2. Towards Human Olfactory DX Realization
Comprehensive deorphanization to determine exactly how all ORs recognize various odors in human olfaction remains challenging for current human OR-expressing cells. First, when the Gα protein in the cells used for OR expression is different from the OSN, the odor molecular response of some ORs may be altered
[6]. Second, single-nucleotide polymorphism (SNP) is found in many human OR genes, which sometimes alters the odor molecular response of OR
[7][8][9]. Third, almost all odor sensors using OR-expressing cells measure the OR response to odors in the liquid phase. In human olfaction, the ORs of OSNs respond to odors in the gas phase via a very small amount of nasal discharge, resulting in the detection threshold being much lower (typically 100- to 1000-fold) compared to the OR-based sensors. Currently, researchers cannot correctly explain this large difference in OR detection thresholds for odors. Indeed, these three issues are major obstacles to unraveling the scientific proposition of how the entire human ORs discriminates various odors with high sensitivity. However, for realizing human olfactory DX (especially in the processes of recording, saving and transmitting to remote locations), an odor sensor does not necessary perfectly reproduce the way human olfaction perceives odors. It is sufficient to always measure odors with a representative OR set (i.e., 388 ORs) under fixed conditions and define them in a 388-dimensional odor matrix.
3. Preliminary Odor Reconstitution
So far, the original odor has been reconstituted by analyzing its components by GC-MS and mixing the major odor molecules in the same proportions. However, odor molecules that cannot be detected by GC-MS often contribute significantly to odor quality, so perfumers had to spend a lot of time for formulation with many odor molecules by trial and error. For the achievement of the reconstitution part of the human olfactory DX, this process would be virtually impossible without preparing in advance all the approximately 400,000 types of odor molecules that exist in the world
[10]. In other words, it is imperative to reduce the number of odor molecules used for the odor reconstitution of human olfactory DX. There is an attempt to reduce the number of odor molecules by GC-MS analysis of essential oils to sort out odor molecules that exhibit the same odor quality and reconstitute the odor quality of any given essential oil
[11], but this method is applicable only to specific categories of odors (e.g., essential oils) and cannot be used to reconstitute a wide range of all odors.
Thus, by taking advantage of human OR sensors, researchers preliminarily reconstituted the odors of dried bonito flakes, rose essential oil, lavender essential oil, and vanilla flavor. Even though each sample was found to contain numerous types of odor molecules (>500, >50, >50, and >50, respectively) by GC-MS, 10 human ORs responded significantly to dried bonito flakes, 4 ORs to rose essential oil, 3 ORs to lavender essential oil, and 7 ORs to vanilla flavor. In addition, several of the responsive ORs were common. Therefore, odor molecules that selectively stimulate these ORs were selected from the odor molecule library and mixed in a ratio that could reproduce the odor matrix. Finally, the odors of dried bonito flakes, rose essential oil, lavender essential oil, and vanilla flavor were reconstituted with seven, four, three, and five odor molecules that were not included in the respective original samples, for which the odor quality was determined by sensory tests on a scale of several dozen people to be nearly reproduced
[12]. This result strongly suggested that no matter how complex an odor is, if an odor matrix is obtained by a human OR sensor and mixed with odor molecules that stimulate each response OR with pinpoint accuracy, it is possible to reconstitute the odor with far fewer types of odor molecules (
Figure 3).
Figure 3. Conceptual diagram of odor reconstitution. First, a lot of basically safe odor molecules (e.g., natural flavors and fragrances, synthetic fragrances, food additives) are measured by a human OR sensor to build an “odor matrix database”. Incorporate odor molecules that selectively stimulate limited numbers of ORs in the database, exclude odor molecules showing exceptional activities described herein from the database, and generate an “odor molecule library for reconstitution”. The minimum group of odor molecules necessary to reproduce this target odor matrix (including changes in each human OR response over time) is selected from the library. As a result, in fragrance and flavor development, aromas and flavors with difficult-to-obtain or complex compositions can be reconstituted with odor molecules of easy-to-obtain or simple compositions. Furthermore, people around the world will be able to share the same odor in real time, which will revolutionize the expression method by introducing olfactory information to the existing visual and entertainment industries, as well as the metaverse in XR, which has until now relied solely on visual and auditory information.
In general, the OSN response to complex odors is essentially a linear relationship with the OSN response of each odor molecule
[13]. Nevertheless, there is also intermolecular interference between odor molecules when multiple odor molecules bind to an OR
[14]. This competitive binding result in a nonlinear relationship between the OR responses of multiple odor molecules
[15]. These are also explained by a model in which odor molecules bind to allosteric pockets on the OR. In addition, there are inverse agonists in which the same odor molecules promote or inhibit OR responses in a concentration-dependent manner
[16]. Human OR sensors capable of measuring even complex odors can detect odor molecules eliciting these exceptional OR responses. Only odor molecules that selectively stimulate limited number of ORs without exceptional activities in ORs should be used to reconstitute odors that have almost the same odor quality as any odor.
4. Future Improvements of Human OR Sensor
The main issue with human OR sensors is that their detection threshold for odor molecules is higher than that of human olfaction. To ameliorate this issue, one should consider suppressing mechanisms that inhibit intracellular cAMP generation (PDE (phosphodiesterase)) and intracellular Ca
2+ mobilization (CaMKII (Ca
2+-calmodulin-dependent protein kinase II), CaM (calmodulin), NCX (Na
+/Ca
2+ exchanger), PMCA (plasma membrane calcium pump))
[17][18][19], utilize extracellular OBPs that promote odor molecule presentation to the ORs
[20], utilize intracellular OMPs (olfactory marker proteins) that increase intracellular cAMP levels
[21][22], and utilize GRK2 (G-protein-coupled receptor kinase 2) inhibitors that suppress the binding of β-arrestins that promote OR internalization
[23].
Recently, as there have been some reports that conventional ORs are not enough to detect all odors, human OR sensors should also be equipped with OSN-specific TAARs (trace amine-associated receptors; TAAR1, TAAR2, TAAR5, TAAR6, TAAR8, TAAR9)
[24][25][26] to detect amine compounds and TRPs (transient receptor potential channels; 6 subfamily, 27 TRPs)
[27][28] to detect odor molecules such as capsaicin and menthol. Since the second messengers of TAARs and TRPs are Ca
2+ ions, the degree of activation of both TAARs and TRPs can be measured by simply diverting the cells used in the OR-expressing cells of human OR sensors.
The sensitivity of human olfaction is known to decline with age
[29], but it is not a problem for daily life. This fact suggests that we should question if all ORs are necessary or not. Recently, not all OR mRNAs have been found in human olfactory epithelium by RNAseq analysis
[30]. All OR mRNAs were not equally expressed and 26 major OR mRNAs accounted for more than 90% of the total OR mRNAs
[31]. These facts indicate that the number of ORs in current human OR sensors may be overrepresented.