EXTENDED TEXT

General description, Anatomy and Distribution: The vomeronasal organ (VNO) or Jacobson's organ is a chemoreceptor organ present in most vertebrates except fish which have a different arrangement of chemoreceptors, or in birds, whose aerial lifestyle may make chemical communication less useful. It is important in intraspecific chemical (pheromone) communication. The paired organs are separate from the main olfactory organ. In most species they are largely enclosed within a capsule formed by the vomer-bone or vomer-cartilage and in mammals are found within a bulge along each side of the base of the nasal septum (see Fig 1). In species with a functional VNO, the vomeronasal cavity (lumen) is partially lined with chemosensory vomeronasal epithelium. This epithelium contains bipolar sensory neurons similar to those of the main olfactory epithelium except that they generally bear microvilli on their exposed surface rather than cilia (there is a report of cilia in dogs and a mix of cilia and microvilli in hamsters). Receptor-molecules capable of binding specifically to vomeronasal stimulus-molecules would be expected in apical membrane, along with the molecular components of the proposed transduction mechanism, which are localized at the microvillar surface (Liman et al 99, Menco et al 2001). Early work (Tucker 1971) suggested a general rather than a highly selective VNO sensitivity, at least at relatively high concentration, but more recent studies in turtles (Brann et al 2001) and mammals suggest that vomeronasal sensory neurons are highly selective to particular substances at very low concentrations; as are the specialist pheromone sensory cells in insects. Sensory neurons are replaced by maturing stem cells (basal cells) if their axons are cut experimentally. There may naturally be a continuous replacement of the sensory neurons, as in the main olfactory epithelium.
Each vomeronasal organ in mammals (Hamster Fig 1) consists of an elongated tube, opening only at the anterior end via a narrow duct into the floor of the nasal cavity (rodents and lagomorphs, Fig 1; humans, Fig 2) or into the nasopalatine canal (carnivores, ungulates, insectivores and new-world primates - not shown). The nasoplatine canal (incisive canal, Stenson's canal) is a passageway through the palate between the nasal and oral cavities, so that in the second group of species, stimuli could reach the organ via either the nose or the mouth (unless the connection of the canal with the mouth is secondarily closed, as in horses). Elephants transfer chemosensory stimuli to the nasopalatine duct opening in the roof of the mouth using the mechanosensory "finger" at the tip of the trunk (Rasmussen et al 1982). In some, probably most mammalian species, access of stimulus molecules to the receptors is regulated by an autonomically controlled vascular pumping mechanism. The pump consists of large blood vessels running alongside the VNO, within the vomer capsule. When the blood vessels are constricted by vasomotor action, the lumen of the organ expands, drawing potential stimuli in through the duct (Meredith 1994). In hamsters, at least, the pump appears to operate in response to novel stimuli and situations that attract the animals' attention. In some ungulates (horses, cattle, antelope) and some carnivores (cats), a behavior called "Flehmen" or "testing", typically seen when males investigate female secretions, has been suggested as an outward manifestation of active vomeronasal stimulation. The stereotyped posture involved (lifted head, curled back upper lip) appears to increase access to the receptor epithelium (at least in goats, Melese d'Hospital and Hart 1985), perhaps by opening the narrow access passageways.
In mammals, reptiles and amphibians, there is evidence that the large non-volatile chemical components of natural secretions may be the fraction involved in some examples of vomeronasal chemical communication. Such substances are presumably picked up, either by nasal contact or on the tongue, during physical contact with the stimulus source. For example, a proteinaceous extract from female vaginal secretions increases mating behavior in intact male hamsters but not in males whose vomeronasal organs have been surgically removed. A protein, "Aphrodisin" coupled with a small lipid molecule may be the effective signal for this behavior (Singer and Macrides 1993). In other species, nonvolatile substances can enter the vomeronasal lumen (Meles d'Hospital et al 1080). Direct demonstration of a vomeronasal neural response to non-volatile substances has been reported for snakes and amphibians but not for mammals. However, biochemical responses are elicited in mammalian vomeronasal tissue extracts suggesting a response to alpha-2-microglobulins (Major Urinary Proteins) in rodent urine (Runnenburger et al 2002), and there is both a biochemical and an immediate-early gene response in hamster accessory olfactory bulb to Aphrodisin (Jang et al 2001). Both olfactory and vomeronasal systems could detect either volatile or non-volatile substances that reach the sensory cells. In both cases the stimulus substance has to traverse a thin overlying aqueous medium. In the case of vomeronasal systems in mammals, the stimulus may also have to be transported into the organ dissolved in an aqueous medium (mucus), so aqueous solubility might be more important there than in the olfactory system. However, proteins produced by nasal glands have been proposed as carriers in both systems (olfactory binding proteins; vomeromodulin: Krishna et al 1995), and it may be the solubility of the bound complex which is important in determining which chemicals can stimulate the organ.
The VNO is prominent in reptiles (other than crocodilians) and especially so in snakes (Fig. 3). The organs open into the roof of the mouth and are stimulated by materials delivered to their ducts by the flicking of the forked tongue. The snake VNO is used for prey trailing and capture as well as intraspecific chemical communication (courtship, mating, and aggregation; Halpern 1987, and in press). In some amphibia the VNO appears involved in reproductive signalling (Wirsig-Wiechmann et al 2002) via peptide/protein chemicals. Cetacean and sirenian marine mammals (whales, dolphins, manatees), birds and fish lack the organ, but fish have microvillar sensory neurons in their main olfactory epithelium which resemble the microvillar sensory neurons of the tetrapod vomeronasal organ, may express V2R-class receptors, and have been implicated as pheromone detectors in goldfish (see Vomeronasal Receptor Genes.

The Human Vomeronasal Organ: The VNO has long been known to be present in human fetuses and has been reported sporadically in adults since the eighteenth century (Bhatnagar and Smith 2003). A closer examination, beginning in the mid-1980s (Johnson et al 1985) revealed the consistent presence of vomeronasal duct openings (VNO pit) in most if not all human adults. The opening (Fig. 2) is somewhat higher up the nasal septum than might be expected from the location in other species, and the organ lacks the characteristic capsule and large blood vessels of other mammals. Both these features suggest that air-borne rather than mucus-borne stimuli would more likely to reach the human VNO. The duct leads into a tubular lumen without an obvious thick sensory epithelium. A very few cells in the lining of the lumen have some similarity with VNO bipolar sensory neurons of other species and can be stained for neuron-specific marker substances (Trotier et al 2000, Witt et al 2002). However, they appear to lack axons, quite unlike sensory neurons of other species. Effective connections from the VNO to the brain, by which these or any other cell-type in the human VNO could convey sensory information have not so far been demonstrated in humans. Nerve fibers are found running below the epithelium but some of these may belong to the Nervus terminalis (see below), or they may be trigeminal or autonomic fibers. In addition, the accessory olfactory bulb, which is the normal termination of vomeronasal sensory-neuron axons, cannot be distinguished in the human brain. Genetic evidence also points to a degradation or elimination of human vomeronasal function over the course of recent evolution. Almost all known vomeronasal receptor (VR) genes in the human genome are non-functional pseudogenes, as is the TRP-2 gene, thought to be essential for electrical responses in vomeronasal sensory neurons (see Transduction). One VR-like gene is expressed in human main olfactory epithelium (see Vomeronasal Receptor (VR) Genes) but its function in coding for a "pheromone" versus a "general odor" receptor is unclear. There appears to be no reason a VR-type receptor could not be functional if expressed in an olfactory epithelial sensory neuron (or displaced VNO-type neuron). However, the special vomeronasal mucous environment might be important. Functional vomeronasal sensory neurons are normally sequestered in the VNO which has its own mucous glands.
The accumulated evidence for loss of important elements of the human VNO would be reasonably convincing of a lack of function except for the physiological evidence for an electrical response (see below). The evidence for different repertoires of vomeronasal genes in distantly related species raises the remote possibility that other vomeronasal receptor genes may await discovery (see Vomeronasal Receptor Genes). However, for the human VNO to be a functional chemoreceptor organ we may have to believe in both unknown receptor genes and an unknown transduction process. This does not mean that pheromone communication is impossible in humans, or even unlikely, just that it is not likely to be mediated by classical vomeronasal sensory neurons using known transduction processes.
Over the last decade, electrical signals recorded from the human VNO have been reported in response to steroid chemicals claimed to be human pheromones (Monti-Bloch et al 1998). Full details of some control experiments have not been given for these experiments so they are difficult to evaluate but so far there are no obvious problems with their methodology. The authors of this work have given the name "vomeropherin" to substances that stimulate the human VNO but this term is not synonymous with "pheromone". They have also demonstrated systemic responses such as changes in heart rate and small decreases in testosterone when "vomeropherins" are delivered to the VNO in humans. Several groups of researchers have independently shown changes in mood after exposure to these chemicals, especially androstadienone (Grosser et al 2000, Jacob and McClintock 2000), even when subjects detected no preceptible, or the odor was masked. In one imaging study, a significant brain response to estratetraenol was detected at concentrations where subjects reported no detectable odor (Sobel et al 1999). Only one of these reports involved stimulation explicitly of the VNO (Grosser et al 2000). The others could have produced their effects through the main olfactory system. The local-electrical response and systemic-physiological response to local stimulation suggest that some information travels from the neighborhood of the VNO to the brain, where cardiovascular, endocrine and emotional modulation can occur. In the absence of a vomeronasal nerve the neural pathways involved are unclear, and the lack of other VNO components means the source of the signals is unlikely to be classical VNO sensory neurons. This question is reviewed in greater detail in a recent review (Meredith 2001). The scientists reporting VNO electrical responses are associated with a commercial group that is promoting the use of "vomeropherins" as therapeutic agents, implying that they are pheromones (Taylor 1994) (see Pheromone). The evidence that these are pheromones evolved for chemical communication seems insufficient. Since behavioral and physiological responses to odors can be conditioned and individual, social and cultural norms for such responses to human odors vary, a response seen in one small group is poor evidence for a fundamental biological mechanism evolved for communication.
There is evidence for chemosensory communication in humans, as in the synchronization of menstrual cycles among women who live together (Stern and McClintock 1998). There is also an intriguing influence of the Major Histocompatibility Complex (MHC) genes of both receiver and donor on the outcome of human-odor preference studies (Jacob et al 2002: see Vomeronasal Receptor Genes for another intriguing connection - between VNO and MHC). Whether these are "pheromone" communications by the more rigorous criteria remains to be demonstrated and there is no evidence to suggest that they are mediated by the VNO. Of course, the existence of a VNO-mediated response does not prove that a stimulus substance is a pheromone. Nor does the existence of pheromone communication imply vomeronasal mediation (see Pheromone).

Pheromone: Definitions of the word "pheromone" vary but all agree that it describes a chemical produced by one member of a species that is detected by another member in which it produces a physiological or behavioral response. An evolutionary basis for use of the special term "pheromone" is recognized in more sophisticated definitions, which include a requirement that the chemical communication be mutually beneficial to sender and receiver, before it should be given a special label (Rutowski 1981, Albone 1984, Meredith 2001). Although the VNO is frequently described as a pheromone detector, there is clear evidence for VNO function (e.g. in snakes) to detect non-pheromone chemicals and clear evidence for pheromone communication not mediated by the VNO. For example:. Newborn rabbits find their mother's nipples by odor cues but this behavior is unimpaired by vomeronasal lesions (Hudson and Distel 1986). Sheep recognize their lambs by odors and this function is also unimpaired by vomeronasal damage (Levy et al 1995). In pigs, VNO duct occlusion did not prevent the female "standing" responses to male pheromone (Dorries et al 1997).

Vomeronasal Receptor (VR) Genes: Two superfamilies of genes (V1Rs and V2Rs) of the seven-transmembrane-domain (7TMD) receptor type are expressed in the vomeronasal organs of rodents (Dulac 2000; a subgroup of V1Rs has also been classified as V3Rs). V1R/V3R genes are expressed in neurons located more superficially in the vomeronasal epithelium and V2Rs are expressed in neurons with deeper cell bodies. In both cases there appears generally to be one VR gene expressed per neuron. Neurons expressing V2R genes also express specific genes of the Major Histocompatability Complex (MHC; Loconto et al 2003, Ishii et al 2003). Neurons expressing the two types of VRs also express different "g-proteins" and project their axons to different subdivisions of the accessory olfactory bulb (see Central Neural Connections). Both features suggest different transduction mechanisms. G-proteins are essential parts the cellular "transduction mechanism", coupling 7TMD-type receptor molecules to electrical responses in the cell. Each VR gene superfamily has on the order of 100 members plus many non-functional "pseudogenes (possibly resulting from an earlier period of divergent evolution; Rodriguez et al 2002). If each complete gene is functional and codes for a highly selective receptor protein, they could allow the identification of about 200 different chemicals. This may be sufficient for species specific discrimination of rodent pheromones conveying many different messages, given that messages in different related species may be carried by blends of the same chemicals in different proportions. Mouse vomeronasal sensory neurons in the V1R/V3R zone of the epithelium respond very sensitively and highly selectively to putative mouse pheromone chemicals (Leinders-Zufall et al 2000). Electrophysiology and calcium imaging indicate that the proportion of responsive neurons appears to be consistent with the activation of one receptor-type by each chemical. Species other than rodents appear to lack the V2R class of vomeronasal receptor genes (although genes of the V2R family are expressed in the olfactory epithelium of fish: Cao et al 1998). It is not clear whether mammals lacking V2R genes have a poorer repertoire of pheromonal communications. In marmosets, none of the genes corresponding to rodent vomeronasal receptor genes appears to be functional, although this species does appear to have a classical vomeronasal organ and to communicate by pheromones (Giorgi and Rouquier, 2002). If all VR receptors are highly selective in the chemicals that activate them, one might expect disparate species to have evolved disparate receptors. Additional vomeronasal receptor gene families may exist in these non-rodent species and have escaped detection by the genomic search methods used so far. Searches within the human genome have revealed only 5 potentially functional VR-type receptor genes, one of which is expressed in the main olfactory epithelium (Rodriquez and Mombaerts 2002). Other VR genes have accumulated mutations over the course of evolution in the lineage of the old-world primates (Giorgi and Rouquier 2002), many of which appear to have no vomeronasal organ..

Transduction: The binding of chemosensory stimuli to vomeronasal receptor molecules triggers a sequence of intracellular events, the transduction cascade, leading to generation of impulses. The necessity for the VR protein has been demonstrated directly in mice where a particular V1R gene was genetically-engineered to also express a fluorescent marker. All marked neurons of that type extracted from the vomeronasal epithelium responded quite selectively to a few closely related chemicals (Boschat et al 2003) but failed to respond to the V1R gene was deleted. Much indirect evidence suggests that the transduction mechanism for VNO sensory neurons involves inositol trisphosphate (IP3) or diacyl-glycerol (DAG) as "second messenger" molecules produced intracellularly when the chemosensory stimulus molecule, (the "first messenger"), binds extracellularly to the VR receptor. Both IP3 and DAG are generated by the enzyme Phospho-Lipase-C (PLC), and may open a surface membrane ion channel allowing extracellular sodium or calcium into the cell. Positively charged sodium or calcium ions can depolarize the cell, initiating impulses that carry the message to the brain. In both reptiles and rodents, there is evidence for intracellular increases in IP3/DAG, and calcium, within vomeronasal sensory neurons in response to chemical stimulation (Wang et al 2002, Runnenburger et al 2002). The best candidiate for the ion channel that could convert an IP3 response into a depolarization is the transient-receptor-2 (TRP-2) channel (Liman et al 1999). It is present at the microvillar surface (Menco et al 2001), as are IP3 receptor molecules, and these may link together to open the TRP-2 channel (Brann et al 2002). TRP-2 is central to one line of argument that humans and other old-world primates have no VNO function (see Human VNO). Deletion of this gene in mice results in dramatic changes in behavior, although not identical with changes produced by surgical removal of VNOs (see Function).

Function: Most of the work on vomeronasal function has been in rodents and snakes. Early experiments involved lesions (damage or removal) of the vomeronasal or olfactory systems in order to reveal deficits in behavior or physiological function. More recently, researchers have deleted genes thought to be involved with vomeronasal function and looked for deficits. In other experiments the activation of brain areas receiving central vomeronasal projections has been investigated with other techniques (see Central Neural Connections).
In rodents, intraspecific chemical communication can produce dramatic effects on reproductive behavior and physiology, and on aggressive behaviors, that depend on chemosensory input from the vomeronasal system (Keverne 1999). For example, female mice housed in groups produce a urinary chemosignal that suppresses estrus in other females. Male mice and other rodents produce chemosignals that accelerate puberty and females may produce chemosignals that delay puberty in immature females of the same species. In all cases, removal of the VNO prevents the response (Wysocki and Meredith 1987) and many clearly involve "pheromone communication" (se Pheromone). The VNO is occasionally removed in humans during surgery to repair a malformation of the nasal cavity but the consequences have not been systematically studied.
In both mice and hamsters, males do not mate when they are experimentally deprived of both olfactory and vomeronasal sensory input. In experienced males, either olfactory or vomeronasal input is sufficient to allow courtship (e.g. ultrasonic calling) and mating in response to female chemical cues, and either input can be removed with little effect. However, in sexually naive animals, the removal of the vomeronasal organs alone produces severe deficits in these behaviors, whereas removal of the olfactory input alone does not (Meredith 1986, Fernandez-Fewell and Meredith 1998). Thus, the vomeronasal input may be critical for reliable preprogrammed behavior (e.g. mating) in naive animals. Later, this behavior may be elicited by olfactory sensory input after a period of association between vomeronasal and olfactory sensory input.
Deletion of a group of VR genes that code for a subset of vomeronasal receptor proteins also disrupts behaviors involving pheromone communciation in mice (DelPunta et al 2002a). These mutant mice had reduced male sexual behavior and reduced maternal aggression. Their VNOs failed to give detectable electrical responses to a subset of the chemicals proposed as pheromones in mice, supporting other evidence that VR proteins can be pheromone receptors and functional in VNO sensory neurons. More dramatic effects (Stowers et al 2002, Leypold et al 2002) occur with deletion of the TRP-2 gene, thought to involved in generating electrical responses in vomeronasal sensory neurons (see Transduction). There was a failure of VNO electrical responses to urine chemosignals and to specific putative pheromone chemicals in these TRP-2 knock-out (KO) mice. However, unlike mice with VNO lesions (or with VR-gene deletion), TRP-2-KO males also failed to distinguish between females and other males as mating partners. TRP-2-KO mice did show reduced aggression, as in mice with vomeronasal lesions, but did not show reduced male mating behavior. Whether these differences between mice with intact but potentially non-functional VNOs and mice without VNOs are due to developmental differences related to the loss of TRP-2, or to collateral damage (or residual function) following VNO surgery in wild-type mice, or to some other cause, is not clear. The proposal that the vomeronasal organ is essential for gender discrimination is not supported unequivocally by these data. However, responses of some second-order neurons in the accessory olfactory bulb, recorded in awake mice investigating other (anesthetized) individuals, do suggest gender is an important category of information obtained through VNO sensory input. Some AOB neurons appear to be highly selective for gender and for breed of mouse investigated, or both (Luo et al 2003). The initial evidence suggests these responses are determined by the "category" (sex and breed) of the stimulus animal, not by differences between individuals.
In mice and hamsters, removal of the vomeronasal organs prevents the hormonal changes (LH and later testosterone) normally observed after exposure of males to female chemosignals (Coquelin et al 1984). The LH changes may indicate an intracerebral release of Gonadotropin Releasing Hormone (GnRH; also called: luteinizing-hormone releasing-hormone; LHRH). Intracerebral injection of GnRH substantially restores deficits in mating behavior produced by VNO removal in sexually naive male hamsters, suggesting that intracerebral liberation of GnRH may be an important component of the response to vomeronasal sensory input (Fernandez-Fewell and Meredith 1995, Westberry and Meredith 2003b). It has been suggested that the vomeronasal system might be functional in utero. It is possible to influence postnatal response to an odor in some rodents by exposing fetuses to that odor in utero (Hudson and Distel 1998). However, such associational capabilities may be more a function of the main olfactory system rather than the VNO. There is also evidence in the rat fetus, for a high 2-deoxyglucose uptake in the vomeronasal nerve-terminal region of the accessory olfactory bulb, possibly indicating neuronal activation (Pedersen et al 1983). However, the VNO duct is not open in utero in some species (Coppola and Millar 1994), so an in utero sensory function may not be essential.

Central Neural Connections: The central neural connections of the vomeronasal system are consistent with its proposed role in initiating social and reproductive responses. The vomeronasal sensory neurons have cell bodies in the VNO epithelium and axons that project to the accessory olfactory bulb (AOB), a distinct structure, usually posterior-dorsal to the main olfactory bulb (Fig. 1). In rodents, vomeronasal neurons expressing V1/3 type receptors project their axons to the rostral subdivision of the AOB. Neurons expressing V2R receptors project to the caudal subdivision, suggesting a difference of function (Fig. 4). In both AOB subdivisions, VN axons end in small clusters called glomeruli where they connect with second order neurons (mitral cells). VN axons expressing a particular VR receptor divide up to innervate several small glomeruli, and AOB mitral cells have multiple dendrites ending in several glomeruli. Initial evidence suggests these connect to glomeruli receiving axons of cells bearing the same VR molecule, so the apparently divergent pathway may come back together (Del Punta et al 2002b). This organization is different from that of the main olfactory bulb (MOB), where the axons from neurons expressing a particular olfactory receptor (OR) all converge onto one glomerulus (on each side of each MOB), synapsing with mitral cells that make connections only with that glomerulus (MOB mitral cells receive afferent input only from one glomerulus). The function of the AOB organization is not clear but might involve selective inhibitory interactions between distant AOB mitral cells to resolve confusions between particularly important stimuli. In contrast, the apparently non-selective lateral-inhibition between MOB mitral cells could sharpen differences among responses to a broad family of chemicals.
Accessory olfactory bulb neurons receiving VN input project to the medial nucleus of the amygdala and to the medial part of the cortical nucleus (Price 1987). The density of projections directly from AOB decreases posteriorly, so much of the vomeronasal-related input to posterior medial amygdala and the posterior cortical nucleus may be relayed from the anterior medial amygdala. There appears to be little or no selective connections between different (rostral/caudal) subregions of the AOB and the different corticomedial amygdala regions. In those species with a VNO, the corticomedial amygdala regions apparently receive no direct projections from the main olfactory bulb, although there are indirect connections from other (lateral cortical) amygdala areas that do receive main olfactory projections. These cross-connections may allow olfactory input to substitute for VNO input to maintain mating in experienced hamsters. In humans, the corticomedial amygdala receives projections from the olfactory bulb but a distinct AOB cannot be distinguished in humans so it is not clear if these projections to the medial parts of the human amygdala have any specialized (e.g. VNO) peripheral inputs. In any case, the human corticomedial amygdala probably evolved as an accessory-olfactory projection area as in other mammals and may have continuing functions in social behavior. In amphibia and reptiles, the AOB projects to structures thought to be equivalent to the amygdala, such as the nucleus sphericus, and again at this level the projection areas are separate from those of the main olfactory system. In mammals, cells in the corticomedial amygdala project to more central preoptic and hypothalamic areas of the brain that are concerned with hormonal control, reproduction and regulatory functions, including feeding. There is apparently no neocortical projection of the vomeronasal system. Partly on this evidence, it has been suggested that vomeronasal sensory input may be unavailable to cognitive processes, i.e. that the recipient of vomeronasal communication, even though showing behavioral or physiological responses, may be unaware of the stimulus. However, an unconscious chemical communication would not have to be via the VNO.

Central neural connections: The central neural connections of the vomeronasal system are consistent with its proposed role in initiating social and reproductive responses. The vomeronasal sensory neurons have cell bodies in the VNO epithelium and axons that project to the accessory olfactory bulb (AOB), a distinct structure, usually posterior-dorsal to the main olfactory bulb (Fig. 1). In rodents, vomeronasal neurons expressing V1/3 type receptors project their axons to the rostral subdivision of the AOB. Neurons expressing V2R receptors project to the caudal subdivision, suggesting a difference of function (Fig. 4). In both AOB subdivisions, VN axons end in small clusters called glomeruli where they connect with second order neurons (mitral cells). VN axons expressing a particular VR receptor divide up to innervate several small glomeruli, and AOB mitral cells have multiple dendrites ending in several glomeruli. Initial evidence suggests these connect to glomeruli receiving axons of cells bearing the same VR molecule, so the apparently divergent pathway may come back together (Del Punta et al 2002b). This organization is different from that of the main olfactory bulb (MOB), where the axons from neurons expressing a particular olfactory receptor (OR) all converge onto one glomerulus (on each side of each MOB), synapsing with mitral cells that make connections only with that glomerulus (MOB mitral cells receive afferent input only from one glomerulus). The function of the AOB organization is not clear but might involve selective inhibitory interactions between distant AOB mitral cells to resolve confusions between particularly important stimuli. In contrast, the apparently non-selective lateral-inhibition between MOB mitral cells could sharpen differences among responses to a broad family of chemicals.
Accessory olfactory bulb neurons receiving VN input project to the medial nucleus of the amygdala and to the medial part of the cortical nucleus (Price 1987). The density of projections directly from AOB decreases posteriorly, so much of the vomeronasal-related input to posterior medial amygdala and the posterior cortical nucleus may be relayed from the anterior medial amygdala. There appears to be little or no selective connections between different (rostral/caudal) subregions of the AOB and the different corticomedial amygdala regions. In those species with a VNO, the corticomedial amygdala regions apparently receive no direct projections from the main olfactory bulb, although there are indirect connections from other (lateral cortical) amygdala areas that do receive main olfactory projections. These cross-connections may allow olfactory input to substitute for VNO input to maintain mating in experienced hamsters. In humans, the corticomedial amygdala receives projections from the olfactory bulb but a distinct AOB cannot be distinguished in humans so it is not clear if these projections to the medial parts of the human amygdala have any specialized (e.g. VNO) peripheral inputs. In any case, the human corticomedial amygdala probably evolved as an accessory-olfactory projection area as in other mammals and may have continuing functions in social behavior. In amphibia and reptiles, the AOB projects to structures thought to be equivalent to the amygdala, such as the nucleus sphericus, and again at this level the projection areas are separate from those of the main olfactory system. In mammals, cells in the corticomedial amygdala project to more central preoptic and hypothalamic areas of the brain that are concerned with hormonal control, reproduction and regulatory functions, including feeding. There is apparently no neocortical projection of the vomeronasal system. Partly on this evidence, it has been suggested that vomeronasal sensory input may be unavailable to cognitive processes, i.e. that the recipient of vomeronasal communication, even though showing behavioral or physiological responses, may be unaware of the stimulus. However, an unconscious chemical communication would not have to be via the VNO.

Central Neural Responses: Areas on the central vomeronasal pathways, such as the medial amygdala, are activated during mating behavior in rodents, as indicated by experiments using expression of immediate-early genes (IEGs) such as c-fos as markers for neural activity. However, only part of the activity in the medial amygdala appears due to vomeronasal input; the rest is associated with the mating performance itself (Fiber et al 1993, Fernandez-Fewell and Meredith 1994, Kollack-Walker and Newman 1997). This mating-related c-fos expression may reflect other sensory input that is an integral part of the process, such as genital stimulation, or may reflect the integrative or motor aspects of performance. The exact distribution of IEG activation in central vomeronasal regions also appears to depend in part on the experience of the animal: Activation is more extensive in experienced animals. The same regions of the amygdala are also activated when male hamsters investigate female pheromones or when investigating other chemosignals from males or females of the same species. Only a subset of the amygdala areas are activated when animals investigate chemosignals from different species. This categorization of the incoming information is different from the way the accessory olfactory bulb responds to the same stimuli (Westberry and Meredith 2003a and unpublished). There the response may reflect the chemical composition of the stimuli rather than, perhaps, their social relevance. Activation in the main olfactory system is different again. The number of activated neurons is not increased above control levels during pheromone stimulation, although the pattern of activation may be different for different stimuli at higher concentrations where the main olfactory system does respond to "pheromones" as "odors".

Sensitivity and Selectivity: Early electrophysiological recordings from the vomeronasal system in the turtle showed responses to volatile chemicals selected to be useful odors for these experiments but having no known vomeronasal functions. These substances also stimulated the main olfactory system, which appeared to be somewhat more sensitive to some of these chemicals and less sensitive to others. The turtle is unusual in having its vomeronasal receptors exposed directly to nasal airflow; but a few recordings from the vomeronasal organ or AOB in mammals also suggest a general rather than a highly selective sensitivity, at least at relatively high concentration. It is important to point out that identified substances known to be involved in the behavioral and physiological responses to vomeronasal input have not generally been available for testing. The vomeronasal receptor neurons may ultimately turn-out to be highly selective to particular substances at very low concentrations; as are the specialist pheromone receptor cells in insects. This has not been demonstrated as a general principle but one recent report showed that some chemicals that have been proposed as mouse pheromones do activate vomeronasal neurons in that species with remarkable selectivity and sensitivity (Leinders-Zufall et al 2000). It is not yet clear whether putative receptor genes of the recently discovered gene families expressed in the vomeronasal organ are actually responsible for all sensitivity to pheromones of rodents, although deleting a subset of these genes does eliminate responses to some putative pheromone chemicals. The two gene families are very different from each other and are not similar to those expressed in the main olfactory system (for which there is evidence of a chemosensory function).
In both mammals and reptiles, there is evidence that the large non-volatile chemical components of natural secretions may be the fraction involved in some examples of vomeronasal chemical communication. Such substances are presumably picked up, either by nasal contact or on the tongue, during physical contact with the stimulus source. For example, a proteinaceous extract from female vaginal secretions increases mating behavior in intact male hamsters but not in males whose vomeronasal organs have been surgically removed. Recent evidence suggests that a protein named "Aphrodisin" coupled with a small lipid molecule may be the effective signal for this behavior. In other species, nonvolatile tracer substances mixed in with the appropriate natural stimulus materials do enter the vomeronasal lumen, after the animal has been allowed to investigate and physically contact the stimulus. Direct electro-physiological demonstration of a vomeronasal response to non-volatile substances has been reported for snakes but not for mammals. It is not yet clear whether the vomeronasal system is specifically adapted to respond preferentially to non-volatile stimuli. Both olfactory and vomeronasal systems are capable of detecting either volatile or non-volatile substances that reach the receptor cells. In both cases the stimulus substance has to traverse a thin overlying aqueous medium. In the case of vomeronasal systems in mammals, the stimulus may also have to be transported into the organ dissolved in an aqueous medium (mucus), so aqueous solubility might be more important there than in the olfactory system. However, proteins produced by nasal glands have been proposed as carriers in both systems (olfactory binding proteins; vomeromodulin), and it may be the solubility of the bound complex which is important in determining which chemicals can stimulate the organ.
Details of the transduction mechanism (conversion of chemical stimulation into electrical response) have not been worked out in detail. Nevertheless, stimulus chemicals presumably generate action potentials in VNO receptor neurons by binding to protein receptor-molecules in the apical membrane of the receptor neurons exposed in the VNO lumen. Recent results suggest that stimulation of VN receptor neurons in both snakes and mammals might involve activation of the enzyme phospholipase-C to generate the intracellular second messenger IP3. This, in turn, can open ion channels to depolarize the cell and generate action potentials that propagate along the VN axons to the brain (AOB). This mechanism has also been proposed for olfactory transduction in some insects and fish. A prime candidate for the channel essential for transduction in VNO receptors is the "transient receptor potential-2" (TRP-2) protein. It is expressed selectively in vomeronasal receptor neurons and the protein product appears to be localized to the microvilar membrane at the apical surface of the cell (Liman et al 1999). Genetic deletion of the TRP-2 gene in mice results in a loss of electrical responses from the VNO and produces dramatic changes in behavior, some of which mimic those produced by VNO removal (Stowers et al 2002, Leypold et al 2002, see above).

Nervus terminalis

Most of the evidence implicating vomeronasal input in social and sexual behavior and in physiology relies on selective lesions of some part of the system, and the subsequent observation of deficits in behavioral and hormonal responses. There is a possibility that some of the deficits observed are due to damage of an alternative neural system, the Nervus terminalis (NT) or terminal nerve, which is probably damaged to some extent in each case of intended vomeronasal lesion. The NT is ubiquitous in vertebrates, including humans (Wirsig-Weichmann et al 2002). It innervates the olfactory and vomeronasal cavities and projects directly to the medial septum and preoptic area (and to the retina in some fish). There may be both sensory and autonomic components. The nerve and its ganglion contain high levels of GnRH-like immunoreactivity in cells that appear to project centripetally into the brain. In some, possibly all mammals and in other species, the GnRH neurons that control pituitary LH release (and any forebrain intracerebral GnRH release) apparently migrate from the olfactory placode into the brain during development, along the NT pathway (Schwanzel-Fukuda and Pfaff 1989). The persistence of NT in all vertebrates suggests some persistent function but that function is not yet clear. However, the release of LH and testosterone in response to chemical stimuli in male hamsters (which may also involve release of GnRH into the brain), apparently does not depend on an intact NT (Wirsig-Wiechmann 1993). The NT has been implicated in pheromonal responses in fish (Demski and Northcutt 1983), but in goldfish, where pheromone responses have been studied most intensely, these appear to be mediated by the main olfactory system (Hanson et al 1998). Electrical stimulation of the nervus terminalis in elasmobranchs results in increased GnRH in the brain but chemosensitivity to pheromonal (or other) substances could not be directly demonstrated (Moeller and Meredith 1998, Meredith and White 1987).