The number of functional VRs varies dramatically between and even within mammalian species 4, 15 and 42]; most humans are likely to have very few, if any [43]. VR expression is largely restricted to sensory neurons in the olfactory system, where they are thought to be specialised to detect chemical signals that provoke behaviour, including pheromones. However, only a handful of VR-ligand pairs have been fully characterised 5, 13•• and 17]. There are two structurally divergent classes of VR, V1Rs and V2Rs, that
bind organic volatiles and peptides, respectively (reviewed in [4]). In rodents, each class is independently expressed in spatially restricted FK228 mouse vomeronasal sensory neuron that project to distinct aspects of the accessory olfactory bulb [23]. V1Rs are expressed in a monogenic manner, so that each neuron is patterned by a single receptor [14]. In contrast, V2Rs are expressed in combinations of two or more per
neuron [44]. The mechanism of VR gene selection and the functional consequence of combining V2Rs are not yet known. This simple non-redundant VNO coding model suffers from a fundamental theoretical challenge: there are over 350 functional VRs encoded in the mouse genome (see Box 2) [4]. While there are certainly sufficient Trichostatin A cost peptides and organic molecules secreted by mice to generate a unique ligand for each, are there 350 distinct social behaviours for each pheromone receptor to mediate? Three complementary studies revealed that this is unlikely, because the VNO has evolved to mediate more than just social behaviours. By recording from the accessory olfactory bulb, the brain region that receives primary inputs from the VNO, Ben-Shaul and colleagues observed patterns of neural activity when urine from predators was applied to the VNO [6]. Moreover, most of the neurons they recorded from responded only to the predator urine, not mouse urine.
This suggested that the VNO was specifically tuned to detect chemical cues derived from predators. One of these cues was independently isolated from rat urine, and shown to directly activate a subset of VSNs [7]. D-malate dehydrogenase Mice responded by displaying stereotypic avoidance and defensive behaviours, proving that the mammalian VNO can also detect behaviour-provoking chemicals from other species. Interestingly the rat-derived signal, MUP13, is a homologue of mouse major urinary proteins (MUPs) which are themselves pheromones with diverse social functions (reviewed in [8]). It is therefore likely that some of VRs have evolved to distinguish between structurally related proteins from the same and differences species, and in turn mediate very different behaviours. This was further reinforced when a homologous protein from a cat, Feld4 was shown to activate an overlapping subset of VSNs and provoke similar defensive behaviours as the rat signal [7].