Fig. 5D). The receptive fields for neurons in primary somatosensory cortex are among the smallest reported for a mammalian skin surface— often less than 1 mm2 in area (Sachdev and Catania, 2002). Cortical receptive fields typically cover only a single appendage, unless the recording electrode is in a border site between appendage representations. In contrast, the receptive fields for neurons in the SC were much larger and usually covered more than one appendage. The differences between the cortical and collicular representations of the star are consistent with the different roles that have been historically proposed for these structures. Large receptive fields relative to cortical representations are a hallmark of visual field representation in the colliculus (McIlwain and Buser, 1968; Sterling and Wickelgren, 1969; Schiller and Koerner, 1971; Goldberg and Wurtz, 1972; and see Stein and Meredith, 1993), and, correspondingly distributed, population coding of sensory representations is central to theories of collicular function (McIlwain, 1976, 1991; Findlay 1982; Findlay and Walker, 1999). The premise is that, although no single neuron or small group of neurons codes for the precise location of a stimulus, positional information can be accurately extracted from the population of active neurons with large, overlapping receptive fields. Similarly, appropriate motor outputs could be coded by the summation of premotor activity across the colliculus (Lee et al., 1988; Sparks et al., 1990). Presumably, information processing at the cortical level is more heavily involved in the analysis of object-specific feature extraction necessitating relatively larger numbers of cells with correspondingly smaller receptive fields, as well as multiple representations of the star to process different facets of touch information. It has also been proposed that large receptive fields in the SC are necessary to maintain the integrative function of the colliculus across sensory modalities that have changing frames of reference relative to one another (Stein and Meredith, 1993). For example, if bimodal neurons that receive inputs from the eyes and ears had small receptive fields, they would presumably lose the additive effects of coordinated inputs when the eyes were shifted off the midline and out of register with the reference frame for the ears (the fixed skull). In support of this hypothesis, Stein and Meredith (1993) point out that small receptive fields are found for the auditory and visual neurons in the optic tectum of barn owls, a species that keeps its ears and eyes in the same reference frame. (They move their entire head to foveate.) Of course this does not explain the large receptive fields for the nasal appendages of the star-nosed mole, as vision is clearly not coordinated with somatosensory information from the star. It is worth considering another possible influence that may be related in many ways to the former functional considerations. Receptive field size in cortex has generally been found to have an inverse correlation with cortical magnification (e.g., Sur et al., 1980). It is clear that the collicular representation of the star is smaller than the cortical representation (1.4 mm2 for the SC and 3.2 mm2

422

S.D. CRISH ET AL.

Figure 5

STAR-NOSED MOLE SUPERIOR COLLICULUS

423

Fig. 6. Highly schematized representation of the star-nosed mole’s body proportions in the superior colliculus (see Fig. 5A for details). The contralateral half of the body is represented in an upright position with the rostral body parts represented at the rostral end of the

colliculus. The star has a greatly magnified representation and takes up approximately half of the somatosensory map. (Figure by Portia Sloan.)

for primary somatosensory cortex). We might therefore expect larger receptive fields in the SC compared with the cortex, based on anatomical considerations.

reduced in thickness compared with more visual mammals. Our receptive field data were collected from superficial layers—approximately 100 –300 ␮m from the SC surface. This raises the question of what layers of the star-nosed mole’s SC are homologous to the collicular layers described for more visual mammals. There has been some disagreement regarding the nomenclature for collicular layers between the cat (Kanaseki and Sprague, 1974) and the rodent. Wiener’s (1986) nomenclature for the mouse provides a detailed anatomical review that is particularly appropriate for comparison with star-nosed moles. The deeper layers of the star-nosed mole SC share clear anatomical similarities to those of the mouse, as outlined by Wiener for cytochrome oxidase-processed material (Fig. 7). A series of large fiber bundles that run rostrocaudally provides an obvious landmark visible in both mice and star-nosed moles (Fig. 7). These appear as small light circles in transverse sections processed for cytochrome oxidase and correspond to layer IVb. Deep to layer IVb, Wiener identifies a thin transitional layer IVc followed by layers V–VII, which consist of different regions of mediolaterally running fibers just superior to the central gray (see Figs. 2, 4a, and 15, Wiener, 1986). All these areas correspond well to the anatomy of the deep layers of the star-nosed mole SC (Fig. 7). However, it is harder to compare the superficial layers (I–III) normally devoted to vision across star-nosed moles and mice. The main question is whether the superficial layers from which our mapping

Anatomy of the star-nosed mole Superior colliculus It is obvious from transverse sections (Figs. 1, 7) that star-nosed moles have a colliculus that is proportionally

Fig. 5. Summary of the results of microelectrode recordings from the star-nosed mole superior colliculus (SC). A: Compilation of the recording points from all eight colliculi investigated in this study. The entire contralateral body surface was represented in the SC in an upright position with the head oriented rostrally and caudal body parts (tail and hindlimb) represented caudally. The star representation dominated the colliculus, occupying approximately half of the responsive area. B: Enlargement of the star representation illustrated in A. Within the star representation, the nasal appendages formed a pinwheel from 1 to 11 in the clockwise direction with appendage 11 at the rostrolateral extreme. Neurons at many penetrations responded to tactile stimulation of more than one appendage (numbers in boxes). C: The representations of different appendages overlapped one another in the SC. For clarity, the extent of the representations of only the odd appendages is illustrated (compiled from all eight cases). Because this summary is a compilation of recording results from all the cases, the degree of overlap is likely to be less than illustrated here (see Fig. 4 for individual cases). D: The sizes of the receptive fields in the SC (top) are much larger than those found in S1 (bottom), the primary somatosensory cortex (modified from Sachdev and Catania, 2002). Aud, auditory stimulation; caud tr, caudal trunk; chst, chest; dor, dorsal; FL, forelimb; hd, head; HL, hindlimb; shd, shoulder; TL, tail; vib, vibrissae.

424

S.D. CRISH ET AL.

Fig. 7. A transverse section through the star-nosed mole superior colliculus compared with that of a mouse. The deep layers of both species are numbered according to the nomenclature of Wiener (1986) for the mouse. Layer IVb is particularly distinctive in both species and provides a frame of reference for homologizing the deeper layers

V–VII. The identity of layers superficial to IVb in star-nosed moles is less certain. One possibility is that layer IVa is hypertrophied in star-nosed moles. Alternatively, layer II or III may have been co-opted for somatosensation in star-nosed moles. CG, central gray.

data were obtained correspond to one of the traditional visual layers, to a thick layer IVa, to a combination of layers, or perhaps to a specialization unique to moles. This issue might be addressed by examining retinotectal inputs from tracer injections in the eye (Cooper et al., 1993a,b). In summary, our results reveal an SC dominated by somatosensory inputs. The representation of the mole’s heavily innervated mechanosensory star is greatly magnified in the collicular map, and it seems likely that the SC plays an important role in the nearly continuous orienting movements made by star-nosed moles as they explore their environment through touch. Although the colliculus is clearly reduced in thickness compared with more visual mammals, this reduction is not uniform, rather, deeper layers have been retained in a normal-looking configuration, and there is some evidence for selective enlargement of particular subsections of the deep colliculus (layer IVa). An alternative possibility is that some superficial layers (II–III) have been retained and now process tactile information in star-nosed moles.

progression of the visual system in a blind subterranean mammal. Nature 361:156 –159. Cooper HM, Herbin M, Nevo E. 1993b. Visual system of a naturally microphthalmic mammal: the blind mole rat, Spalax ehrenbergi. J Comp Neurol 328:313–350. Cynader M, Berman N. 1972. Receptive-field organization of monkey superior colliculus. J Neurophysiol 35:187–201. Dra¨ ger UC, Hubel DH. 1976. Topography of visual and somatosensory projections to mouse superior colliculus. J Neurophysiol 39:91–101. Findlay JM. 1982. Global visual processing for saccadic eye movements. Vision Res 22:1033–1045. Findlay JM, Walker R. 1999. A model of saccade generation based on parallel processing and competitive inhibition. Behav Brain Sci 22: 661–721. Franklin KBJ, Paxinos G. 1997. The mouse brain in stereotaxic coordinates. San Diego: Academic Press. Gaither NS, Stein BE. 1979. Reptiles and mammals use similar sensory organizations in the midbrain. Science 205:595–597. Goldberg ME, Wurtz RH. 1972. Activity of superior colliculus in behaving monkey, I. Visual receptive fields of single neurons. J Neurophysiol 35:542–559. Hamilton WJ. 1931. Habits of the star-nosed mole, Condylura cristata. J Mammal 12:345–355. Hartline PH, Kass L, Loop MS. 1978. Merging of modalities in the optic tectum: infrared and visual integration in rattlesnakes. Science 199: 1225–1229. Huerta MF, Harting JK. 1984. The mammalian superior colliculus: studies of its morphology and connections. In: Vanegas H, editor. Comparative neurology of the optic tectum. New York: Plenum Press. p 687–757. Kanaseki T, Sprague JM. 1974. Anatomical organization of pretectal nuclei and tectal laminae in the cat. J Comp Neurol 158:319 –337. Knudsen EI. 1999. Mechanisms of experience-dependent plasticity in the auditory localization pathway of the barn owl. J Comp Physiol [A] 185:305–321. Knudsen EI, Brainard MS. 1991.Visual instruction of the neural map of auditory space in the developing optic tectum. Science 253:85– 87. Kruger L. 1970. The topography of the visual projection to the mesencephalon: a comparative survey. Brain Behav Evol 3:169 –177. Lee C, Rohrer WH, Sparks DL. 1988. Population coding of saccadic eye movements by neurons in the superior colliculus. Nature 332:357–360. McIlwain JT. 1976. Large receptive fields and spatial transformations in the visual system. In: Porter R, editor. Neurophysiology II, vol 10. Baltimore, MD: University Park Press. p 223–248. McIlwain JT. 1991. Distributed spatial coding in the superior colliculus: a review. Vis Neurosci 6:3–13. McIlwain JT, Buser P. 1968. Receptive fields of single cells in the cat’s superior colliculus. Exp Brain Res 5, 314 –325.

ACKNOWLEDGMENTS Special thanks to Fiona Remple for help with many phases of the research and critical comments on the manuscript and to Erin Henry for help in collecting starnosed moles.

LITERATURE CITED Catania KC. 1999. A nose that looks like a hand and acts like an eye: the unusual mechanosensory system of the star-nosed mole. J Comp Physiol 185:367–372. Catania KC. 2000. Cortical-organization in moles: evidence of new areas and a specialized S2. Somatosens Mot Res 17:335–347. Catania KC, Kaas JH. 1995. Organization of the somatosensory cortex of the star-nosed mole. J Comp Neurol 351:549 –567. Catania KC, Kaas JH. 1997. Somatosensory fovea in the star-nosed mole: behavioral use of the star in relation to innervation patterns and cortical representation. J Comp Neurol 387:215–233. Chalupa LM. 1984. Visual Physiology of the mammalian superior colliculus. In: Vanegas H, editor. Comparative neurology of the optic tectum. New York: Plenum Press. p 775– 818. Cooper HM, Herbin M, Nevo E. 1993a. Ocular regression conceals adaptive

STAR-NOSED MOLE SUPERIOR COLLICULUS Paxinos G, Watson C. 1986. The rat brain in stereotaxic coordinates, 2nd ed. San Diego: Academic Press. Sachdev RN, Catania KC. 2002. Receptive fields and response properties of neurons in the star-nosed mole’s somatosensory fovea. J Neurophysiol 87:2602–2611. Schiller PH, Koerner F. 1971. Discharge characteristics of single units in superior colliculus of the alert rhesus monkey. J Neurophysiol 34:920 – 936. Sparks DL, Lee C, Rohrer WH. 1990. Population coding of the direction, amplitude, and velocity of saccadic eye movements by neurons in the superior colliculus. Cold Spring Harb Symp Quant Biol 55:805– 811. Stein BE. 1981. Organization of the rodent superior colliculus: some comparisons with other mammals. Behav Brain Res 3:175– 88. Stein BE. 1984. Multimodal representation in the superior colliculus and optic tectum. In: Vanegas H, editor. Comparative neurology of the optic tectum. New York: Plenum Press. p. 819 – 841. Stein BE. 1998. Neural mechanisms for synthesizing sensory information and producing adaptive behaviors. Exp Brain Res 123:124 –135 Stein BE, Gaither NS. 1983. Receptive-field properties in reptilian optic tectum: some comparisons with mammals. J Neurophysiol 50:102–24. Stein BE, Meredith MA. 1993. The merging of the senses. Cambridge, MA: MIT Press. Stein BE, Magalhaes-Castro B, Kruger L. 1975. Superior colliculus: visuotopic-somatotopic overlap. Science 189:224 –226.

425 Stein BE, Magalhaes-Castro B, Kruger L. 1976. Relationship between visual and tactile representations in cat superior colliculus. J Neurophysiol 39:401– 419. Stein BE, Spencer RF, Edwards SB. 1983. Corticotectal and corticothalamic efferent projections of SIV somatosensory cortex in cat. J Neurophysiol 50:896 –909. Sterling P, Wickelgren BG. 1969. Visual receptive fields in the superior colliculus of the cat. J Neurophysiol 32:1–15. Sur M, Merzenich MM, Kaas JH. 1980. Magnification, receptive-field area, and “hypercolumn” size in areas 3b and 1 of somatosensory cortex in owl monkeys. J Neurophysiol 44:295–311. Sze´ kely G. 1973. Anatomy and synaptology of the optic tectum. In: Jung BR, editor. Handbook of sensory physiology, vol VII/3, Central visual information. Berlin: Springer-Verlag. p. 1–26. Wallace MT, Stein BE. 1996. Sensory organization of the superior colliculus in cat and monkey. Prog Brain Res 112:301–311. Wiener SI. 1986. Laminar distribution and patchiness of cytochrome oxidase in mouse superior colliculus. J Comp Neurol 244:137–148. Woolsey TA, Van der Loos H. 1970. The structural organization of layer IV in the somatosensory region (SI) of mouse cerebral cortex. The description of a cortical field composed of discrete cytoarchitectonic units. Brain Res 17:205–242. Wong-Riley MT, Carroll EW. 1984. Quantitative light and electron microscopic analysis of cytochrome oxidase-rich zones in V II prestriate cortex of the squirrel monkey. J Comp Neurol 222:18 –37.