Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2002 Mar;200(3):297–308. doi: 10.1046/j.1469-7580.2002.00025.x

NADPH-diaphorase distribution in the rabbit superior colliculus and co-localization with calcium-binding proteins

Juncal González-Soriano 1, Julio Contreras-Rodríguez 1, Pilar Martínez-Sainz 1, Susana Martín-Palacios 1, Pilar Marín-García 1, Elisia Rodríguez-Veiga 1
PMCID: PMC1570688  PMID: 12033734

Abstract

Nitric oxide (NO) and calcium-binding proteins (CaBP) are important neuromodulators implicated in brain plasticity and brain disease. In addition, the mammalian superior colliculus (SC) has one of the highest concentrations of NO within the brain. The present study was designed to determine the distribution of nitric oxide-synthesizing neurons in the SC of the rabbit by enzyme histochemistry for reduced nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d), and its degree of co-localization with CaBP, parvalbumin (PV) and calbindin (CB). NADPH-d-labelled fibres formed dense patches of terminal buttons within the intermediate grey layer and streams of fibres within the deepest layers of SC. Cells expressing NOS constitute a subpopulation of neurons in which practically all cell types are represented. Combined PV/NADPH-d experiments showed a complete lack of co-localization within individual neurons and fibres. On the contrary, double-labelled neurons appeared in CB/NADPH-d-stained sections, only in the superficial layers, and mostly in the SGS and SO. These cells, which were intermingled with other neurons containing either NADPH-d or CB, appear to be a subtype of narrow-field and wide-field vertical cells, and display an anterior–posterior gradient of density. Owing to the involvement of the superficial layers of the SC in the organization and integration of the visual information, it is suggested that these neurons may play a concrete role within the visual circuits. Our data indicate a clear selectivity in the expression of NADPH-d, PV and CB in the SC, and that NO and CB probably serve as co-modulators and/or co-transmitters in the connectivity of the superficial layers of this midbrain structure.

Keywords: calbindin (CB) immunoreactivity, co-localization, neuronal types, nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d), (NADPH-d) histochemistry, parvalbumin (PV) immunoreactivity, superior colliculus (SC)

Introduction

Nitric oxide (NO) has been identified as a neuronal messenger molecule. NO is a membrane-permeate gas with a very short half-life, which precludes its direct detection. For this reason, different methods have been used to identify nitric oxide synthase (NOS), the synthesis enzyme of NO, in order to localize neurons using the signalling molecule throughout the nervous system.

In aldehyde-fixed tissue, NOS is able to catalyse selectively a histochemically detectable nicotinamide adenine dinucleotide phosphate diaphorase (NADPH-d) reaction, and NADPH-diaphorase histochemistry produces dense, Golgi-like staining of the reactive neurons. Thomas & Pearse (1961) first described NADPH-d-containing neurons. Since then, several authors have found a one-to-one correspondence between neurons expressing NOS mRNA or NOS immunoreactivity, and those expressing NADPH-d (Vincent & Kimura, 1992; Giraldi-Guimarães et al. 1999). The demonstration that in transgenic mice lacking a functional neuronal NOS gene there is a complete loss of NADPH-d staining in the nervous system provided definitive evidence for the molecular correspondence between these two activities (Huang et al. 1993). In addition, Cork et al. (2000) examined the distribution of NOS-labelled neurons in the adult mouse to serve as a reference for studying NOS expression in the developing SC. Distributions in NADPH-d and NOS sections were similar with either label. However, their classification of cell morphology was based on NADPH-d labelling, which filled the dendrites of many neurons far better than did the NOS antibody.

The presence of NADPH-d reactive/NOS positive cells and neuropil in the optic tectum/superior colliculus (SC) has been reported in several species, such as the cat, rat and monkey (Mizukawa et al. 1989; González-Hernández et al. 1992; Williams et al. 1994; Satoh et al. 1995; Tenório et al. 1995, 1996, 1998; Cork et al. 1998a). However, no information is available concerning the NADPH-d expression in the adult rabbit SC. Rabbits were chosen for the present study because of several reasons. First, in terms of evolution, they represent species between rodents and carnivores. For example, rabbits have a moderate presence of thalamic GABAergic cells in relation to local GABAergic circuits (Penny et al. 1984; Arcelli et al. 1997). These circuits are also present in most of the cat thalamic nuclei but practically absent in the rat (Penny et al. 1983; Spreafico et al. 1983; Barbaresi et al. 1986; Harris & Hendrickson, 1987; Bentivoglio et al. 1991). According to these data, rabbits could represent an intermediate stage in the evolution of thalamic GABAergic circuits. In addition, unlike the SC of the rat but similar to that of the cat, retinal projections of the rabbit SC are binocularly segregated (Penny et al. 1984). Thus the rabbit SC might be a more favourable system than the rat SC for the future analysis of controversial issues such as the role of nNOS in the topography of retinocollicular projections in amniote vertebrates. Such developmental studies require a precise knowledge of NOS expression in the SC.

A large population of neurons in the SC is found to exhibit strong calcium-binding-protein (CaBP) activity (Mize et al. 1992; Leuba & Saini, 1996; Cork et al. 1998b; González-Soriano et al. 2000). However, the differences between the distribution of CaBP in the SC of the cat, the rat and the rabbit support the idea that CB-expressing collicular cells are functionally different from PV positive neurons. For example, more than 90% of CB immunoreactive neurons in the SGS project to the LGN, whereas only 17% of PV-expressing cells project in this layer (Lane et al. 1997). It is important to consider that a co-localization between CB and NADPH-d has been previously described in different parts of the central nervous system (CNS), for example the basal forebrain (Geula et al. 1993) and the cerebral cortex (Bertini et al. 1996). NADPH-d activity is also co-localized with other neuromodulators and/or neurotransmitters such as, for example, acetylcholinesterase (AChE) (Illing, 1990; Scheiner et al. 2000; Atoji et al. 2001) or calretinin (CR) (Arévalo et al. 1993). On the contrary, it seems that NADPH-d and PV rarely coexist in the central nervous system (Laing et al. 1994; Bertini et al. 1996). It must be pointed out that PV, CB and CR belong to the same family of CaBP (EF-hand). In the light of the previous data it is of interest to investigate the possible co-localization of PV/NADPH-d and CB/NADPH-d that could take place in the SC.

The study had two distinct aims. The first was to provide a description for NADPH-d/NOS expression in the SC of the rabbit. To examine nitrergic neurons, we used NADPH-d histochemistry, because it has been shown that NADPH-d staining provides more information and is a more reliable method than NOS immunohistochemistry (Cork et al. 2000). Second, we attempted to study the relationship of the NADPH-d-containing neurons with PV-IR and CB-IR immunoreactive populations in the SC, with the intention of determining whether there is a co-localization of NADPH-d with PV and CB, the location of these cells, and the possible functional significance.

Materials and methods

Animals

The brains of 10 adult (1- to 2-year-old) pigmented rabbits (Oryctolagus cuniculus, 2–5 kg) were used. The rabbits were all females and obtained from the Department of Animal Production (Veterinary Faculty-UCM). The Ethical Committee of the Veterinary Faculty (UCM) approved all procedures. Rabbits were kept in individual cages, and provided with food and water ad libitum.

Fixation and histological procedures

Animals were anaesthetized by intraperitoneal injection of sodium pentobarbital (50 mg kg−1) and perfused transcardially (at ambient temperature) with: (1) normal saline in 0.1 m phosphate buffer (PB; pH 7.4); (2) 4% paraformaldehyde in PB 0.1 m; and (3) 10% sucrose in PB. Perfusion time was about 1 h. The brains were removed, post-fixed for 2–4 h by immersion in fresh fixative solution, and transferred to 30% sucrose in PB for 1–3 days at 4 °C. Finally, they were frozen and stored at –80 °C. Brains were cut into 50-µm-thick coronal sections on a freezing microtome. Nissl staining with cresyl violet and/or cytochrome oxidase (CO) staining was carried out. Four adjacent series were collected every 200 µm, one for histochemistry, one for PV/NADPH-d, one for CB/NADPH-d double labelling and one for Nissl staining or CO. Four additional brains were cut into 50-µm-thick coronal sections on a freezing microtome and reacted for either PV or CB immunohistochemistry. To inhibit endogenous peroxidase, free-floating sections were treated for 30 min with 1% hydrogen peroxide – 90% methanol in distilled water and washed in phosphate buffer (PB).

NADPH-d histochemistry

Sections were incubated for 40–60 min in a solution containing 0.1 m PB (pH 7.4), with 0.8% Triton X-100 (TX), 1 mm reduced NADPH-d (Sigma) and 0.8 mm Nitroblue Tetrazolium (Sigma), at 37 °C. After incubation, the sections were rinsed in PB (pH 7.4), mounted and cover slipped.

Co-localization of NADPH-d and CaBP

Double labelling experiments for demonstration of NADPH-d and PV/CB were carried out in the following manner. The histochemical demonstration of NADPH-d was completed in free-floating sections as previously described. After rinsing the sections in PB, sections were then reacted with a monoclonal primary antibody (raised in a mouse): parvalbumin (PV, Sigma, PA-235) or calbindin D28k (CB, Sigma, CL-300). The antibodies were diluted in 0.2% TX and 3% normal goat serum in PB. To visualize antibody labelling, series of sections were: (1) pre-incubated for 2 h in 0.2% TX–3% normal goat serum in PB; (2) incubated for 48 h at 4 °C in the mouse PV (diluted 1 : 3000) or CB (1 : 2000); (3) incubated in biotinylated goat anti-mouse IgG (Chemicon 1 : 200–1 : 500) for 2 h at room temperature; (4) incubated in ABC (Vectastain Elite Kit, Vector, Burlingame, CA, USA) (diluted 1 : 100) for 90 min; (5) treated with 0.005% 3,3-DAB (Sigma) and 0.001% H2O2 PB. Sections were rinsed in multiple changes of PB between each step. Individual PV and CB reactions were carried out following the same protocol, with the exception of the NADPH-d staining, which was avoided.

All labelled sections (NADPH-d, PV, CB, and PV/NADPH-d or CB/NADPH-d) were examined with a Leitz Diaplan microscope, and images were obtained with a Leica system MPS 60. Drawings of positive cells were made using a Leitz Laborlux microscope with a drawing tube. The locations of these cells were determined by making drawings at ×4, whereas cell bodies and proximal dendrites were drawn at ×40, or even ×100 if necessary. The montage of cell bodies was constructed by placing the centre of each cell soma at the appropriate distance from the dorsal surface of the SC. Cells were orientated in the montage, so that the angle between the proximal dendrites and the closest part of the dorsal surface of the SC was correct. The calculation of soma size in all layers was carried out on a computer-assisted microscopy system, Kontron Videoplan (Zeiss). To estimate the soma size, the maximal length and width of 40–50 cells of each classified type were measured. Boundaries of the different lamina of the SC in immunocytochemical material were drawn by referring to the adjacent CO or Nissl section and by counterstaining with neutral red.

Control sections

For NADPH-d, the following controls were carried out: (1) omission of the substrate β-NADPH in the incubation medium; (2) omission of the chromogen nitroblue tetrazolium.

For PV and CB, negative control sections were treated identically except for the omission of primary antiserum. In all cases, specific reactivity was not observed anywhere in the control sections.

Results

NADPH-d histochemistry

NADPH-d labelling in the rabbit SC produces a specific pattern, with labelled cells and fibres throughout the rostral–caudal extent of the SC. This pattern includes three lamina: one in the superficial grey (SGS) and optic (SO) layers, one in the intermediate grey layers (basically SGI) and one in the deepest layers. The greatest concentration of labelled cells was found within a single dense band that spans the SGS. The heaviest staining of the neuropil was located in the intermediate layers, followed by the deep layers. In between the superficial layers (SGS and SO) and the heavily stained intermediate layer, there was a weak area with labelled cells and some positive fibres. Within the superficial layers, the fibre labelling was sparse and diffuse. By contrast, NADPH-d positive fibres within the SGI formed discontinuous but dense patches of label, which were much denser in caudal than in rostral sections. These patches of fibres (Fig. 1a) were interspersed with interpatch regions that contained fewer fibres and had a characteristic periodicity, with six or seven patches spread across the medium–lateral extent of the SC. Small- (6 × 9 µm) to medium-sized neurons (9 × 19 µm) were also observed within the SGI, and outside the dense fibre patches (Fig. 1b). Many of the fibres had varicosities and terminal buttons (Fig. 1c). Vertically orientated streams of fibres (Fig. 2a) and cells (Fig. 2b) were found within the deeper layers. Labelled cells within the SGP were darkly stained (Fig. 2c). Most labelled cells in this region were of medium (12 × 17 µm) or large size (up to 40 µm). All the cellular measurements corresponded to the minimum and maximum diameters, respectively.

Fig. 1.

Fig. 1

Open in a new tab

(a) Photomicrograph showing the characteristic periodicity of the fibre patches in the SGI (black arrows). Note that these patches are interspersed with interpatch regions with fewer fibres. Dorsal is to the top and lateral is to the right. Scale bar = 300 μm. (b) Photomicrograph showing small- to medium-sized neurons (white stars) within the SGI, or nearby the labelled patches of NADPH-d positive fibres (black arrows). Scale bar = 100 μm. (c) SGI in detail. Note that most of the fibres have varicosities (white arrows) and terminal buttons. Scale bar = 100 μm.

Fig. 2.

Fig. 2

Open in a new tab

Coronal sections through the rabbit SC showing the distribution of NADPH-d in the deep layers (dorsal to the top, lateral to the right). (a) Vertically orientated streams of cells (black arrows) near the PAG (white arrowheads). (b) Vertically orientated streams of labelled fibres within the deep layers (black arrows). (c) Labelled cells in the SGP. Scale bars = 100 μm.

NADPH-d-labelled cells formed a distinctive wedge in the peri-aqueductal gray (PAG). The apex of the wedge extended nearly to the endothelial lining of the cerebral aqueduct while the base was continuous with the base of the deep grey layer. This distribution appears to represent a distinct subgroup of neurons within the PAG.

Cell morphology

NADPH-d produces a Golgi-like staining of the reactive neurons. This fact allowed us to describe the NOS positive cell types, which were present in the adult rabbit SC. The neurons were classified according to the following criteria: (1) position of the cell; (2) shape of the soma; (3) orientation of soma and processes with respect to the collicular surface; and (4) direction, extension and branching pattern of dendrites and extension of the axon.

Labelled neurons consisted of the following morphologies: marginal cells, pyriform cells, narrow-field vertical cells, wide-field vertical cells and very few horizontal cells. Marginal cells were distributed within the SZ, and can be considered as the most superficial in the SC. The cell bodies were either round or oval, with the main axis orientated perpendicular to the collicular surface, and two or three primary processes projected from the ventral part of the cell body (Figs 3a and 5a). Pyriform cells had small, pear-shaped cell bodies with two or more dendrites extending from the dorsal surface of the cell body towards the pial surface (Figs 3b and 5b). This cell type was mostly within the dorsal SGS; the dendrites did not extend for long distances away from the cell body. Horizontal cells were also found mostly within the SGS. The NADPH-d horizontal cells had fusiform cell bodies and dendrites that extended for considerable distances, parallel to the pial surface of the SC (Figs 3c and 5c). Narrow-field vertical cells had fusiform cell bodies and vertically orientated dendrites that extended both above and bellow the cell body (Figs 3d and 5d). Their dendrites often extended for longer distances. These cells were mostly within the SGS and SO. Wide-field vertical neurons had horizontal fusiform somata, with elaborate dendritic trees (Figs 3d and 5e), eventually curving in a vertical direction, towards the SC surface. Many of these cells were found in close proximity to the SO.

Fig. 3.

Fig. 3

Open in a new tab

Photomicrographs of NADPH-d neurons in the superficial layers of the rabbit SC (dorsal to the top, lateral to the right). (a) Marginal cell in the SZ (black arrow). Note the primary processes from the ventral part of the cell body (white arrows). (b) Pyriform cell in the dorsal SGS (black arrow). Note the dendrites extending from the cell body towards the pial surface (black arrowheads). (c) Horizontal cell in the ventral SGS (black arrow). (d) Narrow-field (black arrow) and wide-field (white arrow) in the upper part of the SO. Scale bars = 100 μm.

Fig. 5.

Fig. 5

Open in a new tab

Camera lucida drawings of NADPH-d positive cells in the different laminae of the rabbit SC. The pial surface is to the top. (a) Marginal cells. (b) Pyriform cells. (c) Horizontal cells. (d) Narrow-field vertical cells. (e) Wide-field vertical cells. (f) Multipolar/stellate cells. Scale bars = 100 μm.

Labelled neurons in both the deep layers and the PAG of the SC consisted of three basic morphologies: stellate (multipolar) cells, narrow-field vertical cells, and a few wide-field vertical cells. Many of the neurons in close proximity to and/or within the PAG had the typical stellate shapes, with medium-sized and irregularly shaped somas, and multiple dendrites that radiated in several directions from the cell body (Figs 4a and 5f). Owing to its irregular shape and several primary dendrites that radiated from the cell body, some of the neurons of this group were considered as multipolar (Figs 4b and 5g). These cells were also present within the deepest layers (SGP and SAP). Narrow-field vertical cells were similar to those seen in the upper layers, with vertical fusiform somas, and one main dendrite extending vertically from each pole of the cell body. There were other secondary dendrites (Figs 4c and 5d). These neurons were found in the SGI, SAI and dorsal SGP. Wide-field vertical cells also had two primary dendrites, arising laterally from the cell body and approximately parallel to the collicular surface (Figs 4d and 5e). Wide-field vertical cells were seen within the SGI, SAI, SGP and SAP. In addition, there were distinct clusters of cells within the most medial part, within or nearby the PAG (Fig. 4e,f). These cells had round somas, with few or no labelled dendrites and could represent a distinct subgroup of NADPH-positive neurons within the SC.

Fig. 4.

Fig. 4

Open in a new tab

Photomicrographs of NADPH-d neurons in the deep layers of the rabbit SC (dorsal to the top, lateral to the right). (a) Stellate cell in the SGP (black arrow). Note the streams of labelled fibres in the same layer (black arrowheads). (b) Multipolar cells near the PAG (black arrow). (c) Narrow-field vertical cells in the SAI (black arrows). Note that the orientation is either perpendicular or parallel to the collicular surface. (d) Wide-field vertical cell in the SGS (black arrow). Note the primary dendrites arising from the cell body. (e) Cells within the PAG (white arrowheads), with no labelled dendrites. (f) Round somata near the PAG (white arrowheads). Scale bars = 300 μm.

The pattern of distribution of NADPH-d positive neurons was as follows: (1) medium-sized marginal cells, small pyriform cells, medium-sized narrow-field vertical cells, medium-sized wide-field vertical cells and very few small- to medium-sized horizontal cells in the superficial layers; (2) medium-sized stellate cells, big narrow-field vertical cells, and few medium-sized to big wide-field vertical cells in the deep collicular layers, including the PAG. For all neuronal types, the mean, the typical deviation and the range were calculated (Table 1).

Table 1.

Statistical analysis of NADPH-d positive cells in the superior colliculus. All values are expressed in µm

Neuronal types Mean SD Range
Marginales 12.76 3.08  9–17
Piriformes  6.47 0.96  5–8
Narrow-field vertical (superficial) 11.71 0.93 10–13
Wide-field vertical (superficial) 16.68 2.46 12–20
Horizontal 12.64 2.62  9–16
Stellate 15.90 3.06 11–21
Narrow-field vertical (deep) 25.66 3.99 19–32
Wide-field vertical (deep) 30.78 5.43 22–40
Open in a new tab

NADPH-d and CaBP (PV and CB) double labelling

Examination of double-labelled sections by NADPH-d and PV revealed different patterns of distribution. The distribution of positive NADPH-d neurons has been described in detail. The distribution of PV positive neurons was as follows: (1) marginal, horizontal, pyriform, narrow-field vertical cells, and stellate cells in the superficial layers; (2) small- to medium-size stellate cells, big narrow-field vertical cells and wide-field vertical cells and large multipolar cells (up to 67 µm) in the deep collicular layers. With reference to the CB immunoreactive cells, the pattern of distribution was: (1) marginal, horizontal, pyriform, narrow-field vertical cells, stellate and wide-field vertical cells in the superficial layers; (2) small stellate cells, medium- to large-size multipolar, narrow-field vertical cells and wide-field vertical cells in the deep collicular layers.

Combined PV and CB immunocitochemistry and diaphorase histochemistry showed differences between both CaBP. PV/NADPH-d revealed a complete lack of co-localization within individual neurons and fibres. Interestingly, there appears to be no overlap between PV-containing cells and NADPH-d positive neurons in the rabbit SC. Although there was a certain similarity between their morphology and distribution within the superficial layers of the SC, we found no cells that were obviously labelled by both markers. In other words, the two stained cell populations were consistently separated. On the contrary, our study provides evidence that neurons in the superficial layers of the SC contained both NADPH-d and CB. Thus, in CB/NADPH-d double stained sections there were neurons, which showed a combination of the two reaction products, with a brownish soma and dark grey or black processes (Fig. 6a,b). In general, their medium-sized fusiform somas and their position in the SGS and SO characterized all these neurons, which seemed to be particular types of narrow-field and wide-field vertical cells. These cells were very different from CB-labelled or NADPH-d-stained neurons (Fig. 6c,d). Double CB/NADPH-d collicular cells were scarce in comparison with the total amount of collicular cells. In double-labelled sections, 10–12% of the positive neurons contained both markers. This percentage was derived from counts of a large sample of neurons found in a total of eight double-labelled sections of five animals, representing the entire rostro-caudal extent of these cell groups. The percentage of cells that co-localized both substances is thus likely to be an accurate estimate of the total population. It must be emphasized that these cells displayed an anteroposterior gradient of density (Fig. 7). Thus, double NADPH-d-CB collicular cells were more numerous at anterior levels, and gradually decreased caudally.

Fig. 6.

Fig. 6

Open in a new tab

Colour photomicrographs of double labelled NADPH-d/CB cells in coronal sections of the rabbit SC (dorsal to the top, lateral to the right). (a) Double labelled cell in the upper part of the SO (black arrow). The neuronal morphology is similar to that of a narrow-field vertical cell. (b) Double labelled cell in the upper part of the SO (black arrowhead). The neuronal morphology is similar to that of a wide-field vertical cell. (c) CB positive narrow-field vertical cell (black arrow). (d) NADPH-d stained cells (black arrows). Scale bars = 100 μm.

Fig. 7.

Fig. 7

Open in a new tab

Distribution of NADPH-d-stained and CB-immunoreactive cells (small dots), and double labelled CB/NADPH-d neurons (large dots) in rostro-caudal sections of the rabbit SC. Note the anteroposterior gradient of the distribution of double labelled cells. Each dot corresponds to one cell.

Discussion

In the present study we demonstrate that positive neurons and fibres form three lamina in the rabbit SC: one in the superficial grey layer and upper part of the optic layer, one in the intermediate grey layer and one in the deepest layers. In addition, the present findings indicate that the expression of NADPH-d is largely separate from that of PV and CB in neurons of the SC of the adult rabbit. However, CB coexists with NADPH-d in a minority of collicular cells, which are always located within the superficial layers of the SC (SGS and SO). These cells seemed to be a concrete type of narrow-field and/or wide-field vertical neurons. Our data also indicate that these double-labelled neurons display an antero-posterior gradient of distribution, being more numerous at anterior levels.

In rat SC, the greatest concentration of NADPH-d neurons was located in the SGS, whereas stained neurons were very rare in SO (González-Hernández et al. 1992; Tenório et al. 1995, 1996). The superficial layers of the cat also revealed NADPH-d-stained neurons (Mizukawa et al. 1989), but no further details were given on their morphology. In Macaque monkeys, NADPH-d positive cells were located mainly in deeper layers and were rare in the SGS (Satoh et al. 1995). Our results show that NADPH-d positive neurons in the rabbit are present practically all over the SC, but qualitative assessment of histochemical reactions shows that those cells occurred in a great concentration in three layers. This pattern is similar to that described in the opossum (Giraldi-Guimarães et al. 1999) and different from that observed in rats, in which NADPH-d expressing neurons are abundant in SGS and nearly absent from SO (González-Hernández et al. 1992; Tenório et al. 1995, 1996). The rabbit pattern appears also to differ from that of visually specialized animals such as cats (Mizukawa et al. 1989). This major concentration of weak immunoreactive cells in the rabbit SO is intriguing in view of the localization of the ipsilateral terminal fields, which are segregated in SZ, rather than overlapping with contralateral fields as in the rat SO. However, it cannot be excluded from consideration that ipsilateral retinocollicular axons synapse onto distal, apical ends of dendrites of immunoreactive neurons with cell bodies in SO. The differences between mammalian species suggest that pattern of NADPH-d/NOS expression in SC might represent the variability that probably exists as well in other structures related to the visual system as the visual cortex or the LGN, and could represent divergences during evolution. In agreement with NADPH-d studies in the cat and the rat (Mizukawa et al. 1989; González-Hernández et al. 1992), and an immunohistochemical study in the monkey (Satoh et al. 1995), there is an apparent continuity between clusters of labelled cells in the deep layers of the SC and the dorsal portion of the PAG, with immunoreactive dendrites crossing the border between the two regions. This continuity might represent a highly conserved feature of the mammalian midbrain.

A prominent anatomical feature of the SGI in the rabbit SC is the patch-like organization of cholinergic fibres first described by Greybiel (1978). These dense fibre patches within contain fibres with a similar morphology, including a thin calibre and terminal buttons. However, they are more varied in diameter than those found within the SGS. The vast majority of these fibres are also thought to co-localize ACh, and to arise from the pedunculopontine and lateral dorsal tegmental nucleus (Hall et al. 1989; Vincent & Kimura, 1992). There are, however, other possible origins of these axons, including the NADPH-d positive neurons within the SGI (Scheiner et al. 2000).

Contrary to the diffuse distribution of NADPH-d positive fibres within the superficial grey layer, NADPH-d positive fibres form vertical ‘streams’ in the deep layers. NADPH-d activity in different layers of the mouse and rat SC was first visualized by Wallace (1986), who showed that interruption of the afferent pathways from the brainstem left the enzyme activity in the superficial layers unaltered while abolishing the patches in the intermediate layers. To our knowledge, the source of the NADPH-d input to the SC has not been identified. However, it is unlikely that retinal ganglion cells could be its source, as the NADPH-d amorphous reaction product in superficial layers is not abolished by bilateral eye enucleation (Tenório et al. 1998) and can even increase contralaterally to a removed eye in the first weeks (Yan et al. 1995).

NADPH-d-labelled cells in the rabbit SC clearly represent several types of collicular neurons. Two of these cell types represent populations that are likely to be inhibitory GABAergic interneurons. It has been shown that both pyriform and horizontal cells within the superficial layers of cat SC accumulate radiolabelled GABA and are labelled by antibodies to both GABA and glutamic acid decarboxylase (GAD) (Mize et al. 1982; Mugnaini & Oertel, 1985; Mize, 1992). Although in the present study we have not confirmed that NADPH-d containing pyriform and horizontal neurons in the rabbit SC contain GABA, this prospect seems very likely as most of these cells are immunoreactive for GABA (Mize, 1992).

The other three cell morphologies, narrow-field vertical cells, wide-field vertical cells and stellate, are likely to be projection neurons. A sizeable population of narrow-field vertical neurons within the lower SGS projects to the lateral geniculate nucleus, whereas many stellate neurons and wide-field vertical cells in this region project to the lateral posterior nucleus of the thalamus (Caldwell & Mize, 1981).

Double-stained CB/NADPH-d sections showed that there are neurons, possibly a specific type of narrow-field and wide-field vertical cells, which have a different aspect from the combination of the two reaction products. The co-localization of NOS with CB in other parts of the CNS as the basal forebrain (Geula et al. 1993) or the cerebral cortex (Bertini et al. 1996) of several mammalian species has been studied previously. In cortical neurons of the rat, the coexistence of NOS with PV is practically null (18 PV/NADPH-d cells out of 775 NADPH-d ones in Dun et al. 1994), whereas the co-localization between NOS and CB is more frequent, especially in the frontal cortex. Interestingly, the coexistence of NOS with calretinin (CR), another CaBP of the EF-hand family, was found to be negligible in neocortical neurons (Dun et al. 1994). In addition, the coexistence of NADPH-d and/or NOS with PV and CR seems remarkably rare also in the subcortical CNS centres in which it has been hitherto investigated. Neurons containing NOS, PV and CR represent three almost entirely separate GABAergic cell populations in the striatum of rats (Kubota et al. 1993). A rare coexistence between NADPH-d histochemical positivity and CR immunoreactivity was found in the hypothalamic magnocellular secretory nuclei (Arévalo et al. 1993) and olfactory bulb (Alonso et al. 1993) of the rat. Finally, NOS and PV do not coexist in spinal neurons of the superficial laminae of the dorsal horn in the rat (Laing et al. 1994). Our study supports previous results indicating that the neurons displaying NADPH-d activity (which is almost entirely equivalent to NOS) do not contain PV. We have demonstrated the lack of coexistence between NOS and PV in the SC of the rabbit. In addition, our findings show that there are double-labelled neurons in the superficial layers of the rabbit SC. In other words, there are neurons of the same type that differ in their staining, and hence probably differ in their intercellular messenger systems. It means that: (1) groups that were originally defined according to their location and classical transmitter phenotype might in fact be composed of multiple, functionally distinct neuronal subsets or (2) the existence of specific visual pathways in the rabbit SC. The final conclusion could be that neurons, which are morphologically similar to other narrow-field and wide-field vertical cells, are physiologically different. Our result points to a certain degree of selectivity in the distribution of these CB/NADPH-d cells, whose highest proportion was found within the SO and the SGS. It is important to consider that this is the region where the visual information is reorganized and integrated. Then it could be suggested that there are specific visual pathways in which these double-labelled neurons play a specific role, different from the NADPH-d and CB populations in the same layers.

The control of Ca2+ levels plays a fundamental role in the mechanisms in which NO has been implicated in the CNS. It has been repeatedly suggested that NO could act as a retrograde messenger in synaptic plasticity models, such as long-term potentiation and long-term depression. There are previous studies which suggest that NO may also be involved in the mechanism of neurotransmitter release mediated by NMDA receptors in, for example, the cerebral cortex (Laing et al. 1994). NO production requires concentrations of free Ca2+ well above those seen in the resting neurons (Bredt & Snyder, 1990) and calcium influx is required for the NMDA receptor-mediated synaptic mechanisms. On the other hand, it has been shown that NO-induced vesicle exocytosis can take place without a raise in Ca2+ levels (Bredt & Snyder, 1990). In addition, our findings recall attention to the possible presence of these double-labelled CB/NADPH-d neurons not only in the rabbit but also in other species. In this complex picture, the selectivity of the expression of NADPH-d and certain CaBP may provide clues to the understanding of the interrelationships between the NO action and calcium-buffering systems in neurons, and of course may facilitate a better understanding of the mammalian collicular organization.

Acknowledgments

We thank Dr Miguel A. Moreno Romo (Departamento de Patología Animal I, Facultad de Veterinaria, Universidad Complutense, Madrid) for his expert assistance in the preparation of this manuscript.

Reference

  1. Alonso JR, Arevalo R, Porteros A, Brinon JG, Lara J, Aijon J. Calbindin D28K and NADPH-diaphorase activity are localized in different populations of periglomerular cells in the rat olfactory. J. Chem. Neuroanat. 1993;200:1–6. doi: 10.1016/0891-0618(93)90002-l. [DOI] [PubMed] [Google Scholar]
  2. Arcelli P, Frassoni C, Regondi MC, De Biasi S, Spreafico R. GABAergic neurons in mammalian thalamus: a marker of thalamic complexity? Brain Res. Bull. 1997;42:27–37. doi: 10.1016/s0361-9230(96)00107-4. [DOI] [PubMed] [Google Scholar]
  3. Arévalo R, Sanchez F, Alonso JR, Rubio M, Aijon J, Vazquez R. Infrequent cellular coexistence of NADPH-diaphorase and calretinin in the neurosecretory nuclei and adjacent areas of the rat hypothalamus. J. Chem. Neuroanat. 1993;6:335–341. doi: 10.1016/0891-0618(93)90008-r. [DOI] [PubMed] [Google Scholar]
  4. Atoji A, Yamamoto Y, Suzuki Y. Innervation of NADPH diaphorase-containing neurons correlated with acetylcholinesterase, tyrosine hydroxylase and neuropeptides in the pigeon cloaca. J. Anat. 2001;198:181–188. doi: 10.1046/j.1469-7580.2001.19820181.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barbaresi P, Spreafico R, Frassoni C, Rustioni A. GABAergic neurons are present in the dorsal column nuclei but not in ventroposterior complex in rats. Brain Res. 1986;382:305–326. doi: 10.1016/0006-8993(86)91340-5. [DOI] [PubMed] [Google Scholar]
  6. Bentivoglio M, Spreafico R, Minciacchi D, MacChi G. GABAergic interneurons and neuropil of the intralaminar thalamus: an immunohistochemical study in the rat and the cat. Exp. Brain Res. 1991;87:85–95. doi: 10.1007/BF00228509. [DOI] [PubMed] [Google Scholar]
  7. Bertini G, Peng Z-C, Bentivoglio M. The chemical heterogeneity of cortical interneurons: nitric oxide synthase vs. calbindin and parvalbumin immunoreactivity in the rat. Brain Res. Bull. 1996;39:261–266. doi: 10.1016/0361-9230(95)02133-7. [DOI] [PubMed] [Google Scholar]
  8. Bredt DS, Snyder SH. Isolation of nitric oxide synthase, a calmodulin-requiring enzyme. Proc. Natl Acad. Sci. USA. 1990;87:682–685. doi: 10.1073/pnas.87.2.682. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Caldwell RB, Mize RR. Superior colliculus neurons which project to the lateral posterior nucleus have varying morphologies. J. Comparative Neurol. 1981;203:53–66. doi: 10.1002/cne.902030106. [DOI] [PubMed] [Google Scholar]
  10. Cork RJ, Baber SZ, Mize RR. Calbindin D28k and parvalbumin-immunoreactive neurons form complementary sublamina in the rat superior colliculus. J. Comparative Neurol. 1998a;394:205–217. [PubMed] [Google Scholar]
  11. Cork RJ, Perrone ML, Bridges D, Wandell J, Scheiner CA, Mize RR. A web-accessible digital atlas of the distribution of nitric oxide syntheses in the mouse brain. Prog. Brain Res. 1998b;118:37–50. doi: 10.1016/s0079-6123(08)63199-4. [DOI] [PubMed] [Google Scholar]
  12. Cork RJ, Calhouin T, Perrone M, Mize RR. Postnatal development of nitric oxide synthase expression in the mouse superior colliculus. J. Comparative Neurol. 2000;427:581–592. [PubMed] [Google Scholar]
  13. Dun NJ, Huang R, Dun SL, Forstermann U. Infrecuent colocalization of nitric oxide synthase and calcium binding proteins immunoreactivity in rat neocortical neurons. Brain Res. 1994;666:289–294. doi: 10.1016/0006-8993(94)90786-2. [DOI] [PubMed] [Google Scholar]
  14. Geula C, Schatz CR, Mesulam M-M. Differential localization of NADPH-diaphorase and calbindin-D28k within the cholinergic neurons of the basal forebrain, striatum and brainstem in the rat, monkey, baboon and human. Neuroscience. 1993;54:461–476. doi: 10.1016/0306-4522(93)90266-i. [DOI] [PubMed] [Google Scholar]
  15. Giraldi-Guimarães A, Tenório F, Brüning G, Mayer B, Mendez-Otero R, Cavalcante LA. Nitric oxide synthase expression in the Opossum superior colliculus: a histochemical, immunohistochemical and biochemical study. Brain Behav. Evol. 1999;199:303–313. doi: 10.1159/000006630. [DOI] [PubMed] [Google Scholar]
  16. González-Hernández T, Conde-Sedín TM, Meyer G. Laminar distribution and morphology of NADPH-diaphorase containing neurons in the superior colliculus and underlying periaqueductal grey of the rat. Anat. Embryol. 1992;186:245–250. doi: 10.1007/BF00174146. [DOI] [PubMed] [Google Scholar]
  17. González-Soriano J, González-Flores ML, Contreras-Rodríguez J, Rodríguez-Veiga E, Martínez-Sainz P. Calbindin D28k and parvalbumin immunoreactivity in the rabbit superior colliculus: an anatomical study. Anat. Rec. 2000;259:334–346. doi: 10.1002/1097-0185(20000701)259:3<334::AID-AR100>3.0.CO;2-K. [DOI] [PubMed] [Google Scholar]
  18. Greybiel AM. A stereometric pattern of distribution of acetylcholinesterase in the deep layers of the superior colliculus. Nature. 1978;272:539–541. doi: 10.1038/272539b0. [DOI] [PubMed] [Google Scholar]
  19. Hall WC, Ftizpatrick D, Klatt LL, Rackowski D. Cholinergic innervation of the superior colliculus in the cat. J. Comparative Neurol. 1989;287:495–514. doi: 10.1002/cne.902870408. [DOI] [PubMed] [Google Scholar]
  20. Harris RM, Hendrickson AE. Local circuit neurons in the rat ventrobasal thalamus. A GABA immunocytochemical study. Neuroscience. 1987;21:229–236. doi: 10.1016/0306-4522(87)90335-6. [DOI] [PubMed] [Google Scholar]
  21. Huang PL, Dawson TM, Bredt DS, Snyder SH, Fishman MC. Targeted disruption of the neuronal nitric oxide synthase gene. Cell. 1993;75:1276–1285. doi: 10.1016/0092-8674(93)90615-w. [DOI] [PubMed] [Google Scholar]
  22. Illing RB. Choline acetyltransferasa-like immunoreactivity in the superior colliculus of the cat and its relation to the pattern of acetylcholinetransferase staining. J. Comparative Neurol. 1990;296:32–46. doi: 10.1002/cne.902960104. [DOI] [PubMed] [Google Scholar]
  23. Kubota Y, Mikawa S, Kawaguchi Y. Neostriatal GABAergic interneurons contain NOS, calretinin or parvalbumin. Neuroreport. 1993;5:205–208. doi: 10.1097/00001756-199312000-00004. [DOI] [PubMed] [Google Scholar]
  24. Laing J, Todd AJ, Heizmann CW, Schmidt HHHW. Subpopulations of GABAergic neurons in laminae I–III of rat spinal dorsal horn defined by coexistence with classical transmitters peptides, nitric oxide synthase or parvalbumin. Neuroscience. 1994;61:123–132. doi: 10.1016/0306-4522(94)90065-5. [DOI] [PubMed] [Google Scholar]
  25. Lane RD, Allan DM, Bennet-Clarke CA, Howell DL, Rhoades RW. Projection status of the calbindin- and parvalbumin-immunoreactive neurons in the superficial layers of the rat’s superior colliculus. Visual Neurosci. 1997;14:277–286. doi: 10.1017/s095252380001141x. [DOI] [PubMed] [Google Scholar]
  26. Leuba G, Saini K. Calcium-binding proteins immunoreactivity in the human subcortical and cortical visual structures. Visual Neurosci. 1996;13:997–1009. doi: 10.1017/s0952523800007665. [DOI] [PubMed] [Google Scholar]
  27. Mize RR, Spencer RF, Sterling P. Two types of GABA-accumulating neuronsin the superficial grey layer of the cat superior colliculus. J. Comparative Neurol. 1982;206:180–192. doi: 10.1002/cne.902060207. [DOI] [PubMed] [Google Scholar]
  28. Mize RR. The organization of GABA in the mammalian superior colliculus. Prog. Brain Res. 1992;90:219–248. doi: 10.1016/s0079-6123(08)63616-x. [DOI] [PubMed] [Google Scholar]
  29. Mize RR, Luo Q, Butler GD, Jeon C-J, Nabors B. The calcium binding proteins parvalbumin and calbindin D-28K form complementary patterns in the cat superior colliculus. J. Comparative Neurol. 1992;320:234–256. doi: 10.1002/cne.903200208. [DOI] [PubMed] [Google Scholar]
  30. Mizukawa K, Vincent SR, McGeer PL, McGeer EG. Distribution and reduced-nicotinamide-adenine-dinucleotide-phosphate diaphorase-positive cells and fibres in the cat central nervous system. J. Comparative Neurol. 1989;279:281–311. doi: 10.1002/cne.902790210. [DOI] [PubMed] [Google Scholar]
  31. Mugnaini E, Oertel WH. An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In: Bjorklund A, Holkfelt T, editors. GABA and Neuropeptides in the CNS, Handbook of Chemical Neuroanatomy. Vol. 4. Amsterdam: Elsevier; 1985. pp. 436–622. (Part 1) [Google Scholar]
  32. Penny GR, Fitzpatrick D, Schemechel DE, Diamond IT. Glutamic acid decarboxylase-immunoreactive neurons and horseradish peroxidase-labeled projections neurons in the ventral posterior nucleus of the cat and Galago Senegalensis. J. Neurosci. 1983;3:1686–1887. doi: 10.1523/JNEUROSCI.03-09-01868.1983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Penny GR, Conley M, Schemechel DE, Diamond IT. The distribution of glutamic acid decarboxylase immunoreactivity in the diencephalons and the mesencephalon of the opossum and the rabbit. J. Comparative Anat. 1984;228:38–56. doi: 10.1002/cne.902280106. [DOI] [PubMed] [Google Scholar]
  34. Satoh K, Arai R, Ikemoto M, Narita M, Nagai T, Oshima H, Kitahama K. Distribution of nitric oxide synthase in the central nervous system of Maccaca fuscate: subcortical regions. Neuroscience. 1995;66:685–696. doi: 10.1016/0306-4522(95)00040-p. [DOI] [PubMed] [Google Scholar]
  35. Scheiner C, Arceneaux R, Guido W, Kratz K, Mize RR. Nitric oxide synthase distribution in the cat superior colliculus and co-localization with choline acetyltransferase. J. Chem. Neuroanat. 2000;18:147–159. doi: 10.1016/s0891-0618(00)00037-5. [DOI] [PubMed] [Google Scholar]
  36. Spreafico R, Schemechel DE, Ellis LC, Rustioni A. Cortical relay neurons and interneurons in the nucleus ventralis posterolateralis of cats. A horseradish peroxidase, electron-microscopic, Golgi and immunocytochemical study. Neuroscience. 1983;9:491–509. doi: 10.1016/0306-4522(83)90168-9. [DOI] [PubMed] [Google Scholar]
  37. Tenório F, Giraldi-Guimarães A, Mendez-Otero R. Developmental changes of nitric oxide synthase in the rat superior colliculus. J. Neurosci. Res. 1995;42:633–637. doi: 10.1002/jnr.490420505. [DOI] [PubMed] [Google Scholar]
  38. Tenório F, Giraldi-Guimarães A, Mendez-Otero R. Morphology of NADPH-diaphorase-positive cells in the retinoceptive layers of the developing rat superior colliculus. Int. J. Dev. Neurosci. 1996;14:1–10. doi: 10.1016/0736-5748(95)00085-2. [DOI] [PubMed] [Google Scholar]
  39. Tenório F, Giraldi-Guimarães A, Santos HR, Cintra WM, Mendez-Otero R. Eye enucleation alters intracellular distribution of NO synthase in the superior colliculus. Neurosci. Report. 1998;9:145–148. doi: 10.1097/00001756-199801050-00029. [DOI] [PubMed] [Google Scholar]
  40. Thomas DJ, Pearse AGE. The fine localization of dehydrogenases in the nervous system. Histochemistry. 1961;2:266–282. doi: 10.1007/BF00736504. [DOI] [PubMed] [Google Scholar]
  41. Vincent SR, Kimura H. Histochemical mapping of nitric oxide synthase in the rat brain. Neuroscience. 1992;46:755–784. doi: 10.1016/0306-4522(92)90184-4. [DOI] [PubMed] [Google Scholar]
  42. Wallace MN. Spatial relationship of NADPH-diaphorase and acetylcholinesterase lattices in the rat and mouse superior colliculus. Neuroscience. 1986;19:381–391. doi: 10.1016/0306-4522(86)90268-x. [DOI] [PubMed] [Google Scholar]
  43. Williams CV, Nordquist D, McLoon SC. Correlation of nitric oxide synthase expression with changing patterns of axonal projections in the developing visual system. J. Neurosci. 1994;14:1746–1755. doi: 10.1523/JNEUROSCI.14-03-01746.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Yan XX, Garey LJ, Liang Y, Von Bussmann KA, Jen LS. Increased expression of NADPH-diaphorase in visual centres after unilateral optic nerve transection in the rat. J. Brain Res. 1995;4:485–488. [PubMed] [Google Scholar]

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES