3-immunoreactive neurons in the cat, rabbit, and hamster superficial superior colliculus

Neuroscience Research 49 (2004) 139–155

Ionotropic glutamate receptor GluR2/3-immunoreactive neurons in the cat, rabbit, and hamster superficial superior colliculus Won-Mee Park, Min-Jeong Kim, Chang-Jin Jeon∗ Department of Biology, College of Natural Sciences, Kyungpook National University, 1370 Sankyuk-dong, Daegu 702-701, South Korea Received 25 September 2003; accepted 5 February 2004

Abstract Ionotropic glutamate receptor (GluR) subtypes occur in various types of cells in the central nervous system. We studied the distribution of AMPA glutamate receptor subtype GluR2/3 in the superficial layers of cat, rabbit, and hamster superior colliculus (SC) with antibody immunocytochemistry and the effect of enucleation on this distribution. Furthermore, we compared this labeling to that of calbindin D28K and parvalbumin. Anti-GluR2/3-immunoreactive (IR) cells formed a dense band of labeled cells within the lower superficial gray layer (SGL) and upper optic layer (OL) in the cat SC. By contrast, GluR2/3-IR cells formed a dense band within the upper OL in the rabbit and within the OL in the hamster SC. Calbindin D28K-IR cells are located in three layers in the SC: one within the zonal layer (ZL) and the upper SGL in all three animals, a second within the lower OL and upper IGL in the cat, within the IGL in the rabbit and within the OL in the hamster, and a third within the deep gray layer (DGL) in all three animals. Many parvalbumin-IR neurons were found within the lower SGL and upper OL. Thus, the GluR2/3-IR band was sandwiched between the first and second layers of calbindin D28K-IR cells in the cat and rabbit SC while the distribution of GluR2/3-IR cells in the hamster matches the second layer of calbindin D28K-IR cells. The patterned distribution of GluR2/3-IR cells overlapped the tier of parvalbumin-IR neurons in cat, but only partially overlapped in hamster and rabbit. Two-color immunofluorescence revealed that more than half (55.1%) of the GluR2/3-IR cells in the hamster SC expressed calbindin D28K. By contrast, only 9.9% of GluR2/3-IR cells expressed calbindin D28K in the cat. Double-labeled cells were not found in the rabbit SC. Some (4.8%) GluR2/3-IR cells in the cat SC also expressed parvalbumin, while no GluR2/3-IR cells in rabbit and hamster SC expressed parvalbumin. In this dense band of GluR2/3, the majority of labeled cells were small to medium-sized round/oval or stellate cells. Immunoreactivity for the GluR2/3 was clearly reduced in the contralateral SC following unilateral enucleation in the hamster. By contrast, enucleation appeared to have had no effect on the GluR2/3 immunoreactivity in the cat and rabbit SC. The results indicate that neurons in the mammalian SC express GluR2/3 in specific layers, which does not correlate with the expression of calbindin D28K and parvalbumin among the animals. © 2004 Elsevier Ireland Ltd and The Japan Neuroscience Society. All rights reserved. Keywords: Immunocytochemistry; Localization; Enucleation; Calbindin; Parvalbumin

1. Introduction Glutamate is the major excitatory neurotransmitter in the central nervous system. Glutamate and its receptors are thought to play an important function in a variety of cellular mechanisms such as synaptogenesis, neural plasticity, neurodevelopment, and neurodegeneration (Sheng and Kim, 2002; Song and Huganir, 2002). Glutamate receptors are classified into two distinct types termed ionotropic and metabotropic receptors. The ionotropic glutamate receptors mediate faster synaptic transmission than metabotropic ∗ Corresponding author. Tel.: +82-53-950-5343; fax: +82-53-953-3066. E-mail address: [email protected] (C.-J. Jeon).

receptors that modulate postsynaptic ion channels by activating G-proteins. Ionotropic glutamate receptors can be further subdivided according to their agonist selectivity. The three subtypes of ionotropic glutamate receptors are AMPA (␣-amino-3-hydroxyl-5-methyl-4-isoxazolepropionate), NMDA (N-methyl-d-aspartate) and kainate receptors. There are at least four forms of AMPA receptor subunits (GluR1–4), five forms of NMDA (NMDAR1 and NMDAR2A–D), and kainate receptor subunits (GluR5–7 and KA1–2) (Hollmann and Heinemann, 1994; Ben-Ari et al., 1997; Michaelis, 1998; Thoreson and Witkovsky, 1999; Meldrum, 2000). These receptor subtypes occur in various types of cells in the central nervous system with a high degree of regional and cellular specialization (WongRiley and Jacobs, 2002; Van Damme et al., 2003).

0168-0102/$ – see front matter © 2004 Elsevier Ireland Ltd and The Japan Neuroscience Society. All rights reserved. doi:10.1016/j.neures.2004.02.009

140

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

The mammalian superior colliculus (SC) is the center of visuo-motor integration. It is a seven-layered structure on the roof of the midbrain and can be divided into superficial and deep layers on a functional and anatomical basis. The three superficial layers (zonal, superficial gray, and optic), which receive their major input from the retina and the visual cortex, are concerned exclusively with the processing of visual information. In contrast, the four deeper layers (intermediate gray, intermediate white, deep gray, and deep white), which receive auditory, somatic, and visual input from numerous cortical and subcortical areas, are concerned with the generation of head, eye, and ear movements (Huerta and Harting, 1984; Grantyn, 1988; Harting et al., 1991; Isa, 2002). One of the principal organizing features of the SC is the topographical distribution of its afferent fibers and efferent cells. Many afferent fibers and efferent cells of the SC are segregated into specific laminae or form patches or clusters. Several neurotransmitters, peptides, and enzymes show horizontal, vertical, or patch-like segregation in the SC. An example is the horizontal laminar segregation of anti-calcium-binding protein-immunoreactive (IR) neurons. Calcium-binding proteins are thought to mediate a wide variety of cellular mechanisms of calcium. Among the many calcium-binding proteins, at least three types of calciumbinding proteins, calbindin D28K, calretinin, and parvalbumin abundantly occur in distinct subpopulations of neurons in the central nervous system (Heizmann et al., 1990; Rogers et al., 1990; Baimbridge et al., 1992; Polans et al., 1996; Schäfer and Heizmann, 1996). Thus, the calcium-binding protein calbindin D28K is found in cells that are located in the three layers of the cat SC: one within the upper one half of the superficial layer, the second bridging the deep optic (OL) and intermediate gray layer (IGL), and the third within the deep gray layer (DGL) (Mize et al., 1991). Similar results have been reported in rabbit, rat, and hamster SC. However, in the rabbit, rat, and hamster SC, the second tier of calbindin D28K-IR neurons was located in the OL (Behan et al., 1992; Mize and Luo, 1992; Schmidt-Kastner et al., 1992; Leuba and Saini, 1996; González-Soriano et al., 2000; Kang et al., 2002). Parvalbumin-IR cells formed a single dense band in the deep superficial gray (SGL) and OL with loosely scattered cells in the deep layers in the cat and rat (Mize et al., 1992; Cork et al., 1998). The distribution of parvalbumin-IR neurons in the cat and rat is thus complementary to that of calbindin D28K-IR neurons. Calretinin forms a dense plexus of immunoreactive fibers in the superficial layers of cat (Hong et al., 2002), rat (Rogers and Resibois, 1992), mouse (Gobersztejn and Britto, 1996), and hamster SC (Kang et al., 2002). However, instead of a dense plexus of calretinin-IR fibers, many calretinin-IR cells are localized in the superficial layers of the rabbit SC (Jeon et al., 1998). These results indicate that there are considerable species differences in the distribution of calcium-binding proteins in the SC. Glutamate-IR neurons also exhibit laminar segregation in the SC. In the cat SC,

there is a dense band of highly anti-glutamate-IR neurons in the deep SGL and upper OL (Jeon et al., 1997a). The retino-collicular and cortico-collicular pathways use glutamate as the major neurotransmitter (Mize and Butler, 1996, 2000; Jeon et al., 1997b; Binns, 1999). The excitatory synaptic transmission of the retino-collicular pathway is mediated by ionotropic glutamate receptors (Lo et al., 1998; Isa et al., 1998). There is substantial evidence regarding the presence of ionotropic glutamate receptors in the mammalian SC (Petralia and Wenthold, 1992; Sato et al., 1993; Petralia et al., 1994; Binns, 1999). Recently, AMPAtype glutamate receptor subunits GluR1 and GluR2 have been identified in the OL of rat SC (Kondo et al., 2000) and NMDA receptors have been observed at the postsynaptic sites in both the retinal and cortical afferents in the cat SC (Mize and Butler, 2000). AMPA receptors are formed by a combination of four or five protein subunits (GluR1–4) and are responsible not only for the rapid synaptic transmission produced by glutamate but also for neuronal development, synaptic plasticity, and neurodegeneration (Michaelis, 1998; Conti and Weinberg, 1999; Sheng and Kim, 2002; Song and Huganir, 2002). In the SC of most mammals, it is not known whether the different subtypes of GluRs segregate into different layers. The differences in the receptor subunit composition are key factors in understanding the distinct functional characteristics of the AMPA receptors. In addition, the differential localization of glutamate receptors is particularly important in the SC in which glutamate serves as the neurotransmitter and the retino-collicular and cortico-collicular pathways use glutamate. Thus, in the present study we investigated the organization of AMPA receptors-IR neurons using commercially available specific antibodies. First, we examined the distribution and morphology of GluR2/3-IR neurons to determine if this receptor subtype is localized in the specific lamina and specific cell types of cat SC as this animal has been most widely used for various anatomical and physiological studies of the SC. As other receptor subtypes (GluR1 and GluR4) were not distinctively localized in collicular neurons in our present study, we focused on GluR2/3IR neurons first. Second, we examined whether GluR2/3 is specifically localized in subpopulations of calcium-binding protein calbindin D28K- and parvalbumin-containing neurons. GluR2 subunit determines the calcium-permeability of the functional glutamate receptor. Calbindin D28K and parvalbumin play important roles in calcium buffering and protection from excitotoxic cell death. In addition, in the cat SC many GluR2/3-IR neurons in the superficial layers are located within the tiers of the calbindin D28K- and parvalbumin-IR cells (Mize et al., 1991, 1992). Third, we investigated whether there were any species differences in the distribution of this receptor as various species are used in anatomical and physiological studies of collicular neurotransmission. It is well-known that great interspecies differences in specialization of neurochemical properties of the brain exist. Understanding those differences is very impor-

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

tant in understanding species diversity. We compared the distribution of GluR2/3 in cat SC with other non-primate species (rabbit and hamster) which are commonly used in various anatomical and physiological studies of SC. Finally, we investigated whether GluR2/3 immunoreactivity is affected by enucleation, especially whether the effect can be produced in all three mammalian species tested. It is wellknown that the chemical properties of neurons can be altered by environmental modifications in adult animals. We wanted to see if there were alterations of the numbers of GluR2/3-IR cells when reducing glutamate activity from retina through the enucleation. Our results show that the organizational features of the GluR2/3 found in mammalian SC are strikingly different among species. In our study, the GluR2/3-IR cells contained abundant calbindin D28K in the hamster, some in the cat, but none in the rabbit. Some of the GluR2/3-IR cells in cat SC contained parvalbumin, while no GluR2/3-IR cells contains parvalbumin in the rabbit and hamster. Enucleation produced a clear reduction of GluR2/3 immunoreactivity in the SC contralateral to the enucleation in the hamster, but not in the cat and rabbit.

141

0.002% calcium chloride added. Following a pre-rinse with phosphate-buffered saline (PBS, pH 7.2, 500 ml for cat and rabbit and 30 ml for hamster) over a period of 1–4 min, each animal was perfused with a fixative (1800–2000 ml for cat and rabbit and 30–50 ml for hamster). This was for 1–2 h for cats and rabbits and for 5–10 min for hamsters via a syringe needle inserted through the left ventricle and aorta. The head was then removed and placed in the fixative for 2–3 h. The brain was then removed from the skull and stored 2–3 h in the same fixative and left overnight in 0.1 M phosphate buffer (pH 7.4) containing 8% sucrose and 0.002% CaCl2 . The SC was removed, mounted onto a chuck, and cut coronally into 50 ␮m thick sections with a vibratome. For every four or five sections, three or four sections were used for immunocytochemistry and one was used for thionin. The thionin stained sections were used to identify collicular layers. The collicular layers were identified by staining and cellular densities and morphologies of SC neurons (Huerta and Harting, 1984). We also identified collicular layers using calbindin and parvalbumin labeled sections where the collicular layers of these proteins have been well described previously (Mize et al., 1991, 1992; Behan et al., 1992; González-Soriano et al., 2000).

2. Materials and methods 2.3. HRP immunocytochemistry 2.1. Animals Adult cats (2.5–3.0 kg, n = 15), adult New Zealand white rabbits (2.5–3.0 kg, n = 15), and golden hamsters (25–30 g, n = 15) were used for these experiments. The animals were divided into two groups. First, intact animals (n = 9) were used to determine the normal distribution of immunoreactivity to GluR2/3 and to the calcium-binding proteins, calbindin D28K, and parvalbumin, in the SC. Second, unilaterally (n = 6) enucleated animals were used to examine the effects of retinal deafferentation. Enucleation was performed (cats, between 60 and 90 weeks of age; rabbit, between 30–50 weeks of age; and hamster, between 15 and 20 weeks of age) under anesthesia with a mixture of ketamine hydrochloride (30–40 mg/kg) and xylazine (3–6 mg/kg), supplemented as needed to maintain anesthesia. Proparacaine HCl (200–300 ␮l) was applied to the cornea to suppress blink reflexes. The enucleated animals were allowed to survive for 10 (n = 3) and 20 (n = 3) days. The National Institute of Health guidelines for the use and care of animals were followed for all experimental procedures. All efforts were made to minimize animal suffering as well as the number of animals used. 2.2. Perfusion and tissue processing All animals were anesthetized deeply with a mixture of ketamine hydrochloride (30–40 mg/kg) and xylazine (3–6 mg/kg) before perfusion. They were perfused intracardially with 4% paraformaldehyde and 0.1–0.3% glutaraldehyde in 0.1 M sodium phosphate buffer (pH 7.4) with

A polyclonal antibody against GluR2/3 (AB1506) was obtained commercially from Chemicon (Temecula, CA, USA). This antibody has been widely used for the specific localization of GluR2/3 in the central nervous system of cat (Qin and Pourcho, 1999), rabbit (Firth et al., 2003), and hamster (Ryoo et al., 2003). Monoclonal antibodies against calbindin D28K (clone CB-955) and parvalbumin (clone PA235) were obtained from Sigma Chemical (St. Louis, MO, USA). The tissue was processed free floating in small vials. For immunocytochemistry, the sections were incubated in 1% sodium borohydride (NaBH4 ) for 30 min. Sections were rinsed 3 × 10 min in 0.1 M phosphate buffer, and incubated in 0.1 M phosphate buffer with 4% normal serum (normal goat serum for GluR2/3 and normal horse serum for calbindin D28K and parvalbumin) for 2 h with 0.5% Triton X-100 added. Sections were then incubated in the primary antiserum in 0.1 M phosphate buffer with 4% normal serum for 48 h with 0.5% Triton X-100 added. The primary antiserum was diluted from 1:100 to 1:1000 (GluR2/3), 1:200 to 500 (calbindin D28K), or 1:500 to 1000 (parvalbumin). Following 3 × 10 min rinses in 0.1 M phosphate buffer, sections were incubated in a 1:200 dilution of biotinylated secondary IgG in 0.1 M phosphate buffer with 4% normal serum for 2 h with 0.5% Triton X-100 added. Sections were then rinsed for 3 × 10 min in 0.1 M phosphate buffer and incubated in a 1:50 dilution of avidin-biotinylated horseradish peroxidase (Vector lab, Burlingame, CA, USA) in 0.25 M Tris for 2 h. The sections were again rinsed in 0.1 M phosphate buffer for 3 × 10 min. Finally, the staining was visualized by reacting with 3,3 -diaminobenzidine tetrahy-

142

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

drochloride (DAB) and hydrogen peroxide in 0.25 M Tris for 3–10 min using a DAB reagent set (Kirkegaard & Perry, Gaithersburg, MD, USA). All sections were then rinsed in 0.1 M phosphate buffer before mounting. As a control, some sections were incubated in the same solution without addition of the primary antibody. The control sections showed no GluR2/3 or calbindin D28K or parvalbumin immunoreactivities. Following the immunocytochemical procedures, the tissue was mounted on Superfrost Plus slides (Fisher, Pittsburgh, PA, USA) and dried overnight in a 37 ◦ C oven. The mounted sections were dehydrated through alcohol, cleared with xylene, and coverslipped with the mounting medium, Permount (Fisher). The tissue was examined and photographed on a Zeiss Axioplan microscope using conventional or differential interference contrast (DIC) optics. 2.4. Fluorescence immunocytochemistry To generate two simultaneous labels, sections were incubated in the primary antiserum by using the steps described above. For detection by immunofluorescence, the secondary antibodies were fluorescein conjugated anti-rabbit IgG (Vector lab) to detect the anti-GluR2/3 antibody and Texas red (Sigma) or Cy5 (Jackson ImmunoResearch Lab.) conjugated anti-mouse IgG to detect the anti-calbindin D28K or antiparvalbumin antibody. Labeled sections were coverslipped with the Vectashield mounting medium (Vector lab). Some tissue sections labeled with fluorescent compounds were examined with a Zeiss Axioplan microscope using a Zeiss filter set 05 (excitation BP 395–440 nm) or set 09 (excitation BP 450–490 nm) for fluorescein and set 15 (excitation BP 546/12 nm) for Texas red fluorescence. Some tissue sections labeled with fluorescent compounds were also examined and photographed on a Bio-Rad MRC-1024 Confocal Microscope. 2.5. Quantitative analysis All analyses were done with a 40X Zeiss PlanApochromat objective. The percentage of double-labeled neurons was determined by using a square counting reticule placed in the binocular of a Zeiss Axioplan Universal microscope. We sampled from five sequential fields, each 310 ␮m × 310 ␮m in area, across the medial-lateral extent of the SC. Each field was positioned at approximately equal intervals. Large blood vessels were also excluded from the measurement fields by moving the field slightly to avoid biasing the measures. We centered each field across the dense band of GluR2/3-IR cells of both sides of SC in six sections (two rostral, two middle, and two caudal SC) from each of two normal animals (total 120 fields). The number of GluR2/3-IR cells and of cells double-labeled with calbindin D28K or parvalbumin and GluR2/3 was counted in each field. The number of double-labeled cells was expressed as a percentage of the total population of GluR2/3-IR cells.

The morphological types of GluR2/3-IR cells were estimated on DAB-reacted sections of normal animals. We also sampled from five sequential fields, each 310 ␮m × 310 ␮m in area, across the dense band of GluR2/3-IR cells of both sides of the SC. Cell types were analyzed from the six (two rostral, two middle, and two caudal SC) best labeled sections from each of two normal animals (total 120 fields). To obtain the best images, we analyzed cells under differential interference contrast optics. Only cell profiles containing a nucleus and at least a faintly visible nucleolus were included in this analysis. Since the goal of the present study was to obtain an estimate of each morphological cell type, no attempt was made to assess the total cell numbers of each neuronal subpopulation. To compare the number of GluR2/3-IR cells of the ipsilateral control and contralateral experimental sides, we also counted labeled cells in five sequential fields, each 310 ␮m × 310 ␮m in area, across the medial-lateral extent of the SC. Each field was centered over the dense band of immunoreactivity in the superficial layers. The fields were positioned so as to avoid large blood vessels that would affect the measurements. Six tissue sections (two rostral, two middle, and two caudal) from each of two animals enucleated for 20 days were measured. We summed the numbers of labeled cells obtained for each animal in order to compare the control (total 60 fields) and experimental sides (total 60 fields). The number of GluR2/3-IR cells on the contralateral experimental side was expressed as a percentage of the labeled cells on the ipsilateral control side. The ratios were evaluated statistically by a paired t-test between the ipsilateral controls and contralateral experimental sides.

3. Results 3.1. Distribution of anti-GluR2/3 immunoreactivity Anti-GluR2/3-immunoreactive (IR) cells were very selectively distributed in the cat, rabbit, and hamster superior colliculus (SC). However, their distribution pattern differed according to species. Fig. 1A shows anti-GluR2/3 immunoreactivity in the normal cat SC. In the cat SC, the labeled cells formed a dense band of highly IR neurons within the lower superficial gray (SGL) and upper optic layers (OL). This tier of highly IR cells could be seen throughout the rostral-caudal extent of the SC. The thickness of this tier was approximately 400 ␮m wide from the central portion of the mid-colliculus. By contrast, outside the dense band the majority of labeled cells were lighter and scattered. The tier of highly GluR2/3-IR cells lies between the first and second layers of calbindin D28K-IR neurons (Fig. 1B), while the tier of GluR2/3-IR cells overlaps the tier of parvalbumin-IR neurons (Fig. 1C). The distribution of calbindin D28K-IR neurons and parvalbumin was described previously for the cat SC. Calbindin D28K-IR cells in the cat SC form three distinct laminar layers: one within the zonal layer (ZL) and

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

143

Fig. 1. Low magnification photomicrographs showing the distribution of: (A) GluR2/3, (B) calbindin D28K- and (C) parvalbumin-immunoreactivity in the cat SC. The GluR2/3-IR cells were concentrated within the lower SGL and upper OL. Calbindin D28K-IR neurons were located in the ZL and upper SGL, and lower OL and upper IGL, while parvalbumin-IR neurons were located in the lower SGL and upper OL. ZL, zonal layer; SGL, superficial gray layer; OL, optic layer; IGL, intermediate gray layer. Bar = 200 ␮m.

the upper SGL, the second bridging the deep OL and upper intermediate gray layer (IGL), and the third within the deep gray layer (DGL) (Mize et al., 1991). ParvalbuminIR neurons were sparsely distributed within the upper SGL but were denser within the lower SGL and upper OL (Mize et al., 1992). In the rabbit SC, the GluR2/3-IR cells also formed a dense band of highly IR neurons. By contrast with the cat SC, however, the tier of highly IR cells was not found within the SGL. In the rabbit SC, the dense band of labeled cells was located within the upper OL. This tier of highly IR cells could be seen throughout the rostral-caudal extent of the SC (Fig. 2A). The thickness of this tier was approximately 300 ␮m wide from the central portion of the mid-colliculus. A few lightly labeled cells were seen in the other layers. The tier of highly GluR2/3-IR cells also lies between the first and second layers of calbindin D28K-IR neurons (Fig. 2B), while the tier of GluR2/3-IR cells overlaps parvalbumin-IR neurons within the lower OL (Fig. 2C). The distribution of calbindin D28K- and parvalbumin-IR neurons was described previously for the rabbit SC. Calbindin D28K-IR cells in

the rabbit SC form three distinct laminar layers: one within the ZL and upper SGL, the second within the IGL, and the third within the DGL (Jeon et al., 1998; González-Soriano et al., 2000). Parvalbumin-IR neurons were sparsely distributed within the upper SGL, but were denser within the lower SGL and upper OL (González-Soriano et al., 2000). In the hamster SC, the tier of highly GluR2/3-IR cells was found within the OL (Fig. 3A). This tier of highly IR cells could be seen throughout the rostral-caudal extent of the SC. The thickness of this tier was approximately 200 ␮m wide from the central portion of the mid-colliculus. Very few labeled cells were seen in the other layers. Thus, the tier of highly GluR2/3-IR cells overlaps the second tier of calbindin D28K-IR cells (Fig. 3B), while the tier of GluR2/3-IR cells overlaps parvalbumin-IR neurons within the lower OL (Fig. 3C). The distribution of calbindin D28K- and parvalbuminIR neurons has been described previously for the hamster SC. Calbindin D28K-IR cells in the hamster SC form three distinct laminar layers: one within the ZL and upper SGL, the second within the OL, and the third within the DGL. The third tier in the DGL did not form a continuous band;

144

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

Fig. 2. Low magnification photomicrographs showing the distribution of: (A) GluR2/3, (B) calbindin D28K- and (C) parvalbumin-immunoreactivity in the rabbit SC. The GluR2/3-IR cells were concentrated within the upper OL. Calbindin D28K-IR neurons were located in the ZL and upper SGL, and in the IGL, while parvalbumin-IR neurons were located in the lower SGL and upper OL. ZL, zonal layer; SGL, superficial gray layer; OL, optic layer; IGL, intermediate gray layer. Bar = 200 ␮m.

rather, the cells formed a cluster in the medial portion of the SC (Behan et al., 1992). Parvalbumin-IR neurons were sparsely distributed within the upper SGL but were denser within the lower SGL and upper OL (Kang et al., 2002). 3.2. Morphology of the anti-GluR2/3-IR neurons The labeled cells in the dense band exhibited a variety of morphologies. In the cat, the majority of GluR2/3-IR cells had small to medium-sized cell bodies. Within the band, we identified at least two major types based on the morphologies of the cell bodies and dendrites. The first one consisted of small to medium-sized stellate cells (856 of 2374 cells or 36.1%). Fig. 4A and B (arrowhead) show some representative small to medium-sized stellate cells. Stellate cells had stellate-shaped cell bodies with dendrites which extended in multiple directions. The second morphological type con-

sisted of small to medium-sized round or oval cells (992 of 2374 cells or 41.8%) (Fig. 4C and D). Some of these cells had bipolar processes. However, vertical fusiform cells (Fig. 4B, arrow, F) and horizontal cells (Fig. 4E) were also found. Vertical fusiform cells (308 of 2374 cells or 12.9%) had distinctive vertical fusiform cell bodies and bipolar dendrites directed dorso-ventrally relative to the surface of the SC. The horizontal cells (218 of 2374 cells or 9.2%) had horizontal fusiform cell bodies and at least two horizontally oriented dendrites. In the rabbit SC, the labeled cells were similar to those of the cat. As in the cat SC, the two major types of the labeled cells were small to medium-sized stellate and round or oval neurons. Fig. 5A shows some small to medium-sized round or oval cells (1335 of 2386 cells or 56.0%). Some of these cells showed a single large dendrite emanating from the soma. Stellate cells (600 of 2386 cells or 25.1%) (Fig. 5B

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

145

Fig. 3. Low magnification photomicrographs showing the distribution of: (A) GluR2/3, (B) calbindin D28K- and (C) parvalbumin-immunoreactivity in the hamster SC. The GluR2/3-IR cells were concentrated within the OL. Calbindin D28K-IR neurons were located in the ZL and upper SGL, OL, and in the deep gray layer, while parvalbumin-IR neurons were located in the lower SGL and upper OL. ZL, zonal layer; SGL, superficial gray layer; OL, optic layer; IGL, intermediate gray layer. Bar = 200 ␮m.

and C, large arrowheads) had stellate-shaped cell bodies with dendrites extending in multiple directions. However, vertical fusiform cells (321 of 2386 cells or 13.5%) with a thick, proximal dendritic stump directed towards the pial surface (Fig. 5B and C (arrows), and D) were also found. The horizontal cells (130 of 2386 cells or 5.4%) (Fig. 5C, small arrowhead) had horizontal fusiform cell bodies and at least two horizontally oriented dendrites. Horizontal cells were rarely found. In the hamster SC, the majority of the GluR2/3-IR cells within the dense band of the hamster SC consisted of small to medium-sized round or oval cells (1256 of 1959 cells or 64.1%). Fig. 6A and C show small to medium-sized round or oval cells. These cells usually possessed a single large dendritic stump emanating from the cell body. However, vertical fusiform cells (399 of 1959 cells or 20.4%) (Fig. 6B) with a thick, proximal dendritic stump directed towards the pial surface were also found. Horizontal cells (112 of 1959 cells or 5.7%) (Fig. 6C, arrowhead) were rarely found. In contrast to the cat and rabbit SC, stellate cells (192 of 1959 cells or 9.8%) (Fig. 6D) were not often found in the hamster SC.

3.3. Effect of monocular enucleation on GluR2/3 expression To determine whether enucleation affects the distribution of the GluR2/3-IR cells in the SC, we performed monocular enucleations in some animals. Fig. 7A shows the cat SC ipsilateral to the enucleation, while Fig. 7B shows the SC contralateral to the enucleation. Fig. 7C shows the rabbit SC ipsilateral to the enucleation, while Fig. 7D shows the SC contralateral to the enucleation. Enucleation appeared to have no effect on the distribution of the GluR2/3-IR cells of the cat and rabbit SC. Overall differences in numbers of labeled neurons were not statistically significant on the contralateral side compared to the ipsilateral controls (paired ttest, P < 0.25 in cat; P < 0.94 in rabbit) (Table 1). Fig. 7E shows the hamster SC ipsilateral to the enucleation, while Fig. 7F shows the SC contralateral to the enucleation. An obvious reduction in immunoreactivity to GluR2/3 within the hamster SC contralateral to the enucleation appeared 10 and 20 days after enucleation. Immunoreactivity was clearly reduced both in the cytoplasm and in proximal dendrites. Quantitatively, after unilateral enucleation there was a significant decrease (1027 versus 1592 cells or 35.8%) in the

146

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

Fig. 4. High magnification differential interference contrast micrographs of anti-GluR2/3-IR cells in the lower SGL and upper OL of the cat SC. (A) Three small to medium-sized stellate cells with multiple dendrites are seen. (B) A stellate cell (arrowhead) and a vertical fusiform cell (arrow). (C) and (D) Small to medium-sized round or oval cells. (E) Horizontal cell with fusiform cell body and horizontally oriented dendrites. (F) Vertical fusiform cells with a thick, proximal dendrite directed towards the pial surface. Bar = 20 ␮m.

number of GluR2/3-IR cells on the contralateral experimental side compared to the ipsilateral controls (paired t-test, P < 0.001). These percentages of change of the labeled cells were relatively constant among the animals, ranging from 33.45 to 38.09% (Table 1). Although we saw some occasional variability in the labeling of the experimental and control SC, there were no systematic differences in antibody labeling after 10 and 20 days of enucleation. 3.4. Co-localization of GluR2/3- and calbindin D28K- and parvalbumin-IR cells To determine whether the GluR2/3-IR cells in the superficial layers co-localize with calbindin D28K or parvalbumin, we labeled GluR2/3 with fluorescein and calbindin D28K or parvalbumin with Texas red or Cy5. Some cells were clearly

labeled by both GluR2/3 and calbindin D28K antibodies in the cat and hamster SC. Other cells were labeled by either one of the antibodies, but not by both (Fig. 8). Some GluR2/3-IR cells were clearly labeled by parvalbumin antibody in cat but not in rabbit and hamster SC (Fig. 9). There was no obvious relationship between cell morphology and whether the cell was single or double labeled. To estimate the percentage of double-labeled cells, we counted the numbers of GluR2/3 and double-labeled cells within the tier of the GluR2/3-IR cells in twelve sections from two animals. More than half of the GluR2/3-IR cells were double-labeled with calbindin D28K in the hamster SC (1550 of 2811 cells or 55.1%), while only about 10% of GluR2/3-IR cells were double-labeled with calbindin D28K in the cat SC (296 of 2996 cells or 9.9%). By contrast, none of the GluR2/3-IR cells was double-labeled

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

147

Fig. 5. High magnification differential interference contrast micrographs of anti-GluR2/3-IR cells in the upper OL of the rabbit SC. (A) Some small to medium-sized round or oval cells are seen. (B) and (C) Stellate cells (large arrowheads) with multiple dendrites and vertical fusiform cells (arrow) with a thick, proximal dendrite directed towards the pial surface. A horizontal cell (small arrowhead) is also seen in (C). (D) Another vertical fusiform cell. Bar = 20 ␮m.

with calbindin D28K in the rabbit SC (Table 2). Only about 5% of GluR2/3-IR cells were double-labeled with parvalbumin in the cat SC (140 of 2925 cells or 4.8%). By contrast, none of the GluR2/3-IR cells was double-labeled with parvalbumin in the rabbit and hamster SC (Table 3). This percentage of double-labeled cells was relatively consistent across sections and among animals (Tables 2 and 3).

4. Discussion Our results indicate that GluR2/3 is contained in a large number of neurons in the cat, rabbit, and hamster SC. However, differences in the pattern of distribution, in the colocalization pattern with calbindin D28K or parvalbumin, and in the effect of enucleation were seen among the species examined.

Fig. 6. High magnification differential interference contrast micrographs of anti-GluR2/3-IR cells in the OL of the hamster SC. (A,C) Some small to medium-sized round or oval cells. A horizontal cell (arrowhead) is also seen in (C). (B) Vertical fusiform cells with a thick, proximal dendrite directed towards the pial surface. (D) A stellate cell. Bar = 20 ␮m.

148

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

Fig. 7. GluR2/3 immunoreactivity in: (A) the cat SC ipsilateral and (B) contralateral, (C) rabbit SC ipsilateral and (D) contralateral, and (E) hamster SC ipsilateral and (F) contralateral to the enucleation. In the hamster SC contralateral to the enucleation (F), there was a marked reduction of anti-GluR2/3 immunoreactivity. The animals had their right eye removed 20 days before being sacrificed. Bar = 100 ␮m.

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

149

Fig. 7. (Continued ).

There is considerable anatomical and physiological evidence that AMPA receptors are present in the superficial SC (Binns, 1999). With respect to the distribution, however, at least two conclusions can be drawn from the present data. First, the present study indicates that GluR2/3-IR neurons have a very specific neurochemical sublaminar distribution and occupy specific areas within the SC. Second, differences in the pattern of distribution are clearly seen in different species. Thus, the dense band of labeled neurons was located within the lower SGL and upper OL in the cat, and within the upper OL in the rabbit and within the OL in the hamster. Recently, Kondo et al. (2000) located GluR2/3-IR neurons

in the rat SC. As in the hamster SC, the dense band of labeled neurons was located in the OL. Unlike in the hamster SC, however, there were many labeled neurons in the SGL, too. Thus, the present and previous studies clearly indicate that there are different expressional patterns of GluR2/3 immunoreactivity in the superficial layers of the SC in different species. The functional significance of the different laminar pattern of GluR2/3-IR neurons is not yet clear. Consistent with the present results, there is significant interspecies variation in neurochemical organization in the SC. For example, the calcium-binding protein calretinin forms a dense plexus of IR fibers in the superficial SC in the cat (Hong et al.,

Table 1 Change in the number of GluR2/3-IR cells in the cat, rabbit, and hamster SC after monocular enucleation

Table 2 Percentage of GluR2/3-IR cells, and double-labeled cells with calbindin D28K in the cat, rabbit, and hamster superior colliculus

Animal no.

Cat 4 Cat 8

No. sections

No. GluR2/3 cells Control

Experimental

Animal no.

Cat total

12

2996

296

9.9

0.02 ± 4.4 0.31 ± 4.06

Rabbit 7 Rabbit 13

6 6

1597 1549

0 0

0 0

0.2

Rabbit total

12

3146

0

0

33.45 ± 1.99 38.09 ± 4.05

Hamster 6 Hamster 13

6 6

1410 1411

755 795

35.8

Hamster total

12

2811

1550

1057 981

1052 972

12

2038

2024

6 6

783 809

520 497

12

1592

1017

12

1655

Rabbit 3 Rabbit 8

6 6

Rabbit total Hamster 4 Hamster 5

Percentage double (mean ± S.E.M.) 9.76 ± 0.28 9.93 ± 0.34

2.9

Cat total

No. double 151 145

1608

791 817

No. GluR2/3 cells 1546 1450

Cat 5 Cat 13

808 847

No. sections 6 6

2.31 ± 2.83 3.46 ± 2.03

6 6

Hamster total

Mean ± S.E.M. (%)

56.57 ± 1.91 53.93 ± 2.30 55.1

150

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

Fig. 8. Fluorescence photomicrographs of: (A) the cat SC immunostained for GluR2/3 and (B) calbindin D28K, and (C) the hamster SC immunostained for GluR2/3 and (D) calbindin D28K. Both cells that are labeled with a single antibody and cells that are double-labeled (arrowheads) are seen. Bar = 20 ␮m.

2002) and hamster SC (Kang et al., 2002), whereas many calretinin-IR cells are localized in the rabbit SC (Jeon et al., 1998). The specific sublaminar distribution of GluR2/3 in all three species reflects again one of the characteristic features of the neurochemical organization of the SC. Segregation into a specific lamina and/or a patch-like organization of many afferent fibers and efferent cells with distinct neurochemical elements is characteristic of the SC (Huerta and Harting, 1984; Illing, 1992; Harting, 2003). For example, glutamate (Jeon et al., 1997a), glutamate receptors

(Cirone et al., 2002), acetylcholine (Graybiel, 1978; Hall et al., 1989; Jeon et al., 1993), nitric oxide (Scheiner et al., 2000; González-Soriano et al., 2002), adhesion molecules (Yamagata et al., 1995), and enkephaline-IR cells and/or fibers (Graybiel et al., 1984; Mize, 1989) are distributed in specific laminae in the SC. The three major calciumbinding proteins calbindin D28K, calretinin, and parvalbumin are also distributed in specific laminae in the SC (Mize et al., 1991, 1992; Behan et al., 1992; Jeon et al., 1998; González- Soriano et al., 2000; Hong et al., 2002).

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

151

Fig. 9. Fluorescence photomicrographs of: (A) the cat SC immunostained for GluR2/3 and (B) parvalbumin, and (C) the hamster SC immunostained for GluR2/3 and (D) parvalbumin. Both cells that are labeled with a single antibody and cells that are double-labeled (arrowheads) are seen in cat. Double-labeled neurons were not found in the hamster SC. Bar = 20 ␮m.

Many afferent systems to the superficial SC are partially segregated within specific layers. Most of the contralateral retinal afferents to the SC are concentrated in the upper part of the superficial layers, while the visual cortical inputs Table 3 Percentage of GluR2/3-IR cells, and double-labeled cells with parvalbumin in the cat, rabbit, and hamster superior colliculus Animal no.

Cat 14 Cat 15

No. sections

No. GluR2/3 cells

No. double

Percentage double (mean ± S.E.M.) 4.84 ± 0.62 4.73 ± 0.42

6 6

1488 1437

72 68

Cat total

12

2925

140

Rabbit 14 Rabbit 15

6 6

1672 1746

0 0

0 0

Rabbit total

12

3418

0

0

Hamster 14 Hamster 15

6 6

1529 1581

0 0

0 0

12

3110

0

0

Hamster total

4.8

are concentrated in the lower portion (Huerta and Harting, 1984). The almost no reduction of GluR2/3-IR cells in the cat and rabbit SC after enucleation may suggest that most of the GluR2/3-IR cells receive non-retinal inputs. By contrast, the clear reduction in hamster SC indicates that many GluR2/3-IR cells receive retinal inputs. Kondo et al. (2000) also suggested that the GluR2 expression in the most OL neurons is regulated by the correct afferentation from retina in rat SC. Axons from ventral lateral geniculate nucleus and many cortical areas such as dorsolateral suprasylvian and posteromedial lateral suprasylvian areas projects into the lower portion of SC (Huerta and Harting, 1984; Harting et al., 1992). Other areas such as pretectum and parabigeminal nucleus also project to the superficial SC. Thus, the GluR2/3-IR cells in the SC may receive input from many afferent sources. The SC efferent cells may be organized into subgroups. There are both ascending and descending groups of efferent neurons in the SC. The ascending pathways send fibers to the thalamus. These include two major segregated pathways. The SC neurons projecting into the lateral geniculate nucleus are largely located within the

152

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

upper superficial SC (Harrell et al., 1982). The SC neurons projecting into the lateral posterior nucleus are largely located in the lower superficial SC (Caldwell and Mize, 1981; Abramson and Chalupa, 1988). In addition to the anatomical segregation, SC efferents also show neurochemical segregation. The calbindin are found in cells that are located in three layers (Mize et al., 1991; Behan et al., 1992; González-Soriano et al., 2000). Similar to the GluR2/3-IR cells, the calbindin D28K-IR neurons in the lower SGL and OL of the cat, rabbit, and hamster included small horizontal, vertical fusiform, pyriform, and stellate-shaped neurons (Mize et al., 1991; Behan et al., 1992; GonzálezSoriano et al., 2000). In the cat SC, the majority of the calbindin D28K-IR neurons were interneurons (Mize et al., 1991). The partial overlap of the GluR2/3-IR cells with calbindin D28K in the present study suggests that at least some of the GluR2/3-IR cells in the cat SC may be interneurons. A single dense band of parvalbumin-IR neurons has been found within the deep SGL and upper OL in the cat, rabbit, and hamster SC (Mize et al., 1992; GonzálezSoriano et al., 2000; Kang et al., 2002). The parvalbuminIR neurons in the SC included vertical fusiform, stellate, and pyriform cells. The partial overlap of GluR2/3-IR cells and parvalbumin-IR cells, and the fact that a large number of parvalbumin-IR neurons in the cat are projection neurons (Mize et al., 1992) suggest that at least some of the GluR2/3-IR cells in the cat SC may be projection neurons. In contrast to cat SC, many of the calbindin D28K-IR cells in the OL of the rat SC are thalamic projection neurons, while parvalbumin-IR neurons are primarily interneurons (Lane et al., 1997). In the present study, more than 50% of the GluR2/3-IR cells in the hamster SC also contained calbindin D28K, while no GluR2/3-IR neurons contained parvalbumin. Thus, these results suggest that a large number of the GluR2/3-IR cells in the hamster SC may be a subset of projection neurons in the hamster SC. Both projection and interneurons in the cerebral cortex (Vissavajjhala et al., 1996) and hippocampus (He et al., 1998) contained GluR2. To determine whether the GluR2/3 -IR cells are projection neurons, the cells should be backfilled with a tracer. In the present study, there were marked interspecies differences in the expression of calbindin D28K or parvalbumin in the GluR2/3-IR cells. The differential co-localization ratio of GluR2/3 with calbindin D28K or parvalbumin indicates that GluR2/3-activated downstream pathways in postsynaptic neurons should be diverse. Along with NMDA receptors, AMPA receptors are also involved in neurotoxicity (Choi, 1988; Michaelis, 1998). Calbindin and parvalbumin are thought to play an important role in calcium buffering and calcium-mediated signal transduction. Impaired regulation of calcium by calbindin or parvalbumin is closely related to many neurodegenerative processes (Heizmann and Braun, 1995; Schäfer and Heizmann, 1996). For example, calbindin D28K containing neurons are resilient to excitotoxic insults in the hippocampal formation

of Alzheimer’s disease (Iritani et al., 2001). Therefore, the present results suggest that vulnerability to glutamateinduced neurotoxicity may be differentially regulated in the SC of different species. Our present results show that deafferentation produces a loss of GluR2/3 labeling in cell bodies and proximal dendrites in the hamster SC. This is similar to that of the rat SC. AMPA-binding was significantly reduced in the rat SC after enucleation (Chalmers and McCulloch, 1991). Recently, Kondo et al. (2000) found a lower expression of GluR2 mRNA and GluR2/3 immunoreactivity in the contralateral rat SC after enucleation. The results in the hamster and rat are substantially different from the results in the cat and rabbit. We found no evidence that enucleation affects GluR2/3 immunoreactivity in the cat and rabbit SC. Although retinal ganglion cells project both to ipsilateral and contralateral SC in the cat, most ganglion cells project contalateral SC in the rabbit, rat, and hamster (Rodieck, 1998). Thus, there seems no direct correlation between the reduction manner of GluR2/3-IR cells in the SC and the projection manner of retinal ganglion cells to the SC. Similar to these results, calbindin D28K decreased in the hamster SC (Kang et al., 2002), but not in the monkey (Mize and Luo, 1992), rabbit (Jeon et al., 1998), and rat SC (Lane et al., 1996). These results suggest that the hamster SC may have GluR2/3and calbindin D28K-containing cells that are more plastic upon enucleation than in the cat and rabbit SC. It is, however, unclear whether the effect is due to an interruption of GluR2/3 metabolism, to the atrophy of cells, or to both. The visual cortical lesion produced a slight reduction in the cell size of glutamate-IR neurons in the cat SC (Jeon et al., 1997b). This result suggests that at least some of the loss of GluR2/3 immunoreactivity in hamster SC is due to early phases of cell degeneration, and not simply to reduced metabolism. In the SC, the AMPA receptors are critical in visual transmission (Binns and Salt, 1994; Binns, 1999) and in maintaining dendritic growth rate and arbor structure in mature neurons (Rajan and Cline, 1998). The subunit composition of the AMPA receptor is critical to its function (Michaelis, 1998; Conti and Weinberg, 1999; Sheng and Kim, 2002; Song and Huganir, 2002). In the present study, the anti-GluR2/3 antibody localized both GluR2 and GluR3 that had nearly identical carboxyl terminal sequences. A previous study in the rat superficial SC, however, showed a low expression of GluR3 mRNA expression, while the expression of GluR2 mRNA was moderate to high (Sato et al., 1993). Thus, the GluR2/3-IR neurons in the present study may primarily correspond to the GluR2 expressing neurons. AMPA glutamate receptors that display very low calcium permeability include the GluR2 subunit (Michaelis, 1998; Conti and Weinberg, 1999). Thus, a large number the GluR2/3-IR neurons in the present study may represent very low calcium-permeable AMPA receptor subtypes. A recent study by Endo and Isa (2001) identified the AMPA receptor with low calcium permeability in the superficial rat SC by a whole-cell patch

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

clamp. The precise role of the GluR2/3 in the SC, however, is unclear. Recent studies have shown that GluR2/3 is crucial for the regulation of synaptic plasticity, especially in hippocampal long-term depression (LTD) (Kim et al., 2001). As both LTP and LTD are present in the SC (Lo and Mize, 2002), it is conceivable that the GluR2/3IR cells in the present study are involved in LTD in the SC. The present investigation indicates that there are significant interspecies variations between cat, rabbit, and hamster in the expression of GluR2/3. Although it is too early to determine the exact functional significance of these differences, one possible reason for these apparent differences could be the discrete differences of the visual field of eyes and the physical attributes for survival. Both rabbits and hamsters have a visual streak that is directed toward the horizon (Rodieck, 1998). The laterally directed eyes of rabbit and hamster provide with a panoramic view, allowing them to see almost the entire horizon. Although their field of view is almost entirely monocular, lateral placement of the eyes is essential to the survival of hunted or herbivorous animals. In the rabbit the binocular field above and in front of the head is very narrow. In contrast to rabbits and hamsters, cats have an area centralis and visual streak (Rodieck, 1998). The frontally directed carnivorous hunter eyes of cats provide them with a wide field of binocular vision to accurately locate their prey. Further research is needed to provide materials to bridge between collicular anatomy and function between monocular herbivorous and binocular carnivorous species. In conclusion, the present results demonstrate that the organizational feature of the GluR2/3 found in the mammalian SC is strikingly different among the species examined. The GluR2/3-IR cells were specifically and differentially localized in the cat, rabbit, and hamster SC. The patterned distribution of the GluR2/3-IR cells closely matched the second tier of calbindin D28K-IR cells in the hamster, but was different in the cat and rabbit. The patterned distribution of GluR2/3-IR cells closely overlapped the tier of parvalbumin-IR neurons in cat, but only partially overlapped in hamster and rabbit. The GluR2/3-IR cells contained many calbindin D28K in the hamster, some in the cat, but none in the rabbit, while GluR2/3-IR cells contained some parvalbumin only in cat. Enucleation produced a clear reduction of GluR2/3 immunoreactivity in the SC contralateral to the enucleation in the hamster, but not in the cat and rabbit. Although the function of the AMPA receptor subunit in mammalian SC is not yet clear, the precise segregation of the GluR2/3-IR neurons in the SC, its differential co-localization with calbindin D28K and parvalbumin, and differential response to enucleation among different species likely reflect the functional diversity of this receptor subtype in visual activity. Further studies are needed to investigate the functional significance of the variations observed among different species.

153

Acknowledgements We thank Profs. Thomas D. Marshall and Robert P. Flaherty, Language Institute of Kyungpook National University, for correcting the English. This work was supported by Korea Research Foundation to C.-J. Jeon (KRF-2000-015DP0421).

References Abramson, B.P., Chalupa, L.M., 1988. Multiple pathways from the colliculus to the extrageniculate visual thalamus of the cat. J. Comp. Neurol. 271, 397–418. Baimbridge, K.G., Celio, M.R., Rogers, J.H., 1992. Calcium-binding proteins in the nervous system. Trends Neurosci. 15, 303–308. Behan, M., Jourdain, A., Bray, G.M., 1992. Calcium-binding protein (calbindin D28K) immunoreactivity in the hamster superior colliculus: ultrastructure and lack of co-localization with GABA. Exp. Brain Res. 89, 115–124. Ben-Ari, Y., Khazipov, R., Leinekugel, X., Caillard, O., Gaiarsa, J.L., 1997. GABAA, NMDA and AMPA receptors: a developmentally regulated ‘menage a trois’. Trends Neurosci. 20, 523–529. Binns, K.E., 1999. The synaptic pharmacology underlying sensory processing in the superior colliculus. Prog. Neurobiol. 59, 129–159. Binns, K.E., Salt, T.E., 1994. Excitatory amino acid receptors participate in synaptic transmission of visual responses in the superficial layers of the cat superior colliculus. Eur. J. Neurosci. 6, 161–169. Caldwell, R.B., Mize, R.R., 1981. Superior colliculus neurons which project to the cat lateral posterior nucleus have varying morphologies. J. Comp. Neurol. 203, 53–66. Chalmers, D.T., McCulloch, J., 1991. Selective alterations in glutamate receptor subtypes after unilateral orbital enucleation. Brain Res. 540, 255–265. Choi, D.W., 1988. Glutamate neurotoxicity and disease of the nervous system. Neuron 1, 623–634. Cirone, J., Sharp, C., Jeffery, G., Salt, T.E., 2002. Distribution of metabotropic glutamate receptors in the superior colliculus of the adult rat, ferret and cat. Neuroscience 109, 779–786. Conti, F., Weinberg, R.J., 1999. Shaping excitation at glutamatergic synapses. Trends Neurosci. 22, 451–458. Cork, R.J., Baber, S.Z., Mize, R.R., 1998. Calbindin D28K- and parvalbumin-immunoreactive neurons form complementary sublaminae in the rat superior colliculus. J. Comp. Neurol. 394, 205– 217. Endo, T., Isa, T., 2001. Functionally different AMPA-type glutamate receptors in morphologically identified neurons in rat superficial superior colliculus. Neuroscience 108, 129–141. Firth, S.I., Li, W., Massey, S.C., Marshak, D.W., 2003. AMPA receptors mediate acetylcholine release from starburst amacrine cells in the rabbit retina. J. Comp. Neurol. 466, 80–90. Gobersztejn, F., Britto, L.R., 1996. Calretinin in the mouse superior colliculus originates from retinal ganglion cells. Braz. J. Med. Biol. Res. 29, 1507–1511. González-Soriano, J., González-Flores, M.L., Contreras-Rodr´ıguez, J., Rodr´ıguez-Veiga, E., Mart´ınez-Sainz, P., 2000. Calbindin D28K- and parvalbumin-immunoreactivity in the rabbit superior colliculus: an anatomical study. Anat. Rec. 259, 334–346. González-Soriano, J., Contreras-Rodr´ıguez, J., Mart´ınez-Sainz, P., MartinPalacios, S., Marin-Garcia, P., Rodr´ıguez-Veiga, E., 2002. NADPHdiaphorase distribution in the rabbit superior colliculus and colocalization with calcium-binding proteins. J. Anat. 200, 297–308. Grantyn, R., 1988. Gaze control through superior colliculus: structure and function. Rev. Oculomot. Res. 2, 273–333.

154

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155

Graybiel, A.M., 1978. A stereometric pattern of distribution of acetylthiocholinesterase in the deep layers of the superior colliculus. Nature 272, 539–541. Graybiel, A.M., Brecha, N., Karten, H.J., 1984. Cluster-and-sheet pattern of enkephalin-like immunoreactivity in the superior colliculus of the cat. Neuroscience 12, 191–214. Hall, W.C., Fitzpatrick, D., Klatt, L.L., Raczkowski, D., 1989. Cholinergic innervation of the superior colliculus in the cat. J. Comp. Neurol. 287, 495–514. Harrell, J.V., Caldwell, R.B., Mize, R.R., 1982. The superior colliculus neurons which project to the dorsal and ventral lateral geniculate nuclei in the cat. Exp. Brain Res. 46, 234–242. Harting, J.K., 2003. Puffs and patches: a brief chronological review. In: Hall, W.C., Moschovakis, A. (Eds.), The Superior Colliculus. CRC Press, Boca Raton, pp. 83–105. Harting, J.K., Huerta, M.F., Hashikawa, T., Van Lieshout, D.P., 1991. Projection of the mammalian superior colliculus upon the dorsal lateral geniculate nucleus: organization of tectogeniculate pathways in nineteen species. J. Comp. Neurol. 304, 275–306. Harting, J.K., Updyke, B.V., Van Lieshout, D.P., 1992. Corticotectal projections in the cat: anterograde transport studies of twenty-five cortical areas. J. Comp. Neurol. 324, 379–414. He, Y., Janssen, W.G., Morrison, J.H., 1998. Synaptic coexistence of AMPA and NMDA receptors in the rat hippocampus: a postembedding immunogold study. J. Neurosci. Res. 54, 444–449. Heizmann, C.W., Braun, K., 1995. Calcium Regulation by CalciumBinding Proteins in Neurodegenerative Disorders. Springer–Verlag, New York. Heizmann, C.W., Rohrenbeck, J., Kamphuis, W., 1990. Parvalbumin, molecular and functional aspects. Adv. Exp. Med. Biol. 269, 57–66. Hollmann, M., Heinemann, S., 1994. Cloned glutamate receptors. Annu. Rev. Neurosci. 17, 31–108. Hong, S.K., Kim, J.Y., Jeon, C.J., 2002. Immunocytochemical localization of calretinin in the superficial layers of the cat superior colliculus. Neurosci. Res. 44, 325–335. Huerta, M.F., Harting, J.K., 1984. The mammalian superior colliculus: studies of its morphology and connections. In: Vanega, H. (Ed.), The Comparative Neurology of the Optic Tectum. Plennum Press, New York, pp. 687–772. Illing, R.B., 1992. Association of efferent neurons to the compartmental architecture of the superior colliculus. Proc. Natl. Acad. Sci. U.S.A. 89, 10900–10904. Iritani, S., Niizato, K., Emson, P.C., 2001. Relationship of calbindin D28Kimmunoreactive cells and neuropathological changes in the hippocampal formation of Alzheimer’s disease. Neuropathology 21, 162–167. Isa, T., 2002. Intrinsic processing in the mammalian superior colliculus. Curr. Opin. Neurobiol. 12, 668–677. Isa, T., Endo, T., Saito, Y., 1998. The visuo-motor pathway in the local circuit of the rat superior colliculus. J. Neurosci. 18, 8496–8504. Jeon, C.J., Spencer, R.F., Mize, R.R., 1993. Organization and synaptic connections of cholinergic fibers in the cat superior colliculus. J. Comp. Neurol. 333, 360–374. Jeon, C.J., Gurski, M.R., Mize, R.R., 1997a. Glutamate containing neurons in the cat superior colliculus revealed by immunocytochemistry. Vis. Neurosci. 14, 387–393. Jeon, C.J., Hartman, M.K., Mize, R.R., 1997b. Glutamate-like immunoreactivity in the cat superior colliculus and visual cortex: further evidence that glutamate is the neurotransmitter of the corticocollicular pathway. Vis. Neurosci. 14, 27–37. Jeon, C.J., Pyun, J.K., Yang, H.W., 1998. Calretinin and calbindin D28Kimmunoreactivity in the superficial layers of the rabbit superior colliculus. Neuroreport 9, 3847–3852. Kang, Y.S., Park, W.M., Lim, J.K., Kim, S.Y., Jeon, C.J., 2002. Changes of calretinin, calbindin D28K- and parvalbumin-immunoreactive neurons in the superficial layers of the hamster superior colliculus following monocular enucleation. Neurosci. Lett. 330, 104–108.

Kim, C.H., Chung, H.J., Lee, H.K., Huganir, R.L., 2001. Interaction of the AMPA receptor subunit GluR2/3 with PDZ domains regulates hippocampal long-term depression. Proc. Natl. Acad. Sci. U.S.A. 98, 11725–11730. Kondo, M., Sumino, R., Okado, H., 2000. Expression of AMPA receptors in rat superior colliculus and effect of orbital enucleation. Brain Res. 883, 238–242. Lane, R.D., Allan, D.M., Bennett-Clarke, C.A., Rhoades, R.W., 1996. Differential age-dependent effects of retinal deafferentation upon calbindin- and parvalbumin-immunoreactive neurons in the superficial layers of the rat’s superior colliculus. Brain Res. 740, 208–214. Lane, R.D., Allan, D.M., Bennett-Clarke, C.A., Howell, D.L., Rhoades, R.W., 1997. Projection status of calbindin- and parvalbuminimmunoreactive neurons in the superficial layers of the rat’s superior colliculus. Vis. Neurosci. 14, 277–286. Leuba, G., Saini, K., 1996. Calcium-binding proteins immunoreactivity in the human subcortical and cortical visual structures. Vis. Neurosci. 13, 997–1009. Lo, F.S., Mize, R.R., 2002. Properties of LTD and LTP of retinocollicular synaptic transmission in the developing rat superior colliculus. Eur. J. Neurosci. 15, 1421–1432. Lo, W., Molloy, R., Hughes, T.E., 1998. Ionotropic glutamate receptors in the retina: moving from molecules to circuits. Vision Res. 38, 1399–1410. Meldrum, B.S., 2000. Glutamate as a neurotransmitter in the brain: review of physiology and pathology. J. Nutr. 130, 1007S–1015S. Michaelis, E.K., 1998. Molecular biology of glutamate receptors in the central nervous system and their role in excitotoxicity, oxidative stress and aging. Prog. Neurobiol. 54, 369–415. Mize, R.R., 1989. Enkephalin-like immunoreactivity in the cat superior colliculus: distribution, ultrastructure, and colocalization with GABA. J. Comp. Neurol. 285, 133–155. Mize, R.R., Butler, G.D., 1996. Postembedding immunocytochemistry demonstrates directly that both retinal and cortical terminals in the cat superior colliculus are glutamate immunoreactive. J. Comp. Neurol. 371, 633–648. Mize, R.R., Butler, G.D., 2000. The NMDAR1 subunit of the N-methyl-daspartate receptor is localized at postsynaptic sites opposite both retinal and cortical terminals in the cat superior colliculus. Vis. Neurosci. 17, 41–53. Mize, R.R., Luo, Q., 1992. Visual deprivation fails to reduce calbindin 28 kD or GABA immunoreactivity in the rhesus monkey superior colliculus. Vis. Neurosci. 9, 157–168. Mize, R.R., Jeon, C.J., Butler, G.D., Luo, Q., Emson, P.C., 1991. The calcium-binding protein calbindin-D28K reveals subpopulations of projection and interneurons in the cat superior colliculus. J. Comp. Neurol. 307, 417–436. Mize, R.R., Luo, Q., Butler, G., Jeon, C.J., Nabors, B., 1992. The calciumbinding proteins parvalbumin and calbindin-D28K form complementary patterns in the cat superior colliculus. J. Comp. Neurol. 320, 243–256. Petralia, R.S., Wenthold, R.J., 1992. Light and electron immunocytochemical localization of AMPA-selective glutamate receptors in the rat brain. J. Comp. Neurol. 318, 329–354. Petralia, R.S., Yokotani, N., Wenthold, R.J., 1994. Light and electron microscope distribution of the NMDA receptor subunit NMDAR1 in the rat nervous system using a selective anti-peptide antibody. J. Neurosci. 14, 667–696. Polans, A., Baehr, W., Palczewski, K., 1996. Turned on by Ca2+ ! The physiology and pathology of Ca(2+)-binding proteins in the retina. Trends Neurosci. 19, 547–554. Qin, P., Pourcho, R.G., 1999. Localization of AMPA-selective glutamate receptor subunits in the cat retina: a light- and electron-microscopic study. Vis. Neurosci. 16, 169–177. Rajan, I., Cline, H.T., 1998. Glutamate receptor activity is required for normal development of tectal cell dendrites in vivo. J. Neurosci. 18, 7836–7846.

W.-M. Park et al. / Neuroscience Research 49 (2004) 139–155 Rodieck, R.W., 1998. The First Steps in Seeing. Sinauer Associates, Sunderland. Rogers, J.H., Resibois, A., 1992. Calretinin and calbindin-D28K in rat brain: patterns of partial co-localization. Neuroscience 51, 843–865. Rogers, J., Khan, M., Ellis, J., 1990. Calretinin and other CaBPs in the nervous system. Adv. Exp. Med. Biol. 269, 195–203. Ryoo, S.R., Ahn, C.H., Lee, J.Y., Kang, Y.S., Jeon, C.J., 2003. Immunocytochemical localization of neurons containing the AMPA GluR 2/3 subunit in the hamster visual cortex. Mol. Cells. 16, 211–215. Sato, K., Kiyama, H., Tohyama, M., 1993. The differential expression patterns of messenger RNAs encoding non-N-methyl-d-aspartate glutamate receptor subunits (GluR1–4) in the rat brain. Neuroscience 52, 515–539. Schäfer, B.W., Heizmann, C.W., 1996. The S100 family of EF-hand calcium-binding proteins: functions and pathology. Trends Biochem. Sci. 21, 134–140. Scheiner, C., Arceneaux, R., Guido, W., Kratz, K., Mize, R., 2000. Nitric oxide synthase distribution in the cat superior colliculus and colocalization with choline acetyltransferase. J. Chem. Neuroanat. 18, 147–159. Schmidt-Kastner, R., Meller, D., Eysel, U.T., 1992. Immunohistochemical changes of neuronal calcium-binding proteins parvalbumin and

155

calbindin-D-28K following unilateral deafferentation in the rat visual system. Exp. Neurol. 117, 230–246. Sheng, M., Kim, M.J., 2002. Postsynaptic signaling and plasticity mechanisms. Science 298, 776–780. Song, I., Huganir, R.L., 2002. Regulation of AMPA receptors during synaptic plasticity. Trends Neurosci. 25, 578–588. Thoreson, W.B., Witkovsky, P., 1999. Glutamate receptors and circuits in the vertebrate retina. Prog. Retin. Eye Res. 18, 765–810. Van Damme, K., Massie, A., Vandesande, F., Arckens, L., 2003. Distribution of the AMPA2 glutamate receptor subunit in adult cat visual cortex. Brain Res. 960, 1–8. Vissavajjhala, P., Janssen, W.G., Hu, Y., Gazzaley, A.H., Moran, T., Hof, P.R., Morrison, J.H., 1996. Synaptic distribution of the AMPA-GluR2 subunit and its colocalization with calcium-binding proteins in rat cerebral cortex: an immunohistochemical study using a GluR2-specific monoclonal antibody. Exp. Neurol. 142, 296–312. Wong-Riley, M.T., Jacobs, P., 2002. AMPA glutamate receptor subunit 2 in normal and visually deprived macaque visual cortex. Vis. Neurosci. 19, 563–573. Yamagata, M., Herman, J.P., Sanes, J.R., 1995. Lamina-specific expression of adhesion molecules in developing chick optic tectum. J. Neurosci. 15, 4556–4571.