J Neurophysiol 117: 1293–1304, 2017. First published December 21, 2016; doi:10.1152/jn.00802.2016.

RESEARCH ARTICLE

Sensory Processing

Single-neuron responses to intraoral delivery of odor solutions in primary olfactory and gustatory cortex Joost X. Maier Department of Neurobiology and Anatomy, Wake Forest School of Medicine, Winston Salem, North Carolina Submitted 11 October 2016; accepted in final form 15 December 2016

NEW & NOTEWORTHY Food perception is inherently multisensory, taking into account taste, smell, and texture qualities. However, the neural mechanisms underlying flavor perception remain unknown. Recording neural activity directly from the rat brain while animals consume multisensory flavor stimuli, we demonstrate that information about odor, taste, and mouthfeel of food converges on primary taste and smell cortex. The results suggest that processing of naturalistic, multisensory information involves an interacting network of primary sensory areas. flavor; multisensory; electrophysiology; olfactory cortex; gustatory cortex ANYONE WHO HAS EVER suffered a common cold knows that experience of food relies heavily on olfaction: an impaired sense of smell makes food taste bland. This intuition is supported by experimental research that highlights the role of smell in flavor perception and food choice (Shepherd 2006). Rozin (1982) suggested that flavor-related odor processing is shaped by the specific context provided by consumption. Odorants released inside the mouth during ingestion and mastication stimulate the nasal epithelium retronasally when they are exhaled though the nostrils via the nasopharynx and typically occur in combination with multisensory (taste and oral somato-

Address for reprint requests and other correspondence: J. X. Maier, Dept. of Neurobiology and Anatomy, Wake Forest School of Medicine, Gray 4137, Medical Center Blvd. Winston Salem, NC 27157 (e-mail: jmaier @wakehealth.edu). www.jn.org

sensory) signals. This idea is supported by behavioral findings in humans and animals demonstrating that orally sourced odors readily interact with taste signals to enhance flavor perception (Green et al. 2012; Veldhuizen et al. 2010b) and enable the formation of food preferences and aversions (Bouton et al. 1986; Chapuis et al. 2007; Holman 1975; Sclafani and Ackroff 1994). At the neural level, evidence for convergence of orally sourced olfactory, gustatory, and somatosensory inputs comes mainly from whole brain functional imaging studies in humans. This line of research has shown that orally sourced odors activate a widespread network including olfactory regions (piriform cortex/amygdala), taste areas (insula/operculum), and the somatosensory mouth area (Cerf-Ducastel and Murphy 2001; Small et al. 2005). Thus, in parallel with the multisensory nature of flavor perception, these findings suggest that processing of the sensory cues associated with consumption involves a multisensory network of processing systems. With the use of electrophysiological recordings in rodents, recent work has started to uncover the mechanisms by which brain areas involved in flavor processing interact with each other at the network and cellular level. The results point to a role for primary olfactory (OC) (Carlson et al. 2013; Maier et al. 2012; Wesson and Wilson 2010) and gustatory cortex (GC) (Katz et al. 2001; Maier et al. 2015; Vincis and Fontanini 2016) in processing multisensory information. For example, Maier et al. (2012) demonstrated that single neurons in primary OC respond to taste stimuli. Subsequent work revealed that gustatory input reaches OC via GC (Maier et al. 2015). However, it remains unclear how gustatory input to OC relates to orally sourced olfactory and somatosensory input and whether GC receives reciprocal orally sourced olfactory input. Here, we characterize response dynamics of single neurons in OC and GC to intraoral delivery of odor solutions and compare intraoral olfactory responses to plain water, gustatory and orthonasal odor responses. We find that neurons in both areas exhibit nonselective responses to fluid delivery, as well as olfactory- and gustatory-specific responses, demonstrating convergence of multisensory flavor signals in both OC and GC. Moreover, olfactory responses in OC lead those in GC; gustatory responses follow the reverse pattern, suggesting bidirectional interactions between primary OC and GC. The results provide novel insight into the neural basis underlying multisensory flavor processing.

0022-3077/17 Copyright © 2017 the American Physiological Society

1293

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Maier JX. Single-neuron responses to intraoral delivery of odor solutions in primary olfactory and gustatory cortex. J Neurophysiol 117: 1293–1304, 2017. First published December 21, 2016; doi: 10.1152/jn.00802.2016.—Smell plays a major role in our perception of food. Odorants released inside the mouth during consumption are combined with taste and texture qualities of a food to guide flavor preference learning and food choice behavior. Here, we built on recent physiological findings that implicated primary sensory cortex in multisensory flavor processing. Specifically, we used extracellular recordings in awake rats to characterize responses of single neurons in primary olfactory (OC) and gustatory cortex (GC) to intraoral delivery of odor solutions and compare odor responses to taste and plain water responses. The data reveal responses to olfactory, oral somatosensory, and gustatory qualities of intraoral stimuli in both OC and GC. Moreover, modality-specific responses overlap in time, indicating temporal convergence of multisensory, flavor-related inputs. The results extend previous work suggesting a role for primary OC in mediating influences of taste on smell that characterize flavor perception and point to an integral role for GC in olfactory processing.

1294

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

METHODS

Subjects A total of eight adult female Long-Evans rats (http://www.criver. com), weighing between 250 and 300 g at the time of surgery served as subjects. All rats were individually housed and kept on a 12:12-h light-dark cycle. Experiments were conducted during the light cycle and complied with the Wake Forest Baptist Medical Center Institutional Animal Care and Use Committee guidelines (Protocol No. A15-116). Surgery

Histology Electrodes were labeled with a drop of Vybrant DiI cell-labeling solution (http://www.thermofisher.com/us/en/home.html/us/en/home. html), applied with a needle tip before implantation, allowing post mortem reconstruction of the implant location. To do so, rats were perfused transcardially with saline and 10% formalin, and their brains were extracted and placed in 30% sucrose for 2–5 days. Brains were then frozen; coronal sections were cut around the implant location using a sliding microtome, mounted on glass slides in DAPI Fluoromount-G medium (https://www.southernbiotech.com), and a coverslip was applied. Epifluorescence microscopy was used to visualize DiI and DAPI. Figure 1 shows representative examples of electrode placement in piriform OC and insular GC. Electrophysiological Recording The continuous extracellular signal recorded from each electrode was amplified, digitized and stored for offline analysis at a sampling rate of 20 –30 kHz using INTAN RHD2000 hardware and acquisition software (http://www.intantech.com). Further processing and analysis

Respiration Measurements Nasal cannulas were attached to a pressure sensor (24PCAFA6G, http://sensing.honeywell.com/) using PE-50 tubing and a 22-gauge metal tube inserted into the sensor casing, terminating just in front of the sensing membrane. Any excess space in the sensor casing was filled with superglue to reduce turbulence. The signal from the sensor was amplified using a differential AC amplifier (https://www. adinstruments.com/), digitized, and stored using INTAN RHD2000 hardware and software (see above) for offline analysis. Data analysis Offline analyses were performed using Matlab (www.mathworks. com). Time-averaged analysis. Responses were averaged over 2 s following stimulus delivery (stimulus period) and 1 s preceding stimulus delivery (baseline period). Stimulus-specific responses were obtained by subtracting average responses to control stimuli (i.e., plain water or plain water in combination with plain air, see below) from average responses to odor or taste stimuli. Time-resolved analysis. Responses were averaged in a sliding window over time (window size: 250 ms; step size: 25 ms) and normalized by converting to z-score: average activity during the baseline period was subtracted, and the difference was divided by the standard deviation (over trials) of time-averaged baseline activity. Stimulus-specific response magnitude was obtained by subtracting the z-score of the response to control stimuli from the z-score of the response to taste or odor stimuli. For all analyses, we used the absolute value of the z-score and difference in z-score to account for the fact that responses in some instances constituted increases and in others decreases relative to baseline and/or plain water. Significant portions of time-varying single-neuron responses were detected by applying a threshold of z ⫽ 2.75 to the smoothed (moving average method), z-transformed response. Latency was estimated as the point in time following stimulus onset where the smoothed z-score reached 75% of maximum (similar to Siegel et al. 2015). Respiration signal obtained from the pressure sensor was low-pass filtered (15 Hz, 2nd order Butterworth filter) and analyzed in the frequency domain using the Chronux toolbox (http://chronux.org/) (Maier et al. 2008; Mitra and Bokil 2008). Differences in response measures were tested using t-tests (withinneuron measures) as well as nonparametric tests (between-areas

Fig. 1. Electrode implant location. DAPIstained (blue) coronal brain sections showing the reconstructed location of electrodes implanted in olfactory (piriform) cortex (bregma ⫽ ⫺1.4 mm; A) and gustatory (insular) cortex (bregma ⫽ ⫹1.6 mm; B), indicated by DiI labeling (pink).

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Stereotaxic surgery was performed under ketamine/xylazine anesthesia. Multielectrode assemblies (8 –32 electrodes/assembly, https:// www.microprobes.com and https://www.neuronexus.com) were implanted into left posterior piriform OC: 1.4 mm posterior to bregma, 5.2–5.6 mm lateral to the midline, and 6.4 –7.4 mm ventral from the surface of the brain; or left insular GC: 1.6 mm anterior to bregma, 4.6 –5.2 mm lateral from the midline, and 4.8 –5.5 mm ventral from the surface of the brain (Paxinos and Watson 1986). Intraoral cannulas (IOC) were implanted into the oral cavity (Maier and Katz 2013). Intranasal cannulas (22 gauge infusion cannulas, http://www.plastics1. com/) were implanted into the left nasal cavity, protruding 2.4 mm below the surface of the skull (Maier et al. 2012; Verhagen et al. 2007). Animals recovered in their home cage with ad libitum access to water and mashed rat chow for 4 –5 days after surgery before starting data collection.

was performed using Matlab (www.mathworks.com). In short, waveforms exceeding at least 5 standard deviation units from the high-pass filtered signal (400 Hz) were extracted and clustered using the WaveClus toolbox (http://www2.le.ac.uk/centres/csn/research-2/spike-sorting) to obtain single neuron spike time stamps (Quiroga et al. 2004).

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

measures). All tests were two-tailed and unpaired, with ␣ ⫽ 0.05, unless otherwise noted. Stimuli

Stimulus Presentation and Recording Procedures Animals were habituated to the recording setup and stimulus delivery apparatus before recording/training by presenting drops of water through IOC in the recording arena. Airborne odorants were cleared by a continuously running fan mounted in the ceiling of the recording arena. Passive presentation. Intraoral stimuli were delivered with a random intertrial interval (ITI) ranging from 30 to 45 s (see Fig. 2A), allowing sufficient time for animals to swallow the fluid and clear their mouth. To encourage consumption of stimuli, animals were water deprived for 2– 6 h before passive recording sessions. Active acquisition. In sessions where odors were presented both intraorally and orthonasally, stimuli were delivered immediately upon rats triggering an infrared beam in a nose poke. Animals were water-deprived and trained for 2–3 days before recording sessions to keep their nose in the odor port for 2 s. During recording sessions, animals were rewarded for successful trials (30 ␮l plain water, presented through IOC, 3 s after stimulus delivery). Availability of a new trial occurred at a random ITI of 10 –15 s and was indicated by back-lighting the nose poke with LEDs. To control for differences in nasal and oral somatosensory stimulation between orthonasal and

Fig. 2. Stimulus presentation protocols. A: during passive presentation, drops of fluid are delivered into the oral cavity via intraoral cannulas (IOC) at random intertrial intervals (ITI). B: during active acquisition, illuminating LEDs in a nose poke signals the availability of a stimulus (cue light). Stimulus delivery occurs upon the animal breaking an infrared beam in the nose poke, and consists of simultaneous presentation of 1) an orthonasal stimulus; and 2) a drop of fluid into the oral cavity via IOC. A water reward (delivered via IOC) is presented when the animal holds its nose in the nose poke for the duration of the orthonasal stimulus. After completion of a trial, the nose poke is unresponsive for a random ITI, indicated by turning off the cue light. Two trials are shown for each protocol, separated by dashes lines.

intraoral odor stimuli, orthonasal odor stimuli were always presented simultaneously with intraoral plain water; intraoral odor stimuli were always presented simultaneously with orthonasal plain air (see Fig. 2B). In addition to 5–15 ml of fluid consumed during recording sessions, animals received 10 –20 ml of water in their home cage at the end of each day. All animals (n ⫽ 8 total, n ⫽ 5 and 3 with implants in OC and GC, respectively) underwent one to three passive recording sessions (n ⫽ 17 total, n ⫽ 11 and 6 sessions recording from OC and GC, respectively), during which one to three novel odorants were presented (10 repetitions of each stimulus). A single taste stimulus was presented during 15 sessions in 8 animals. Respiration was recorded during three sessions in two animals. A subset of animals (n ⫽ 4 total, n ⫽ 2 and 2 with implants in OC and GC, respectively) underwent two active sessions (8 sessions total), during which one or two odorants were presented both intraorally and orthonasally (5–10 repetitions of each stimulus). Plain water (passive sessions) or plain water in combination with plain air (active sessions) was presented as a control stimulus (10 repetitions). All stimuli during all sessions were presented in random order. Each session (total duration ⬍1 h) yielded between 1 and 27 (mean ⫽ 11.6) single neurons. A single electrode contact typically yielded action potentials from at most one neuron. RESULTS

Olfactory and Gustatory Cortex Responses to Intraoral Odorants We recorded spiking activity from single neurons in primary olfactory (piriform) cortex (OC) and primary gustatory (insular) cortex (GC) of naïve, awake rats in response to intraoral odor stimuli. Odorants were passively delivered in aqueous solution directly into the oral cavity at random interstimulus intervals (Fig. 2A), producing somatosensory and retronasal olfactory stimulation associated with natural flavor stimuli. OC and GC neurons (n ⫽ 121 and 77) were first screened for overall responsiveness to fluid delivery (i.e., stimulus-responsiveness, regardless of modality). Visual inspection indicated that responses are modulated over the course of ~2 s following stimulus delivery (single neuron and population response dynamics to intraoral stimuli can be seen in Figs. 3, 4, 7, and 8). Stimulus responsiveness was determined by comparing average firing rate during the stimulus period (2 s following stimulus delivery) to average firing rate during the baseline period (1 s preceding stimulus delivery) for all trials (i.e., all available plain water, odor and taste stimuli combined) using a paired samples t-test. A total of 34 OC neurons (28%) exhibited a significant response to fluid delivery. Responsiveness to intraoral fluid stimuli in GC was more prevalent, consistent with its well-known role in signaling intraoral stimuli (Katz et al. 2001): n ⫽ 33 GC neurons (43%) responded significantly to intraoral stimulus delivery (␹2 test comparing the proportion of stimulus-responsive neurons in OC vs. GC: ␹2 ⫽ 4.58, P ⬍ 0.05). These data demonstrate that single neurons in primary OC and GC respond to intraoral delivery of fluid stimuli. All analyses reported below were limited to neurons that exhibited significant responses to fluid delivery (i.e., stimulus-responsive neurons). As noted above, intraorally delivered odor solutions have both olfactory and somatosensory qualities, and neurons may receive both olfactory and nonolfactory sensory inputs (Carlson et al. 2013; Katz et al. 2001; Maier et al. 2012; Schoenbaum and Eichenbaum 1995; Wesson and Wilson 2010;

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Odor stimuli were exemplars of monomolecular odorants, e.g., amyl acetate, methyl valerate, 2-hexanone, and ethyl benzoate. Intraoral stimuli consisted of small aliquots (30 ␮l) of odorant solution (presented at a single concentration of either 0.01 or 0.02% in distilled water), delivered through IOC directly into the oral cavity via syringe pumps (delivery time ⬍100 ms). Orthonasal stimuli consisted of saturated vapor (created in N2, diluted ~10 times in zero grade air) diffusing out of a nose poke in front of the animal. Plain water and air were used as control stimuli. Taste stimuli were exemplars of basic taste qualities: sucrose (100 mM), sodium chloride (100 mM), citric acid (100 mM), and saccharin (100 mM) in distilled water. Taste stimuli were delivered intraorally.

1295

1296

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

Zinyuk et al. 2001). We evaluated odor specificity as differential responsiveness to odor solutions vs. plain water. Figure 3 shows examples of single neuron responses in both brain areas to plain water and odor solutions. Time-averaged analysis (see METHODS) indicated that neurons exhibit nonspecific fluid responses to plain water (Fig. 3, i, ii, and iv); t-test comparing responses to plain water during the stimulus period vs. baseline), as well as odor-specific responses to odor solutions (Fig. 3, i–iv); t-test comparing responses to odor solutions vs. plain water). Visual inspection revealed that odor-specific responses may vary in latency and duration. To capture this variability in the population, we quantified responses using a continuous measure of response magnitude (z-normalized firing rate) in a sliding window over time (see METHODS) for individual neuronodor pairs (n ⫽ 74 and 66 in OC and GC, respectively). The average response magnitude in response to plain water and odor solutions is shown in Fig. 4, A and B. The magnitude of

the OC population response was clearly larger for odor solution as compared with plain water. In GC, no differences in the magnitude of the population response were obvious, suggesting that adding odor to the stimulus solution did not consistently increase or decrease the firing rate of GC neurons. We therefore further characterized odor specificity in individual neuron-odor pairs using the difference in response magnitude between plain water and odor conditions. Significant odorselective response portions were detected using a threshold of z ⫽ 2.75 and plotted in Fig. 4C. This threshold was highly conservative and chosen to exclude all spurious responses (as measured by the absence of significant responses during baseline). This analysis revealed that individual responses in OC and GC are equally likely to exhibit odor-specific modulations (n ⫽ 11 and 12 in OC and GC, respectively, ␹2 ⫽ 0.57, P ⬎ 0.05). In OC, odor-specific responses typically manifest as response increases relative to plain water; in GC, olfactory stimulation caused bidirectional modulations of responses to

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Fig. 3. Single olfactory (OC) and gustatory cortex (GC) neuron responses to intraoral delivery of odor solutions. Example single trial raster diagrams (i and ii) and responses profiles (iii and iv) of single OC (blue) and GC (red) neuron responses to odor solutions (colored lines) and plain water (gray lines) relative to stimulus delivery. Profiles are averaged (⫾SE) over trials (n ⫽ 10 per condition). Insets: randomly selected action potential waveforms (n ⫽ 100) for each neuron.

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

1297

plain water (see also Fig. 3). We further quantified odorspecific response magnitude and latency in the set of responses that showed significant odor-specific modulations. The occurrence of significant odor-specific responses as a function of time is summarized in Fig. 4D. Average odor-specific response magnitude over time for this set of responses is shown in Fig. 4E. Overall, magnitude of the odor-specific responses in OC did not differ from GC (Wilcoxon rank sum test, z ⫽ 0.30, P ⬎ 0.05). However, odor-specific responses in OC occurred at shorter latency (see METHODS) as compared with GC [Kolmogorov-Smirnoff (K-S) test comparing odor-specific latency in OC vs. GC: K-S statistic ⫽ 0.57, P ⬍ 0.05]. We next quantified odor selectivity of responses (i.e., the degree to which responses of the same neuron to different

odorant solutions differ). Figure 5 shows responses to all pairs of odorants recorded from the same neuron (n ⫽ 49 and 33 pairs obtained from 33 and 24 neurons in OC and GC, respectively; responses plotted relative to water, averaged over the stimulus period). Overall, the response of a given neuron to one odorant reliably predicted its response to another odorant (r ⫽ 0.47, P ⬍ 0.001), indicating nonselective responses. However, single OC neurons in particular were sometimes highly selective, eliciting different responses to different odorants (indicated by data points that fall off the diagonal in Fig. 5). Higher odor selectivity in OC vs. GC was confirmed by directly comparing the correlation between pairs of responses in the two areas (OC: r ⫽ 0.36; GC: r ⫽ 0.68; one-tailed Fisher z-test: z ⫽ 1.95, P ⬍ 0.05). Results did not differ when considering only neurons that exhibited significant periods in Fig. 5. Intraoral odor-selectivity of single OC and GC neurons. A and B: firing rate (A) and z-normalized response magnitude (B) during the stimulus period (relative to water) for all pairs of intra-orally delivered odorants recorded from the same neurons in OC (dark gray, n ⫽ 49) and GC (light gray, n ⫽ 33). Each circle represents a pair of responses: filled circles indicate pairs for which both responses showed a significant odor-specific response period; dotted circles indicate pairs for which only one of the responses showed a significant odor-specific response period; open circles indicate pairs for which none of the responses showed a significant odor-specific period. Diagonal line indicates identical responses to both odors.

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Fig. 4. Population responses to intraoral delivery of odor solutions. A and B: absolute magnitude of the response to odor solutions (colored lines) and plain water (black lines), averaged (⫾SE) over all stimulus-responsive neuron-odor pairs in OC (blue, n ⫽ 74; A) and GC (red, n ⫽ 66; B). C: occurrence of significant odor-specific responses for individual neuron-odor pairs in OC and GC. D: proportion (out of all stimulus responsive neuron-odor pairs) of neuron-odor pairs exhibiting significant odor-specific modulations over time in OC and GC. E: absolute magnitude of the difference in response to water and odor solution, averaged (⫾SE) over all responses that exhibited significant odor-specific modulations in OC (n ⫽ 11) and GC (n ⫽ 12).

1298

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

odor stimuli evoked an increase in coherence (average coherence between 2 and 8 Hz during 1.5 s following stimulus delivery) relative to plain water [paired t-test comparing coherence in response to odor solutions vs. plain water: t(23) ⫽ 2.73, P ⬍ 0.05]. Coherence induced by intraoral odor stimuli was also significantly higher than coherence during the baseline period [paired t-test comparing average coherence during 1.5 s following stimulus delivery vs. 1.5 s preceding stimulus delivery: t(23) ⫽ 2.73, P ⬍ 0.05]. A control analysis computing the coherence between random pairings of spike trains and respiration on a trial-by-trials basis yielded significantly reduced coherence [paired t-test comparing coherence obtained from observed data vs. shuffled data: t(23) ⫽ 2.24, P ⬍ 0.05], indicating that coherence was not due to mere amplitude covariations. These findings demonstrate that respiration reliably predicts responses to intraoral odor stimuli. Taken together, these data demonstrate that neurons in both OC and GC respond in a modality-specific manner to orally sourced olfactory stimulation. Comparison of Intraoral Odor and Taste Responses Previous work has demonstrated that single neurons in both OC (Maier et al. 2012, 2015) and GC (e.g., Katz et al. 2001) respond to intraoral delivery of taste solutions. To characterize how intraoral odor responses relate to taste responses, we recorded responses to intraoral delivery of taste stimuli in a subset of recordings (n ⫽ 23 and 33 stimulus-responsive neurons in OC and GC, respectively). Figure 7, A–C, shows taste-specific response portions (characterized in the same manner as odor-specific response above) of individual OC and GC responses and the average taste-specific response. Tasteselective responses in OC occur at longer latencies as compared with GC (K-S test comparing taste-specific latency in OC vs. GC: K-S statistic ⫽ 0.69, P ⬍ 0.05), consistent with previous reports comparing taste responses in OC and GC

Fig. 6. OC responses to intraoral odor solutions are modulated by respiration. A: examples of respiration and spiking activity of a single OC neuron recorded during 2 trials during which odor solution was delivered intraorally. B: frequency spectra of respiration and spiking activity during the stimulus period, averaged over trials for all recording sessions (n ⫽ 3) and stimulus-responsive neuron-odor pairs (n ⫽ 24), respectively. C: coherence as a function of frequency and time relative to intraoral delivery of odor solutions, averaged over all stimulus-responsive neuronodor pairs. D: coherence between respiration and spiking activity during 1.5 s following intraoral delivery of odor solutions, averaged (⫾SE) over all stimulus-responsive neuron-odor pairs.

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

response to at least one odorant (OC: n ⫽ 13, r ⫽ 0.43, Fisher z-test comparing significant pairs vs. all pairs: z ⫽ 0.24, P ⬎ 0.05; GC: n ⫽ 9, r ⫽ 0.86; z ⫽ 1.04, P ⬎ 0.05). Similar results were obtained when using z-scores instead of firing rates (Fig. 5B). To further characterize intraoral odor responses in OC neurons, we monitored animals’ respiration during a subset of recording sessions using an intranasal cannula connected to a pressure sensor (n ⫽ 24 stimulus-responsive neuron-odor pairs obtained from 10 neurons in OC only). Previous reports on OC neuron responses to orthonasal odorants have shown that odor-specific responses are modulated by respiration: every time air-containing odorants passes through the nasal cavity, OC neuron firing rates peak (Miura et al. 2012). Sniffing behavior showed complex variations from trial to trial. We used spectral analysis to test for the presence of respirationrelated modulations. Figure 6A shows the average power spectra of the respiration signal and spiking activity during the stimulus period. Respiration is modulated at frequencies between 2 and 8 Hz, consistent with other reports of respiration activity during orthonasal odor stimulation in awake rats (Wesson et al. 2008). The average power spectrum of OC neuron spiking activity in response to odor solutions during the same period exhibits modulations at overlapping frequencies. To test the hypothesis that respiration drives OC spiking activity, we calculated the coherence between spiking activity and respiration for each neuron-odor combination. Coherence measures the consistency of the relation between two signals, in this case respiration and OC neuron spiking activity. The resulting coherogram (i.e., coherence as a function of time and frequency, averaged over all neuron-odor pairs) is shown in Fig. 6B. Coherence between respiration and spiking activity of OC neurons in the 2 to 8-Hz frequency band increases after delivery of odor solutions. Figure 6C shows coherence during the stimulus period between respiration and spiking activity in response to odor solutions and plain water. Intraoral delivery of

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

1299

(Maier et al. 2012, 2015). Direct comparison of taste- and odor-specific response latencies revealed that taste- and odorspecific responses in OC and GC differ in their temporal dynamics: in OC, odor responses lead taste responses (K-S test comparing odor-specific vs. taste-specific latency in OC: K-S statistic ⫽ 0.70, P ⬍ 0.05); in GC, taste responses lead odor responses (K-S test comparing odor-specific vs. taste-specific latency in GC: K-S statistic ⫽ 0.52, P ⬍ 0.05). This pattern of results demonstrates that responses to intraoral odors cannot be explained by potential gustatory qualities of odor solutions. Instead, our findings suggest that OC and GC responses to taste and odor stimuli reflect converging gustatory and olfactory input, respectively. Comparison of Intraoral and Orthonasal Odor Responses The results presented above demonstrate that OC and GC neurons respond in an odor-selective manner to intraoral delivery of odor solutions. However, it is not clear whether responses are specific for intraoral odor presentation. Previous studies have suggested that retronasal olfactory stimulation resulting from intraoral odorant release is processed differently from orthonasally presented odors (Gautam and Verhagen 2012a; Heilmann and Hummel 2004; Small et al. 2005). Here, we tested whether retro- and orthonasal modes of odor stimulation elicit different responses in OC and GC neurons by directly comparing responses to the same odorants presented intraorally and orthonasally in a separate set of recordings (n ⫽ 26 and 32 stimulus-responsive neuron-odor pairs obtained from 13 and 16 neurons in OC and GC, respectively). To control for differences in somatosensory stimulation, orthonasal odorants were presented simultaneously with a drop of plain water into the oral cavity; intraoral odorants were presented simultaneously with orthonasal plain air. Control stimuli consisted of simultaneously presented water and air. Animals actively approached a nose poke to trigger stimulus delivery (see Fig. 2B). Figure 8A shows responses of example OC and GC neurons to the same odors presented intraorally and orthonasally. Timeaveraged analysis revealed that all responses shown were odor specific (t-test comparing responses during the stimulus period

in orthonasal and intraoral modes pooled vs. control). Some responses differed in magnitude between intraoral and orthonasal modes (Fig. 8Ai), others did not (Fig. 8, Aii–Aiv; t-test comparing responses to the same odorants during the stimulus period in intraoral vs. orthonasal modes). Figure 8B shows average response during the stimulus period (relative to baseline) in intraoral and orthonasal modes for each neuron-odor pair. Overall, responses in intraoral and orthonasal modes did not differ [paired t-test comparing average z-score during the stimulus period relative to water for intraoral vs. orthonasal modes, OC and GC combined: t57 ⫽ 0.57, P ⬎ 0.05; OC only: t25 ⫽ 0.21, P ⬎ 0.05; GC only: t31 ⫽ 0.63, P ⬎ 0.05)] Moreover, responses in intraoral and orthonasal modes were positively correlated (OC and GC combined: r ⫽ 0.47, P ⬍ 0.001), indicating that the response of a given neuron to a given odorant presented intraorally predicted its response to the same odorant presented orthonasally (Fig. 8, B and C). Directly comparing response similarity in the two modes between OC and GC revealed no differences in correlation (OC: r ⫽ 0.54; GC: r ⫽ 0.41; Fisher z-test: z ⫽ 0.6, P ⬎ 0.05), indicating that mode of stimulation evoked equally similar responses in OC and GC. The finding that overall responses did not differ between modes of presentation suggests that at the population level OC and GC neurons are equally sensitive to olfactory stimulation driven by retro- and orthonasal airflow. However, the example responses in Fig. 8A suggest that intraoral and orthonasal responses may differ for individual neurons, either in overall responsiveness (e.g., as suggested by Fig. 8, Ai–Aiii) or temporal dynamics (e.g., as suggested by Fig. 8Aiv). Similar results were obtained when using z-scores instead of firing rates (Fig. 8C). DISCUSSION

Smell forms an integral part of our experience of flavor. Coherent flavor perception requires multisensory integration of orally sourced odor cues with oral somatosensory and gustatory cues provided by a food (Small and Green 2012). Previous behavioral work has suggested that rodents are a promising model system for investigating multisensory interactions underlying flavor perception: rodents are sensitive to intraoral

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Fig. 7. Population response of OC and GC neurons to taste solutions. A: occurrence of significant taste-specific responses for individual neuron-taste pairs in OC (blue) and GC (red). B: proportion of neuron-taste pairs (out of all stimulus responsive neuron-taste pairs, n ⫽ 23 and 33 in OC and GC, respectively) exhibiting significant taste-specific modulations over time in OC and GC. C: absolute magnitude of the difference in response to water and taste solution, averaged (⫾SE) over all responses that exhibited significant taste-specific modulations in OC (n ⫽ 5) and GC (n ⫽ 13).

1300

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

olfactory stimulation (Gautam and Verhagen 2012a; Rebello et al. 2015) and combine orally sourced odor and taste cues to form flavor preferences (Holman 1975; Maier et al. 2015; Sclafani and Ackroff 1994). The present study constitutes a logical next step in this line of research by demonstrating multisensory converge of consumption-related olfactory, somatosensory and gustatory inputs onto single neurons in primary olfactory (OC) and gustatory (GC) cortexes of the rat. Our findings support and extend a model in which primary sensory cortex plays an integral role in multisensory flavor processing (Ghazanfar and Schroeder 2006; Maier et al. 2015; Scott and Plata-Salamán 1999; Veldhuizen et al. 2010a). To our knowledge, this is the first study to characterize single neuron responses to intraoral odor solutions in OC. Our findings of odor-specific and selective OC neuron responses that are modulated by respiration are generally consistent with previous reports on responses to orthonasally presented odors in the olfactory system of awake rats (Cury and Uchida 2010; Miura et al. 2012; Shusterman et al. 2011). Reports on olfactory cortical neuron responses in awake animals are rare (Calu

et al. 2007; Chen et al. 2011; Gire et al. 2013; Miura et al. 2012; Roesch et al. 2007; Schoenbaum and Eichenbaum 1995; Zhan and Luo 2010; Zinyuk et al. 2001), and most previous studies have either focused on responses during the first sniff after stimulus onset only (Miura et al. 2012) or averaged responses over the duration of the stimulus (Calu et al. 2007; Chen et al. 2011; Roesch et al. 2007; Schoenbaum and Eichenbaum 1995; Zhan and Luo 2010; Zinyuk et al. 2001). The present results demonstrate that OC neuron responses to intraorally delivered odors exhibit variable, extended, dynamics over the course of 2 s following stimulus delivery. The fact that odorants were delivered in a naturalistic context to awake animals likely contributed to the observed response variability. Intraoral fluid delivery evokes complex mouth and sniffing movements that may shape single neuron responses to odor solutions in different ways. First, sniffing controls odorant flow to the nasal epithelium and shapes neural responses (Verhagen et al. 2007) and variability in sniffing frequency and amplitude observed here could contribute to variability in magnitude and temporal profile of odor-driven responses. Similarly, mouth movements could influence odorant release, shaping response

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

Fig. 8. Comparison of intraoral and orthonasal odor presentation mode. A: examples of single OC (i and ii) and GC (iii and iv) neuron responses to the same odorants presented intraorally (dark gray) and orthonasally (light gray) relative to stimulus delivery, averaged (⫾SE) over trials (n ⫽ 5-10 per condition). Insets: randomly selected action potential waveforms (n ⫽ 100) for each neuron. B and C: firing rate (B) and z-normalized response magnitude (C) during the stimulus period relative to control, in intraoral vs. orthonasal modes, for all stimulus-responsive neuron-odor pairs in OC (dark gray, n ⫽ 26) and GC (light gray, n ⫽ 32).

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

in mediating interactions between orally sourced, multisensory flavor information. The functional significance of gustatory input interacting with olfactory processing is suggested by several behavioral studies. Psychophysical work in humans has shown that taste influences odor quality ratings (Dalton et al. 2000; Green et al. 2012), and behavioral work has demonstrated that rats and humans form preferences for the smell component of a flavor after experiencing that smell in mixture with a palatable taste (Holman 1975; Sclafani and Ackroff 1994; Stevenson et al. 1998), suggesting that taste stimuli can imbue olfactory representations with gustatory characteristics. The present study also demonstrates olfactory sensitivity of GC neurons. This finding is consistent with imaging studies that suggested flavor-related olfactory input to gustatory insular regions (Small et al. 2005; Veldhuizen et al. 2010a) and with a recent report showing GC neuron responses to orthonasal odors (Vincis and Fontanini 2016). In contrast to OC responses to intraoral odor stimuli, olfactory-specific GC responses constitute modulations (i.e., either decreases or increases in firing rate) of responses to fluid stimuli. Several previous studies have demonstrated olfactory modulations in gustatory neurons in both the nucleus of the solitary tract and the parabrachial nucleus (Di Lorenzo and Garcia 1985; Escanilla et al. 2015; Van Buskirk and Erickson 1977). Escanilla et al. (2015) suggested that olfactory signals may originate from the olfactory bulb and reach the gustatory brainstem via corticofugal projections. Such a pathway is consistent with the present finding that odor-specific responses in OC lead odorspecific responses in GC and supported by anatomical data revealing bidirectional connectivity between insular and piriform cortex (Datiche et al. 1996; Krushel and van der Kooy 1988; Reep and Winans 1982; Shi and Cassell 1998; Shipley and Geinisman 1984; Yasui et al. 1991). Previous studies hved already demonstrated that gustatory signals reach OC via GC (Maier et al. 2015; Maier et al. 2012); olfactory signals could travel the reverse route: from OC to GC and from there cause modulation of taste responses in the gustatory brainstem via feedback projections. The functional significance of olfactory input to GC is unknown. As noted above, taste-smell interactions in flavor processing typically involve taste influencing smell and not the other way around (Green et al. 2012). One possibility is that GC plays an integral role in odor processing, even in the absence of taste stimuli. Thus olfactory input to GC might not reflect olfactory signals interacting with gustatory representations per se but instead an active role for GC in processing olfactory signals. A recent study showed that inactivation of GC changes how OC neurons respond to odor stimuli in the absence of taste stimuli (Maier et al. 2015). Such influences are paralleled by findings that rats are unable to properly recognize odors in unisensory olfactory tasks (Fortis-Santiago et al. 2010; Maier et al. 2015) and consistent with case studies in human patients demonstrating that lesions of the insular cortex affect odor intensity and pleasantness ratings (Mak et al. 2005). It is not clear whether involvement of GC in odor processing is specific for consumption-related odor signals. We found that olfactory input modulates GC responses around one second following stimulus delivery. Previous work on the dynamics of gustatory processing in GC has demonstrated that GC neuron activity during this time period reflects the palatability of taste stimuli (Katz et al. 2001; Sadacca et al. 2012) and is involved

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

profiles. Monitoring these factors in future research will provide insight into how behavior controls stimulus processing in naturalistic settings. We did not observe a consistent effect of mode of stimulation (intraoral vs. orthonasal) on the spatial pattern of OC and GC neuron odor responses. Overall response magnitude in both areas was similar for intraoral and orthonasal odor stimuli. This finding is in agreement with two lines of previous research. First, imaging studies have shown that retro- and orthonasal odors evoke highly similar glomerular activation patterns in the olfactory bulb of anesthetized rodents (Gautam et al. 2014; Gautam and Verhagen 2012b). Second, work on cortical coding of odor information has suggested that odor identity is encoded in the spatial pattern of OC neuron responses, independent of context or concentration (Choi et al. 2011; Illig and Haberly 2003; Stettler and Axel 2009). However, behavioral work has suggested that orally sourced odors are processed differently from externally sourced (orthonasal) odors (Bouton et al. 1986; Chapuis et al. 2009; Chapuis et al. 2007; Gautam and Verhagen 2012a; Heilmann and Hummel 2004). Moreover, using fMRI in humans, Small et al. (2005) demonstrated differential activation of various brain areas in response to retro- vs. orthonasal odor stimulation, including piriform and insular cortex. Several factors likely contribute to the discrepancy between physiological findings (including the present report) on the one hand and the results from behavioral and imaging studies on the other. First, it is possible that oral somatosensory stimulation that accompanied orthonasal stimulation in the present study interacts with olfactory input to produce responses that are similar to the intraoral odor response. Extra-olfactory input has been shown to interact with olfactory input in shaping OC neuron responses (Calu et al. 2007; Gire et al. 2013; Roesch et al. 2007; Schoenbaum and Eichenbaum 1995; Zinyuk et al. 2001). Second, it is possible that the difference between ortho- and retronasal olfaction is reflected in the temporal dynamics of OC neuron responses. Previous work has demonstrated that information about odor context is encoded in temporal aspects of OC neuron firing (Blumhagen et al. 2011; Gire et al. 2013). Differences in sampling behavior in orthonasal and intraoral contexts may contribute to different temporal response patterns. For example, intraoral odors cause olfactory stimulation upon exhalation, potentially evoking neural responses during different phases of the sniff cycle. Future work will directly measure responses to intraoral and orthonasal odors in conjunction with respiration to further investigate how consumption-related contextual factors such as direction of airflow and oral somatosensory cues shape spatiotemporal odor coding in olfactory networks. The second main finding of the present study is that OC neurons respond to multisensory aspects of oral stimuli. OC neuron sensitivity to gustatory stimulation is consistent with previous work from our laboratory. The fact that OC neurons showed nonspecific responses to plain water suggests somatosensory input, although it has been suggested that intraoral delivery of water causes gustatory stimulation (Stapleton et al. 2006). Here, we demonstrate temporal overlap of olfactory, gustatory, and oral somatosensory inputs. Spatiotemporal overlap of multisensory inputs is considered a prerequisite for performing multisensory computations (Stein and Meredith 1993). Thus the present results support a possible role for OC

1301

1302

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

that involve multiple sensory systems (Ghazanfar and Schroeder 2006; Maier et al. 2015). ACKNOWLEDGMENTS Many thanks to Chad Collins for excellent technical assistance and to Meredith Blankenship and Don Katz for valuable discussions. GRANTS This work is supported by National Institute on Deafness and Other Communication Disorders Grant R03-DC-014017 and the Johnston Family Endowment Fund. AUTHOR CONTRIBUTIONS J.X.M. performed experiments; J.X.M. analyzed data; J.X.M. interpreted results of experiments; J.X.M. prepared figures; J.X.M. drafted manuscript; J.X.M. edited and revised manuscript; J.X.M. approved final version of manuscript. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). REFERENCES Barnes DC, Chapuis J, Chaudhury D, Wilson DA. Odor fear conditioning modifies piriform cortex local field potentials both during conditioning and during post-conditioning sleep. PLoS One 6: e18130, 2011. doi:10.1371/ journal.pone.0018130. Blumhagen F, Zhu P, Shum J, Schärer YP, Yaksi E, Deisseroth K, Friedrich RW. Neuronal filtering of multiplexed odour representations. Nature 479: 493– 498, 2011. doi:10.1038/nature10633. Bouton ME, Jones DL, Mcphillips SA, Swartzentruber D. Potentiation and overshadowing in odor-aversion learning–role of method of odor presentation, the distal-proximal cue distinction, and the conditionability of odor. Learn Motiv 17: 115–138, 1986. doi:10.1016/0023-9690(86)90006-8. Calu DJ, Roesch MR, Stalnaker TA, Schoenbaum G. Associative encoding in posterior piriform cortex during odor discrimination and reversal learning. Cereb Cortex 17: 1342–1349, 2007. doi:10.1093/cercor/bhl045. Carlson KS, Xia CZ, Wesson DW. Encoding and representation of intranasal CO2 in the mouse olfactory cortex. J Neurosci 33: 13873–13881, 2013. doi:10.1523/JNEUROSCI.0422-13.2013. Cerf-Ducastel B, Murphy C. fMRI activation in response to odorants orally delivered in aqueous solutions. Chem Senses 26: 625– 637, 2001. doi:10. 1093/chemse/26.6.625. Chapuis J, Garcia S, Messaoudi B, Thevenet M, Ferreira G, Gervais R, Ravel N. The way an odor is experienced during aversive conditioning determines the extent of the network recruited during retrieval: a multisite electrophysiological study in rats. J Neurosci 29: 10287–10298, 2009. doi:10.1523/JNEUROSCI.0505-09.2009. Chapuis J, Messaoudi B, Ferreira G, Ravel N. Importance of retronasal and orthonasal olfaction for odor aversion memory in rats. Behav Neurosci 121: 1383–1392, 2007. doi:10.1037/0735-7044.121.6.1383. Chapuis J, Wilson DA. Bidirectional plasticity of cortical pattern recognition and behavioral sensory acuity. Nat Neurosci 15: 155–161, 2011. doi:10. 1038/nn.2966. Chen CF, Barnes DC, Wilson DA. Generalized vs. stimulus-specific learned fear differentially modifies stimulus encoding in primary sensory cortex of awake rats. J Neurophysiol 106: 3136 –3144, 2011. doi:10. 1152/jn.00721.2011. Choi GB, Stettler DD, Kallman BR, Bhaskar ST, Fleischmann A, Axel R. Driving opposing behaviors with ensembles of piriform neurons. Cell 146: 1004 –1015, 2011. doi:10.1016/j.cell.2011.07.041. Cury KM, Uchida N. Robust odor coding via inhalation-coupled transient activity in the mammalian olfactory bulb. Neuron 68: 570 –585, 2010. doi:10.1016/j.neuron.2010.09.040. Dalton P, Doolittle N, Nagata H, Breslin PA. The merging of the senses: integration of subthreshold taste and smell. Nat Neurosci 3: 431– 432, 2000. doi:10.1038/74797. Datiche F, Litaudon P, Cattarelli M. Intrinsic association fiber system of the piriform cortex: a quantitative study based on a cholera toxin B subunit

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

in driving palatability-related behaviors (Li et al. 2016; Sadacca et al. 2016). It is thus possible that GC is involved in signaling the palatability of oral stimuli, regardless of sensory modality. Our findings that GC neuron odor responses show little selectivity for specific odorants is in line with this idea, given that our experiments used neutral odors in naïve animals. The present finding that GC neuron responses to intraoral stimuli are modulated by both intraoral and orthonasal odors demonstrates that the involvement of GC in olfactory processing is not limited to retronasal odors, as suggested previously (Chapuis et al. 2009; Small et al. 2005). Finally, as noted above, the findings of multisensory convergence identifies OC and GC as potential substrates for flavor processing. However, behavioral work on flavor perception and food choice shows that this process is heavily shaped by individual experience with specific foods (Gautam and Verhagen 2010; Holman 1975; Shepard et al. 2015; Small et al. 2004; Stevenson et al. 1998; Yeomans et al. 2006). Previous work supports a role for primary OC and GC in mediating experience-dependent formation of multisensory flavor representations by showing that neural processing in both GC and OC exhibits extensive experience-dependent plasticity (Barnes et al. 2011; Chapuis and Wilson 2011; Chen et al. 2011; Moran and Katz 2014; Vincis and Fontanini 2016; Wilson 2010). Several lines of evidence suggest that neural activity observed in response to odor solutions reflects olfactory stimulation. Behavioral work has shown that the concentrations of odorant used here cannot be detected by rats after ablation of the olfactory bulb, suggesting they are below the threshold for stimulating the trigeminal or gustatory system (Rebello et al. 2015; Slotnick et al. 1997). Moreover, our data show that responses to odor solutions in both OC and GC differ in latency from responses to taste solutions that OC responses are driven by respiration and correlated with responses to the same odorants presented orthonasally. These findings argue against gustatory stimulation being the source of responses to intraoral odorants, although the involvement of the trigeminal system cannot be completely ruled out, as is the case in most studies involving odor stimuli (Lübbert et al. 2013; Silver and Moulton 1982). Finally, it is possible that odor-specific mouth movements contribute to the observed neural responses. Recent work on GC has shown that sensory- and motor-related activity overlaps in the response of single neurons (Li et al. 2016). In summary, using electrophysiological recordings in awake rats, the present study demonstrates convergence of consumption-related olfactory, oral somatosensory, and gustatory inputs in primary OC and GC. From a functional perspective, multisensory interactions in primary chemosensory cortex may play a role in mediating flavor perception and food choice behaviors. The present study provides a systematic characterization of single neuron responses to multisensory intraoral stimuli, which constitutes a necessary step toward elucidating the mechanisms underlying multisensory flavor interactions. The rodent model provides an excellent opportunity for future research along these lines by allowing controlled investigation into how experience with specific multisensory flavor cues shapes neural activity in primary chemosensory cortex. From a network perspective, our findings provide further support for the idea that primary OC in particular, and primary sensory systems in general, functions as part of distributed networks

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

Quiroga RQ, Nadasdy Z, Ben-Shaul Y. Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering. Neural Comput 16: 1661–1687, 2004. doi:10.1162/089976604774201631. Rebello MR, Kandukuru P, Verhagen JV. Direct behavioral and neurophysiological evidence for retronasal olfaction in mice. PLoS One 10: e0117218, 2015. doi:10.1371/journal.pone.0117218. Reep RL, Winans SS. Efferent connections of dorsal and ventral agranular insular cortex in the hamster, Mesocricetus auratus. Neuroscience 7: 2609 – 2635, 1982. doi:10.1016/0306-4522(82)90087-2. Roesch MR, Stalnaker TA, Schoenbaum G. Associative encoding in anterior piriform cortex versus orbitofrontal cortex during odor discrimination and reversal learning. Cereb Cortex 17: 643– 652, 2007. doi:10.1093/cercor/bhk009. Rozin P. “Taste-smell confusions” and the duality of the olfactory sense. Percept Psychophys 31: 397– 401, 1982. doi:10.3758/BF03202667. Sadacca BF, Mukherjee N, Vladusich T, Li JX, Katz DB, Miller P. The behavioral relevance of cortical neural ensemble responses emerges suddenly. J Neurosci 36: 655– 669, 2016. doi:10.1523/JNEUROSCI. 2265-15.2016. Sadacca BF, Rothwax JT, Katz DB. Sodium concentration coding gives way to evaluative coding in cortex and amygdala. J Neurosci 32: 9999 –10011, 2012. doi:10.1523/JNEUROSCI.6059-11.2012. Schoenbaum G, Eichenbaum H. Information coding in the rodent prefrontal cortex. I. Single-neuron activity in orbitofrontal cortex compared with that in pyriform cortex. J Neurophysiol 74: 733–750, 1995. Sclafani A, Ackroff K. Glucose- and fructose-conditioned flavor preferences in rats: taste versus postingestive conditioning. Physiol Behav 56: 399 – 405, 1994. doi:10.1016/0031-9384(94)90213-5. Scott TR, Plata-Salamán CR. Taste in the monkey cortex. Physiol Behav 67: 489 –511, 1999. doi:10.1016/S0031-9384(99)00115-8. Shepard TG, Veldhuizen MG, Marks LE. Response times to gustatoryolfactory flavor mixtures: role of congruence. Chem Senses 40: 565–575, 2015. doi:10.1093/chemse/bjv042. Shepherd GM. Smell images and the flavour system in the human brain. Nature 444: 316 –321, 2006. doi:10.1038/nature05405. Shi CJ, Cassell MD. Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. J Comp Neurol 399: 440 – 468, 1998. doi:10.1002/(SICI)1096-9861(19981005)399:4⬍440::AID-CNE2⬎3.0.CO;2-1. Shipley MT, Geinisman Y. Anatomical evidence for convergence of olfactory, gustatory, and visceral afferent pathways in mouse cerebral cortex. Brain Res Bull 12: 221–226, 1984. doi:10.1016/0361-9230(84)90049-2. Shusterman R, Smear MC, Koulakov AA, Rinberg D. Precise olfactory responses tile the sniff cycle. Nat Neurosci 14: 1039 –1044, 2011. doi:10. 1038/nn.2877. Siegel M, Buschman TJ, Miller EK. Cortical information flow during flexible sensorimotor decisions. Science 348: 1352–1355, 2015. doi:10.1126/ science.aab0551. Silver WL, Moulton DG. Chemosensitivity of rat nasal trigeminal receptors. Physiol Behav 28: 927–931, 1982. doi:10.1016/0031-9384(82)90216-5. Slotnick BM, Westbrook F, Darling FM. What the rat’s nose tells the rat’s mouth: long delay aversion conditioning with aqueous odors and potentiation of taste by odors. Anim Learn Behav 25: 357–369, 1997. doi:10.3758/ BF03199093. Small DM, Gerber JC, Mak YE, Hummel T. Differential neural responses evoked by orthonasal versus retronasal odorant perception in humans. Neuron 47: 593– 605, 2005. doi:10.1016/j.neuron.2005.07.022. Small DM, Green BG. A proposed model of a flavor modality. In: The Neural Bases of Multisensory Processes (Murray MM, Wallace MT, eds.). Boca Raton, FL: CRC Press 2012. 22593893 Small DM, Voss J, Mak YE, Simmons KB, Parrish T, Gitelman D. Experience-dependent neural integration of taste and smell in the human brain. J Neurophysiol 92: 1892–1903, 2004. doi:10.1152/jn.00050.2004. Stapleton JR, Lavine ML, Wolpert RL, Nicolelis MA, Simon SA. Rapid taste responses in the gustatory cortex during licking. J Neurosci 26: 4126 – 4138, 2006. doi:10.1523/JNEUROSCI.0092-06.2006. Stein BE, Meredith MA. The Merging of the Senses. Cambridge, MA: MIT Press, 1993, p. xv. Stettler DD, Axel R. Representations of odor in the piriform cortex. Neuron 63: 854 – 864, 2009. doi:10.1016/j.neuron.2009.09.005. Stevenson RJ, Boakes RA, Prescott J. Changes in odor sweetness resulting from implicit learning of a simultaneous odor-sweetness association: an example of learned synesthesia. Learn Motiv 29: 113–132, 1998. doi:10. 1006/lmot.1998.0996. Van Buskirk RL, Erickson RP. Odorant responses in taste neurons of the rat NTS. Brain Res 135: 287–303, 1977. doi:10.1016/0006-8993(77)91032-0.

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

tracing in the rat. J Comp Neurol 376: 265–277, 1996. doi:10.1002/ (SICI)1096-9861(19961209)376:2⬍265::AID-CNE8⬎3.0.CO;2-1. Di Lorenzo PM, Garcia J. Olfactory responses in the gustatory area of the parabrachial pons. Brain Res Bull 15: 673– 676, 1985. doi:10.1016/03619230(85)90219-9. Escanilla OD, Victor JD, Di Lorenzo PM. Odor-taste convergence in the nucleus of the solitary tract of the awake freely licking rat. J Neurosci 35: 6284 – 6297, 2015. doi:10.1523/JNEUROSCI.3526-14.2015. Fortis-Santiago Y, Rodwin BA, Neseliler S, Piette CE, Katz DB. State dependence of olfactory perception as a function of taste cortical inactivation. Nat Neurosci 13: 158 –159, 2010. doi:10.1038/nn.2463. Gautam SH, Short SM, Verhagen JV. Retronasal odor concentration coding in glomeruli of the rat olfactory bulb. Front Integr Nuerosci 8: 81, 2014. doi:10.3389/fnint.2014.00081. Gautam SH, Verhagen JV. Evidence that the sweetness of odors depends on experience in rats. Chem Senses 35: 767–776, 2010. doi:10.1093/chemse/ bjq075. Gautam SH, Verhagen JV. Direct behavioral evidence for retronasal olfaction in rats. PLoS One 7: e44781, 2012a. doi:10.1371/journal.pone.0044781. Gautam SH, Verhagen JV. Retronasal odor representations in the dorsal olfactory bulb of rats. J Neurosci 32: 7949 –7959, 2012b. doi:10.1523/ JNEUROSCI.1413-12.2012. Ghazanfar AA, Schroeder CE. Is neocortex essentially multisensory? Trends Cogn Sci 10: 278 –285, 2006. doi:10.1016/j.tics.2006.04.008. Gire DH, Whitesell JD, Doucette W, Restrepo D. Information for decisionmaking and stimulus identification is multiplexed in sensory cortex. Nat Neurosci 16: 991–993, 2013. doi:10.1038/nn.3432. Green BG, Nachtigal D, Hammond S, Lim J. Enhancement of retronasal odors by taste. Chem Senses 37: 77– 86, 2012. doi:10.1093/chemse/bjr068. Heilmann S, Hummel T. A new method for comparing orthonasal and retronasal olfaction. Behav Neurosci 118: 412– 419, 2004. doi:10.1037/ 0735-7044.118.2.412. Holman E. Immediate and delayed reinforcers for flavor preferences in rats. Learn Motiv 6: 91–100, 1975. doi:10.1016/0023-9690(75)90037-5. Illig KR, Haberly LB. Odor-evoked activity is spatially distributed in piriform cortex. J Comp Neurol 457: 361–373, 2003. doi:10.1002/cne.10557. Katz DB, Simon SA, Nicolelis MA. Dynamic and multimodal responses of gustatory cortical neurons in awake rats. J Neurosci 21: 4478 – 4489, 2001. Krushel LA, van der Kooy D. Visceral cortex: integration of the mucosal senses with limbic information in the rat agranular insular cortex. J Comp Neurol 270: 39 –54, 1988. doi:10.1002/cne.902700105. Li JX, Maier JX, Reid EE, Katz DB. Sensory cortical activity is related to the selection of a rhythmic motor action pattern. J Neurosci 36: 5596 –5607, 2016. doi:10.1523/JNEUROSCI.3949-15.2016. Lübbert M, Kyereme J, Rothermel M, Wetzel CH, Hoffmann KP, Hatt H. In vivo monitoring of chemically evoked activity patterns in the rat trigeminal ganglion. Front Syst Neurosci 7: 64, 2013. doi:10.3389/fnsys.2013. 00064. Maier JX, Blankenship ML, Li JX, Katz DB. A multisensory network for olfactory processing. Curr Biol 25: 2642–2650, 2015. doi:10.1016/j.cub. 2015.08.060. Maier JX, Chandrasekaran C, Ghazanfar AA. Integration of bimodal looming signals through neuronal coherence in the temporal lobe. Curr Biol 18: 963–968, 2008. doi:10.1016/j.cub.2008.05.043. Maier JX, Katz DB. Neural dynamics in response to binary taste mixtures. J Neurophysiol 109: 2108 –2117, 2013. doi:10.1152/jn.00917.2012. Maier JX, Wachowiak M, Katz DB. Chemosensory convergence on primary olfactory cortex. J Neurosci 32: 17037–17047, 2012. doi:10.1523/JNEUROSCI.3540-12.2012. Mak YE, Simmons KB, Gitelman DR, Small DM. Taste and olfactory intensity perception changes following left insular stroke. Behav Neurosci 119: 1693–1700, 2005. doi:10.1037/0735-7044.119.6.1693. Mitra P, Bokil H. Observed Brain Dynamics. Oxford, New York: Oxford University Press, 2008. Miura K, Mainen ZF, Uchida N. Odor representations in olfactory cortex: distributed rate coding and decorrelated population activity. Neuron 74: 1087–1098, 2012. doi:10.1016/j.neuron.2012.04.021. Moran A, Katz DB. Sensory cortical population dynamics uniquely track behavior across learning and extinction. J Neurosci 34: 1248 –1257, 2014. doi:10.1523/JNEUROSCI.3331-13.2014. Paxinos G, Watson C. The Rat Brain Atlas. San Diego, CA: Academic, 1986.

1303

1304

MULTISENSORY FLAVOR RESPONSES IN PRIMARY SENSORY CORTEX

Veldhuizen MG, Nachtigal D, Teulings L, Gitelman DR, Small DM. The insular taste cortex contributes to odor quality coding. Front Hum Neurosci 4: 58, 2010a. doi:10.3389/fnhum.2010.00058. Veldhuizen MG, Shepard TG, Wang MF, Marks LE. Coactivation of gustatory and olfactory signals in flavor perception. Chem Senses 35: 121–133, 2010b. doi:10.1093/chemse/bjp089. Verhagen JV, Wesson DW, Netoff TI, White JA, Wachowiak M. Sniffing controls an adaptive filter of sensory input to the olfactory bulb. Nat Neurosci 10: 631– 639, 2007. doi:10.1038/nn1892. Vincis R, Fontanini A. Associative learning changes cross-modal representations in the gustatory cortex. eLife 5: e16420, 2016. doi:10.7554/eLife.16420. Wesson DW, Carey RM, Verhagen JV, Wachowiak M. Rapid encoding and perception of novel odors in the rat. PLoS Biol 6: e82, 2008. doi:10.1371/ journal.pbio.0060082. Wesson DW, Wilson DA. Smelling sounds: olfactory-auditory sensory convergence in the olfactory tubercle. J Neurosci 30: 3013–3021, 2010. doi: 10.1523/JNEUROSCI.6003-09.2010.

Wilson DA. Single-unit activity in piriform cortex during slow-wave state is shaped by recent odor experience. J Neurosci 30: 1760 –1765, 2010. doi: 10.1523/JNEUROSCI.5636-09.2010. Yasui Y, Breder CD, Saper CB, Cechetto DF. Autonomic responses and efferent pathways from the insular cortex in the rat. J Comp Neurol 303: 355–374, 1991. doi:10.1002/cne.903030303. Yeomans MR, Mobini S, Elliman TD, Walker HC, Stevenson RJ. Hedonic and sensory characteristics of odors conditioned by pairing with tastants in humans. J Exp Psychol Anim Behav Process 32: 215–228, 2006. doi:10. 1037/0097-7403.32.3.215. Zhan C, Luo M. Diverse patterns of odor representation by neurons in the anterior piriform cortex of awake mice. J Neurosci 30: 16662–16672, 2010. doi:10.1523/JNEUROSCI.4400-10.2010. Zinyuk LE, Datiche F, Cattarelli M. Cell activity in the anterior piriform cortex during an olfactory learning in the rat. Behav Brain Res 124: 29 –32, 2001. doi:10.1016/S0166-4328(01)00212-1.

Downloaded from http://jn.physiology.org/ by 10.220.32.246 on March 16, 2017

J Neurophysiol • doi:10.1152/jn.00802.2016 • www.jn.org