The olfactory receptor (OLFR) gene family is the largest in the mammalian genome (Buck and Axel, 1991), suggesting relevance of olfaction for survival and Evolution [see (Snyder et al., 1988; Dryer and Berghard, 1999; Gaillard et al., 2004) for review]. Despite the expression of OLFRs in different systems, their function outside the olfactory system is unknown. OLFR nomenclature for a given organism, for instance mouse, is different from that for others (see https://www.orthodb.org/?ncbi=1&query=259049). The proteins encoded by the OLFR genes belong to the well-studied superfamily of G protein-coupled receptors (GPCRs). However, there are two main reasons that explain the paucity of data relative to OLFRs in contrast to the large amount of information for the remaining members of class A GPCRs: i) the lack of pharmacological tools (agonists/antagonists), and ii) the difficult expression in heterologous systems (Alexander et al., 2021). The lack of tools has delayed the demonstration of whether OLFRs may be involved in receptor-receptor interactions as has been confirmed for other GPCRs, namely, for class C, class A, and taste receptors (Gines et al., 2000; Duthey et al., 2002; Hillion et al., 2002; Kniazeff et al., 2011; Kuhn and Meyerhof, 2013; Brugarolas et al., 2014; Fujita et al., 2014; Franco et al., 2016; Pin and Bettler, 2016).

Adenosine A_2A receptor (A_2AR), a class A GPCR, is considered a target for neuroprotection. In the central nervous system, the receptor is widely expressed in striatal neurons, although it is also present in neurons in other brain regions as well as glia. In the microglia surrounding brain lesions of Alzheimer’s disease patients, the receptor is upregulated (Angulo et al., 2003). Relevant to therapeutic drug development has been the approval of istradefylline, a selective A_2AR antagonist, as a first-in-class drug for the treatment of Parkinson’s disease (Jenner et al., 2009; Jenner et al., 2021; Mizuno et al., 2013; Suzuki et al., 2017; Berger et al., 2020; Mori et al., 2022). Istradefylline targets the A_2AR which often interacts with other GPCRs on the cell surface as dopamine and cannabinoid receptors (see https://www.gpcr-hetnet.com/ for more interactions involving the receptors; accessed on 13 November 2022). The formation of such complexes is physiologically relevant since heteromers display functional properties that are not possible when receptors act as monomers/homomers (Ciruela et al., 2006; Fuxe et al., 2007; Brugarolas et al., 2014; Franco et al., 2018; Franco et al., 2021; Gallo et al., 2019).

On the homo/heteromerization potential of OLFRs, this sentence in a 2005 paper still holds true: “Given the large number of OLFRs in mammals, the specificity and functional importance if heteromerization of the vast majority of receptors in this family remain to be determined” (Prinster et al., 2005). The present article, whose objective was to search for interactions involving OLFRs, has taken advantage of the results of a transcriptomic study (data in preparation) showing that an OLFR, coded by the Olfr-78 gene (mouse nomenclature), is differentially expressed when mouse primary activated microglial cells are treated with A_2AR antagonists. Therefore, we reasoned that a first approach to the possibility that OLFRs could form receptor-receptor complexes was to determine whether this specific OLFR was capable of interacting with the A_2AR in a heterologous system. Although the expression and relevance of OLFRs in general and of Olfr-78 in particular outside the olfactory system has not been widely studied, it was reported three years ago that the Olfr-78 gene product is expressed in the intestine and is involved in the response to colitis (Kotlo et al., 2020). Here, we have used bioluminescence energy transfer (BRET) assays to discover, in a heterologous expression system, a direct interaction between the human version of the A_2AR and the Homo sapiens ortholog of Olfr-78, namely, the olfactory 51E2 receptor (OR51E2). In both cells expressing the A_2AR and cells expressing the A_2A-51E2 receptor heteromer, the effect of two odorants whose specific receptors have yet to be identified, menthol and 1,8-cineole, was studied by measuring intracellular cAMP levels.

Despite constituting a highly populated family of the mammalian proteome, OLFRs are poorly characterized due to the lack of pharmacological tools. The sequences of the genes and the primary structure of OLFRs show that these proteins belong to the GPCR superfamily. Also of note, the heterotrimeric G proteins that mediate olfaction are known as “G_olf” (“olf” for olfactory). G_olf proteins were first discovered in sensory neurons of the olfactory epithelium (Jones and Reed, 1989); shortly thereafter they were identified in the basal ganglia where they were supposed to mediate dopaminergic neurotransmission (Drinnan et al., 1991). Interestingly, the A_2AR is enriched in the basal ganglia where it colocalizes and activates G_olf (Kull et al., 2000). In fact, there are two G proteins to which the A_2AR may couple: G_olf and G_s (Alexander et al., 2019). Given the findings of this study, it is tempting to speculate whether the coupling of A_2AR to G_olf is direct or mediated by heteromerization with G_olf-coupled OLFRs.

Unlike other GPCRs for which the physiological agonist is well known, there is little indication of physiological agonists and/or volatile compounds capable of activating OLFRs. One of the unknown aspects of these receptors is whether they are able to establishing receptor-receptor interactions, as it is the case with members of other GPCR families (Borroto-Escuela et al., 2014). In fact, the relevance of GPCR heteromerization to understand the actual function of OLFRs, under physiological conditions, and their actual potential as targets for therapy is becoming clearer. The results in this paper show that neither menthol nor 1,8-cineole can engage the putative G_olf/s that could be coupled to the OR51E2 (Figure 3). The minimal changes observed by the action of menthol or 1,8-cineole were observed both in cells expressing the A_2AR and in cells expressing the A_2AR and the OR51E2. Therefore, these two odorants are unlikely to directly interact with the OR51E2, unless the receptor is not coupled to G_olf but to another signal transduction pathway.

The main result of this paper is the demonstration that the human A_2AR can interact with the human OR51E2 in a heterologous system. The discovery of this interaction is relevant as it shows, for the first time, that OLFRs can participate in receptor-receptor interactions as has already been shown for rhodopsin-like class A, for class C and for taste GPCRs (Borroto-Escuela et al., 2014). However, understanding the physiological significance of the finding will require further investigation. Our results prove that heteromer formation correlates with a sharp reduction of the G_s-mediated signaling triggered by A_2AR agonists. In some heteromeric settings, activation of one of the receptors is required for crosstalk between receptors and regulation of signal transduction. Within the 51E2/A_2A receptor heteromer, adenosinergic signaling is reduced even in the absence of OLFR activation. This result indicates that there are marked conformational changes in the structure of the A_2AR when the OR51E2 is present. There are examples of regulation of signaling simply through the coexpression of two receptors. One is the heteromer formed by A_2AR and another adenosine receptor, A₃, which couples to G_i. Within the A_2A-A₃ receptor heteromer, the A₃R-mediated signaling is blocked when the two receptors are coexpressed; blockade is detectable both in a heterologous expression system and in primary striatal neurons from mouse brain (Lillo et al., 2020).

If the interaction has physiological relevance, there are two different scenarios that are not mutually exclusive. One involves the cells in which the A_2AR might be expressed in the olfactory system. In the upper respiratory tract, both A_2A and A_2B adenosine receptors have been shown to regulate mucociliary clearance in the nose, where OLFRs are primarily expressed (Hua et al., 2013). The second scenario would consist of a non-olfactory system in which A_2AR and OR51E2 are jointly expressed in a certain cell type. Hans Hatt is known for being the scientist who best understood the relevance of “ectopic” OLFRs, that is, receptors located outside the nose that exert non-olfactory actions. After intense work over decades, he and his colleagues have reviewed research suggesting roles of OLFRs outside the nose and speculating about their potential as therapeutic targets (Maßberg and Hatt, 2018; Lee et al., 2019). Using kidney cells, Hatt and others suggested that the mouse ortholog of OR51E2, Olfr78, might respond to propionate, which would then become an odorless physiological ligand for the receptor (Pluznick et al., 2013). Noteworthy, the Olfr-78 is highly expressed in the gastrointestinal system and its expression is reduced in two experimental models of colitis (Kotlo et al., 2020). In the same study it was proven that OR51E2 is expressed in human colon (Kotlo et al., 2020). Recently, this OLFR has been found in a variety of mouse central nervous system cells, including microglia and vasopressin/oxytocin neurons (Nakashima et al., 2022).

To the best of our knowledge, the OLFRs involved in the sense of odor produced by menthol or 1,8-cineole have not been identified. The experiments reported here with odorants also provided limited but relevant information. First, menthol and 1,8-cineole reduced heteromer formation and/or affected the conformation of the receptor-receptor complex, as detected by a significant decrease in BRET_max (Figure 1). This effect can be direct, on one of the receptors. As discussed above, menthol or 1,8-cineole do not appear to interact with the OR51E2. Conversely, it cannot be completely ruled out that odorants interact with the A_2AR, as both were able to markedly reduce the G_s-mediated response elicited by the A_2AR agonist. Supporting this view is the negative regulation occurring in both cells expressing the A_2AR and in cells coexpressing A_2A and 51E2 receptors. In summary, menthol and 1,8-cineole do not seem to activate the OR51E2, but they block G_s-mediated signaling suggesting that a direct interaction of these odorants with the adenosine receptor is possible. Otherwise, these odorants could interact with the membrane components surrounding the heteromer, thus affecting its overall structure.