A looongie but goodie. I know I didn’t have the stamina to read through it when Jasmin originally assigned it to me as essential reading, but nevertheless it’s really really good!
Check the link here: Smell - Neurobiology of Sensation and Reward - NCBI Bookshelf
All credit given where credit is due. We thank the authors for their exhaustive work on the subject and post it here in hopes to raise awareness for this work. --JSLV
Chapter 5 Smell
Jay A. Gottfried and Donald A. Wilson.
To speak generally then, things that have been cooked, delicate things, and things which are least of an earthy nature have a good odour, (odour being a matter of exhalation), and it is obvious that those of an opposite character have an evil odour.
** Theophrastus c. 300 BC **
Go to:
5.1 INTRODUCTION
Smell is arguably the most ancient sense. Consider for example a bacterial prokaryote, monocellular, anuclear, flagella-rotating, as it tumbled through the Archaean (Archaeozoic) seas roughly 2?3 billion years ago. For such an organism, chemotaxis?the ability to redirect its movements in the presence of chemical gradients?was synonymous with survival. This most rudimentary sense of smell was critical for finding chemoattractants like organic nutrients (Figure 5.1a; Adler 1969) and evading chemorepellents like excreted waste (Figure 5.1b; Tso and Adler 1974; Adler 1978). In this manner, the sense of smell was rooted at the earliest evolutionary stages with the machinery of affective processing, to the extent that chemical sensing was indistinguishable from the sensing of biological imperatives.
http://www.ncbi.nlm.nih.gov/books/NBK92786/bin/ch5_f01.gifFIGURE 5.1
Olfactory hedonics in bacteria. (a) Chemosensory attraction. E. coli bacteria are using chemo-reception to migrate towards the open end of a capillary tube that has been filled with aspartate, an amino acid with nutritive value (Adler 1969). (Reprinted (more…)
Interestingly, the key features of bacterial chemotaxis presage many of the same principles guiding olfactory processes in higher animals (Koshland 1980; Kleene 1986; Baker, Wolanin, and Stock 2006). First, the bacterial chemoreceptor complex is highly localized into clusters at the cell poles (Figure 5.2; Maddock and Shapiro 1993). This anatomical polarity effectively segregates sensory detection into spatially discrete regions of the organism, an arrangement that reaches its full eukaryotic expression in the form of antennae, noses, and olfactory epithelia.
http://www.ncbi.nlm.nih.gov/books/NBK92786/bin/ch5_f02.gifFIGURE 5.2
A primordial nose. Immunoelectron microscopy shows that the chemoreceptor complex of E. coli bacteria is clustered at a discrete polar location of the cell’s inner membrane (arrow) (Maddock and Shapiro 1993). This cellular sequestration of chemoreceptive (more…)
Second, bacteria adapt to their chemical environment. Directional movement through a chemical gradient relies on concentration changes, rather than on absolute concentration per se, so that chemotaxis ceases once the bacterium finds itself in an isotropic solution. Reduced responsiveness to an unvarying chemical background closely parallels sensory adaptation in the mammalian olfactory system, which, as described later in this chapter, is an important mechanism underlying odor discrimination.
Third, bacteria learn from experience. Wild-type Escherichia coli show exuberant chemotaxis toward maltose when grown in a medium containing maltose, but respond minimally to the same attractant when grown in a maltose-free medium (Adler, Hazelbauer, and Dahl 1973). As a basic example of sensory-specific plasticity, this experience-dependent gain in chemical sensitivity is also instantiated in vertebrate and invertebrate olfaction, whereby learning and experience robustly modify how smells are perceived. Remarkably, human neonates respond selectively to the odor of familiar (vs. unfamiliar) amniotic fluid (Schaal, Marlier, and Soussignan 1998) and they prefer food flavors that had been experienced prenatally by way of maternal consumption (Mennella, Jagnow, and Beauchamp 2001). That human fetuses ?grown? in carrot-containing amniotic fluid are attracted to carrot flavor postnatally (Mennella, Jagnow, and Beauchamp 2001) curiously mirrors the growth-dependent inducibility of chemotaxis in prokaryote species. The role of experience in modulating odor coding is addressed in detail in Section 5.4.
Fourth, bacterial chemotaxis is synthetic. In other words, the decision to swim or tumble is based on the integration of chemical inputs. Simultaneous exposure to two different chemicals often elicits chemotactic behaviors that cannot be simply predicted from the mere sum of the individual components (Rubik and Koshland 1978). Even in Salmonella typhimurium, which contains a humble armament of five chemoreceptor genes (Wadhams and Armitage 2004; Hazelbauer, Falke, and Parkinson 2008), response potentiation (supra-additivity) to a combination of chemoattractants can be observed. The idea that behavioral non-linearities emerge in the presence of chemical mixtures has interesting implications for odor quality perception, multisensory integration, and flavor processing in rodents and humans, topics that are examined in Section 5.4, and throughout the sensory chapters of this book.
Fifth, and last, bacteria have a memory for smells, albeit on an extremely short timescale. The very fact that a 2-?m bacterium successfully swims through a chemical gradient to pinpoint a food source suggests that it must be able to retain a memory for what just came before (Macnab and Koshland 1972). Without the means to compare chemical concentrations from adjacent time frames (i.e., temporal integration), the bacterium would soon founder. This feature is not meant to imply that the olfactory systems of ?higher? animals have adopted the same molecular cellular strategies to remember smells. Rather the point is that already from a very primordial moment in animal evolution, the same functional strategies were in place to maximize contact with behaviorally beneficial, or rewarding, chemical stimuli.
It is thus perhaps no accident that the human sense of smell remains so closely aligned with emotional processes, anatomically, physiologically, and psychologically. To our predecessors the bacteria, chemical sensation was fundamentally intertwined with the acquisition of rewards and the avoidance of threats. The operating principles by which bacteria achieved better living through chemistry?anatomical segregation of chemical detectors, response adaptation to static environments, sensory plasticity, signal integration, and chemical memory?continue to define and constrain the ways that the olfactory systems of higher organisms, including humans, contend with the odor landscape.
An olfactory enthusiast might plausibly go so far as to state that all biological subsystems are essentially a spin-off of the chemical detector apparatus. Since Buck and Axel first identified a large multigene family of G-protein-coupled receptor genes on rat olfactory sensory neurons (OSNs) (Buck and Axel 1991), it has become apparent that the G-protein-coupled conformation of the olfactory receptor shares much in common with receptors that bind a great many other critical biological ligands, including neurotransmitters, neuromodulators, tastants, light, hormones, growth factors, chemokines, cell adhesion molecules, and chemotactic peptides (Dryer 2000; Fredriksson et al. 2003). Laurence Dryer (2000) has nicely captured the idea that olfaction is a virtual microcosm of the human nervous system:This functional and structural diversity [within the olfactory receptor gene family] is not surprising if one bears in mind that there is no essential difference between neurotransmitter binding and odorant detection. Whether they are expressed at the synapse or in sensory cells, chemoreceptors serve the same function, that is, the chemical detection of a ligand carried by extracellular or external space.
Go to:
5.2 THE ODOR STIMULUS SHAPES THE OLFACTORY SYSTEM
If a sensory system is to extract meaning from a sensory stimulus, what is the natural form of that stimulus in the real-world environment that a brain is likely to confront? For the olfactory system, a general answer to this question would be to state that olfactory percepts arise from physical and chemical properties of odor molecules, just as visual percepts arise from wavelengths of light.li This is correct, though troublingly vague, and indeed much of our current scientific understanding of olfactory systems in insects, rodents, and primates stems from research using monomolecular odorants (L-carvone, amyl acetate, etc.) with well-defined physical features, helping to ensure experimental control over the sensory input and to promote experimental replicability across different laboratories and species.[/li]But it is critical to consider that the ecological context of an odor source defines what type of odor signal will be transmitted to a receiver. (See Dusenbery 1992 for an extensive and insightful overview of the sensory ecology of chemical signals, as well as thermal, light, sound, and mechanical signals.) The important biological implication is that, as an outcome of natural selection, how an odor is realistically encountered in the environment will ultimately shape how an olfactory nervous system evolves to optimize feature detection (Gottfried 2009).
The first point to keep in mind is that most real-world odors are not monomolecular compounds courtesy of the Sigma-Aldrich Flavors & Fragrances catalog. Single-molecule, single-meaning odorants are extremely rare in nature. The majority of emitted odors making contact with a chemical biosensor are mixtures of tens to hundreds of different odorant molecules. For example, microwave-popped popcorn contains at least 56 distinct volatile compounds (Walradt, Lindsay, and Libbey 1970), pressure-cooked pork liver liberates 179 volatile components (Mussinan and Walradt 1974), and the smell of chocolate is a blend of over 600 chemical constituents (Counet et al. 2002). Even the release of a potent moth sex attractant in the wild is often packaged together with many additional volatile components (Linn, Campbell, and Roelofs 1987; Dusenbery 1992). The compound nature of most environmental odors suggests that an olfactory system with the ability to encode information about the overall configuration of an emitted chemical stimulus would have a sensory discriminatory advantage. This issue is discussed at length in the final section of this chapter.
The second point is that odors are highly subject to the whims of their natural settings. The same odor may be experienced in the context of different background smells, potentially distorting how the odor is perceived. The direction or strength of the wind may make all the difference to an organism using airborne odor cues to locate its lunch. The length of time that has elapsed since the origination of an odor at its source will determine whether that odor cue still contains reliable predictive information, for example, if the odor source has run away (in the case of an animal), washed away (in the case of rain), or chemically transformed (in the case of a decomposing fruit). Each of these ecological scenarios means that an animal may have partial or corrupted access to an olfactory signal. Therefore, an olfactory system that reconstructs odor meaning?i.e., species, gender, sexual receptive state, social dominance, edibility?from sensory fragments would provide maximal adaptability to an organism.
The last point is that earthly transmission of an odor message markedly differs from the transmission of the other distance senses. In vision and audition, spatial patterns in the external world provide important sources of information that the brain can sample, extract, and codify in the form of spatial patterns of peripheral receptor activation. Critically, it is the spatial fidelity of visual and auditory stimuli?all the way from point source to receiver?that makes possible the use of spatial ?codes? within these sensory modalities. However, in contrast to sights and sounds, smells diffuse rather less predictably and less quickly from their sources. As mentioned above, air currents rapidly shear an odor message away from its spatial site of origin. Indeed, this carries certain advantages: olfactory signals can defy mundane physical obstacles, such as trees, bushes, hilltops, that would quickly extinguish a visual signal unable to bend around these foes. Thus, with the possible exception of localizing smells to one nostril or the other (Wilson 1997; Porter et al. 2007), spatial information is not a feature with which the olfactory system has evolved to contend. It is true that, as described in the next section, odorant-specific spatial patterns of neural activity are evoked within the olfactory bulb, but these patterns probably have more to do with encoding of odor identity, as opposed to odor space per se. Actually the spatial organization of receptor projections and bulb activity patterns are lost en route to olfactory cortex, and at this level temporal codes may be more critical for information processing (Haberly 2001).
In concluding this section, a proviso is in order. Tremendous advances in our neuroscientific understanding of the visual and auditory systems have been possible because the fundamental physical ?primitives? underlying visual and auditory processing have been precisely identified. That is, receptive fields of color in vision or pitch in audition can be mapped precisely by varying the wavelength of light or sound along a linear dimension.li However, a precise characterization of odorant receptive fields remains elusive, given that the specific metric along which olfactory space is measured is not clear. For example, the physicochemical dimensions along which piperonyl acetate or 4-hydroxy-3-methoxy-benzaldehyde lie are still rather obscure; of equal perplexity is the lack of consensus regarding the perceptual dimensions along which the smell of cherry or vanilla lie. In spite of these uncertainties, much contemporary olfactory research, with modest successes, has placed an emphasis on hydrocarbon chain length and chemical functional group as two putative dimensions of odorant space, which will become evident from the following discussion.[/li]
Go to:
5.3 THE OLFACTORY SYSTEM: ANATOMY AND PHYSIOLOGY
The olfactory system shares many functional properties with its counterpart sensory systems, yet is organizationally unique among mammalian sensory pathways. Broadly speaking, the functional anatomy of all sensory systems begins with an analysis of complex information streams into component features, local circuit enhancement of contrasts between features, and subsequent synthesis of those features into perceptual objects. Especially for complex organisms, the neurobiological assembly of sensory inputs into sensory ?outputs? (i.e., perceptually meaningful objects) is optimized when sensory systems have an opportunity to integrate ascending and descending information flow, contextual and state-dependent modulatory effects, and multisensory interactions. Current evidence suggests that olfaction shares these functional properties with each of the other sensory systems.
A canonical mammalian sensory nervous system includes peripheral receptors, initial local circuit processing (e.g., retinal bipolar cells, auditory brainstem nuclei, dorsal root ganglia), a thalamic nucleus processing center, and a primary sensory cortex with reciprocal corticothalamic connections. Higher-order cortical areas may be specialized for specific, behaviorally relevant information content within one sensory modality, for example, faces, speech, or pain. Many sensory systems include parallel processing streams, wherein different aspects of the same stimulus are simultaneously captured and processed through anatomically distinct subsystems. Finally, it is becoming apparent that the opportunity for multimodal sensory convergence occurs at least as early as primary sensory cortex, an area traditionally considered to be a unimodal processing stream (e.g., Lakatos et al. 2007). Again, olfaction shares many of these anatomical characteristics with the other sensory systems, but there are important differences, as outlined below.
The anatomy of the initial stages of the olfactory pathway suggests a hierarchical, combinatorial processing of odorants, with initial feature extraction in the periphery and subsequent convergence and blending of features through subsequent cortical stages. The following sections trace the encoding and transfiguration of an odor message as it moves through the olfactory system, beginning with the nose, olfactory epithelium, and OSNs, then pausing at the olfactory bulb, glomeruli, and mitral and tufted cells (the second-order neurons), and finishing at the olfactory cortex, including piriform cortex, and higher-order brain regions. The reader is referred to many comprehensive reviews (Price 1990; Carmichael, Clugnet, and Price 1994; Shipley and Ennis 1996; Haberly 1998; Cleland, Linster, and Doty 2003; Wilson, Sullivan, and Doty 2003; Zelano and Sobel 2005; Gottfried 2006; Gottfried, Small, and Zald 2006a) for in-depth discussions of olfactory system anatomy and physiology across different species.
5.3.1 Peripheral Considerations
The mammalian olfactory system begins at the olfactory receptor sheet, where it lines the deeper recesses along the nasal passage. The sheltered location of the olfactory epithelium means that stimulus sampling is tied to airflow, if not absolutely dependent on it: in the ?orthonasal? direction, inspiration draws in odorous molecules (odorants) from the outside environment and across the receptor layer; in the ?retronasal? direction, expiration forces out odorant molecules from the inside environment (typically, food-based substances residing in the oral cavity) across the receptor layer. Thus, as Rozin (1982) has argued, olfaction is the only dual sensory system, operating both as a distance sense and as a contact sense.
The route of stimulus access can affect stimulus perception (e.g., King et al. 2006) and central brain circuit activation (Small et al. 2005). Detection thresholds and perceptual qualities for a given odorant are both dependent on the pathway by which that stimulus is delivered. These route-specific differences may arise because of variations in stimulus access to different zones within the olfactory receptor sheet (Scott and Brierley 1999; Scott et al. 2007) or because of gas chromatographic effects as the odorant molecules diffuse through the mucus overlying the sensory receptor cilia (Mozell 1991). Odor pleasantness, irritancy, and intensity, as well as state effects related to arousal and attention, can influence sniffing rate and volume (Mainland and Sobel 2006; Wesson et al. 2008), with further potential for modulating olfactory perception. Odor sampling behavior itself has an impact on olfactory coding: high-frequency sniffing vs. low-frequency passive respiration has been shown to filter out static background odors, enhancing an animal’s ability to detect changes in the olfactory landscape (Verhagen et al. 2007).
5.3.2 Contact: Olfactory Sensory Neurons
Initial contact between an odor stimulus and the olfactory system occurs at the receptor endings of OSNs. Each OSN expresses one specific G-protein-coupled receptor gene (Buck and Axel 1991), of which there are approximately 1000 in rodents, 350 in humans, and 100 in fish (Firestein 2001). These receptors are studded along 20?30 long dendritic cilia that poke into the overlying mucus of the olfactory epithelium. Interestingly, OSNs expressing the same receptor gene do not segregate together within the epithelium, but instead are widely dispersed throughout different epithelial zones (Ressler, Sullivan, and Buck 1993). The lack of a systematic topographical organization within the olfactory receptor sheet already marks a fundamental difference between olfaction and the other distance senses (vision and audition).
The G-protein-coupled receptors may play a role not only in stimulus binding on the receptor cilia, but also in helping target the sensory neuron axons to discrete locations in the olfactory bulb called glomeruli (Vassalli et al. 2002). It is important to bear in mind that olfactory receptors do not bind and transduce ?smells? (e.g., banana, rose), or even entire molecules (e.g., isoamyl acetate, phenyl ethyl alcohol), but sub-molecular moieties (Araneda, Kini, and Firestein 2000). Furthermore, an individual odorant molecule can serve as a full agonist of an olfactory receptor, a partial agonist, or even an antagonist (Gentilcore and Derby 1998). Thus, some interaction between odorant molecules occurs at the receptor sheet.
Once bound to receptor proteins, odorants evoke neural activity in OSNs, which transmit that activity into the central nervous system. Like neurons in other sensory systems, OSNs and their central targets respond to only a subset of all possible odorants (Duchamp-Viret, Chaput, and Duchamp 1999; Malnic et al. 1999), which is termed their molecular receptive range or, in the vernacular of sensory neuroscience, their odorant receptive field. The receptor ligand binding site on the OSN appears to be selective for molecular constituents of an odorant. In consequence, a given monomolecular odorant may bind to multiple different odorant receptors and a given receptor may bind multiple different odorants. Therefore, a natural scent, potentially composed of many different monomolecular odorants, will activate a unique combination of OSNs. Despite these guiding principles, recent work suggests that, at least in Drosophila, some individual receptor neurons may exhibit a highly narrow tuning profile for biologically salient odors such as pheromones (Schlief and Wilson 2007).