CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector Dispatch R157

Sensory Neuroscience: Taste the classic flavor model to include convergence in anterior insular Responses in Primary Olfactory cortex (Figure 1B). It also suggested to Maier et al. [5] Cortex that taste inputs might influence olfactory coding as early as the primary olfactory cortex. To test this hypothesis A new electrophysiological study in rodents demonstrates that taste–odor they recorded activity from single convergence occurs in posterior piriform olfactory cortex and calls for neurons in posterior olfactory cortex a reformulation of classic models of the central representation of flavor. of awake rats while presenting basic taste solutions directly to the tongue. Dana M. Small1,2,3, To be fair, even within classical This particular region of piriform cortex Maria G. Veldhuizen1,3, models, the chemical senses were receives dense projections from the and Barry Green1,4 always considered somewhat of insula [14], as well as from the a special case, as it was clear that the amygdala, prefrontal entorhinal and Fifteen years ago, the pathway from synaptic distance separating taste and perirhinal cortex, prompting Johnson sensation to cognition was elegantly smell was shorter than for audition and et al. [14] to suggest that it functions laid out as a series of hierarchical and vision [1]. Whereas visual information like association cortex in other sensory parallel circuits [1]. Sensory maintained fidelity through four systems. Consistent with this proposal, information entered the cortex at synapses, the primary gustatory and approximately half of the 221 neurons primary unimodal sensory zones and olfactory cortices were separated by recorded showed a significant was further elaborated by higher-order a single synapse in classic diagrams modulation of their firing rate by taste unimodal zones before eventually of flavor processing [6] (Figure 1A). solutions. Some neurons responded converging with input from other This early convergence seemed selectively to bitter stimuli while others modalities in multisensory zones reasonable given the intimate responded selectively to sweet stimuli. where perceptual objects were created relationship between taste and smell Modulation of firing rate was observed and then ‘‘woven into the fabric of in producing flavor; however, the more frequently for the unpalatable cognition’’[1]. A critical feature of this idea that taste could influence olfactory tastes of citric acid and quinine. conceptually appealing view was the processing in primary olfactory Next, Maier et al. [5] set out to verify maintenance of sensory fidelity — and cortex or that odors could influence that the taste responses were in fact in service of fidelity was the idea gustatory processing in primary of gustatory origin. This is an issue that no connectivity existed among gustatory cortex was considered because mixing pure gustatory sensory cortical areas. Over the past just as unlikely as for the auditory stimuli — such as citric acid and 15 years this well-established view and visual modalities [6]. sucrose — can produce volatile of brain organization has dissolved But then, just as happened with the compounds that activate olfactory in the face of discoveries of clear auditory and visual systems, evidence receptors to produce odor sensations. sensory–sensory connectivity and began to mount that called this old view As a first step, they tested the influence direct influences of one modality into question. It became increasingly of a topical anesthetic applied to the upon primary sensory cortex of another clear that taste–odor integration occurs tongue on the taste-evoked piriform [2]. For example, we now know earlier. Human neuroimaging studies responses. As is observed in gustatory that there are direct connections frequently reported that odors cortex [13], the anesthetic significantly from the auditory core and parabelt activated a region of insular cortex that reduced the taste response so that regions to visual areas V1 and V2 in looked a lot like primary taste cortex, firing rates were indistinguishable from the monkey [3], and in humans there a contention subsequently proved by baseline. In contrast, taste responses is evidence of sounds influencing a meta-analysis of gustatory and were still present following deciliation visual information processing in V1 [4]. olfactory studies [7]. Supra-additive of the nasal epithelium, which has been These and similar findings directly responses to taste-odor mixtures, shown to effectively abolish olfactory challenge the notion that primary which are a hallmark of multisensory responses. Finally, unlike olfactory sensory cortices are strictly unimodal integration [8], were then reported responsive neurons in the piriform [2]. In a recent paper, Maier et al. [5] in the anterior ventral insula, strongly cortex, taste-responsive neuronal take this re-conceptualization suggesting that flavor perception firing was found to be unrelated to a step further by showing that begins to emerge in the insula [9]. respiration. These results indicate that a significant portion of ‘primary’ Patients with insular lesions were gustatory stimulation is sufficient to posterior piriform olfactory cortex found to display both gustatory and drive a subset of posterior piriform neurons responds selectively to taste. olfactory sensory deficits [10,11], and cortex neurons. The finding is especially intriguing a role for the insular cortex in coding The next critical question was because it goes beyond demonstrating the ‘taste-like’ properties of odors determining whether taste-odor that gustatory information merely [10,12] and in olfactory learning was convergence occurred within single influences olfactory responses in described, with inactivation of taste neurons. Another 41 neurons were piriform cortex, further suggesting cortex blocking the ability of rodents tested for responses to tastes and that the gustatory system might to learn to use an odor cue to guide odors. Thirty-two percent responded have its own real estate in primary food preference [13]. Collectively this to odors only, 17% to tastes and olfactory cortex. work suggested a revision of odors, and 22% only to taste. Thus, Current Biology Vol 23 No 4 R158

et al. [12] examined responses to sweet ABVPMpc Insula/ VPMpc Insula/ NST thalamus operculum NST thalamus operculum odors and sweet tastes and found Amygdala TASTE Amygdala TASTE that while insular cortex showed overlapping responses to odors and tastes, the piriform cortex responded very selectively to odors. One SMELL Orbitofrontal SMELL possibility is that piriform taste cortex Orbitofrontal Olfactory Piriform Olfactory Piriform cortex representation is biased towards bulb cortex bulb cortex Current Biology unpalatable tastes; another is that functional imaging (fMRI) taste studies Figure 1. Schematic diagram of taste and olfactory pathways and their convergence. may sometimes report piriform (A) The classical hierarchical view. Information from taste receptors in the mouth is conveyed responses as arising from the to the nucleus of the solitary tract (NST), which then projects to the ventral posterior medial amygdala, which is immediately nucleus (VPMpc) of the thalamus, which in turn projects to primary gustatory cortex then adjacent to piriform cortex. An informal insula and overlying operculum. Information from olfactory receptors in the nasal cavity is re-perusal of our own data certainly conveyed to the olfactory bulb, and from there to the primary olfactory cortex in piriform. suggests that this is a reasonable Both gustatory neurons in insula/operculum and olfactory neurons in piriform cortex project possibility. A third possibility is to amygdala and orbitofrontal cortex, the first neural level at which they converge. (B) The emerging view representing early, but oft ignored research showing the existence of olfactory inter-species differences, which are neurons in the NST, thalamus, and insula and the new evidence for taste odor-convergence in very pronounced in the gustatory the insula (reviewed in [20]) and the existence of taste neurons in the piriform cortex provided neuroaxis between rodents and by Maier and colleagues. Dotted blue lines indicate that olfactory responses likely arise from primates. cortico-fugal projections from insula/operculum. Dotted orange lines suggest possible path- Whatever the case, it is our hope ways for taste to reach piriform cortex. Contrasting with the classic view, the emerging model that the report by Maier et al. [5] strikes clearly suggests highly integrated systems. (Panel A adapted with permission from [6].) the final blow against the dogma that gustatory and olfactory information convergence occurred and the They also open the door to a whole new must reach the orbitofrontal cortex possibility of unimodal gustatory set of questions about the nature of before interaction occurs. neurons in primary olfactory cortex convergence. was established. Determining whether What is the source of these piriform References these taste-responsive neurons are taste responses? Anatomical 1. Mesulam, M.M. (1998). From sensation to cognition. Brain 121, 1013–1025. truly unimodal will be of considerable projections to posterior piriform have 2. Driver, J., and Noesselt, T. (2008). Multisensory interest. Studies of multisensory been identified from the orbitofrontal interplay reveals crossmodal influences on ‘sensory-specific’ brain regions, neural interactions with odors have shown cortex, insula and amygdala; however, responses, and judgments. Neuron 57, 11–23. that the response in chemosensory no one has truly probed olfactory 3. Falchier, A., Clavagnier, S., Barone, P., and cortex depends upon congruence cortex for its taste representation Kennedy, H. (2002). Anatomical evidence of multimodal integration in primate striate cortex. [9,15]. Psychophysical studies have (Figure 1B). Maier et al. [5] favor the J. Neurosci. 22, 5749–5759. also shown that the ability of taste to amygdala source since the firing 4. Watkins, S., Shams, L., Tanaka, S., Haynes, J.D., and Rees, G. (2006). Sound alters enhance retronasal odor [16] and patterns of piriform taste responses activity in human V1 in association with illusory to cause retronasal odors to be referred are more similar to the firing patterns visual perception. Neuroimage 31, 1247–1256. to the mouth [17] are both strongly of amygdala neurons than insular taste 5. Maier, J.X., Wachowiak, M., and Katz, D.B. (2012). Chemosensory convergence on dependent on the taste-odor neurons. primary olfactory cortex. J. Neurosci. 32, congruence. These findings raise the What is the role of these taste 17037–17047. 6. Rolls, E.T. (2001). The rules of formation of the possibility that all or a majority of neurons? As alluded to above, flavor olfactory representations found in the taste-responsive neurons in olfactory perception is the obvious possibility; orbitofrontal cortex olfactory areas in primates. cortex may in fact be bimodal, and but some odors are perceived to Chem. Senses 26, 595–604. 7. Verhagen, J.V., and Engelen, L. (2006). The that the repertoire of odorants tested have taste-like qualities even in the neurocognitive bases of human multimodal by Maier et al. [5] was simply not absence of taste. For example, vanilla food perception: sensory integration. Neurosci. Biobehav. Rev. 30, 613–650. extensive enough to capture bimodal is often described as sweet and lemon 8. Stein, B.E. (1998). Neural mechanisms for responses to congruent tastes and odor as sour, yet no gustatory synthesizing sensory information and odors. For example, unpalatable tastes receptors are activated, causing producing adaptive behaviors. Exp. Brain Res. 123, 124–135. tended to drive the piriform taste people to confuse olfactory stimulation 9. Small, D.M., Voss, J., Mak, Y.E., responses, whereas the odors were for taste [18]. Could these taste Simmons, K.B., Parrish, T.B., and Gitelman, D.R. (2004). Experience-dependent all palatable sweet food odors neurons play a role in shaping the neural integration of taste and smell in the (for example, cherry). taste-like qualities of odors as has been human brain. J. Neurophysiol. 92, 1892–1903. Even if these neurons turn out to be suggested for insular odor neurons? 10. Stevenson, R.J., Miller, L.A., and Thayer, Z.C. (2008). Impairments in the perception of bimodal, however, the intriguing Alternatively, they may be important for odor-induced tastes and their relationship to findings of Maier et al. [5] clearly show integrating taste and odor cues in the impairments in taste perception. J. Exp. Psychol. Hum. Percept. Perform. 34, that the gustatory and olfactory formation of flavor aversions, in which 1183–1197. systems converge earlier than in the the amygdala has been directly 11. Mak, Y.E., Simmons, K.B., Gitelman, D.R., and orbitofrontal cortex, as originally implicated [19]. Small, D.M. (2005). Taste and olfactory intensity perception changes following left insular posited (Figure 1A). Rather, they point A final mystery raised by these data stroke. Behav. Neurosci. 119, 1693–1700. to a more integrated model rife with the is why it is that neuroimaging studies 12. Veldhuizen, M.G., Nachtigal, D., Teulings, L., Gitelman, D.R., and Small, D.M. (2010). The possibility of interaction and recurrent of gustation fail to report piriform insular taste cortex contributes to odor quality influences as suggested in Figure 1B. responses. For example, Veldhuizen coding. Front. Hum. Neurosci. 4, pii:58. Dispatch R159

13. Fortis-Santiago, Y., Rodwin, B.A., Neseliler, S., 16. Green, B.G., Nachtigal, D., Hammond, S., and 1The John B Pierce Laboratory, Piette, C.E., and Katz, D.B. (2010). State Lim, J. (2012). Enhancement of retronasal odors 290 Congress Avenue, New Haven, dependence of olfactory perception as by taste. Chem. Senses 37, 77–86. CT 06519, USA. 2Interdepartmental a function of taste cortical inactivation. Nat. 17. Lim, J., and Johnson, M.B. (2011). Neurosci. 13, 158–159. Potential mechanisms of retronasal odor Neuroscience Program, Yale University, 14. Johnson, D.M., Illig, K.R., Behan, M., and referral to the mouth. Chem. Senses 36, SHM L-200, New Haven, CT 06520, USA. Haberly, L.B. (2000). New features of 283–289. 3Department of Psychiatry, Yale University connectivity in piriform cortex visualized 18. Murphy, C., Cain, W.S., and Bartoshuk, L.M. School of Medicine, 300 George Street, by intracellular injection of pyramidal cells (1977). Mutual action of taste and olfaction. New Haven, CT 06511, USA. 4Department of suggest that "primary" olfactory cortex Sens. Process 1, 204–211. functions like "association" cortex in 19. Bermudez-Rattoni, F., Rusiniak, K.W., and Surgery, Yale University School of Medicine, other sensory systems. J. Neurosci. 20, Garcia, J. (1983). Flavor-illness aversions: 800 Howard Avenue, New Haven, 6974–6982. potentiation of odor by taste is disrupted by CT 06519, USA. 15. Gottfried, J.A., and Dolan, R.J. (2003). The nose application of novocaine into amygdala. Behav. E-mail: [email protected] smells what the eye sees: crossmodal visual Neural Biol. 37, 61–75. facilitation of human olfactory perception. 20. Small, D.M. (2012). Flavor is in the brain. Neuron 39, 375–396. Physiol. Behav. 107, 540–552. http://dx.doi.org/10.1016/j.cub.2012.12.036

Mechanotransduction: Vinculin Vinculin is a bipolar protein with a head domain that binds talin, the Provides Stability when Tension integrin activator and integrin–F-actin linker protein, and a tail domain that Rises binds to F-actin. A high-affinity interaction between its head and tail domains, however, prevents isolated By beautiful imaging and state-of-the-art experiments, vinculin is established interactions with either the head or the to be a central switch in mechanotransduction at integrin-based focal tail from taking place [9]. It is likely that adhesions. Cycles of tension-regulated vinculin switching control focal this intramolecular interaction is adhesion dynamics and signaling to enable polarized cell migration and tunable to ‘activate’ vinculin and allow alignment. its presence and function in focal adhesions. Carisey et al. [8] have used Emma Spanjaard and Johan de Rooij vinculin. Vinculin’s presence at a set of mutations to tune vinculin’s adhesion complexes is force activity and regulation in a controlled The extracellular environment and dependent [2] and loss of vinculin leads manner: a point mutant that affects the its biophysical properties control to a reduction in adhesion-dependent head–tail interaction (Vinc-T12) fundamental cellular processes during cytoskeletal stiffening [3]. Vinculin produces constitutively active vinculin tissue development and homeostasis. has a similar function in [10]; a deletion mutant (Vinc-880) that The extracellular matrix (ECM) is cadherin-dependent cell–cell junctions lacks the tail domain produces active connected to the intracellular [4]. A common model proposes that vinculin that is uncoupled from actin actomyosin cytoskeleton through forces expose vinculin-binding sites [11]; and a minimal fragment of vinculin large integrin-based multi-protein in upstream proteins — a-catenin in capable only of binding talin (Vinc-258) complexes called focal adhesions. cadherin junctions [5] and talin in uncouples talin binding from any other Cytoskeletal contraction and ECM integrin adhesions [6]. Thus, vinculin effects of vinculin. These constructs stiffness or deformations produce is the common effector of several are dominant over endogenous vinculin tension across these focal adhesions, mechanosensitive systems. Vinculin in cells. Previously, the Ballestrem lab which respond by growing when itself, however, has proven a hard nut had shown that expression of these tension rises or by disassembling to crack: it is essentially a closed ball constructs leads to a strong when tension drops. Conversely, when purified in vitro [7] and stabilization of focal adhesions and focal adhesions induce signaling by mechanisms of conformational hypothesized that this was due to the the Rho-family small GTPases to regulation to sustain its interactions uncoupling of focal adhesions from control actomyosin organization and functions in cells are not well their normal regulation by actomyosin and contraction, thus creating understood. Also, the downstream [11]. Here this hypothesis is proven, by a feedback cycle between cytoskeletal molecular events driven by vinculin in a set of beautiful imaging experiments tension and integrin–ECM adhesion. cells are still unknown and, with that, its (watch the supplemental movies!), Controlling focal adhesion dynamics precise role in mechanotransduction which show that active vinculin and signaling is essential for proper has remained unclear. In this issue of constructs protect a subset of focal cell migration. Thus, focal adhesions Current Biology, Carisey et al. [8] now adhesion proteins from disassembly, are mechanosensitive structures provide new insights into both even when actomyosin structure or that transduce physical cues from regulation and downstream contractility is disrupted by drugs. the environment into cellular functionality that argue that vinculin Inactivation of vinculin is needed behavior [1]. is a central, tension-regulated for the release of these proteins One of the proteins with a key switch and master regulator of and the full disassembly of focal involvement in the mechanical mechanically controlled focal adhesions in response to decreased regulation of focal adhesions is adhesion dynamics. tension (Figure 1).