Physiologically, odorant receptors are activated by odorant molecules inhaled in the vapor phase. However, most in vitro systems utilize liquid phase odorant stimulation. Here, we present a method that allows real-time in vitro monitoring of odorant receptor activation upon odorant stimulation in vapor phase.
Olfactory perception begins with the interaction of odorants with odorant receptors (OR) expressed by olfactory sensory neurons (OSN). Odor recognition follows a combinatorial coding scheme, where one OR can be activated by a set of odorants and one odorant can activate a combination of ORs. Through such combinatorial coding, organisms can detect and discriminate between a myriad of volatile odor molecules. Thus, an odor at a given concentration can be described by an activation pattern of ORs, which is specific to each odor. In that sense, cracking the mechanisms that the brain uses to perceive odor requires the understanding odorant-OR interactions. This is why the olfaction community is committed to "de-orphanize" these receptors. Conventional in vitro systems used to identify odorant-OR interactions have utilized incubating cell media with odorant, which is distinct from the natural detection of odors via vapor odorants dissolution into nasal mucosa before interacting with ORs. Here, we describe a new method that allows for real-time monitoring of OR activation via vapor-phase odorants. Our method relies on measuring cAMP release by luminescence using the Glosensor assay. It bridges current gaps between in vivo and in vitro approaches and provides a basis for a biomimetic volatile chemical sensor.
The sense of smell allows terrestrial animals to interact with their volatile chemical environment to drive behaviors and emotions. Fundamentally, the odor detection process begins with the very first interaction of odorant molecules with the olfactory system, at the level of odorant receptors (ORs)1. In mammals, ORs are individually expressed in olfactory sensory neurons (OSNs) located in the olfactory epithelium2. They belong to the G-protein coupled receptor (GPCR) family and more precisely to the rhodopsin-like sub-family (also called class A). ORs couple with the stimulatory G protein Golf whose activation leads to cAMP production followed by the opening of cyclic nucleotide gated channels and the generation of action potentials. It is accepted that an odor percept relies on a specific pattern of activated ORs3,4 and therefore odor recognition follows a combinatorial coding scheme, where one OR can be activated by a set of odorants and one odorant can activate a combination of ORs. And through such combinatorial coding, it is postulated that organisms can detect and discriminate between a myriad of volatile odor molecules. One of the keys to understanding how odors are perceived is to understand how and which ORs are activated by a given odor.
In an attempt to elucidate odorant-OR interactions, in vitro functional assays have played an essential role. The identification of agonist odorous ligands for orphan ORs (OR de-orphanization) has been a very active field for the past twenty years, through the use of various in vitro, ex vivo and in vivo functional assays5,6,7,8,9,10,11,12,13,14,15,16,17.
In vitro assay systems are best suited for the detailed functional characterization of ORs, including identifying the functional domains and critical residues of ORs, as well as potential engineering applications. However, further development of valuable in vitro systems for ORs has been a challenge, in part due to difficulty with culturing OSNs and functional expression of ORs in heterologous cells. The first challenge had been to establish protocols that allowed for the cell surface expression of functional ORs in the mapping of odorant-OR interactions. A number of independent groups have utilized various approaches5,6,7,8,9,10,11,12,14,18,19,20. One of the earliest achievements was made by Krautwurst et al. in tagged the N-terminus of ORs with a shortened sequence of rhodopsin (Rho-tag) and observed an improved surface expression in human embryonic kidney (HEK) cells13. Variations made to the tag attached to the OR sequence is still a path explored for improving OR expression and functionality19,21. Saito et al. then identified receptor-transporting protein 1 (RTP1) and RTP2 which facilitate OR trafficking.22 A shorter version of RTP1, called RTP1S, has also been shown to be even more effective than the original protein23. The development of a cell line (Hana3A) which stably expresses Golf, REEP1, RTP1, and RTP2 24, coupled with the use of cyclic adenosine monophosphate (cAMP) reporters has enabled identification of odorant-OR interactions. The mechanism by which the RTP family of proteins promotes cell surface expression of ORs remains to be determined.
One caveat of these established methods is that they rely on odorant stimulation in liquid phase, meaning that odorants are pre-dissolved into a stimulation medium and stimulate cells by replacing the medium. This is very distinct from the physiological conditions where odorant molecules reach the olfactory epithelium in vapor phase and activate ORs by dissolution into the nasal mucosa. To more closely resemble physiologically relevant stimulus exposure, Sanz et al.20 proposed an assay based on vapor stimulation by applying a drop of odorant solution to hang beneath the inner face of a plastic film placed on the top of cell wells. They recorded the calcium responses by monitoring fluorescence intensity. This method was the first to use air-phase odorant stimulation, but it did not allow a large screening of OR activation.
Here, we developed a new method that enables real-time monitoring of in vitro OR activation via vapor phase odorant stimulation by the Glosensor assay (Figure 1). This assay has been used previously in the context of liquid odorant stimulation18,19,25,26,27,28,29,30,31. The monitoring chamber of the luminometer is first equilibrated with vaporized odorant prior to plate reading (Figure 1A). Odorant molecules are then solvated into the buffer, bathing Hana3A cells expressing the OR of interest, RTP1S and the Glosensor proteins (Figure 1B). If the odorant is an agonist of the OR, the OR will switch to an activated conformation and bind the Golf, activating the adenylyl cyclase (AC), and ultimately cause cAMP levels to rise. This rising cAMP will bind to and activate the Glosensor protein to generate luminescence catalyzing luciferin. This luminescence is then recorded by the luminometer and enables OR activation monitoring. This method is of high interest in the context of OR deorphanization as it brings in vitro systems closer to the natural perception of odors.
1. Hana3A Cells Culture
2. Preparation of the Cells for Transfection
3. Plasmid Transfection
4. Substrate Incubation
5. Odorant Stimulation
6. Removal of remaining odorant inside the Luminometer
7. Data Analysis
We screened the response of three mouse ORs, Olfr746, Olfr124 and Olfr1093 using cinnamaldehyde vapor stimulation (Figure 3). Simultaneously, we used an empty vector control (Rho-pCI) to assure that the odorant-induced activities of the tested ORs were specific (Figure 3A). The real-time activation of the ORs upon vapor odorant stimulus was monitored over 20 measurement cycles. The data for each well were first normalized to the empty vector control averaged value for each cycle (Figure 3B). It is important to note that ORs can show variable levels of basal activity in this assay in an OR-dependent manner and it can be important to also normalize their response to this parameter (Figure 3C). The average value of an OR can be divided by its response at t = 0 s, allowing to compare between ORs. Single activation values for each OR can also be computed by calculating the area under the curve (AUC) for each OR by summing all measurement cycle values (Figure 3D).
Additionally, dose-dependent responses can be measured using this method using increasing odorant doses. We present the response of Olfr1377 to acetophenone stimulation recorded following the same procedure (Figure 4). Acetophenone volatility can be evaluated by its vapor pressure, which is equal to 0.44 mm Hg at 25 °C. At a given temperature, an odorant possessing a higher vapor pressure is more volatile than an odorant with a lower vapor pressure. Acetophenone is, in consequence, a moderately volatile odorant and is exposed to the cell pure and diluted at 10-2, 10-4, 10-6 and 10-8. Only the three higher concentrations (pure, 10-2 and 10-4) are able to activate Olfr1377 (Figure 4A). We can also notice that the pure compound stimulation shows a tendency to decrease the OR response during time, likely due to cell toxicity, as it has been shown for eugenol in a recent publication32. Nevertheless, we can observe a typical dose dependency behavior of Olfr1377 (Figure 4B), showing that this method can also be used to determine the EC50 of OR to volatile compound.
Note that when dose-dependent experiments are performed, we do not vacuum clean the luminometer chamber. However, experiments are always performed from low to high, increasing odorant doses to minimize contaminations. Further, since the real concentration of odorant molecules in the cell media is not known, the EC50s obtained by this method are not necessarily comparable to those obtained by liquid stimulation. The EC50 values determined by our method take into account the odorant volatility, the kinetics of odorant dissolution into the medium, the odorant’s solubility, and the affinity between the odorant and the OR, which are closer to the natural perception of odor. For our example of Olfr1377 stimulation by acetophenone, the EC50 value from our vapor dose-response is 161 µM (0.001874 %), which is around 50 fold higher than the liquid stimulation (3.28 µM from Saito et al.6). This difference between liquid and vapor stimulations was also reported by Sanz et al.20 in their vapor stimulation assay where vapor stimulation gave 100 to 1000 fold lower EC50 values than the liquid stimulation.
Figure 1: Principle of real-time monitoring of odorant receptor activation by vapored odorant. (A) The 96-well plate was placed into the already equilibrated with odor (violet cloud) luminometer chamber. (B) Vaporized odorant molecules (violet) at the surface of the cell media buffer dissolved into it to reach the OR cavity, at the Hana3A cell membrane surface. The accessory protein RTP1S (gray) was transfected as well, to favor the cell surface expression of the OR. If the odorant molecule was an OR agonist, the OR switched to an activated state and bound the Golf, triggering the activation of adenylyl cyclase (AC) and the production of cyclic AMP (cAMP; green). The Glosensor protein (luciferase) possesses a binding site for cAMP that, once bound, allows the protein to bind its ligand, the luciferin (yellow). The final complex Glosensor protein/luciferin/cAMP produced the luminescence that was recorded to monitor the OR activation. Please click here to view a larger version of this figure.
Figure 2: Schematic protocols for real-time monitoring of odorant receptor activation by vapored odorant. The odorant (violet) was diluted at the desired concentration in mineral oil (A). The solution was then plated into a new 96-well plate, which was placed into the luminometer chamber for 5 min to equilibrate the volume with vapored odorant before monitoring the transfected 96-well plate (B). The odorant solution was pipetted into each space between the wells of the transfected 96-well plate (see zoom). The plate was then read for 20 cycles in the luminometer chamber equilibrated vapor odorant (C). Please click here to view a larger version of this figure.
Figure 3: Example of normalization that can be performed after data recording. (A) An empty vector (negative) control was inserted in the transfection plan to be able to identify any potential non-OR specific odor activations of cells. It also provided information on the background luminescence of the plate. (B) Each OR response was then normalized by dividing the emission value of each well by the average value of the control vector in each plate. The luminescence values of each OR were then averaged along the measurement cycles. (C) OR responses can also then be normalized to their basal activity by further dividing each cycle value by the average value of the 1st cycle of measurement (t = 0 s). (D) The AUC can be calculated by summing the emission values of each measurement cycle. For B, C and D, error bars represent the standard error of the mean (SEM, n = 3). Please click here to view a larger version of this figure.
Figure 4. Example of dose dependent responses obtained with the method. (A) The response of Olfr1377 to acetophenone was recorded for five different dilutions in mineral oil (100, 10-2, 10-4, 10-6, 10-8), and without odorant (0), and normalized following the normalization protocol shown in Figure 3. (B) The data during the measurement cycles was translated into an AUC value for each dilution. This figure has been modified from Kida et al.32. This figure is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0). Error bars represent the standard error of the mean (SEM, n = 3). Please click here to view a larger version of this figure.
Per 96-well plate | |
MEM | 500 µL |
pGlosensor | 10 ng |
RTP1S | 5 ng |
OR | 75 ng |
Table 1: Mix for transfection. Quantities of plasmids (Glosensor protein, RTP1S and OR) to add to MEM to transfect to one 96-well plate.
The perception of odor is fundamentally dependent on the activation of ORs. Consequently, understanding of their functionality is required to crack the complex mechanisms that the brain use to perceive its volatile chemical environment. However, the understanding of this process has been hampered by the difficulties in establishing a robust method to screen the OR repertoire for functionality against odorants in vitro. Cell surface and heterologous expression of ORs has been partially solved by the creation of tagged receptors13,19 and by the discovery and optimization of the receptor-transporting proteins (RTPs) expressed in OSNs22,23. The first screening studies then appeared, bringing new insights to our understanding of OR function such as their odorant recognition pattern6,7,26,33, activation mechanism34,35, theoretical tridimensional structure36,37, evolution38,39, odor space40,41,42, and implications in odor detection43,44,45,46. We believe that improving the in vitro methods could boost the deciphering of odor coding. As such, we here developed a new functional assay to allow the monitoring of OR activation by vaporized odorant molecules.
The success of this method depends on several critical steps. Although improved in the recent years, heterologous expression of some ORs remains difficult, which may influence OR responsiveness to odorant. The expression of ORs at the cell membrane can be evaluated in a parallel experiment using flow cytometry19,24. Notice that low levels of surface expression can still present robust responses in heterologous cell systems35. Another critical point is to avoid odorant contamination. Given the sensitivity of this assay, odorant molecules from any perfume, food, or previous assay can pollute the experiment by inducing uncontrolled OR responses. This is why we advise to vacuum the chamber of the luminometer for at least 2 h, preferably overnight, before performing an assay with a different odorant. It is also important to consider a potential cytotoxicity of the odorant used in the experiment. A potential decrease of the response of the OR during the 30 min monitoring can show that the odorant itself has a negative effect on the cell health. The odorant cytotoxicity can be evaluated by performing a cell viability assay after exposing the cells to the odorant. To mitigate this problem, it is possible to consider only the first 10 min of the recording time for the analysis of data, cropping the end. We noticed that odorant toxicity mainly occurs when the odorant is tested pure or at very high concentrations. The sensitivity of our assay allows the detection of odorant diluted in mineral oil. The optimal concentration to elicit maximal response varies from one odorant to another, but we observed that a 1% dilution is enough to elicit a saturation of an OR response32. However, some odorant molecules can possess low vapor pressure (and therefore low volatility) and may need some adjustments to be made to the presented protocol. The incubation time of the odorant in the luminometer chamber here is set to 5 min. We assumed that this amount of time is sufficient to equilibrate the chamber with odorant volatile molecules, but low volatility odorants may require longer incubation times.
Apart from these critical points, this method brings many new possibilities to exploring the structure-function relationships of odorant-OR interactions. The real-time monitoring aspect of this method also allows for the understanding of the kinetics of events that occur during an odor perception. As an example, we used the protocol to explore the functionality of a metabolite enzyme, the carboxyl esterase 1d (Ces1d), found to be expressed in the olfactory mucosa of mammals47. This enzyme is known to convert esters to carboxylic acid and alcohol48. The co-transfection of Ces1d showed a modulation of in vitro OR responses to ester compounds,32 demonstrating that this new protocol is efficient in exploring the importance of metabolic enzymes in odorant detection. Furthermore, using this platform to investigate odorant mixtures, and the way odors are presented in settings that are more natural, will enable future study in understanding more complex odorant interactions. Finally, detection of odorant molecules by ORs is also of high interest in the development of an odor sensor. Having shown that our system can detect odorant molecules presented in a vapor phase, this method is a first step in the development process of a miniaturized biosensor.
The authors have nothing to disclose.
This work was supported by grants from NIH (DC014423 and DC016224) and the Defense Advanced Research Project Agency RealNose Project. YF stayed at Duke University with financial support from JSPS Program for Advancing Strategic International Networks to Accelerate the Circulation of Talented Researchers (R2801). We thank Sahar Kaleem for editing of the manuscript.
0.05 % trypsin-EDTA | Gibco | 25300-054 | 0.05% Trypsin – EDTA (1x), phenol red – store at 4°C |
100 mm cell culture dish | BD Falcon | 353003 | 100 mm x 20 mm cell culture dish |
15 mL tube | BD Falcon | 352099 | 17 mm x 120 mm conical tubes |
96-well plate | Corning | 3843 | 96 well, with LE lid white with clear bottom Poly-D-lysine coated Polystyrene |
Amphotericin | Gibco | 15290-018 | Amphotericin B 250 µg/mL – store at 4°C |
centrifuge machine | Jouan | C312 | Centrifuge machine with swinging bucket rotor for 15 mL |
Class II Type A/B3 fumehood | NUAIRE | NU-407-500 | fumehood for cell culturing |
FBS | Gibco | 16000-044 | Fetal Bovine Serum – store at -20°C |
GloSensor cAMP Reagent | Promega | E1290 | GloSensor cAMP Reagent luminescent protein substrate – store at -20°C |
Incubator 37 °C; 5 % CO2 | Fisher Scientific | 11-676-604 | Incubator for cell culturing |
Lipofectamine 2000 reagent | Invitrogen | 11668-019 | Lipofectamine 2000 Reagent 1mg/ml transfection reagent – store at 4°C |
Luminometer POLARstar OPTIMA | BMG LABTECH | discontinued | 96 well plate reader for luminescence |
Mineral oil | Sigma | M8410 | Solvent for odorants – store at room temperature |
Minimum Essential Medium (MEM) | Corning cellgro | 10-010-CV | Minimum Essential Medium Eagle with Earle’s salts & L-glutamine – store at 4°C |
Penicillin/Streptomycin | Sigma Aldrich | P4333 | Penicillin-Streptomycin solution stabilized with 10,000 U of penicillin and 10 mg streptomycin – store at -20°C |
pGlosensor | Promega | E2301 | pGloSensor-22F cAMP luminescent protein plasmid – store at 4°C |
phase contrast microscope | Leica | 090-131.001 | phase contrast microscope with x4, x10, x20 objectives |
RTP1S | H. Matsunami lab | – | 100 ng/µL plasmid – store at 4°C |