Research 795Ž. 1998 204±214

Research report immunoreactivity in the rat main olfactory bulb

JesusÂÄ G. Brinon a, Rosario Arevalo a, Carlos Crespo a, Ignacio G. Bravo a, Katsuo Okazaki b, Hiroyoshi Hidaka b, Jose Aijon a, Jose  R. Alonso a,) a Dpto. Biologõa Celular y Patologõa, UniÕersidad de Salamanca, Salamanca E-37007, Spain b Department of Pharmacology, Nagoya UniÕersity School of Medicine, Nagoya 466, Japan

Accepted 17 March 1998

Abstract

The morphological characteristics and distribution of neurocalcinŽ. NC -immunoreactive elements were studied in the rat main olfactory bulbŽ. OB using a polyclonal antibody and the avidin±biotin immunoperoxidase method. NC-positive elements were abundant in the glomerular layerŽ. GL , where numerous immunostained external tufted cells and periglomerular cells were detected. Other less abundant NC-immunolabeled populations included middle and internal tufted cells, Van Gehuchten cells, horizontal cells, vertical cells of Cajal, deep short-axon cells and granule cells. This study demonstrates the presence of NC immunoreactivity in subsets of different neuronal types in the rat main OB. This -binding has been found in interneurons, and no evidence of immunoreactivity to NC is detected in projecting . Despite the large population of labeled external tufted cells, most of them belong according to morphological criteria to the local circuit group and some others to those with interbulbar andror intrabulbar connections. The identification of neuronal subpopulations expressing NC provides a further characterization and shows the existence of biochemical differences within morphologically identical neurons. Thus, this marker may be a useful tool in unravelling the circuitries of the rodent OB in both normal and experimental conditions. The exact physiological function of NC in the olfactory system remains unknown. On the basis of similarities to , it could be involved in mechanisms responsible for sensory . Additionally, its calcium-binding abilities may contribute to improve the temporal precision of stimuli transmission, or be concerned with general calcium-related events occurring in specific interneuronal groups. q 1998 Elsevier Science B.V. All rights reserved.

Keywords: Calcium-binding protein; Neurocalcin; Olfactory bulb; Specific cell marker

1. Introduction understood, however, since they are expressed in particular neuronal groups. When immunohistochemically detected, Ž.Ž . they provide a selective staining of distinct neuronal popu- Neurocalcin NC mol.wt. 23±24 kDa, pI 5.3±5.5 is a wx calcium-binding protein isolated from bovine brain where lations throughout the brain 9,10,30,32 . In addition to this four isoforms have been identifiedwx 36 . This protein shares neurochemical characterization, the study of their distribu- similarities with visinin and recoverin, and has been in- tion patterns in anatomically well-known structures pro- cluded into the neural calcium-sensor group de- vides a basis to interpret their functional roles. The olfac- fined by the -specific protein, recoverin tory bulbŽ. OB , with its highly laminated structure, the wx15 . presence of several types of fully characterized output Several types of calcium-binding proteins are present in neurons and interneurons, and its well-known connections the nervous system where they mediate the effects of is an ideal region to study the general principles of neural calcium ions in neurons. Their exact function is poorly and chemical organization. Immunoblot analysis demonstrated that NC is present in the OB, cerebrum, cerebellum, brainstem, spinal cord, retina, pituitary, and adrenal glandwx 5,24,25 . Immunohisto- chemical studies have shown NC-immunoreactivity in ) wx Corresponding author. Fax: q34-923-294549; E-mail: amacrine and ganglion cells of the bovine retina 24 , and [email protected] in different regions of the olfactory system, such as the

0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S0006-8993Ž. 98 00311-4 J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214 205 receptors of the olfactory epithelium and vomeronasal ethanol series, cleared in xylene and coverslipped with organwx 6,14 , and in the main and accessory OBs wx 5,28 . In EntellanŽ. Merck . the accessory OB, several neuronal types, including tufted cells, periglomerular cells, Van Gehuchten cells and verti- 2.2. Controls cal cells of Cajal demonstrated NC-immunoreactivitywx 28 ; whereas in the main OB, the NC-immunopositive elements The primary antibody used has been fully characterized were described as groups of external tufted cells located by immunoblottingwx 24 and does not cross-react with other Ž. both in the glomerular layer GL , and in the external brain proteins. Its specificity was tested by antibody pread- Ž.wx plexiform layer EPL 5 . However, in addition to these sorption with native NCŽ. 2.5 mgrml , obtaining a com- NC-containing elements, we have found other different plete abolition of the immunohistochemical staining. In NC-immunopositive neuronal types distributed throughout addition, three controls of the immunohistochemical proce- all layers of the rat main OB, except in the olfactory nerve dure were carried out by incubating sections in the same Ž. layer ONL . Therefore, the aim of the present study is to media omitting:Ž. 1 the primary antibody; Ž. 2 the biotinyl- complete the description of NC-containing elements in the ated immunogammaglobulin, andŽ. 3 the avidin peroxidase rat main OB in order to obtain a full characterization of all complex. No residual reaction was observed. neuronal types expressing this calcium-binding protein. 2.3. Analysis

2. Materials and methods Sections were analyzed under a Zeiss III microscope and drawings were made through a camera lucida adapted Five adult male Wistar rats, weighing between 200 and to this microscope. Photographs were taken on a Zeiss III 250 g and grown and kept under standard laboratory photomicroscope. Positive neurons were counted with 16= conditions were used. The animals were deeply anaes- and 40= planapochromatic objectives. Quantitative as- thetized with ketamineŽ. Ketolar 50 mgrkg body weight sessments of the relative density of NC-immunoreactive and perfused through the ascending aorta with 100 ml neurons were performed by analyzing coronal sections at Ringer solution followed by 400 ml fixative solution made nine levels in each animal. The most rostral section was up of 4% paraformaldehyde and 15% saturated picric acid that where all bulbar layers were clearly delineated. Eight in 0.1 M phosphate buffer, pH 7.3Ž. PB . After being additional levels were chosen each 300 mm throughout the removed, the OBs were dissected and postfixed at 48C for rostro-caudal axis of the OB. All labeled neurons, regard- two additional hours in the same fixative. After several less of their staining intensity, were included in the evalua- washes in PB, 35-mm thick vibratome sections were ob- tion. The number of immunostained cells for each neuronal tained or, after cryoprotection in 30% sucrose in PB at type was referred to the total area of the section, and the 48C, 25-mm thick sections were cut on a cryostat. density is given as number of cells per mm2 of fixed tissue. ANOVA was used for the statistical analysis of the 2.1. Immunohistochemistry results. Fisher PLSD and Sheffe F-tests did not detect statistically significant differences in the density of each Free-floating sections were washed in PB and incubated neuronal type neither throughout the rostro-caudal axis of for 30 min in 10% normal goat serum in PB. After being the OB nor between animals. washed in PB they were incubated for 48 h at 48Cin Individual labeled cells with their focusing plane in the anti-NC primary antibody, raised in rabbit against rat brain middle of the section thickness were drawn with a camera = NCwx 24 diluted 1:8000 in PB. Thereafter, the sections lucida using the 100 oil immersion objective, and their were processed according to the avidin±biotin immunoper- cell-body diameters were plotted on a digitizer tablet con- oxidase methodwx 13 . The sections were washedŽ 3=10 nected to a semiautomatic image analysis systemŽ MOP- min. in PB and incubated for 1 h at room temperature in Videoplan, Kontron. . For each neuronal type, 50 cells biotinylated anti-rabbit immunogammaglobulinŽ Vecta- from each analyzed section in all five animals were mea- stain, Vector Laboratories, Burlingame, USA. diluted 1:250 sured, with the exception of those neuronal types which in PB. After washing in PBŽ. 3=10 min , sections were were rarely observed. In this case, all observable cells transferred to Vectastain ABC reagent diluted 1:200 in PB were measured. for two additional hours. The antibody±enzyme complex was visualized by incubating the sections in a solution of X 0.05% 3,3 -diaminobenzidine tetrahydrochlorideŽ. Sigma 3. Results with 0.003% hydrogen peroxide in 0.2 M Tris±HCl buffer, pH 7.6, for 5±10 min, until the desired staining intensity After NC-immunohistochemistry, most immunopositive was reached. Sections were then rinsed in PB, mounted on neurons displayed densely stained cell bodies, dendrites gelatin-coated slides, air dried, dehydrated through graded and lengthy axons. Weakly labeled neurons were also 206 J.G. BrinonÄ et al.rBrain Research 795() 1998 204±214 observed, mainly located in the inner layers. The results OB, except the ONL, demonstrated NC-immunopositive were constant among all animals used and they are sum- elementsŽ. Fig. 2a . marized in Fig. 1 and Table 1. The different NC-im- 3.1. Olfactory nerÕe layer munopositive neuronal types were identified following the criteria of previous classifications based on Golgi impreg- The ONL was completely negative. Olfactory fibers nationswx 20,21,27,33 . The NC-immunostaining was mainly coursing from the olfactory receptor cells and all ONL located in the GL and the EPL, although all layers of the intrinsic cells were NC-immunonegativeŽ. Fig. 2a .

Fig. 1. Camera lucida drawings showing the different neuronal types displaying NC-immunoreactivity in the rat OB. Glomeruli and mitral cell profiles are marked with dashed lines. Scale bar: 100 mm.Ž. a±d Periglomerular cells. Ž. e±h External tufted cells in the GL. Ž. i External tufted cell in the EPL. Ž. j Van Gehuchten cell.Ž. k Internal tufted cell. Ž l,m . Displaced granule cells. n: Horizontal cell. Ž o,p . Vertical cells in the IPL. Ž q±u . Granule cells. Ž v,w . Deep short-axon cells in the innermost region of the GCL and in the WM. ONL: olfactory nerve layer. GL: Glomerular layer. EPL: External plexiform layer. MCL: Mitral cell layer. IPL: Internal plexiform layer. GCL: Granule cell layer. WM: White matter. J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214 207

Table 1 NC-positive neuronal types in the rat olfactory bulb Cell type Location Max. diameter mmŽ. mean"S.E.M. Frequencya Staining intensity Periglomerular cell GL 9.0"0.2 qqq low External tufted cell GLrEPL 12.3"0.3 qqq high Van Gehuchten cell EPL 13.2"0.5 qq high Middle tufted cell EPL 15.7"0.6 qq low Internal tufted cell EPL 18.3"0.5 q low Displaced granule cell EPLrMCL 9.0"0.3 qq medium Horizontal cell MCLrIPL 14.4"0.4 qq high Vertical cell of Cajal MCLrIPL 12.0"0.3 q medium Granule cell GCL 8.6"0.5 qqq low Deep short-axon cell GCLrWM 14.9"0.7 q medium a Frequency: q: 1 to 5 cells per section; qq: 5 to 20 cells per section; qqq: more than 20 cells per section.

3.2. Glomerular layer ramifying close to the soma were also observed. More rarely, two primary dendrites arose from different poles of The most prominent staining was obtained in the GL the cell body and entered either a single glomerulus or where two different neuronal populations were clustered adjacent glomeruli. A thin axon arose from the opposite around the glomeruliŽ. Fig. 1a±h, Fig. 2a . One of these side of the cell body coursing towards the EPL. In most populations was formed by neurons of 10±15 mmof cases, the axons did not penetrate deeper into the EPL; but maximum diameter, with spherical or fusiform somata, and they extended into the superficial region of this layer, strong immunostaining intensity. They surrounded the around adjacent glomeruli. In accordance with these mor- glomeruli although they were most abundant on their inner phological features, these cells were identified as half, that is, close to the EPLŽ. Fig. 1e±h, Fig. 2a±c . The periglomerular cells. external cell pole gave rise to a single, thick primary dendrite that divided close to the soma, and each sec- 3.3. External plexiform layer ondary dendrite subsequently branched again, producing a dense tuft. The terminal dendritic branches lacked spines NC-immunopositive elements appeared throughout the and varicosities, although because of the strong staining of whole extent of this layer. Both NC-immunoreactive neu- the glomerular neuropil it was sometimes difficult to ob- rons and fibers were observed. Fibers displayed variable serve the complete arborization of each individual den- thicknesses and trajectories, and most of them coursed drite. The tuft originating from each was normally perpendicular or slightly obliquely to the OB layeringŽ Fig. located in the lateral sides of the glomerulus rather than in 2e. . NC-immunopositive neurons appeared in all regions its central core. In general, the axons of these NC-im- of this layer although they were more numerous at the munopositive neurons arose from the opposite pole from border with the MCLŽ. Fig. 2a . They were morphologi- which the dendrite originated. These thin, finely beaded cally heterogeneous, corresponding to different neuronal axons were orientated in two directions. Some of them types that are described according to their location, from coursed parallel or slightly obliquely to the OB layering, the most external to the most internal ones. that is through the GLrEPL boundary, giving collaterals, Close to the border with the GL, NC-immunoreactive when observed, at the inner side of neighboring glomeruli. neurons had fusiform cell bodies ranging in size from 11 In other cases, the axons were oriented in a descending to 14 mm. They showed a thick and smooth primary fashion, and some of them could be followed deep into the dendrite directed towards the GL that entered into a single EPL and the mitral cell layerŽ. MCL . In accordance with glomerulus forming a characteristic tuft of terminal these characteristics, these cells were classified as external branchesŽ. Fig. 1i . Whereas most primary dendrites ex- tufted cells. tended perpendicular to the OB layeringŽ. Fig. 2d , in a few The other NC-immunoreactive neuronal population cases the main dendrite approached the glomerulus follow- found in the GL was formed by smallerŽ. 6±10 mm, ing oblique or almost horizontal trajectoriesŽ. Fig. 2e . faintly stained neurons with ovoid or round cell bodies Frequently, the opposite pole of the cell body gave rise to Ž.Fig. 2b,c . They were also located around the glomeruli a thinner secondary dendrite that extended parallel to the and showed some variability, including monodendritic and GLrEPL border for long distances without entering any bidendritic cells, as well as monoglomerular and glomerulusŽ. Fig. 1f . These secondary dendrites did not biglomerular onesŽ. Fig. 1a±d . The most frequent dendritic branch in their initial portion, but only in their distal region pattern consisted of a long spiny primary dendrite that where they terminate in several thin branches. In some entered and ramified within a single glomerulus forming a cases, a short axon-like process emerged close to the base wide terminal field. Neurons with this primary dendrite of the secondary dendrite or directly from it, extending 208 J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214 J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214 209 radially inwards in the EPL. All these morphological fea- tions. Most labeled cells occurred in both the MCL and the tures are in accordance with those of tufted cells. IPL, and had morphological characteristics of horizontal Another distinctive type of NC-immunopositive neuron cells. The presence of a continuous row of these cells at of the EPL resembled the morphological characteristics the level of the MCL was a constant feature in the rat OB and location of Van Gehuchten cells. They had piriform or after NC-immunostainingŽ. Fig. 2a . They displayed vari- slightly ovoid somata with sizes between 10 and 15 mmin able morphological shapesŽ. round, piriform, ovoid and maximum diameter. The cell body gave rise to a variable were medium-sizedŽ. 12±15 mm maximum diameter . Den- number of thick and varicose dendrites that branched drites arose from opposite poles of the somata and ex- elaborately close to itŽ. Fig. 1j, Fig. 2e, Fig. 3a,b . Addi- tended throughout the MCL, branching into the adjacent tional thinner dendrites emerged from variable somatic regions of this layerŽ. Fig. 1n, Fig. 3e±g . locations and extended towards superficial regions, branch- The other NC-immunopositive neuronal type observed ing close to the GLŽ. Fig. 3a . No NC-immunostained in these layers had the morphological characteristics of the axon-like process was observed emerging from these cells. vertical cells of CajalŽ. Fig. 1o±p . They had spherical, The most abundant population of NC-positive neurons piriform or ovoid neuronal bodies with maximum diame- in this layer was located at the border with the MCL, and ters between 10 and 13 mm. In addition to the variability they showed the morphological features of granule cells. of the somal shape, variations in the dendritic branching These neurons had spherical or slightly ovoid cell bodies pattern were detected. Most of them exhibited two main with maximum diameters ranging from 8 to 11 mm. They dendritic trunks emerging from opposite poles of the so- exhibited two dendrites arising from opposite poles of the mata, and were radially orientated. Frequently, the dendrite somata that extended radially in opposite directions. One that extended towards the inner layers of the OB was main dendritic trunk coursed towards the external region thicker than the one that crossed the EPL and branched of the EPL branching before the border with the GL was close to the GLŽ. Fig. 4a . More rarely, only one dendrite reachedŽ. Fig. 1l,m, Fig. 3c . The opposite dendritic trunk directed towards either the EPL or the GCL was observed branched profusely close to the cell body, originating a in these cellsŽ. Fig. 4b,c . small dendritic field either in the internal plexiform layer Ž.IPL or in the granule cell layer Ž GCL . . In the same region, close to the border with the MCL, a 3.5. Granule cell layer and white matter few NC-immunostained internal tufted cells were ob- served. They showed fusiform, faintly stained, large cell The largest population of NC-immunopositive cells in bodies, with maximum diameters ranging from 17 to 20 the GCL was composed of small-sized neuronsŽ. 8±11 mm mm. Two thick smooth dendrites arose from their poles identified as granule cells. They had a segregated distribu- coursing parallel to the MCL for distances longer than 400 tion, being only observed in the most external region of the mm, branching into the inner half of the EPLŽ Fig. 1k, Fig. GCLŽ. Fig. 1q±u . No immunostained granule cells were 3d. . Another dendrite emerged from one side of the neu- detected in the middle or inner region of this layer. NC-im- ronal body coursing perpendicular to the OB layering, munoreactive granule cells displayed weak staining re- reaching and entering one glomerulusŽ. Fig. 1k . The stricted to the cell body, whereas dendrites lacked branching pattern of these dendrites, as well as recogniz- immunostaining or were faintly labeled only in their proxi- able axon-like processes, were difficult to observe because mal regionŽ. Fig. 4d . Although the weak staining in this of the weak labeling. neuronal group could be interpreted as doubtful, it disap- peared when the antibody was preincubated with native 3.4. Mitral cell layer and internal plexiform layer NCŽ. not shown . In the internal region of the GCL and in the white The mitral cells did not display NC-immunoreactivity, matter, a few NC-immunopositive fusiform cells with max- although their immunonegative cell bodies were frequently imum diameters ranging from 13 to 16 mm were observed. distinguishable because they were surrounded by strongly According to their size, shape and location, they presum- positive profilesŽ. Fig. 2a . The somal staining in the MCL ably belonged to the deep short-axon cell group, although and IPL was classified into two different neuronal popula- they were difficult to classify with certainty since only the

Fig. 2. Photomicrographs of coronal sections of the rat OB after NC-immunohistochemistry. Scale bar inŽ. a 100 mm; in Ž b±e . 50 mm. Ž. a View of a coronal section of the main olfactory bulb showing NC-immunoreactive elements. The strongest staining can be observed in the GL, although scattered immunostained cellsŽ. open arrows are observed in deeper layers. Ž. b,c : NC-immunostaining in the superficial layers, periglomerular cells Ž. open arrows and external tufted cellsŽ. solid arrows surround the olfactory glomeruli. Ž. d External tufted cell in the outer region of the EPL. Note the thick dendrite reaching the glomerulus.Ž. e Van Gehuchten cell Ž open arrow . and external tufted cell Ž solid arrow . in the border of the EPL and the GL. Obliquely orientated fibers were often observed in this regionŽ. arrowheads . 210 J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214

Fig. 3. Photomicrographs of NC-positive neurons in the EPL, MCL and IPL of the rat OB. Scale bar: 50 mm for all figures.Ž. a,b Van Gehuchten cells in the EPL displaying different branching patterns in their dendrites.Ž. c Characteristic bipolar cell in the deep region of the EPL with morphological features of granule cells.Ž. d Internal tufted cell Ž solid arrow . in the internal region of the EPL, showing thick processes arising from opposite poles of the somata. Note the difference in size related to the horizontal cell beside itŽ.Ž. open arrow . e Horizontal bipolar cells in the inner region of the EPL with dendrites directed towards the MCL.Ž. f Fusiform cell in the MCL with a main dendritic trunk running parallel to the OB lamination.Ž. g Horizontal cells in the MCL showing slightly stained axonsŽ. arrows that can be followed towards superficial layers. J.G. BrinonÄÂ et al.rBrain Research 795() 1998 204±214 211

Fig. 4. NC-immunostained neurons in the deep layers of the OB. Scale bar: 50 mm.Ž. a,b Vertical cells in the MCL with long dendritic branches reaching the superficial EPLŽ. a or the deep GCL Ž. b . Ž. c Vertical cell in the IPL arborizing towards both the MCL and the GCL. Ž. d Weakly stained granule cells with stained apical dendritesŽ.Ž. arrows . e Deep short-axon cell with varicose dendrite in the GCL. Ž. f Bipolar short-axon cell in the WM. 212 J.G. BrinonÄ et al.rBrain Research 795() 1998 204±214 cell body and the proximal region of their primary varicose branching pattern of their dendrites, as previously de- dendrites were stainedŽ. Fig. 1v±w, Fig. 4e,f . scribedwx 27 . A few NC-positive tufted cells were also observed in the EPL. These neurons were identified as middle tufted cells in accordance with their location and 4. Discussion the presence of long secondary dendrites extending through the central region of the EPLwx 20 . The present study has clearly established the distribu- Tufted cells differ in several aspects, such as the pres- tion of NC-immunoreactive neurons in the rat main OB. ence and morphological features of secondary dendrites, The general distribution pattern of these elements is differ- the size and density of the intraglomerular tuft, the pres- ent from those reported for other calcium-binding proteins, ence, extent, and distribution of the axonal projection, the such as D-28k, , or wx degree of responsiveness to electrical stimulation of the 7,8,17,30 . Thus, the immunohistochemical olfactory nerve, and in their chemical contentwx 21 . Of detection of NC constitutes an excellent neuroanatomical particular interest in relation with our observations is the tool to characterize biochemical subpopulations within classification established according to their target regions. groups previously defined on the basis of morphological wx Internal tufted cells and most middle tufted cells have criteria by using Golgi impregnation 27,33 or immuno- projections that are largely extrinsic to the OBwx 35 , whereas cytochemistry for other calcium-binding proteins wx external tufted cells and the middle tufted cells located 1,4,7,8,17,30 . near the GL are local circuit neurons or they are involved in intra- or interbulbar connectionswx 21,35 . Although dou- 4.1. Methodological considerations ble-labeling studies with retrograde tracers would be nec- Our data confirm and extend previous observationswx 5 , essary to be conclusive, our results suggest that most, if in which only external tufted cells were reported as NC- not all, tufted cells projecting into high brain areas are NC immunopositive. Both studies have been carried out using immunonegative. the same type of animalsŽ. Wistar rats and the same Within the external tufted cell group, neurons located at primary antibodywx 24 . A possible explanation for the the GL±EPL boundary project to the contralateral OB, or different number and characteristics of the stained struc- connect medial and lateral regions of the same OB, whereas cells occurring in the superficial two-thirds of the GL tures could be a loss in the protein antigenicity during the wx chemical treatment of the tissue. In fact, Rogers and appear to be local circuit neurons 21,34,35 . These groups Resibois wx 32 reported a poor staining for calretinin in can be distinguished on the basis of morphological criteria: Helly-fixed paraffin-embedded , the same method neurons engaged in interbulbar and intrabulbar circuits wx have secondary dendriteswx 34 , whereas most local circuit used by Bastianelli et al. 5 . Therefore, the paraffin em- wx bedding procedure could reduce the sensitivity in the neurons do not 21 . Accordingly, most NC-stained tufted detection of NC in comparison with vibratome or cryostat cells are local circuit neurons. Therefore, NC is an excel- sections, as used in the present study. lent and reliable marker of external tufted cells categorized as local circuit neurons, and its immunocytochemical local- 4.2. Neurocalcin-positiÕe elements in the superficial layers ization provides a method for the identification and demar- cation of this group, and the study, through combined The NC-immunoreactive elements were found in all techniques, of its hodological and physiological peculiari- layers of the rat OB except in the ONL where all olfactory ties. fibers were negative as previously describedwx 5 . Although 4.3. Neurocalcin-positiÕe elements in the deep layers NC is expressed by numerous olfactory receptor cellswx 6 , their axons are NC-negative when they reach the OB. This Granule cells were classified into three morphological lack of NC-immunoreactivity in the distal olfactory axons groups on the basis of the location of their somata and could indicate a heterogeneous intracellular distribution of gemmuleswx 22 . This laminar organization offers a basis for this protein. differential synaptic interactions with deep vs. superficial The largest number of stained elements was observed in principal neuronswx 16,20,22,26 . The different granule cell the GL, where subpopulations of two different neuronal types are of functional significance, and each functional types displayed NC-immunoreactivity. They were clearly group appears to be determined by its position in the GCL distinguishable by their size and staining intensity. Cells wx38 . The existence of these distinct groups with particular displaying strong immunoreactivity were larger and showed location in the GCL is suggestive regarding the distribu- a smooth dendritic branch with a characteristic tuft inside tion pattern of those expressing NC. Since, according to the glomeruli, which are typical features of external tufted their location, most, if not all NC-immunopositive granule cellswx 27 . The lightly labeled neurons were identified as cells are able to interact with superficial tufted cells, the periglomerular cells and they were differentiated from the expression of this protein could identify neurons that are external tufted cells on the basis of their smaller cell-body interconnected as part of a specific circuit, as suggested for size, the presence of dendritic spines and the different other calcium-binding proteinswx 10 . In addition, the identi- J.G. BrinonÄ et al.rBrain Research 795() 1998 204±214 213 fication of this biochemically different subset of granule involved tend to preserve temporally distributed informa- cells could constitute a strategy to differentiate a physio- tion across neuronal populationswx 35 . logical subtype within the entire granule cell population. Considering the remarkable parallels between chemo- Concerning the second-order interneurons, the expres- and photoreceptionwx 3 , some points of analogy could be sion of NC is much more scarce. The NC-positive Van detected. The amacrine cells, like the granule cells, are Gehuchten cells were identified on the basis of their believed to produce prolonged GABA-mediated inhibition location, somal morphology and size and the characteristi- through signals which propagate mainly within local por- cally elaborated branching pattern of their varicose den- tions of their arborizationswx 11,12,38 . In these cells, NC, driteswx 33 . Although initially this neuronal type was only like other calcium-binding proteinswx 31 would be needed reported in the cat OBwx 37 , subsequent studies have to limit substantially the spread of calcium signal produced detected the presence of neuronal populations showing the by excitatory amino acidswx 23 . Interestingly, NC has been same morphological features in other mammalian species, detected in amacrine cellswx 24 , thus arguing for a possible such as hamster, hedgehog and ratwx 8,19,33 . relationship between these similar functional properties Most NC-positive secondary interneurons of the deep and the expression of NC. layers were identified as horizontal cells, whereas a few neurons showed the characteristics of vertical cells of Cajalwx 29,33 . A few deep short-axon cells also displayed Acknowledgements NC immunoreactivity. Nevertheless, it was not possible to classify them with total certainty within any particular The authors express their gratitude to Mr. G.H. Jenkins subtype of this group since their dendrites were poorly for revising the English version of the manuscript. This stained and only their proximal region was observable, work was supported by grants from the DGICyTŽ PB94- whereas this further classification is based upon the com- 1388. and the `Junta de Castilla y Leon'. plexity and distribution of their dendritic arborization and the density and distribution of dendritic spineswx 33 . References 4.4. Functional implications wx1 J.R. Alonso, R. Arevalo,ÂÄ A. Porteros, J.G. Brinon, J. Lara, J. Aijon, Our understanding of the possible functions of NC is Calbindin D-28k and NADPH-diaphorase activity are localized in different populations of periglomerular cells in the rat olfactory bulb, still limited; therefore, its physiological role in neurons J. Chem. Neuroanat. 6Ž. 1993 1±6. remains speculative. This calcium-binding protein has been wx2 C. Andressen, I. Blumcke,È M.R. Celio, Calcium-binding proteins: categorized as a calcium-sensor protein involved in modu- selective markers of nerve cells, Cell Tissue Res. 271Ž. 1993 latory and eventswx 15 . Since NC oc- 181±208. wx curred in the vast majority of olfactory receptors, a signifi- 3 R. Anholt, Molecular neurobiology of olfaction, Crit. Rev. Neuro- biol. 7Ž. 1993 1±22. cant function in olfactory signal transduction has been wx wx 4 K.G. Baimbridge, J.J. Miller, Immunohistochemical localization of hypothesized 14 . Additionally, sensory adaptation in ol- calcium-binding protein in the cerebellum, hippocampal formation factory receptors is thought to be mediated by a calcium- and olfactory bulb of the rat, Brain Res. 245Ž. 1982 223±229. binding protein similar to recoverinwx 18 . The similarities wx5 E. Bastianelli, K. Okazaki, H. Hidaka, R. Pochet, Neurocalcin between recoverin and NC, and its predominant expression immunoreactivity in rat olfactory bulb, Neurosci. Lett. 581Ž. 1993 585±588. by juxtaglomerular neurons, would argue for an involve- wx6 E. Bastianelli, A.S. Polans, H. Hidaka, R. Pochet, Differential ment of NC in calcium related events concerned with distribution of six calcium-binding proteins in the rat olfactory olfactory sensory adaptation. Nevertheless, the widespread epithelium during postnatal development and adulthood, J. Comp. distribution of NC also support their involvement in gen- Neurol. 354Ž. 1995 395±409. wx eral mechanisms rather than, or in addition to, such spe- 7 S. Biffo, L. Verdun di Cantogno, A. Fasolo, Double labeling with non-isotopic in situ hybridization and BrdU immunohistochemistry: cialized functions. calmodulinŽ. CaM mRNA expression in post-mitotic neurons of the Other mechanisms where NC would be involved may olfactory system, J. Histochem. Cytochem. 40Ž. 1992 535±540. be those proposed for other calcium-binding proteins that wx8 J.G. Brinon,Ä J.R. Alonso, R. Arevalo,  E. Garcõa-Ojeda,  J. Lara, J. include short-term calcium buffering, redistributing cal- Aijon, Calbindin D-28k-positive neurons in the rat olfactory bulb. cium within the neuron, and protecting the cell against the An immunohistochemical study, Cell Tissue Res. 269Ž. 1992 289± wx 297. damaging effects of excessive calcium influx 2 . Among wx9 M.R. Celio, Calcium-binding proteins in the brain, Arch. Ital. Anat. them, the prevailing expression of NC in superficial tufted Embriol. 94Ž. 1989 227±236. cells could be related with the control of endogenous wx10 M.R. Celio, Calbindin D-28k and parvalbumin in the rat nervous oscillations of cytosolic calcium originated as normal re- system, Neuroscience 35Ž. 1990 375±475. wx sponses to olfactory stimuli. This would contribute to 11 J. Dowling, The Retina: An Approachable Part of the Brain, Belk- r nap, Cambridge, 1987. shortening action potentials and or transmitter release, wx12 P. Ezeh, D. Wellis, J. Scott, Organization of inhibition in the rat thus improving the temporal precision of the stimuli trans- olfactory bulb external plexiform layer, J. Neurophysiol. 70Ž. 1993 mission. In fact, circuits where superficial tufted cells are 263±274. 214 J.G. BrinonÄ et al.rBrain Research 795() 1998 204±214 wx13 S.M. Hsu, L. Raine, H. Fanger, The use of avidin±biotin peroxidase wx25 A. Nakano, M. Terasawa, M. Watanabe, K. Okazaki, S. Inoue, M. complexŽ. ABC in immunoperoxidase techniques. A comparison Kato, Y. Nimura, N. Usuda, T. Morita, H. Hidaka, Distinct regional between ABC and unlabeled antibodyŽ. peroxidase procedures, J. localization of neurocalcin, a Ca2q-binding protein, in the bovine Histochem. Cytochem. 29Ž. 1981 577±590. adrenal gland, J. Endocrinol. 138Ž. 1993 283±290. wx14 S. Iino, S. Kobayashi, K. Okazaki, H. Hidaka, Neurocalcin-im- wx26 E. Orona, J.W. Scott, E.C. Rainer, Different granule cell populations munoreactive receptor cells in the rat olfactory epithelium and innervate superficial and deep regions of the external plexiform vomeronasal organ, Neurosci. Lett. 191Ž. 1995 91±94. layer in rat olfactory bulb, J. Comp. Neurol. 217Ž. 1983 227±237. wx15 M. Ikura, Calcium-binding and conformational response in EF-hand wx27 A.J. Pinching, T.P.S. Powell, The neuronal types of the glomerular proteins, Trends Biochem. Sci. 21Ž. 1996 14±17. layer of the olfactory bulb, J. Cell Sci. 9Ž. 1971 347±377. wx16 K. Kishi, K. Mori, H. Ojima, Distribution of local axon collaterals of wx28 A. Porteros, J.G. Brinon,Ä C. Crespo, K. Okazaki, H. Hidaka, J. mitral, displaced mitral, and tufted cells in the rabbit olfactory bulb, Aijon, J.R. Alonso, Neurocalcin immunoreactivity in the rat acces- J. Comp. Neurol. 225Ž. 1984 511±526. sory olfactory bulb, Brain Res. 729Ž. 1996 82±89. wx17 K. Kosaka, C.W. Heizmann, T. Kosaka, Calcium-binding protein wx29 S. RamonÂÁ y Cajal, Histologie du Systeme Nerveux de l'Homme et parvalbumin-immunoreactive neurons in the rat olfactory bulb: 1. des Vertebres, Maloine, Paris, 1911. Distribution and structural features in adult rat, Exp. Brain Res. 99 wx30 A. Resibois, J.H. Rogers, Calretinin in rat brain: an immunohisto- Ž.1994 191±204. chemical study, Neuroscience 46Ž. 1992 101±134. wx18 R.H. Kramer, S.A. Siegelbaum, Intracellular Ca2q regulates the wx31 J.H. Rogers, Two calcium-binding proteins mark many chick sen- sensitivity of cyclic nucleotide-gated channels in olfactory receptor sory neurons, Neuroscience 31Ž. 1989 697±709. neurons, Neuron 9Ž. 1992 897±906. wx32 J.H. Rogers, A. Resibois, Calretinin and calbindin D-28k in rat wx19 L. Lopez-Mascaraque, J.A. De Carlos, F. Valverde, Structure of the brain: patterns of partial co-localization, Neuroscience 51Ž. 1992 olfactory bulb of the hedgehog Ž.Erinaceus europaeus : a Golgi 843±865. study of the intrinsic organization of the superficial layers, J. Comp. wx33 S.P. Schneider, F. Macrides, Laminar distributions of interneurons in Neurol. 301Ž. 1990 243±261. the main olfactory bulb of the adult hamster, Brain Res. Bull. 3 wx20 F. Macrides, B.J. Davis, The olfactory bulb, in: P.C. EmsonŽ. Ed. , Ž.1978 73±82. Chemical Neuroanatomy, Raven Press, New York, 1983, pp. 391± wx34 T.A. Schoenfeld, F. Macrides, Topographic organization of connec- 426. tions between the main olfactory bulb and pars externa of the wx21 F. Macrides, T.A. Schoenfeld, J.E. Marchand, A.N. Clancy, Evi- anterior olfactory nucleus in the hamster, J. Comp. Neurol. 227 dence for morphologically, neurochemically and functionally hetero- Ž.1984 121±135. geneous classes of mitral and tufted cells in the olfactory bulb, wx35 T.A. Schoenfeld, J.E. Marchand, F. Macrides, Topographic organi- Chem. Senses 10Ž. 1985 175±202. zation of tufted cell axonal projections in the hamster main olfactory wx22 K. Mori, K. Kishi, H. Ojima, Distribution of dendrites of mitral, bulb: an intrabulbar associational system, J. Comp. Neurol. 235 displaced mitral, tufted, and granule cells in the rabbit olfactory Ž.1985 503±518. bulb, J. Comp. Neurol. 219Ž. 1983 339±355. wx36 M. Terasawa, A. Nakano, R. Kobayashi, H. Hidaka, Neurocalcin: a wx23 C. Murphy, Olfactory psychophysics, in: T. Finger, W. SilverŽ. Eds. , novel calcium-binding protein from bovine brain, J. Biol. Chem. 267 Neurobiology of Taste and Smell, Wiley, New York, 1987, pp. Ž.1992 19596±19599. 251±273. wx37 A. Van Gehuchten, I. Martin, Le bulbe olfactif chez quelques wx24 A. Nakano, M. Terasawa, M. Watanabe, N. Usuda, T. Morita, H. mammiferes, La Cellule 7Ž. 1891 205±237. Hidaka, Neurocalcin, a novel calcium-binding protein with three wx38 D.P. Wellis, J.W. Scott, Intracellular responses of identified rat EF-hand domains, expressed in retinal amacrine cells and ganglion olfactory bulb interneurons to electrical and odor stimulation, J. cells, Biochem. Biophys. Res. Comm. 186Ž. 1992 1207±1211. Neurophysiol. 64Ž. 1990 932±947.