Morphological and immunohistochemical study of the rabbit vomeronasal organ

Abstract

The characterization of the rabbit mammary pheromone, which is sensed by the main olfactory system, has made this species a unique model for the study of pheromonal communication in mammals. This discovery has brought attention to the global understanding of chemosensory communication in this species. Chemocommunication is mediated by two distinct organs located in the nasal cavity, the main olfactory epithelium and the vomeronasal organ (VNO). However, there is a lack of knowledge about the vomeronasal system in rabbits. To understand the role of this system, an exhaustive anatomical and histological study of the rabbit VNO was performed. The rabbit VNO was studied macroscopically by light microscopy, and by histochemical and immunohistochemical techniques. We employed specific histological staining techniques (periodic acid‐Schiff, Alcian blue, Gallego's trichrome), confocal autofluorescence, histochemical labelling with the lectin Ulex europaeus agglutinin (UEA‐I), and immunohistochemical studies of the expression of the Gαi2 and Gαo proteins and olfactory marker protein. The opening of the vomeronasal duct into the nasal cavity and its indirect communication with the oral cavity through a functional nasopalatine duct was demonstrated by classical dissection and microdissection. In a series of transverse histological sections, special attention was paid to the general distribution of the various soft‐tissue components of this organ (duct, glands, connective tissue, blood vessels and nerves) and to the nature of the capsule of the organ. Among the main morphological features that distinguish the rabbit VNO, the presence of a double envelope, which is bony externally and cartilaginous internally, and highly developed venous sinuses stand out. This observation indicates the crucial role played in this species by the pumping mechanism that introduces chemical signals into the vomeronasal duct. The functional properties of the organ are also confirmed by the presence of a well‐developed neuroepithelium and profuse glandular tissue that is positive for neutral mucopolysaccharides. The role of glycoconjugates was assessed by the identification of the α1‐2 fucose glycan system in the neuroepithelium of the VNO employing UEA‐I lectin. The pattern of labelling, which was concentrated around the commissures of the sensory epithelium and more diffuse in the central segments, is different from that found in most mammals studied. According to the expression of G‐proteins, two pathways have been described in the VNOs of mammals: neuroreceptor cells expressing the Gαi2 protein (associated with vomeronasal receptor type 1); and cells expressing Gαo (associated with vomeronasal receptor type 2). The latter pathway is absent in most mammals studied. The expression of both G‐protein families in the rabbit VNO places Lagomorpha together with rodents and insectivores in a small group of mammals belonging to the two‐path model. These findings support the notion that the rabbit possesses a highly developed VNO, with many specific morphological features, which highlights the significance of chemocommunication in this species.

Introduction

Olfaction is the primary sensory modality for most mammals, and allows them to interact adequately with each other and their environment (Müller‐Schwarze, 2006). This sense is mediated by two distinct organs located in the nasal cavity, the main olfactory epithelium (MOE) and the vomeronasal organ (VNO). The first organ is closely related to the limbic system, and is therefore intimately associated with memories, emotions and behaviour (Reep et al. 2007). The second organ is more linked to reproduction behaviour, such as sexual attraction and maternal aggression (Martín‐Sánchez et al. 2015). The anatomy and the central neuroanatomical projections into the brain are well distinguished in both systems (Wysocki, 1979; Halpern & Martinez‐Marcos, 2003). However, these two systems are not thoroughly independent with the presence of vomeronasal receptors expressed in the olfactory mucosa and vice versa. In addition, the two systems have overlapping projections into the same amygdala subregion (Baxi et al. 2006; Pro‐Sistiaga et al. 2007).

The vomeronasal system is formed by the VNO, which projects into the accessory olfactory bulb through the vomeronasal nerves (McCotter, 1912). The vomeronasal amygdala and the nerves and tracts by which these structures are connected are also included in the system. The VNO consists of three components: a duct, whose lumen is covered laterally and medially by a non‐sensory and sensory epithelium, respectively; a parenchyma made up of veins, arteries, nerves, glands and connective tissue; and a capsule formed by cartilage or bone, which encloses all the structures mentioned above (Halpern, 1987). Furthermore, the vomeronasal system comprises two gene families, VR1 and VR2, which are coupled with the Gαi2 and Gαo proteins, respectively (Dulac & Axel, 1995; Herrada & Dulac, 1997; Matsunami & Buck, 1997; Ryba & Tirindelli, 1997). Additionally, formyl peptide receptors are a novel family of vomeronasal receptors whose function is associated with the identification of pathogens or pathogenic states (Riviere et al. 2009). Beyond this general description, each species has its own unique characteristics, not only anatomic but also related to G‐protein expression and vomeronasal receptor families. Thus, each species should be studied independently, so as to avoid extrapolating information from the published data (Salazar et al. 2016).

The VNO of mammals plays an important role in the perception of chemical stimuli, such as pheromones. These molecules are secreted by one individual and perceived by another individual of the same species, which then undergoes a specific physiological or behavioural reaction (Karlson & Lüscher, 1959), allowing individuals to communicate and interact among themselves (Wyatt, 2014). In mammals, chemical communication has been studied extensively in the order Lagomorpha (consisting of two families: Leporidae and Ochotonidae). Thus far, rabbits are the only mammalian species in which a pheromone that is emitted by lactating females and affects neonatal behaviour has been described. The mammary pheromone (MP; 2‐methylbut‐2‐enal) has been fully characterized (Schaal et al. 2003). Carried in the female's milk, MP awakens rabbit neonates and causes the nipple‐sucking reflex (Coureaud et al. 2010). The cellular activation in response to the MP pheromone has been studied in rabbits using cFos detection on the day of birth, revealing that MP induces increased cFos expression in the posterior piriform cortex (Schneider et al. 2016). This pheromone has also been studied on postnatal day 4, when it induces an increase in neuronal activity in olfactory areas that include the main olfactory bulb (Charra et al. 2013). These studies suggest that MP detection is mediated by the main olfactory system. Nevertheless, greater knowledge of the vomeronasal system may prove important for understanding the role of chemosensory communication in rabbits and for understanding how both systems, which are seemingly distinct but are much more integrated than previously thought according to studies in mice (Kelliher, 2007), could work in communicating between individuals of the same species.

As a prey species with specific behavioural patterns – cues for submission and dominance – rabbits require large areas to escape and avoid aggressive interactions. To date, the most common techniques for improving welfare are based on environmental enrichment, not only on farms but also for research or pet animals (Poggiagliolmia et al. 2011). However, rabbit social houses do not appear to be a necessity, and sometimes result in injury and deleterious effects on wellbeing (Szendro et al. 2013). Research into pheromones could help improve rabbit welfare either on farms or among pets. Previous studies in rats have demonstrated that pheromones play a pivotal role in the stress response (Masini et al. 2010). However, little work has been conducted to study how these molecules could be used with this aim in rabbits, and these studies have been restricted to farming, without a significant scientific impact (Bouvier & Jacquinet, 2008). In fact, as one of the best models for studying chemocommunication in mammals (Schneider et al. 2018), rabbit physiological and behavioural information has not been matched by an exhaustive anatomical and histological study of the vomeronasal system. There are few studies on the rabbit VNO, and those studies have focused mostly on the neuroepithelial ultrastructure (Luckhaus, 1969; Taniguchi & Mochizuki, 1983; Elgayar et al. 2014). There is also a lack of data regarding the expression in the rabbit vomeronasal system of G‐proteins and other markers that are differentially expressed in the olfactory and vomeronasal systems, such as the lectin Ulex europaeus agglutinin (UEA‐I) and the olfactory marker protein (OMP).

To fill the main gaps in our knowledge of the rabbit VNO, the aim of the present work is to focus attention on the general morphological characteristics of the rabbit VNO. Various tissue dissection and microdissection techniques were used, as well as general and specific histological staining, histochemical labelling with the lectin UEA‐I, and immunohistochemical studies of the expression of the OMP, Gαi2 and Gαo proteins.

Materials and methods

Five healthy rabbits of both sexes that were between 67 and 70 days old were provided by an abattoir. The rabbits were humanely killed under current legislation [Council Regulation (EC) 1099/2009], and the heads were separated from the carcasses in the slaughtering line. In the same way, 15 male or female healthy rabbits, 60–70 days old, were euthanised in the Department of Pathology of the Faculty of Veterinary Sciences of Lugo for use in pathological research. The intact heads were donated to the authors. In addition, two adult singly‐housed BALB/c mice bred in the Department of Pharmacy of the same faculty were processed. All procedures followed the guidelines for housing and handling provided by the Bioethical Committee of the University of Santiago de Compostela, and conformed to European legislation (EU directive 2010/63/EU) and Spanish legislation (RD 53/2013).

Two of the rabbit heads were dissected fresh for macroscopic studies of the anatomy of the vomeronasal system. The remaining heads were dissected only superficially and, in some cases, the lateral walls of the nasal cavity were opened. Then, the specimens were either fixed by immersion in 10% buffered formalin or kept in Bouin's fixative for 24 h and then transferred to 70% alcohol.

After fixation, most of the VNOs were dissected out to be processed without decalcification. However, to study the topographical anatomy of the VNO, six nasal cavities were processed entirely from the incisive papilla to the caudal termination of the vomeronasal cartilage. These samples were decalcified by immersion in ‘Shandon TBD‐1 Decalcifier’ (Thermo, Pittsburgh, PA, USA) for 24 h with agitation.

Most of the VNOs and the decalcified nasal cavities were embedded in paraffin wax and cut into serial transverse sections 3–6 μm thick for examination of the VNO and nasal cavity topography (Fig. S1), but one sample was cut in the sagittal plane (Fig. S2). The sections were stained with haematoxylin‐eosin, Gallego's trichrome (Ortiz‐Hidalgo, 2011), periodic acid‐Schiff (PAS) and Alcian blue.

For the autofluorescence imaging study of the VNO, two unembedded samples were cryoprotected by immersion overnight in 30% sucrose in 0.1 m phosphate‐buffered saline (PBS) at 4 °C and cut into 40‐μm transverse sections using a freezing microtome.

Lectin histochemistry protocol

The lectin UEA‐I was obtained as a biotin conjugate from Sigma (St Louis, MO, USA) and detected as follows. (i) Endogenous peroxidase activity was blocked to avoid interference with the final development step. To this end, the slides were incubated in 3% hydrogen peroxide in distilled water. (ii) The samples were incubated for 30 min at room temperature with 2% bovine serum albumin (BSA) in 0.1 m Tris buffer (pH 7.2). (iii) The samples were incubated for 1 h at room temperature with the lectin UEA (Vector Laboratories, Burlingame, CA, USA) diluted 1 : 50 with 0.5% BSA. (iv) The samples were washed for 3 × 5 min in PBS (0.1 m, pH 7.2). (v) For 12 h, the samples were incubated with an immunoglobulin (Ig) against UEA conjugated to peroxidase (Dako, Denmark). (vi) The samples were washed in PBS. (vii) Peroxidase activity was visualized 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). DAB was the chromogen and developed into a brown precipitate.

Controls were run without lectin and with preabsorption of lectin by an excess amount of the corresponding sugar (Fig. S3). In addition, in the absence of a positive control from rabbits, we replicated the whole histochemical procedure with mouse tissues known to express the proteins of interest (Fig. S4).

Immunohistochemical protocol

The immunohistochemistry studies required the blocking of both endogenous peroxidase activity and non‐specific binding with hydrogen peroxide and BSA, respectively. Then, the primary antibody was added (Table 1) and incubated overnight. The next day, the cells were incubated for 1.5 h in the corresponding biotinylated secondary antibody and for 1.5 h at room temperature in Vectastain ABC reagent (ABC, Vector Laboratories, Burlingame, CA, USA). The detection of peroxidase activity was performed in the same manner described for labelling with lectins, using DAB as the chromogen and thus obtaining the same brownish coloration. Finally, the slides were dehydrated through alcohols, cleared in xylene and coverslipped. Samples for which the primary antibody was omitted were used as negative controls (Fig. S5).

Table 1.

Antibodies and lectins used

Lectin*/Antibody	1st Ab Species/Dilution	1st Ab Catalogue number	2nd Ab Species/Dilution	2nd Ab Catalogue number
UEA*	1 : 50	Vector L‐1060	Rabbit 1 : 50	DAKO P289
Anti‐Go	Rabbit 1 : 50–1 : 100	MBL 551	Goat 1 : 250	Vector BA‐2000
Anti‐Gi2α	Rabbit 1 : 50–1 : 100	Santa Cruz Biotechnology sc‐7276	Goat 1 : 250	Vector BA‐1000
Anti‐OMP	Goat 1 : 400	Wako 544‐10001	Horse 1 : 250	Vector BA‐9500

Species of elaboration, dilutions, manufacturer, catalogue number.OMP, olfactory marker protein.

Confocal laser‐scanning microscopy

Autofluorescence imaging of elastic fibres of the VNO was performed by confocal laser‐scanning microscopy. Transverse sections with a 40 μm thickness were collected in 0.1 m PBS, pH 7.2 and stored at 4 °C until free‐floating processing. For the staining of nuclear DNA, the sections were incubated for 15 min in the nucleic acid‐specific dye TO‐PRO‐3 iodide (1 μm; Molecular Probes). After three washes in PBS, the sections were mounted in a drop of the anti‐bleaching mounting medium Slow‐Fade solution (Molecular Probes Cat. # S7461). A Bio‐Rad MRC 1024 ES setup for dual‐channel fluorescence using green autofluorescence and TO‐PRO‐3 iodide (far‐red fluorescence) filter settings (100 mW Ar laser, wavelength 488 nm, and He‐Ne laser, wavelength 633 nm, respectively) were used. Images were taken using a 20 × lens (NA 0.45) on a Nikon TE 2000 inverted microscope, with separate emission filtering for each type of fluorescence (RIAIDT, University of Santiago de Compostela, Spain).

Acquisition of images and digital processing

The digital images were captured with a Karl Zeiss Axiocam MRc5 digital camera coupled to a Zeiss Axiophot microscope. If needed, Adobe Photoshop CS4 (Adobe Systems, San Jose, CA, USA) was used to adjust brightness and contrast, balance light levels, and crop or resize images for presentation. Photomicrographs of the entire nasal cavity were taken as a mosaic of several photographs merged with image‐stitching software (PTGui Pro).

Results

The rabbit VNO is a bilateral formation located in the ventral part of the nasal cavity. The VNO exhibits little prominence under the respiratory mucosa, which covers its lateral surface. This characteristic, together with its moderate size and hidden orientation in the ventral part of the nasal septum, makes it macroscopically invisible without careful dissection (Fig. 1A–C). However, in some cases, the profuse vasculature of the organ allows for easy recognition of its projection area under the mucosa (Fig. 1D). The innervation of the organ comprises two main components: the sensorial unmyelinated vomeronasal nerve that leaves the organ caudodorsally; and the myelinated branches of the nasal caudal nerve that reach the organ caudoventrally (Fig. 1E).

Figure 1.

Superficial dissection of the nasal septum showing the vomeronasal organ (VNO; fresh specimens). (A) Lateral view of the intact septum. There is a lack of recognizable landmarks of the organ. (B) VNO dissected out after removing the respiratory mucosa. (C) Macroscopic appearance of the VNO after transverse sectioning. The tubular shape and parenchymatous tissue are recognizable (arrows). (D) Specimen with the VNO vessels filled with blood. The projection area of the VNO is much clearer (arrows). (E) Main bundle of the vomeronasal nerve detached from the submucosa (black arrow) and vomeronasal branch of the nasal caudal nerve (white arrow). Scale bars: 2 cm (A, B, D); 1 cm (C, E).

Transverse sections of the nasal cavity allow us to more clearly appreciate the topographic relationships of the VNO (Fig. 2A–D). The ridges formed by the alveolar part of the maxillary bones define two independent chambers, one dorsal for the conchae nasalis and the other ventral for the ventral part of the septum. The VNOs are located in the latter chamber, surrounded by two deep furrows. In this way, the VNOs remain quite independent of the turbinate airflow. As the two organs lie against each other side by side, both give shape to the ventral part of the nasal septum (Fig. 2B–D).

Figure 2.

(A–D) Transverse sections of the nasal cavity at four different levels, ordered from rostral to caudal. The vomeronasal organs (VNOs; black arrowheads) give shape to the ventral part of the septum and are located dorsally to the palatine process of the incisive bone (Pp). (E) VNO dissected out, showing its main three components: capsule (black arrow), duct (arrowhead) and parenchyma (asterisk) containing wide vascular sinuses (white arrow). Tb, turbinates; Pp, palatine process of the incisive bone; Ms, maxillary sinuses; white arrowhead, nasolacrimal duct. Scale bars: 1 cm (A–D); 2 mm (E).

The peculiar organization of the rostral bones of the cranium in rabbits, especially the presence of a large caudal palatine process of the incisive bone, means that in this species, the vomer bone is displaced caudally; therefore, the VNO rests over the palatine process (Fig. 2B–D). After dissecting out the organ and sectioning it transversally, its main three components are recognized macroscopically: the capsule, the vomeronasal duct and the parenchyma between them (Fig. 2E).

The vomeronasal ducts open straight into the nasal cavity by means of a very tiny hole located in the rostral part of the mucosa (Fig. 3B,C). Additionally, the organ establishes indirect communication with the oral cavity due to the presence of a functional nasopalatine duct, which provides a connection between the oral and nasal cavities (Fig. 3A,B). The nasopalatine ducts start in the incisive papilla as two barely visible slits situated posterior to the first maxillary incisors (Fig. 4B,C). Enclosed by their own cartilage, the ducts pass through the wide palatine fissures (Fig. 4A) to finally reach the inner region of the nasal cavity (Fig. 4D).

Figure 3.

(A) Schematic drawing of the rabbit cranium showing the projection area of the vomeronasal organ (VNO) and the location of both transverse (upper left, shown in more detail in Fig. 6A) and sagittal (centre left) histological sections of the whole nasal cavity. (B) Drawing of the nasal cavity showing a schematic of the VNO, its opening into the nasal cavity (arrowhead), and the nasopalatine duct (ND) passing dorsocaudally to open into the ventral furrow of the nasal cavity. (C) Dissection of the rostral part of the nasal cavity showing the VNO and the opening of the vomeronasal duct (arrowhead). I2, 2nd incisor; NS, nasal septum; NDC, nasopalatine duct cartilage; NDO, nasopalatine duct opening; PM, palatal mucosa; RM, respiratory mucosa; VNC, vomeronasal cartilage; Asterisk, vomeronasal duct. Scale bar: 500 μm (A and C).

Figure 4.

(A) Ventral view of the rostral part of the skull showing the palatine process of the incisive bone (Pp) and the palatine fissures (Pf). (B) Ventral view of the palatine region. (C) Inset from (B): incisive papilla (Pi), located just between the incisors (Pi). (D) Cannulation of the nasopalatine duct showing the functional communication between the oral and nasal cavities (arrowhead). I1, 1st incisor; I2, 2nd incisor. Scale bars: 1 cm (A,B); 2 cm (D).

The VNO cartilage is formed by a unique piece of cartilage, but it varies in shape according to the level considered. For the sake of simplicity, we selected five representative levels from the whole histological series (Figs 5 and S1). In the rostral‐most part of the VNO, the cartilage is thick and laterally broken to allow the opening of the duct (Fig. 5I). Moving backwards, the whole parenchyma is enclosed by the cartilage (Fig. 5II). In the central segment of the organ, the cartilage is U‐shaped, with the lateral side incomplete (Fig. 5III). This dorsal opening allows the entrance and exit of nerve fascicles. Caudally, the size and thickness of the cartilage and the extent of the parenchyma progressively decrease (Fig. 5IV,V).

Figure 5.

Transverse sections of the vomeronasal organ (VNO) at five different levels numbered from rostral (I) to caudal (V), which show the changes in the conformation of the vomeronasal cartilage. Scale bar: 100 μm.

The microscopic study of decalcified nasal cavities (Fig. 6A) confirmed all the topographical relationships described above and showed how in the caudal third of the organ, the vomeronasal cartilage is enveloped by a thin capsule of bone that grows from the palatine process of the incisive bone (Fig. 6B).

Figure 6.

Gallego's trichrome staining in decalcified transverse sections of the nasal cavity. (A) Section at the anterior part of the organ (level II) showing the opening of the nasopalatine ducts (ND) in the ventral furrows of the nasal cavity, and the relationships of the organ with the ventral concha (Tb), the nasal septum (NS) and the palatine process of the incisive bone (Pp). (B) Section from the caudal part of the organ (level IV) showing both the cartilaginous (white arrows) and bony (black arrows) envelopes of the organ, the blind end of the vomeronasal duct (asterisk), the basal membrane of the epithelium (arrowheads) and the venous sinuses (Vv). Scale bar: 500 μm.

Specific histological staining of the dissected organs allows a more detailed study of their components (Fig. 7). The vomeronasal duct is moved slightly to the medial part of the organ. Over its entire length, the duct is covered in its lateral part by the typical respiratory epithelium (ciliated and pseudostratified; Fig. 8B), and in its medial part and its dorsal and ventral commissures, the duct is covered by a neuroreceptor epithelium (Fig. 7A,B). This sensory epithelium presents the following features: (i) scarce basal cells; (ii) a wide stratum of large bipolar cells; and (iii) a covering layer of sustentacular cells; which provide support to (iv) the superficial layer formed by the apical processes of the receptor cells, which in turn form the luminal microvilli of the epithelium (Fig. 8A).

Figure 7.

(A) The parenchyma of the organ at low magnification shows its main components: the central duct (VND), the lateral vascular compartment (Vv) and the dorsomedial vomeronasal glands (VNG). (B) At higher magnification, the vomeronasal nerves (VNN) may be seen below the basal membrane (arrowheads) of the neuroepithelium. The main two strata of the epithelium are illustrated: the sustentacular cell (SuC) and neuroreceptor cell (NrC) layers. (C) The walls of the venous sinuses are reinforced by thick bundles of smooth muscle (asterisk). Stain: Gallego's trichrome. Scale bars: 250 μm (A); 100 μm (B,C).

Figure 8.

(A) Sagittal section of the sensory epithelium showing its main microscopic features: basal cells (arrowheads), neuroreceptor (NrC) and supporting (SuC) cells, and the outer layer comprising the microvillar processes of both cell types (MvL). (B) Transverse section of the respiratory epithelium. Stain: haematoxylin‐eosin. Scale bar: 50 μm (A,B).

The vascular component of the organ completely fills the lateral part of the parenchyma. This component mainly comprises large venous sinuses, which are enclosed by bundles of smooth muscle (Fig. 7C). Sagittal sections show the large size of these vascular compartments (Fig. 9C). Conversely, the arteries are very small and scant. They are more easily recognized in the confocal images, which reveal the elastic fibres of the tunica intima (Fig. 9A,B). In the transverse series, branches of the vomeronasal nerves are located dorsal, medial and ventral to the vomeronasal duct (Figs 7B and 9A). The rami of these nerves contain unmyelinated fibres. Connective tissue is most patent close to the vomeronasal cartilage and around the vomeronasal duct. This tissue is reinforced in the basal membrane of the epithelium (Figs 6B and 7B). The vomeronasal glands are located dorsomedially and are not very conspicuous. Morphologically, these glands are of the compound acinar type and are exclusively serous (Fig. 10A). The secretion from these glands is PAS‐positive and Alcian blue‐negative (Fig. 10B,C).

Figure 9.

Vasculature of the rabbit vomeronasal organ (VNO). (A) Autofluorescence transverse section of the VNO in which all the components of the organ are identified: vomeronasal cartilage (VNC), vomeronasal nerves (VNN), vomeronasal glands (VNG), venous sinuses (Vv) and neuroreceptor cells (Nrc). The most prominent artery is identified by an arrowhead. (B) Higher magnification image of the artery indicated in (A). Note the tunica intima (arrow). (C) The wide VNO venous sinuses shown in a sagittal section stained by haematoxylin‐eosin. Inset: sagittal section of the rabbit head showing the location of the venous sinuses. Scale bars: 250 μm (A); 25 μm (B); 100 μm (C).

Figure 10.

Rabbit vomeronasal glands. (A) Sagittal section of the parenchyma of the organ stained with haematoxylin‐eosin that shows the conspicuous compound acinar vomeronasal glands. (B) Periodic acid‐Schiff (PAS)‐Alcian blue staining in a transverse section of the dorsolateral part of the vomeronasal organ (VNO). The acid mucopolysaccharides are restricted to the cartilage matrix. Additionally, Alcian blue reveals the presence of collagen fibres in the parenchyma. (C) PAS staining reveals strong reactivity in the acinar glands of the parenchyma. VNN, vomeronasal nerves. Scale bar: 100 μm.

In sections histochemically stained with lectin from U. europaeus, the labelling is heavily concentrated in the commissures of the neuroepithelium, but the lectin also stains in a variable way the central segment of the neuroepithelium. In both cases, the receptor cells and their processes are evident (Fig. 11). The antibody against Gαi2 labels a subpopulation of neuroreceptor cells whose somas are located in the apical zone of the epithelium, immediately beneath the supporting cell layer (Fig. 12A). In addition to cell bodies, the antibody against Gαi2 labels the microvillar surface, dendrites and axons of these receptor neurons, as well as the nerve bundles in the lamina propria. Very few Gαi2‐positive cell bodies can be found in the deep layer or near the basal lamina.

Figure 11.

UEA lectin immunohistochemical labelling of the vomeronasal organ (VNO). (A) UEA labels the epithelia of the organ, the vomeronasal nerves (arrowhead) and the respiratory mucosa of the nasal cavity. In the neuroreceptor epithelium, the labelling is concentrated and stronger in its commissures, whereas in the rest of the epithelium there are patches of unlabelled cells (arrows). (B) Higher magnification of the dorsal commissure area (asterisk in A) showing UEA patches in the neuroepithelium. Scale bars: 1 mm (A); 200 μm (B).

Figure 12.

Immunohistochemical labelling of the vomeronasal sensory epithelium. (A) The anti‐Gαi2 antibody labels a subpopulation of neuroreceptor cells in the apical zone of the neuroepithelium (black arrowheads), whereas wide patches of the basal area remain unlabelled (asterisks). (B) The Gαo antibody gives a broader pattern, although it does not label all the neurons in the neuroepithelium (white arrowheads). (C) The olfactory marker protein (OMP) antibody immunoreacts at variable intensity with the whole epithelium and the vomeronasal nerves (arrows). Scale bar: 100 μm (A–C).

The antibody against Gαo (Fig. 12B) stains a population of neurons whose cell bodies are located in both the basal and apical parts of the epithelium. The microvillar surface, dendrites and axons of these neurons are also labelled. Most of the neuroreceptor cell bodies that are unlabelled by the anti‐Gαo antibody are located in the central zone of the epithelium (white arrowheads in Fig. 12B). Immunohistochemical labelling of OMP marks the whole epithelium, including the vomeronasal nerve bundles (Fig. 12C).

Discussion

Recent years have seen considerable advancement in our knowledge of the function of the vomeronasal system (Buck, 2000; Dulac, 2000; Pantages & Dulac, 2000; Zufall & Munger, 2001; Chamero et al. 2007; He et al. 2008; Leinders‐Zufall et al. 2014). Unfortunately, these advances have not always been matched by parallel morphological information. The domestic rabbit, Oryctolagus cuniculus, is a case in point. Although this species has become a unique model for the study of chemocommunication in mammals, there has been very little research concerning the anatomy of the rabbit VNO.

Since the classical macroscopic description by Klein (1881), there have been very few related studies, and those studies did not consider all the structural features of the organ. In contrast, they focused almost exclusively on the ultrastructure of the epithelium (Luckhaus, 1969; Taniguchi & Mochizuki, 1983; Elgayar et al. 2014). The pioneering contributions of Mori must be added based on the use of monoclonal antibodies against lactoseries carbohydrates in the epithelium of the rabbit VNO (Mori, 1987; Mori et al. 1987).

In view of this paucity, and given the phylogenetic proximity between Lagomorpha and Rodentia, it has been tempting to extrapolate based on information from widely studied mice and rats. This approach ignores the fact that even in the macrosmatic group of mammals, morphological diversity is an important characteristic of the VNO (Salazar & Sanchez‐Quinteiro, 2009). Such variability is in fact consistent with the involvement of the VNS in pheromonal communication because pheromones are, by definition, species‐specific, and different species have different behavioural and reproductive strategies (Keverne, 2005; Brennan & Zufall, 2006).

Our observations confirm that the rabbit possesses a highly developed VNO with remarkable morphological differences from those of rodents and other macrosmatic mammals, which will be the object of this discussion.

The first point concerns the covering of the whole VNO, a lamina made either of bone in rodents or cartilage in most mammals, such as ungulates and carnivores. However, there has been some controversy regarding the composition of the capsule in rabbits. Whereas Taniguchi & Mochizuki (1983) and Wöhrmann‐Repenning (1981) only describe the occurrence of a cartilaginous capsule, Luckhaus (1969) and Elgayar et al. (2014) note the presence of a double envelope, bony externally and cartilaginous internally. Unlike these authors, we have made whole‐organ transverse histological serial sections. This approach may resolve these discrepancies because we noticed in our series that the lateral bony envelope starts to grow from the palatine process of the incisive bone in only the central segment of the organ (Level 3). On the anterior levels, where the vomeronasal cartilage encloses the vomeronasal parenchyma (Levels 1 and 2), this bony capsule is lacking. Along the caudal levels (4 and 5), the bony envelope remains, covering and reinforcing the cartilaginous capsule.

Both cartilage and bone play important roles in the pumping mechanism, which allows fluids carrying stimuli to be sucked into the lumen of the organ (Meredith, 1994; Døving & Trotier, 1998). The vomeronasal pump is the physiological mechanism by which stimuli molecules are transported to vomeronasal receptor neurons. Fluids are moved into and out of the vomeronasal duct by cyclical vasomotor movements within the vomeronasal capsule. Cartilage and bone provide the resistance needed during the venous expansion of the parenchyma that compresses the epithelium of the duct to empty the lumen. On the other hand, to suck fluids from outside the duct, vasoconstriction of the veins in the organ creates a negative pressure inside the duct. Therefore, a strong envelope will protect the parenchyma against collapse (Salazar et al. 2008).

In addition, the morphology of the capsule has been considered an invaluable feature for the purpose of phylogenetic classification (Wöhrmann‐Repenning, 1984). In fact, the shape of the vomeronasal cartilage is specific for major taxa. This observation is true for the Monotremata, Marsupialia, Edentata, Insectivora, Carnivora, Chiroptera, Rodentia and Primates. Our findings in rabbits support this contention for Lagomorpha. This group constitutes a singular example because of the double envelope of bone plus cartilage. In this respect, there is only a similarity with Rodentia, which, at least in very young individuals, show this double capsule. However, after the age of a few postnatal weeks, the cartilaginous capsule disappears completely, and only the bony envelope remains (Salazar & Sanchez‐Quinteiro, 1998).

Regarding the opening of the vomeronasal duct, there are two models in mammals. One is the opening of the duct at the base of the nasal septum, independent of the nasopalatine duct. This formation is observed for Rodents and Marsupialia (Vaccarezza et al. 1981; Schneider et al. 2008). As our histological series has shown, Lagomorpha belongs to this group. The second model comprises those species in which the vomeronasal duct merges into the nasopalatine ducts inside the palate, as occurs in Artiodactyla and Carnivora (Salazar et al. 2000, 2008; Tomiyasu et al. 2017). In both cases, the presence of a functional nasopalatine duct allows the access of pheromones from the mouth into the nose. In fact, rabbits produce rhythmic movements of the upper lips, analogous to the flehmen behaviour described in other mammals. The inspired air stream generated by these movements traverses the nasal cavities, aspirating particles from the mouth through the nasopalatine ducts, which are reinforced by a large semicircular cartilage.

The extensive venous vasculature is another distinctive feature of the rabbit VNO. In the present study, histological sections of the organs demonstrated the magnitude of the venous sinuses, which recall the network of large irregular sinuses typical of the erectile tissues. Taking advantage of the autofluorescence of the tunica intima, we have observed the arterial vessels. In comparison to the venous pathway, the arteries are scarce, small in diameter and have thin walls. This paucity of the arterial vessels has also been described in the opossum (Poran, 1998) and cats (Salazar et al. 1996). However, neither of those species has a venous network comparable to that found in rabbits. In fact, the vascular pattern of the rabbit VNO is quite different from that of any other group of mammals, including the closely related rodents, whose venous return system relies on a large central vein (Vaccarezza et al. 1981; Barrios et al. 2014). Together, the presence of two capsules around the rabbit vomeronasal parenchyma and the prominent development of the venous sinuses suggest a very active role of the pumping mechanism for normal stimulation of the VNO in rabbits, as Meredith et al. (1980) found in hamsters.

Much attention has been devoted to the vomeronasal glands, as their secretion is required for both receptor function and the access of molecules to the vomeronasal duct. In rabbits, these glands are relatively abundant in the dorsolateral part of the parenchyma, throughout the whole length of the organ, although they are more conspicuous in the central levels. In other species, such as opossum and wallaby, the glands are largely situated caudally past the blind end of the vomeronasal duct (Poran, 1998; Schneider et al. 2008).

As occurs in mammals such as cats (Salazar et al. 1996), horses (Lee et al. 2016) and lemurs (Roslinski et al. 2000), the rabbit glands stained for neutral mucopolysaccharides with PAS but not for acid mucopolysaccharides with Alcian blue. Other species either produce mostly Alcian blue‐positive serous components, as observed for the opossum (Poran, 1998), or share both types of secretions, as observed for bears (Tomiyasu et al. 2017) and cows (Salazar et al. 2008). This notorious variation in gland characteristics (particularly frequency, distribution with respect to the vomeronasal duct, and PAS and Alcian blue staining intensity) among species presumably relates to the variable nature of the pheromone–receptor interaction. Pheromones comprise a wide range of compounds of varied chemical nature that range from small‐molecule volatile compounds to high‐molecular‐weight, water‐soluble, non‐volatile peptides and proteins (Wyatt, 2014). The variable properties found in the fluid secreted inside the lumen of the organ reflect the different nature of the pheromone–receptor interaction in each species (Krishna et al. 1994).

There is wide agreement that two different types of epithelium line the vomeronasal duct, receptor and respiratory (Halpern & Martinez‐Marcos, 2003). The rabbit is no exception to this pattern. The medial wall is lined along the whole vomeronasal duct with a thick layer of neurosensory epithelium (receptor), whereas the lateral wall is covered by a thinner non‐sensory ciliated epithelium (respiratory). Although the sensory epithelium represents the most studied part of the VNO in rabbits, there is very little information about its immunohistochemical features.

We have studied the expression of G‐proteins in the sensory epithelium because Gαi2 and Gαo are associated with the two main families of vomeronasal receptor genes: VR1 and VR2 (Halpern & Martinez‐Marcos, 2003; Suarez et al. 2011). Both proteins have been comprehensively examined in the sensory epithelium of a range of mammals, but have not yet been investigated in rabbits. According to these studies, three different patterns of vomeronasal receptor projections have been described in the VNO, based on the differing expression patterns of G‐proteins. In mammals such as mice and rats (Jia & Halpern, 1996) and in the opossum (Halpern et al. 1995), it has been demonstrated that the apical part of the sensory epithelium of the VNO is Gαi2‐positive, whereas the basal part is Gαo‐positive. In other species, such as goats (Takigami et al. 2000) and cats (Salazar & Sanchez‐Quinteiro, 2011), the immunoreactivity in the sensory epithelium is negative for Gαo and positive for Gαi2. Finally, Schneider et al. (2012) observed recently in their study of the tammar wallaby VNO that the sensory epithelium expressed immunoreactivity to only the Gαo protein, and no immunoreactivity was observed for Gαi2.

Our results in rabbits show labelling by both G‐protein antibodies in the sensory epithelium of the VNO in a pattern very similar but not identical to that described in rodents and the opossum. The Gαi2 antibody immunostains a subpopulation of neuroreceptor cells whose cell bodies are located in the apical part of the epithelium, whereas Go labels a complementary subpopulation of neurons that follow a broader disposition in the epithelium. Thus far, the analysis of the vomeronasal receptor repertoire and expression has not been performed in rabbits, but these immunohistochemical results can be extrapolated to the effective expression of the two main families of vomeronasal receptors, VR1 and VR2, in the same way as in rodents.

Although this is the only published study to investigate G‐protein expression in neuroreceptor cells in the rabbit VNO, our results can be contrasted to those obtained in the same species with antibodies against lactoseries carbohydrates (Mori, 1987; Mori et al. 1987). These authors found that the antibody 4C9 identified a subset of vomeronasal receptor cells, which appear in the images to be mostly located away from the basal membrane. On the other hand, antibody 2C5 produced broader labelling without comprising the whole population of vomeronasal receptor cell bodies. In both cases, a pattern very similar to that found by our group with G‐protein labelling was observed. Although they were not able to establish the functional meaning of the specific carbohydrate expression patterns of these subsets of neurons, posterior studies in mice have established a correspondence between glycoconjugate expression and the pattern of labelling determined by antibodies against the Gαo and Gαi2 proteins. In fact, the basal‐apical zonation established by G‐proteins can be emulated with the histochemical labelling of carbohydrates with the lectin Lycopersicon esculentum agglutinin LEA (Barrios et al. 2014). Taken together, our observations and those published by Mori et al. suggest that the rabbit vomeronasal sensory epithelium follows a pattern of zonation that is slightly different from the basal‐apical organization described in rodents.

Employing UEA‐I, we have also studied the expression in the VNO of the α1‐2 fucose glycan, which mediates neuronal changes, such as neurite outgrowth and synaptic plasticity. The labelling we found in the vomeronasal glands and the non‐sensory epithelium is similar to that found in mice (Barrios et al. 2014; Kondoh et al. 2017, 2018). However, the pattern of labelling of the sensory epithelium is quite different. In mice, this labelling follows a uniform pattern, with stronger staining in the apical zone along the whole epithelium; however, in rabbits, the labelling is strongly concentrated around the commissures of the vomeronasal duct, and is much more diffuse and irregular in the central segments of the epithelium. This difference could reflect stronger activity in the commissures of the duct, where most of the proliferation of new receptor cells occurs (Barber & Raisman, 1978). Our observations in rabbits complement those made in the VNO of other Euarchontoglires, one of the four major subclades within the clade Eutheria, which includes Rodentia, Lagomorpha, Scandentia, Primates and Dermoptera, mainly those in the tree shrew (Malz et al. 1999) and Primates (Kinzinger et al. 2005). Although each of the species studied shows a stereotypical distribution, there is a consistency that contrasts with the remarkable differences in labelling in other groups of mammals, which ranges from negative in sheep to uniformly positive in pigs and cats (Salazar et al. 2000; Salazar & Sanchez‐Quinteiro, 2011). This observation would imply that the expression of glycoconjugates in the vomeronasal system is highly specific and more complex than is usually considered.

We have also examined the expression of OMP as a neuronal marker of terminally differentiated olfactory or vomeronasal receptor neurons (Farbman & Margolis, 1980). The labelling of the VNO neuroepithelium confirms the maturity of the neuroreceptor cells in rabbits. As occurs in mice (Barrios et al. 2014), the neuroepithelium does not show differential zonation or patterning, as observed, for instance, in the primate tamarin (Smith et al. 2011). Although the function of OMP remains unknown, it is considered to play a role in olfactory and vomeronasal chemoreception (Bock et al. 2006).

In conclusion, the findings herein support the notion that the rabbit O. cuniculus possesses a highly developed VNO, with many specific morphological features. Together with recent behavioural data, these results highlight the significance of chemocommunication in this species. To consider this species a suitable model for the study of vomeronasal chemoreception in mammals, anatomical studies should be extended to other parts of the vomeronasal system, namely, the accessory olfactory bulb and the vomeronasal amygdala.

Author contributions

P.S.Q. and P.V. designed the research. P.S.Q., P.V., J.M.C. and P.F.T. performed the work, analysed and discussed the results, and wrote the paper.

Supporting information

Fig. S1. Representative serial transverse sections of the nasal cavity showing the topographical relationships of the VNO.

Fig. S2. Representative mid‐sagittal serial sections of the nasal cavity showing the topographical relationships of the VNO.

Fig. S3. Negative control: UEA‐I preabsorbed with 0.1 m l‐fucose.

Fig. S4. Mouse VNO stained with UEA with an identical protocol as that followed in rabbit VNO.

Fig. S5. Negative control (primary antibody omitted).