The mechanisms through which the mammalian VNO transduces chemical signals, such as pheromones and kairomones, into electrical responses that regulate sexual, hormonal, and reproductive function has not yet been fully clarified. During attempts to understand this process, the expression patterns of the α-subunit of the Gi2 and Go proteins were characterized in the rat AOB (Shinohara et al. 1992), leading to the identification of two VR families, V1R (Dulac & Axel 1995) and V2R (Herrada & Dulac 1997; Ryba & Tirindelli 1997; Matsunami & Buck 1997). Gαi2 is considered to serve as an unequivocal marker of the V1R-positive cells of the VNO, whereas the Gαo protein co-expresses with V2R cells. The ubiquitous nature of G protein expression along the NVN extends this labeling along the vomeronasal axons until they reach the superficial layers of the AOB. In the AOB, the NVNs express both G proteins, Gαi2 and Gαo, according to a topographical anteroposterior zonation, in which the anterior region expresses the Gαi2 protein, and the posterior region expresses the Gαo protein. This segregated model of G protein subunit expression was first shown in laboratory rodents (Shinohara et al. 1992; Jia & Halpern 1996) and the opossum (Halpern et al. 1995) and was later confirmed in wild rodents, including degus (Suarez et al. 2009a) and capybara (Suarez et al. 2011b; Torres et al. 2020), tenrecs (Suarez et al. 2009b), and rabbits (Villamayor et al. 2018, 2021). In parallel, the study reported by Takigami et al. (2000) in the goat VNS showed the absence of the Gαo-positive pathway in some mammals for the first time. This deterioration of the Gαo/V2R pathway was confirmed in species as diverse as shrews, horses, dogs, marmosets (Takigami et al. 2004), cats (Salazar & Sanchez-Quinteiro 2011), squirrels, hyrax (Suarez et al. 2011a), and sheep (Salazar et al. 2007). In these species, the NVN and glomeruli only express Gαi2 throughout the AOB, presenting a uniform model of vomeronasal information transduction.
This scenario was complicated by the 2012 publication of a study on the VNS in the tammar wallaby, Macropus eugenii, which unexpectedly revealed that this animal might constitute an alternative model of vomeronasal organization, with vomeronasal neuroreceptor cells along the VNO and AOB expressing only the Gαo protein (Schneider et al. 2012), indicating the existence of a third accessory olfactory type, intermediate to the segregated and uniform types previously described.
However, the authors of the study in tammar wallaby warned that their findings should be taken with caution, as some of their observations prevented the drawing of definite conclusions. Very atypically for a uniform model of G protein subunit expression, the anti-Gαo protein was only expressed in a subpopulation of neuroreceptor cells in the vomeronasal epithelium rather than in the whole population of VR neurons. To explain this low number of Gαo positive cells, the authors hypothesized a reduced affinity of the antibody used for the tammar Gαo; however, no other examples of any similar selectivity deficits in response to the use of anti-Gαo antibodies have been described in the literature.
More critically, immunolabeling of the tammar wallaby AOB using anti-Gαo has only been demonstrated by this single study (Fig. 6E in Schneider et al. 2012). Their staining protocol resulted in very faint labeling, which was restricted to a very small area of the superficial layers of the AOB receiving projections from the vomeronasal neuroreceptor cells. By contrast, in all other studied species, in both the segregated and uniform models, both Gαo and Gαi2 antibodies are associated with positive immunolabeling that can be observed throughout the entire thickness of the superficial layers, including the NVN and glomerular layers. Regrettably, the authors do not specify the antibody employed that resulted in a Gαi2 immunonegative pattern in both the VNO and AOB. Neither the commercial source nor the antibody lot number was reported, precluding any comparisons with other studies of Gαi2 expression in the VNS using the same antibody.
To date, no further studies have been reported that either reinforces or rejects the potential existence of this third model of organization for the vomeronasal transduction pathways. Therefore, we have performed a morphofunctional study of the VNS of the Bennett’s wallaby (Macropus rufogriseus) to characterize the expression patterns of G proteins in this species to shed light on the proposed presence of a third model of vomeronasal transduction in macropodids.
Our immunohistochemical study of G proteins in the Bennet’s wallaby revealed a canonical immunohistochemical labeling pattern for both Gαi2 and Gαo in the VNO and AOB of all samples studied. Moreover, the pattern of labeling observed in the Bennett’s wallaby vomeronasal neuroepithelium using anti-Gαo staining (Fig. 10A–C) was identical to that described for the tammar wallaby (Fig. 3A and B in Schneider et al. 2012), in which only a small fraction of neuroreceptor cells were labeled. The nuclei of these cells were located at the base of the epithelium, with very strong staining in the somas and the knob-like structures protruding into the VNO lumen from the receptor cell endings. However, whereas Schneider et al. did not obtain positive immunolabeling with their anti-Gαi2 antibody in the VNO, we observed an abundant number of immunopositive cells in the neuroepithelium (Fig. 10D) using our anti-Gαi2 antibody. The strong immunopositivity identified in the vomeronasal axons in both the lamina propria and the nasal mucosa confirmed the neuroreceptor nature of these cells (Fig. 10E).
By extending immunohistochemical characterization of G protein expression to the AOB, we were able to verify the existence of the anterior-posterior zonation pattern that is typical of mammalian species belonging to the segregated model. The vomeronasal axons reaching the anterior zone of the AOB only expressed Gαi2, whereas those reaching the posterior zone only expressed the Gαo subunit, and the immunolabeling of these two G proteins revealed a complementary pattern. The employment of the UEA lectin provided additional evidence for the segregation of vomeronasal information in the wallaby, revealing the selective labeling of the anterior zone of the AOB. The affinity of the UEA lectin for the anterior AOB has been reported for all the species belonging to the segregated model in which this histochemical marker has been investigated, including hamster hamster (Taniguchi et al. 1993), mouse (Salazar et al. 2001; Salazar & Sanchez-Quinteiro 2003; Kondoh et al. 2017), rat (Salazar & Sanchez Quinteiro 1998), and capybara (Torres et al. 2020). However, studies using UEA lectin staining in species belonging to the uniform model, including pig (Salazar et al. 2000), cat (Salazar & Sanchez-Quinteiro 2011), and dog (Salazar et al. 2013), and studies that have utilized a very broad panel of lectins, as in the goat (Mogi et al. 2007), have not described any evidence of zonation in the AOB.
The expression patterns observed for both Gαi2 and Gαo proteins in the VNO neuroepithelium, the NVN, and the AOB, including the establishment of a clear zonation pattern that was confirmed by UEA lectin staining, contradict the existence of a hypothetical third model of vomeronasal information processing, as described by Schneider et al. (2012). Although Schneider et al. performed their study in the tammar wallaby, Macropus eugenii, whereas our study was performed in Bennett’s wallaby, Macropus rufogriseus, this difference is unlikely to explain the observed differences in immunohistochemical labeling, as differences in both the structure and G protein expression patterns of the AOB have never been reported between species belonging to the same genus or family (Meisami & Bhatnagar 1998; Halpern & Martinez Marcos 2003).
Moreover, the main morphological and histological features of the VNS described for Macropus eugenii are indistinguishable from those found by us in Macropus rufogriseus. Thus, comparing the observations of Schneider et al. for the VNO (2008) and AOB (2012) of M. eugenii with our observations in M. rufogriseus reveals that both species share important aspects, such as the opening of the VNO to the nasopalatine duct, the semilunar shape of the VND, with the presence of a remarkable mushroom body, the layering and cellularity of the neuroreceptor epithelium, and the presence of many PAS-positive vomeronasal glands in the lateral parenchyma. Both species share a comparable arrangement of large blood vessels surrounding the VND along its medial and lateral planes, and a similar presence of profuse dorsal, medial, ventral, and ventrolateral unmyelinated innervation is described for both species (Fig. 3B in Schneider et al. 2008), transmitting the information collected by the vomeronasal sensory neuroepithelium to the AOB. This arrangement of vomeronasal axons in the parenchyma is very atypical and has not previously been described in the VNO of any other mammal species apart from these two Macropodidae.
In both species, all of the nerve bundles leave the VNO through a dorsolateral opening in the vomeronasal capsule, which is cartilaginous in both species. Another feature common to the VNO of both macropodids, which was observed by Schneider et al. (2008) and Sanchez-Villagra (2001) in the tammar wallaby and which we have verified in our histological series characterizing Bennett’s wallaby, is the failure of the caudal VNC enclose the entire VNO, such that the posterior portion of the VNO is free of surrounding cartilage. A similar finding has been reported for other marsupials, such as Notoryctes (Sweet 1904) and Caenolestes (Broom 1926).
The information available for the AOB of the tammar wallaby is limited to the study of its lamination using hematoxylin–eosin and Nissl stains and the expression pattern of Gα subunits (Schneider et al. 2012). Our work extends these histological observations through the additional use of Tolivia staining, which allows for the characterization of the AOB lamination with greater definition, including its relationship with the lateral olfactory tract and the organization of mitral cells, all of which appeared comparable in the Bennet’s wallaby to species with highly developed AOB structures, such as rodents (Meisami & Bhatnagar 1998) and Lagomorpha (Villamayor et al. 2020). Similarly, our immunohistochemical study allowed for the characterization of UEA and LEA lectin affinity and the immunohistochemical characterization using various marker proteins, including OMP, MAP2, GAP43, and GFAP, which were performed for the first in this marsupial family.
The UEA labeling, we previously discussed as a differential marker for the anterior zone of the AOB, with a distribution pattern analogous to that of Gαi2, is not specific to the VNS of the Bennett’s wallaby, as staining of the superficial layers of the MOB was also observed, similar to the distribution observed in rodents, such as the capybara (Torres et al. 2020), or in pigs (Salazar et al. 2000). LEA lectin stains all mammals, which has been investigated in the vomeronasal and olfactory neuroepithelia (Park et al. 2012; Lee et al. 2016) and the nerve and glomerular layers of both the AOB and MOB compartments in the mouse (Salazar et al. 2001), sheep, pig (Salazar et al. 2000), and rabbit (Villamayor et al. 2020). In the wallaby, unlike UEA, LEA produces a labeling pattern without zonation, similar to the pattern observed for OMP, a marker of mature olfactory and vomeronasal cells (Bock et al. 2006), and identical to the LEA pattern observed in the rabbit AOB, a species in which the zonation determined by G proteins cannot be discriminated using LEA and OMP (Villamayor et al. 2020).
Anti-GAP-43 is one of the best-characterized markers of growing and regenerating neuronal processes (Ramakers et al. 1992) and was able to identify growing axons in the Bennet’s wallaby AOB, with no difference observed between the anterior and posterior zones. However, the antibody against GFAP showed stronger immunolabeling of glial components in the posterior portion of the AOB relative to the anterior segment, a pattern that has not been previously reported for other studies of this marker in the mammalian AOB (Salazar et al. 1994; 2006) and should be examined in more detail in future studies.
CB and CR are expressed in the entire VNS, along the VNO neuroepithelium, the NVN, and AOB. For both markers, the immunostaining of the VNO comprises the soma and dendrites in a pattern similar to that described in the mouse; however, in the mouse, both CB and CR also stain basal cells (Kishimoto et al. 1993). Additionally, in rats, CR labels most of the neuroreceptors cells, whereas CB only labels a neuronal subpopulation without displaying a generalized labeling pattern (Jia & Halpern 2003).
The distribution of CB and CR immunoreactivity in the Bennett’s wallaby’s AOB concentrates on the vomeronasal fibers and glomeruli. The mitral-plexiform and granular layers are stained diffusely with both markers, although this staining is much weaker. This pattern is analogous to that found in other species, such as the rabbit (Villamayor et al. 2020) or the capybara (Torres et al. 2020). However, in both the rabbit and capybara, CB labels the periglomerular cells, and CR labels the mitral cells, which was not observed in the Bennet’s wallaby. In the case of other marsupials, such as the opossum, striking differences in the labeling patterns for CB and CR in the AOB were observed compared with the pattern found in the wallaby. In the opossum AOB, CB-labeled neurons were found in all layers except the nerve layer and the periglomerular cells, whereas CR showed a pattern analogous to that found in the wallaby, although in opossum CR also labels mitral cells and discriminates an anteroposterior zonation, with more intense staining in the posterior AOB than in the anterior AOB zone (Jia & Halpern 2004). Limited knowledge is currently available regarding the functions of calcium-binding proteins in the AOB; therefore, the formulation of hypotheses regarding the importance of these differences between species is challenging; however, CB and CR staining can provide valuable neuroanatomical information in the VNS.
Overall, the VNS of the Bennett’s wallaby shows a degree of differentiation and histochemical and neurochemical diversity comparable to species with greater VNS development; as the main consequence of our study, the existence of a third intermediate type of vomeronasal information processing is not supported. We confirm the presence of effective expression of the two primary VR families in the VNS of the Bennett’s wallaby, and our contribution expands the morphological, histochemical, and immunohistochemical information available on the VNS of Macropodidae and marsupials in general.