SEARCH

SEARCH BY CITATION

Keywords:

  • vomeronasal system;
  • accessory olfactory bulb;
  • growth;
  • anatomy;
  • mouse

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The vomeronasal system is currently a topical issue since the dual functional specificity, vomeronasal system-pheromones, has recently been questioned. Irrespective of the tools used to put such specificity in doubt, the diversity of the anatomy of the system itself in the animal kingdom is probably of more importance than has previously been considered. It has to be pointed out that a true vomeronasal system is integrated by the vomeronasal organ, the accessory olfactory bulb, and the so-called vomeronasal amygdala. Therefore, it seems reasonable to establish the corresponding differences between a well-developed vomeronasal system and other areas of the nasal cavity in which putative olfactory receptors, perhaps present in other kinds of mammals, may be able to detect pheromones and to process them. In consequence, a solid pattern for one such system in one particular species needs to be chosen. Here we report on an analysis of the general morphological characteristics of the accessory olfactory bulb in mice, a species commonly used in the study of the vomeronasal system, during growth and in adults. Our results indicate that the critical period for the formation of this structure comprises the stages between the first and the fifth day after birth, when the stratification of the bulb, the peculiarities of each type of cell, and the final building of glomeruli are completed. In addition, our data suggest that the conventional plexiform layers of the main olfactory bulb are not present in the accessory bulb. Anat Rec Part A, 288A:1009–1025, 2006. © 2006 Wiley-Liss, Inc.

Odors are detected in the majority of mammals by means of two systems, the main olfactory system (MOS) and the vomeronasal system (VNS) (Halpern,1987; Brennan,2001; Halpern and Martínez-Marcos,2003), which, despite some differences, basically share a common pattern of organization (Mori,1987). Each consists of a sensory epithelium, two olfactory bulbs, and different telencephalic structures. This similarity results in the olfactory information being processed in a like manner in both systems, although there are also some specific morphological differences, and at molecular level different genes are involved in the transduction mechanism (Buck and Axel,1991; Dulac and Axel,1995; Mombaerts,1999).

The main olfactory (MOB) and the accessory olfactory bulbs (AOB), which are the first relay stations in each system, play a key role in the physiological mechanism of olfaction. Within the bulbs, glomeruli organize the corresponding input and output information and are considered the true functional unit of the bulbs (Graziadei and Monti-Graziadei,1986; Valverde et al.,1992; Hildebrand and Shepherd,1997; Bozza et al.,2002).

The MOB is a structure that has been very well studied at different morphological and physiological levels (Shepherd and Greer,1998; López Mascaraque and de Castro,2002), but this is not quite the case for the AOB. Albeit during the last 2 decades, progress in the knowledge of the VNS in general and of the AOB in particular has been considerable (Hayashi et al.,1993; Taniguchi and Kaba,2001; Araneda and Firestein,2006), there are still some specific aspects of the AOB that require further exploration. Among the issues in need of precision is the exact morphology of the AOB, especially when it is obtained from mice. No authors with the exception of Hinds (1968a,1968b) have dealt with the AOB of the mouse as a whole, although some researchers have studied partial aspects of its morphology in this species (Wilson and Raisman,1981; von Campenhausen et al.,1997; del Punta et al.,2002).

Other experimental animals, such as rats, guinea pigs, and rabbits, have in fact been used to define the general characteristics of the AOB (Crosby and Humphrey,1939; Allison,1953; Mori,1987; Takami and Graziadei,1991). The whole AOB has also been studied during development in the opossum (Shapiro et al., 1995), a marsupial mammal in which an extensive reliance on chemosensory cues is presumed. However, when choosing a morphological pattern for the AOB, the selected species could be the mouse because its genome has been completed (http://www.ensembl.org), and the possibility of working with transgenic animals represents a wonderful challenge.

In an attempt to fill, at least in part, the possible gaps in our knowledge of the AOB in mice, the aim of the present work is to focus attention on the study of the general morphological characteristics of AOB as a whole. The transition from the AOB during growth to the adult AOB has been considered of special interest, as it provides a detailed description of the more relevant events that take place during the critical period that covers the last gestational days and the first days after birth. Once the characteristic layers of the AOB have appeared and strengthened, few changes, if any, have been observed in relation to the AOB of the adult.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Animals

Female BALB/c mice were bred, housed singly, and mated in the Animal House of the Veterinary Faculty (University of Santiago de Compostela; registry no. 15003AE). Embryonic age was calculated from the time at which the vaginal plug was evident (embryonic day 0). On embryonic days 15–19, pregnant females were placed under deep anesthesia and their fetuses removed. Newborn mice were also examined on postnatal days 0–12, as were a number of adult males and females. In what follows, “En” and “Pn” respectively denote embryonic and postnatal ages of n days. For each developmental age, at least three specimens were examined.

All experiments were performed in accordance with the regulations and laws of the European Union (86/609/EEC) and Spain (RD 223/1998) for the care and handling of animals in research. All efforts were made to minimize animal suffering and limit the number of animals used.

Processing of Olfactory Bulbs

Paraffin.

The fetuses were fixed directly over 24 hr in 4% paraformaldehyde or Bouin's fluid, while newborn, juvenile, and adult mice were washed and perfused with the same fixatives before their heads were removed and the whole brain dissected out and immersed in the fixative. After microdissection (when necessary), the tissue of interest was embedded in paraffin wax. In the case of smaller specimens aged E15–E19, the half rostral part of the brain was cut into transverse or sagittal sections 10 μm thick.

Cryosectioning.

After 4% paraformaldehyde perfusion, brains were postfixed for a further 2 hr and then cryoprotected by immersion overnight in 30% sucrose in O.1 M phosphate buffer saline (PBS) at 4°C. A freezing microtome was used to cut 40 μm sagittal sections through the rostral half of the brain. The sections were collected in PB 0.1 M, pH 7.2, and stored at 4°C until their free-floating processing.

Lectin Histochemistry Protocol

Two lectins, Ulex europaeus agglutinin I (UEA-I) and Lycopersicum esculentum agglutinin (LEA), were obtained as biotin conjugates from Sigma (St. Louis, MO) and detected using a Vectastain avidin-biotin peroxidase complex (ABC) biotin detection kit (Vector, Burlingame, CA). L-fucose and N-acetylglucosamine are the sugar specificity to UEA-I and LEA, respectively. Both lectins are considered very good markers for the study of the AOB in mice; UEA-I labels exclusively the vomeronasal nerve layer and glomerular layer of the AOB, while LEA labels the same layers in both bulbs, but with the special feature that in the AOB only the anterior part is labeled (Salazar et al.,2001; Salazar and Sánchez Quinteiro,2003).

Paraffin sections were dewaxed and transferred to phosphate buffer (PB) and then incubated for 30 min at room temperature with 2% bovine serum albumin in 0.1 M Tris buffer (pH 7.2); incubated for 24 hr at 4°C with lectin at various dilutions in 0.1 M Tris buffer containing 2% bovine serum albumin; washed for 2 × 10 min in PB; incubated for 90 min at room temperature with Vectastain ABC reagent (1:250 in PB); washed for 10 min in PB; and washed for 5 min with Tris-HCl (pH 7.6).

Visualization of peroxidase activity by incubation in a solution containing 0.05% 3,3-diaminobenzidine (DAB) and 0.003% H2O2 in 0.2 M Tris-HCl buffer (pH 7.6) was monitored under a microscope; the reaction was quenched with 0.2 M Tris-HCl buffer. All lectinless control sections were unlabeled and no specific binding was observed in controls with presaturated lectins.

Immunohistochemistry Protocol

The following specific markers were used: olfactory marker protein (OMP; generously provided by Dr. F. Margolis, University of Maryland), growth-associated protein 43 (GAP-43; Sigma), microtubule-associated proteins 2 (MAP-2; Sigma), and glial fibrillary acidic protein (GFAP; Dako; Table 1).

Table 1. Specific information related to the antibodies used
AntibodiesSourceCatalogue numberType*/Mode**CloneBinding/References
  • *

    antibody

  • **

    species in which the secondary antibody is raised. VNra, Vomeronasal receptor axons; Npk, Neuroperikarya.

Anti-OMPDr. F.MargolisPoliclonal raised in goat VNra/Margolis (1972)
Anti-GAP-43SIGMAG9264Monoclonal7B10VNra/Skene (1989)
Anti-MAP-2SIGMAM4403MonoclonalHM-2Dendrites npk/Matus (1988)
Anti-GFAPDAKOZ0334Policlonal raised in rabbit Astrocytes/Bignami et al (1972)
Biotinylated Anti-Mouse IgG (H+L)VECTORBA-2000Raised in horse  
Biotinylated Anti-Rabbit IgG (H+L)VECTORBA-1000Raised in goat  

Dilutions and Incubations

The final working dilutions used in the paraffin-embedded sections were OMP, 1:500; GAP-43, 1:8,000; MAP-2, 1:300; GFAP, 1:250. Free-floating sections were incubated at higher dilutions: OMP, 1:10,000; GAP-43, 1:10,000; MAP-2, 1:500; GFAP, 1:1,000.

The paraffin sections were transferred after dewaxing to PB. Then sections were incubated for 30 min at room temperature with 5% horse normal serum containing 2% bovine serum albumin in PB. Cryostat sections were incubated for 30 min at room temperature with 0.2% Triton X-100 containing 2% bovine serum albumin in PB. From that point, the immunohistochemistry was similar for slide-mounted and free-floating sections.

In both cases, sections were sequentially incubated in primary antibody for 24 hr at 4°C, biotinylated secondary antibody for 1 hr, ABC (Vectastain reagent) for 2 hr, and visualized either with DAB or Vector VIP (SK-4600; Burlingame, CA) following standard procedures for the visualization of horseradish peroxidase complex (HRP).

For double immnunostaining, a sequential twice-repeated enzyme-labeled method was employed. Between both immunolabelings, the sections were subjected to treatment with 0.01 M glycine solution (pH 2.2) for 5 min. In selecting the most suitable dye for the HRP-H2O2 in order to visualize the immunoreaction, products DAB, Vector SG (SK-4700) and VIP were combined. No nonspecific staining was observed in control sections incubated with either normal horse serum or without primary antibodies.

In most of the cases, sections embedded in paraffin were alternatively stained by Nissl and labeled by lectins and other specific markers, and also some sections were lightly counterstained with cresyl violet.

Image Acquisition and Processing

Digital microscopy images were captured by using a Karl Zeiss Axiocam MRc5 digital camera. All images were processed using Adobe Photoshop 6.0 (Adobe Systems, San Jose, CA). When necessary, images were adjusted for contrast and brightness to equilibrate light levels, cropped, resized, and rotated for purpose of presentation.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

In order to facilitate the following description of the stratified structure of the AOB, reference will be made to the conventional arrangement of strata as seen in the MOB, despite the variations between both bulbs. The classical nomenclature usually employed for the MOB is nerve layer (NL), which is substituted here by vomeronasal nerve layer (VNNL), glomerular layer (GlL), external plexiform layer (EPL), mitral/tufted cell layer (M/TcL), internal plexiform layer (IPL), and granular layer (GrL).

Development and Growth

Fetus, newborn and juvenile: Nissl sections.

Given the objectives of this work, we may begin our comments at stage E15. Figures 1–3 show a sequence of events that characterizes the progressive organization of the AOB. With regard to the topography, from the beginning the accessory bulb occupies the same position as in adults, while its size and form show only slight differences. The size is proportionally bigger in young samples than in older (proportionally to the size of the MOB), the external contour changes from convexity to concavity, and the internal appearance is determined by the number and distribution of its cells. Measurements and quantification have been studied separately (data not shown).

thumbnail image

Figure 1. Nissl-stained sagittal sections of the rostral telencephalon of fetuses (left anterior, right posterior-orientated; AH) give precise information about the characteristic form and topography of the AOB (frames; left column). Higher-power photomicrographs (right column) of previous sections showing the beginnings of stratification in the developing AOB. Future M/T cells (asterisks) lead the process of layer formation, and close to them are the two intermediates zones, IsZ and IdZ (curve arrows, in F and H). I and J: General view of the comparative layers of the AOB (I) and the MOB (J) at birth. Scale bars = 100 μm (B and D); 250 μm (all others).

Download figure to PowerPoint

thumbnail image

Figure 2. Nissl-stained sagittal sections of the AOB (left anterior, right posterior-orientated) from birth to postnatal day 5. AC: The thickness of intermediate zones is indicated by broken lines. D: Limits of IsZ are established by basal periglomerular cells (arrowheads) and M/T cells (asterisk). E and F: The vomeronasal nerve and granular layers are easily identified; the topography of the basal periglomerular cells are indicated by double arrows; the IdZ evolves into the LOT. G and H: Higher magnifications of E and F allow the verification of the size and typical morphology of the periglomerular cells (arrowheads) and M/T cells (asterisks). E and G are frozen sections; the rest are paraffin sections. Scale bars = 100 μm (G and H); 250 μm (all others).

Download figure to PowerPoint

thumbnail image

Figure 3. Different images of Nissl-stained sagittal sections of the AOB (left anterior, right posterior-orientated), once the main changes have occurred. A and B: Basal periglomerular cells (double arrows) are considered an interesting reference to establish limits. C: Same image as B, in which the layers of the AOB are indicated. D: Higher-power photomicrograph showing details of different cells present in the AOB, except those of the GrL. E: At 12 postnatal day, the structure of the AOB is close to the adult bulb. A is frozen section; the rest are paraffin sections. Scale bars = 100 μm (D); 250 μm (all others).

Download figure to PowerPoint

E15–P0 (Fig. 1).

The most relevant morphological detail in these stages is the presence of a prominent mass of cells, the future mitral/tufted (M/T) cells, bounded by two zones, which we will call the intermediate superficial zone (IsZ) and the intermediate deep zone (IdZ), where very few cells are present. Externally, the developing VNNL is beginning to appear, while internally a densely stained group of proliferating cells, smaller than those of the main mass, occupies an extensive zone adjacent to the ventricular cavity.

Progressively, the VNNL shows greater definition, and granular cells replace the group of cells situated in the deep area of the bulb. Meanwhile, the distribution of the mass of cells previously mentioned begins to show the appearance of the M/T cells. Nevertheless, few changes take place in the two intermediate zones in this phase of development.

P1–P5 (Fig. 2).

These stages are considered the transitional period in the morphological organization of the laminar architecture of the AOB, as they establish the transition from immature to mature AOB. The vomeronasal nerve and granular layers are differentiated. The VNNL mainly consists of small, elongated, and randomly oriented cells showing considerable thickness, especially in the middle of the bulb. The cells of the GrL are small, round, and densely stained and packed and are distributed in rows. The M/T cells, whose perikarya are quite evident, are lightly stained and still numerous, albeit scattered and arranged so as to form a slightly concave curve.

Moreover, intermediate zones are also affected by the general changes to the bulb. Both zones are especially evident in overstained cryostat sections because of the different degree of stain of the whole bulb. In the IsZ, the presence of a group of small and intensely stained cells can be observed, the periglomerular basal cells. Superficially to them is the future GlL, characterized by the presence of a reduced number of small and medium cells that never outline glomeruli. Deeply to the periglomerular basal cells would be the EPL, although it is very difficult, if not impossible, to establish clear limits between the periglomerular basal cells and M/T cells, because in this narrow and irregular space there are usually sparse and scattered, small, medium, and large cells. The M/TcL is better defined now than in previous stages.

Similarly, the lateral olfactory tract (LOT) fully occupies the IdZ, where there are only a few cells of small and medium size, with the GrL in the following deep stratum. No other space between the M/T cells and GrL is observed.

P6–P12 (Fig. 3).

Changes previously described are corroborated or confirmed at P6 and later stages, and the general organization of the AOB is close to that of the adult (Fig. 8). Frozen sections of the whole AOB (Fig. 3A) clearly show the IsZ, which will be occupied by most of the glomeruli; other glomeruli are intermingled in the VNNL due to the glomeruli not being organized into a defined row. It is also evident that periglomerular cells are sparse and do not outline glomeruli, albeit a considerable increase in the number of periglomerular basal cells establish a sharp limit between the GlL and the rest of the bulb (Fig. 3A, B, and E). In the meanwhile, progressive reduction of the M/TcL is evident, with its cells sparse and not forming true rows. The LOT remains invariable. The existence of plexiform layers is quite doubtful, although the same narrow, irregular, and heterogeneous space as before is also possible to identify in some sections at these stages.

Lectins and inmmunohistochemistry sections.

With regard to the transitional period considered previously, lectins, OMP, and GAP-43 showed a similar pattern of reactivity in the VNNL and GlL. Then sections with these markers were selected indistinctly. Dendrites were identified using MAP-2, and the population of astrocytes was observed by GFAP.

Vomeronasal axons
PO (Fig. 4A).

At birth, GAP-43 immunoreactivity clearly establishes the separation between the future VNNL and GlL and the rest of the AOB.

thumbnail image

Figure 4. VNNL and GlL stained for different markers in sagittal sections of the AOB (left anterior, right posterior-orientated) from birth to postnatal day 5. A: Lower-power photomicrograph of the AOB stained with GAP-43; other structures adjacent to the AOB are also immunoreactive. Note the glomeruli of the MOB, open arrow. B: The anterior part of the AOB is the only tissue of this bulb stained by LEA lectin. C: Higher magnification of B, in which some vomeronasal axons end in the glomeruli (arrowheads). D: Double-labeled UEA-I and MAP-2 of the AOB, with indication of the M/T cell layer. E: Extension and arrangement of the VNNL and GlL stained by OMP, with the shade of some glomeruli more intensely labeled (arrows). F: Glomeruli (arrows) are filled with dense axons in the MOB (same preparation as E). GI: GAP-43 immnunoreactivity showing the general distribution of the vomeronasal axons ending in the anterior (H) and posterior (I) parts of the AOB. JL: GAP-43 immunoreactivity in sagittal sections of the AOB (J and K) and MOB (L); data useful to compare the different images of glomeruli in both bulbs (arrows). Scale bars = 50 μm (C, F, and L); 100 μm (D, E, H, I, and K); 250 μm (A, B, G, and J).

Download figure to PowerPoint

P1 (Fig. 4B and C).

LEA, which is characterized by labeling the VNNL and GlL only in the rostral part of the AOB as well as the same layers of the MOB, labeled some glomeruli-like structures, not very well isolated. The VNNL cannot easily be separated from the GlL.

P2 (Fig. 4D).

Both the separation of the VNNL and GlL from the rest and a discrete population of glomeruli are intensively labeled by UEA-I, an excellent marker specifically for the study of the AOB.

P3 (Fig. 4E and F).

Although OMP labels the VNNL and GlL of the AOB clearly, it is interesting to point out the different immnunoreactivity between the corresponding areas stained in the AOB and MOB. This difference is observed, on the one hand, because of the extension of the vomeronasal receptor neurons axons (AOB; Fig. 4E) and the olfactory receptor neurons axons (MOB; Fig. 4F), and on the other hand, the different appearance of both glomeruli (Fig. 4E and F).

P4 (Fig. 4G–I)/P5 (Fig. 4J–L).

At these two stages of the transitional period, GAP-43 stains not only the VNNL and GlL but also the corresponding axons enclosed in the LOT (Fig. 4G and J); the wide unstained space is the location for the M/TcL and theoretically for the plexiform layers. Albeit some glomeruli are distinguishable in the deep glomerular zone (Fig. 4H, I, and K), there is not a sharp limit between the VNNL and the GlL, as is quite evident in the MOB (Fig. 4L). These immunohistochemistry data corroborate that the glomeruli are neither organized into a defined row nor perfectly isolated from the VNNL, as Nissl material showed.

Dendrites
P0/P4 (Fig. 5).

The two predominant expressions of MAP-2, neuron perykarya and dendrites (Matus,1988), are stained in the AOB at these stages of development. Just below the VNNL, that is to say in the IsZ, dendritic processes form a plexus that grows gradually according to age (Fig. 5A, C, and E). The growth of this plexus means a bigger extension over the VNNL and simultaneously the organization of individual units, with the typical club-shaped appearance of glomeruli (Fig. 5B, D, F).

thumbnail image

Figure 5. Lower (left column) and higher (right column) magnifications of sagittal sections of the AOB (left anterior, right posterior-orientated), from birth to postnatal day 4, showing MAP-2 immunoreactivity. A, C, and E: Note the dendritic process plexus in the IsZ (between double arrows). B and D: Dendritic processes do not form glomerular structures (arrows) at first stages. F: At P4, dendritic glomeruli (arrows) are more evident. Soma of M/T cells are also shown (arrowheads; B, D, and F). Scale bars = 50 μm (B and D); 100 μm (F); 250 μm (A, C, and E).

Download figure to PowerPoint

P5 (Fig. 6).

Neuron perikarya, dendrites, and their processes are clearly immunoreactive for MAP-2 (Fig. 6A and B). At this age, the contribution of the dendritic processes to the formation of glomerular structures is quite evident (Fig. 6B). Because LEA lectin labels only the anterior part of the VNNL and GlL in the AOB, double-stained LEA/MAP-2 gives precise information about the projection of the vomeronasal receptor neuron axons (anterior half) and the extension of the dendritic processes (posterior part; Fig. 6C). Double-immunostained MAP-2/GAP-43 shows the whole AOB stained and allows the conjunction of axons and dendritic processes to be seen (Fig. 6D). Higher magnification of the two previous figures are also useful to delimitate the corresponding layers of this bulb (Fig. 6E and F).

thumbnail image

Figure 6. Sagittal sections of the AOB (left anterior, right posterior-orientated) at the critical stage of P5 (AD). A and B: MAP-2 immunoreactivity showing soma of the M/T cells (arrowheads), their dendrites (open arrows), and dendritic glomeruli (arrows). C: Double-labeled LEA/MAP-2 demonstrated the areas affected by the vomeronasal axons and the dendritic projection. D: Double immnunostaining GAP-43/MAP-2 photomicrograph showing the definitive participation of the vomeronasal axons and dendrites in the formation of the glomeruli. E and F: Higher magnifications of A and B. Scale bars = 50 μm (B); 150 μm (E and F); 250 μm (A, C, and D).

Download figure to PowerPoint

Astrocytes
P0 (Fig. 7A and B).

Sections of the AOB from newborn mice express GFAP in a few astrocytes located in the granular layer, some of them with short processes, in the deep part of the IsZ, which are occasionally intermingled in the mass of cells that characterizes the typical image of the AOB at this stage.

thumbnail image

Figure 7. GFAP immunoreactivity at lower (left column) and higher (right column) magnifications of sagittal sections of the AOB (left anterior, right posterior-orientated) from birth to postnatal day 5. A and B: Immature astrocytes (arrows) mainly located in the thickness of the IsZ. CF: Topography of astrocytes as before and processes of some beginning to form glomeruli (arrowheads). G and H: Astrocytes are densely stained and show the typical appearance of glomeruli in the AOB (arrowheads). Scale bars = 100 μm (B, D, F, and H); 250 μm (A, C, E, and G).

Download figure to PowerPoint

P1/P2 (Fig. 7C and D).

Not only does the number of astrocytes increase but their size and processes are more and more evident, while their distribution remains the same as before; the most superficial branches of astrocytes organize small and globular structures.

P3/P4 (Fig. 7E and F).

Progressively, GFAP-positive astrocytes show neat cell bodies and branches, whose highest immunoreactivity corresponds to the base of the GlL, where the populations of periglomerular cells are highest.

P5 (Fig. 7G and H).

Densely labeled astrocytes and their processes are identified in the GlL and also in other layers of the accessory bulb; projections of these astrocytes close to the glomeruli are typical at these stages. No significant changes are seen in older samples, apart from a logical increase in the number of these cells.

Adults.

In adults, the AOB is a structure that is semilentil in shape, located in the superior, central, and most posterior part of the MOB, and consequently covered by the frontal pole of the brain. It measures about 650 × 700 × 1,100 microns in height, width, and length, respectively, values that have been calculated in postfixed material. Transverse and sagittal sections clearly show its topography and other general anatomical details (Fig. 8A and B). Vomeronasal axons run between the two main bulbs, reach the AOB laterally, in the central part of it, and then branch out (Fig. 8C).

thumbnail image

Figure 8. Transverse (A) and sagittal (B) sections of the most rostral pole of the telencephalon in adult mice (the AOB is framed), with the corresponding magnifications of the AOB (C and D). Nissl-stained. Schematic representation of the arrival of the vomeronasal nerves at the AOB (E, left side) and of the distribution of vomeronasal axons within it, shown in a partial sagittal section of the AOB, limited by periglomerular basal cells (arrows; E, right side). FC, frontal cortex; VNNs, vomeronasal nerves. Scale bars = 250 μm (C and D); 1000 μm (A and B).

Download figure to PowerPoint

Although the base of the GlL is well established by the presence of a considerable number of periglomerular cells, and consequently the separation between this layer and the next is generally quite evident, the glomeruli of the AOB do not follow the typical pattern displayed in the MOB due to the scarcity and the disorganized distribution of the periglomerular cells and even, in some cases, their absence around the glomeruli (Fig. 9). As a result of the heterogeneity of the periglomerular cells, the boundary of the glomeruli is neither clearly nor uniformly defined in this bulb. Nevertheless, when using different markers, it is possible to identify the contribution of the terminals of the vomeronasal nerves, of the dendrites of the M/T cells, and of periglomerular cells that participate in the organization of true glomeruli (Fig. 10).

thumbnail image

Figure 9. Nissl-stained sagittal sections of the AOB (A) and the MOB (B) showing the different number and distribution of periglomerular cells around glomeruli (arrows) in adult mice. Scale bars = 250 μm.

Download figure to PowerPoint

thumbnail image

Figure 10. Sagittal sections of the AOB in adult mice. A: Anterior part of the AOB labeled with LEA showing glomeruli (arrows) more densely stained than the adjacent tissue and their disordered distribution. B: MAP-2-labeled dendrites; their superficial process form glomeruli (arrows). C: GFAP-labeled astrocytes somata (arrows) and processes, more densely stained in the proximal (solid arrowheads) than in the distal (open arrowheads) parts. Scale bars = 100 μm (C); 250 μm (A and B).

Download figure to PowerPoint

Mitral/tufted cells, which in all cases show a large soma, are sparse and dispersed, and consequently M/T cells are not organized into rows (Fig. 11A and B); for the same reason, it is also complicated to define a limit between the M/TcL and the EPL/IPL; in fact, it seems that some M/T cells invade these strata. It would probably be better to refer to undefined spaces or bands rather than true plexiform layers. Similarly to the P6–P12 stages, the presumptive EPL is present with the same characteristics as before, and in thin paraffin sections this layer is easier to identify in some preparations (Fig. 8). Dendrites from M/T cells that cross the EPL, and which are typical in the MOB, are only occasionally found in the AOB, in sections stained by the Nissl method (Fig. 11C and D). Finally, the LOT, as a relatively wide band of white matter, goes through the AOB and is located between the M/TcL and GrL.

thumbnail image

Figure 11. Sagittal sections of the AOB (left column) and MOB (right column) in adult mice. A and B: Comparison between form, size, and distribution of the M/T cells of the AOB (brackets) and the mitral cells (Mcs) of the MOB (brackets). C: Dendrites from the M/T cells in the AOB (arrows) are sparse. D: Dendrites of the mitral cells are numerous and uniformly orientated (arrowheads). Scale bar = 250 μm.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

Techniques and Methodology

It is noteworthy that the expression of different markers, lectins and antibodies, shows variations depending on the processing of the olfactory bulb, that is to say, paraffin or frozen sections, as well as in relation to the fixatives employed, in our case paraformaldehyde and Bouin. Moreover, small variations introduced in the protocols employed could also modify the results obtained.

The differences between paraffin and cryostat sections are very well known in morphology (Baker,1966), and these differences are also applicable to the case of the AOB of mice. However, because of the reduced size of the structure in this animal, it is not possible to obtain enough sections to be stained alternatively with different techniques. In this situation, we have opted to use only frozen material for GFAP, frozen brains and material embedded in paraffin for Nissl, and samples embedded in paraffin for the rest. We have also verified that using Bouin as fixative to lectins, GAP-43, and OMP gives better results, while paraformaldehyde was selected for Nissl and MAP-2.

With regard to the use of lectins, it is interesting to remember that in some cases lectins with the same putative sugar affinity can result in different labeling patterns (Shapiro et al., 1995). In the same study, the authors mentioned that “care must be taken in the interpretation of results from different lectins, different species, and different systems,” a statement with which we agree. Specifically, in the study of the AOB in rodents, numerous technical details can modify the degree of labeling, as was largely discussed previously (Salazar and Sánchez Quinteiro,1998).

In relation to the antibodies employed in the present work, we have tested three different chromogens (DAB, VIP, and SG) and no significant differences were detected in the corresponding sections. Moreover, we have verified that the typical expression for GFAP (Bignami et al.,1972) and MAP-2 (Matus,1988) was regularly shown in the AOB at all ages studied.

Nevertheless, according to the protocol that we followed, our results show that the immunoreactivity for the other two antibodies differs a little from what was expected. GAP-43 stains systematically the same tissue in all postnatal samples, an interesting issue because this antibody is especially recommended for the identification of growing axons (Skene,1989), albeit literature informs that other brain tissue is also reactive in particular circumstances (Verhaagen et al.,1989; Mosevitsky,2005).

Albeit OMP was first considered a “gold standard marker” exclusively for mature olfactory receptor neurons (Bailey et al.,1999), it has also been applied to vomeronasal receptor neurons in the opossum (Shnayder et al.,1993) and in adult mice and rats (Jia and Halpern,1996). Our results corroborate that OMP labels the two most superficial layers of both bulbs after birth, but with considerable difference in the degree of intensity between AOB and MOB, even at high concentrations (Fig. 4E and F). As similar differences in the immunoreactivity were observed in adults (results not shown) and neurogenesis of AOB precedes that of the MOB (Hinds,1968a; Brunjes and Frazier,1986), it suggests that the low density in the VNNL and GlL of the AOB is not related to age but to the peculiarities in the organization of the mentioned layers.

Form and Topography of AOB

During growth, the form of the AOB changes from an external convexity to a concavity and this variation forces cells to be closer, and consequently it is quite difficult for cells to align themselves in rows. This is particularly evident in the M/T cells but also in other populations of the AOB, which organize strata or layers, as in the case of GlL (see transversal sections of the bulb).

Another interesting issue concerns the topography of the AOB and the characteristic arrival of the vomeronasal axons to it, which could be related to the morphological and functional division of this bulb. Using different methodologies and strategies, it has been demonstrated that there is a functional/topographical relationship between the sensory epithelium of the VNO and the AOB (Imamura et al.,1985; Jia and Halpern,1996; Sugai et al.,2000; Dulac and Torello,2003). The vomeronasal receptors situated in the apical epithelium of the VNO send information to the anterior AOB, while those receptors located in the basal part send their axons to the posterior AOB, that is to say, between both structures, VNO/AOB, there is a zone-to-zone projection. Recently, an elegant experiment has been published about the same issue (Knöll et al.,2003). This topographical or zonal organization was previously established between the sensory epithelium of the main olfactory system and the MOB, although in this case four main zones were defined: I, II, III, and IV (Vassar et al.,1993; Ressler et al.,1994; Yoshihara et al.,1997; Mori et al.,2000).

As a consequence of this kind of arrangement, it was thought that there could be a corresponding expression for the olfactory information during development. Some authors demonstrated that the formation of the olfactory bulb (olfactory glomeruli) occurred following a distinct rostrocaudal gradient in the MOB (Bailey et al.,1999). Others authors suggest that the same rostrocaudal gradient is applicable to the AOB in the rat (Schwarting et al.,1992) and even in the mouse (von Campenhausen et al.,1997).

Our present and previous (Salazar et al.,2001; Salazar and Sánchez Quinteiro,2003) studies have confirmed the division of the AOB into rostral and caudal parts, but this fact does not necessarily imply its formation (glomeruli) following the same gradient during development. It is true that LEA labels only the anterior AOB but other lectins, the BSI-B4 for example, stain exclusively the posterior division at the same stages. Moreover, dendritic processes revealed by MAP-2 extend superficially both in the rostral and caudal part, while the information obtained by Nissl material is uniform and does not reveal anything conclusive.

Microscopy and Stratification of AOB

Hinds's publications (1968a,1968b) were fundamental to a better understanding of some aspects of the development of the MOB and AOB in the mouse. Moreover, his publications benefited from the consideration of the bulbs in this species as a whole. In our case, we agree with Hinds's results with regard to the formation of the AOB from a mass of cells, which appear early in development and later evolve into M/T cells and also the stratified organization of its final structure. However, there are discrepancies between Hinds's work and our study, mainly due to the difference in our objectives, and the use of different methodologies and strategies.

By means of autoradiography, at the time an innovative technique, Hinds (1968a) focused his study on the definition of the time of origin of neuroblasts and neurons, and glioblasts and neuroglia in both bulbs. Hinds did not pay special attention to either the morphological characteristics of the cells or their distribution into organized layers.

As regards the presence or not of mitral and/or tufted cells in the AOB of mice, there is clear evidence of a mass of cells in the middle of the AOB, although it is not easy to fit these cells into one group or the other. Cajal (1902) thought that their morphology suggested that these cells were typical tufted cells, while Hinds (1968a) was of the opinion that the same cells, which had a time of origin more similar to the mitral cells of the MOB than to the tufted cells, were in fact mitral cells. By means of a simple comparison between this group of cells in the AOB and the mitral cells of the MOB in the same samples, we consider this group of cells to be closer to being tufted (size, arrangement, and distribution) than mitral (topography) cells. In dealing with this controversy, most authors have opted for the term “mitral/tufted cells.” Further morphological information will address the true nature of these cells.

In any case, it may be stated that there is general agreement that the so-called M/C cells send multiple dendrites to different glomeruli (Cajal,1902; Allison,1953; Takami and Graziadei,1991). We have corroborated this in mice (results not shown).

Likewise, Cajal (1902) considered the glomeruli of the AOB to be characterized by “the lack of definition, indistinctness, and small size of the glomerular territories, which, rather than independent plexus, form a continuous plexiform band, outlined in festoons or unevenness on the deep side.” He added that this organization of the glomeruli was quite similar to the same structures in some reptiles and amphibians.

It is true that the most relevant detail of the glomeruli morphology in the AOB of the mouse is its lack of definition, although not exclusively due to their size, but to the paucity of the periglomerular cells, most of which are located on the basal part of the GlL. Our data suggest that some glomeruli of the AOB have the same or similar size as those of the MOB (data not shown).

Finally, the last comment of the present discussion is devoted to the stratification of the AOB, a controversial issue mainly when attempts are made to generalize for all mammals. The diversity in the mammalian AOB is so huge (Meisami and Bhatnagar,1998) that it is inadequate to extrapolate data from one species to another. For example, concerning the plexiform layers, unpublished results of Meisami and Bhatnagar (1998) indicate that in mammals with large and well-developed AOBs, both plexiform layers do exist. On the contrary, Cajal (1902) concluded that the mentioned layers were absent. Classical comparative studies of the olfactory system (Crosby and Humphrey,1939; Allison,1953) included the EPL in its description of the AOB but not the IPL. Another interesting example, which affects the stratification of the AOB, is the topography of the LOT, situated through or under the accessory bulb.

It is noteworthy that, the obvious morphological diversity of the AOB among mammals, corroborated histochemically and immnunohistochemically (Halpern et al.,1998), is unresolved and could have functional implications.

In mice, a meticulous observation of the paraffin and cryostat series of the AOB reveals sections in which there is a narrow irregular space between the periglomerular basal cells and the M/T cells, the EPL, and other sections in which it is completely absent. Due to the variability of this space, the existence of the EPL is, in our opinion, a subjective issue. In this sense, to enclose both presumptive plexiform layers in the M/TcL would be an alternative to be considered.

The most evident limits in the general stratification of the AOB in mice are represented by the important population of periglomerular basal cells and by the presence of the LOT.

Conclusions

In spite of its size, the only hindrance to the accurate study of the AOB in mice, we are convinced that it is possible, and necessary, to build a consistent morphological model of the AOB in this species.

Furthermore, given the general characteristics previously considered, we are led to the conclusion that the AOB in mice is an undefined structure in terms of cell characteristics and distribution, more markedly so if it is compared with the MOB. It is noteworthy, however, that the general stratification of the AOB in mice may be more clearly understood following the steps in its transition from nonnatum to juvenile and adult mice (Fig. 12). Likewise, at stage P5, the definitive building of glomeruli is concluded.

thumbnail image

Figure 12. Schematic representation of the development of layers formation in the AOB in mice. A: Before birth, E16. B: After birth, P3. C: Adult. fVNNL, future vomeronasal layer; fGrL, future granular layer. Scale bars = 250 μm.

Download figure to PowerPoint

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED

The authors thank M. Otero, L. Bellón, and J. Castiñeira for their excellent technical assistance, and J. Curtin for the revision of the English text. Dr. N. Vandenberghe provided helpful comments on this manuscript and contributed to valuable discussions. The comments of two anonymous reviewers greatly improved the clarity of the manuscript.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED