Download Gustatory processing is dynamic and distributed Donald B

448
Gustatory processing is dynamic and distributed
Donald B Katz*, Miguel AL Nicolelis† and Sidney A Simon†
The process of gustatory coding consists of neural responses
that provide information about the quantity and quality of food,
its generalized sensation, its hedonic value, and whether it
should be swallowed. Many of the models presently used to
analyze gustatory signals are static in that they use the average
neural firing rate as a measure of activity and are unimodal in
the sense they are thought to only involve chemosensory
information. We have recently elaborated upon a dynamic
model of gustatory coding that involves interactions between
neurons in single as well as in spatially separate, gustatory and
somatosensory regions. We propose that the specifics of
gustatory responses grow not only out of information
ascending from taste receptor cells, but also from the cycling
of information around a massively interconnected system.
Addresses
*Volen Center for Complex Systems, Department of Psychology,
Brandeis University, MS 062, Waltham, MA 02454-9110, USA;
e-mail: [email protected]
† Departments of Neurobiology and Neuroengineering, Duke University
Medical Center, Durham, NC 27710, USA
Correspondence: Donald B Katz
Current Opinion in Neurobiology 2002, 12:448–454
0959-4388/02/$ — see front matter
© 2002 Elsevier Science Ltd. All rights reserved.
Published online 12 July 2002
Abbreviations
AFP
across-fiber patterns
CN
central nucleus
GABA γ-amino butyric acid
GC
gustatory cortex
LL
labeled lines
NST
nucleus of the solitary tract
PbN
parabrachial nuclei of the pons
PSTH
peri-stimulus time histogram
Functional anatomy of the gustatory system
Figure 1a shows a schematic diagram of the principle
gustatory pathways [2,3]. Transduction of chemical information occurs in the oral cavity when chemicals make
contact with taste receptor cells [4••,5••]. Primary gustatory
neurons course within the CNS cranial nerves VII, IX and
X [6] to the nucleus of the solitary tract (NST), which in turn
transmits information to the parabrachial nuclei of the pons
(PbN). From the brainstem, taste information is transmitted
to the thalamocortical system, amygdala and hypothalamus.
Such descriptions of the circuitry usually ignore the fact
that the gustatory system is made-up of networks of
feedforward and feedback pathways. Figure 1b presents a
simple reconceptualization of the system with the goal
of assisting the reader in understanding the dynamic and
distributed nature of gustatory processing. The gustatory
system is separated into interacting taste areas (e.g. gustatory cortex [GC] and NST). In each area, the presence of
local connections among primary neurons and interneurons
imply that between-neuron interactions should occur within
each region. Feedback pathways, meanwhile, suggest that
interactions should occur between, for example, the amygdala
and GC, and between each of these forebrain areas and
both the NST and PbN (Figure 1a). Finally, projections
from somatosensory neurons throughout the oral cavity,
and from visceral neurons in the gut, intermingle with
gustatory neurons in the NST [7] and GC [8], suggesting
that interactions should occur between taste and other
systems. Because of such interactions, gustation might be
expected to involve time-varying responses, as does processing in other sensory systems that contain convergence
and divergence of pathways (e.g. [9–11]).
Currently debated models of taste coding are
static and noninteractive
Introduction
The gustatory system has evolved to detect and discriminate
between foods, to select nutritious diets, and to initiate,
sustain and terminate ingestion [1]. These processes
evolve over several seconds and involve the integration of
multiple sources of information. In this review, we discuss
the neural system that underlies taste behaviors, focusing
particularly on the time-varying nature of taste neural
responses and the neural interactions that may give rise to
such time-varying responses. First, we explain how the
convergence of distributed pathways throughout the
gustatory neural circuitry leads to dynamic responses.
Next, we show that the currently debated static models of
taste coding cannot account for the dynamic and interactive
aspects of gustatory neural processing. Finally, we present
recent data confirming that gustation is dynamic and
distributed, and argue that these data require a more
general systemic theory of gustatory coding.
The long-standing debate over the nature of gustatory
neural coding has primarily addressed the question ‘Are
taste stimuli coded by labeled lines (LL), by across-fiber
patterns (AFP), or by some combination of the two?’
Discussions of these models focus on whether dedicated
sets of neurons signal the presence of a particular taste
component, or whether tastes are represented in the ‘landscape’ of responses across an entire neural population (see
[3]). Both of these models have admirably described how
patterns in ‘taste space’ are arranged (by the clumping of
similar tastes) and how they may change under pharmacological (e.g. amiloride, gymnemic acid) and physiological
(e.g. emesis) manipulations [12–14].
The difference between the LL and AFP theories is less
than might be supposed. Consider a hypothetical ensemble
of 27 neurons whose activity is described in terms of mean
firing rates, and that are assumed to represent the gustatory
Gustatory processing is dynamic and distributed Katz, Nicolelis and Simon
449
Figure 1
(b)
(a)
Cortex
(GC, OFC)
VPMpc
thalamus
Gustatory system
Main taste neuroaxis
Taste area
one
(e.g. GC)
Taste area
two
(e.g. NST)
Amygdala
Brain stem
(PbN, NST)
Hypothalamus
Somatosensory/visceral systems
Information from the oral cavity and gut
(cutaneous, thermal, nociceptive)
CN
VII
CN
IX
CN
X
Current Opinion in Neurobiology
The gustatory system viewed as interacting distributed neural networks.
(a) A schematic of gustatory neural circuitry, showing input to the brain
stem nuclei from cranial nerves (CNs) VII, IX and X, and the most
prominent CNS pathways. Note the many ‘double-headed’ arrows
denoting reciprocal (i.e. both feedforward and feedback) connectivity.
OFC, orbitofrontal cortex; VPMpc, ventroposteriomedial thalamus
parvocellular region. (b) The gustatory system as depicted simply in terms
of intra-area and between-area interactions. Each gustatory nucleus
(in this example GC and NST) contains networks of interconnected
neurons that can be expected to modulate each others firing rates
through time. The two nuclei are themselves reciprocally interconnected,
so that the outputs of each nucleus can be expected to affect activity in
the other. Finally, nuclei comprising the gustatory system interact at all
levels with inputs from the somatosensory and visceral systems.
system’s response to a lingual administration of a tastant
(e.g. 0.1 M NaCl). According to AFP theory (Figure 2a,b),
the tastant is identified as 0.1 M NaCl because of the
particular distribution of responses across the entire
ensemble. According to LL theory, the subset of the
neurons having the strongest activation (neurons 8, 9, 10
and 17) will reflect the particular taste of 0.1 M NaCl. In
other words, the simplest version of LL theory is a special
case of AFP theory, in which an arbitrary threshold causes
one set of fibers to be described as ‘on’ and the others to
be described as ‘off’ (compare Figure 2b,c).
Time-varying gustatory responses
The two theories are also similar in that, in their simplest
forms, neither incorporates two related aspects of gustatory
electrophysiology. The first is the occurrence of timevarying responses. As both models are based on firing rate
averages across 3–5 s, neither takes into account information
in the spike trains, such as adaptation, bursting, or poststimulus response dynamics. In addition, neither theory
allows for the interactions between neurons that would be
expected to induce temporal structure in the neural
responses (for recent reviews, see [15,16]). To the extent
that these phenomena occur, we must seek models that are
dynamic and interactive at multiple levels and timescales.
Recent evidence (given below) indicates that gustatory
neural responses are dynamic and affected by interactions
with other neurons within the same nucleus, with other gustatory nuclei and with ‘non-gustatory’ systems (Figure 1b).
Dynamic properties have recently been recognized in the
responses of neurons in several sensory systems, including
some, such as gustation, in which responses had previously
been described in terms of a single value [17••,18,19]. The
investigation of the role of time in gustatory responses
dates back to the recognition that the firing rates of primary
gustatory neurons peak and then decrease as the system
adapts to prolonged application of gustatory stimuli
[20–22]. By the early 1980s, however, researchers also
began considering the possibility that tastant-specific timecourses might provide information for the identification of
tastants. Di Lorenzo and Schwartzbaum [23] showed that
moderate increases in the information content of NST
spike trains in responses to NaCl and sucrose can be traced
to differences in the time-courses of these neurons’ responses.
More recently, researchers have used mathematical
procedures related to fuzzy set theory to identify a set of
temporal components in NST gustatory responses, showing that membership in a taste category may be based on
‘family resemblance’ related to the placement of taste
responses on multidimensional continua, rather than on
‘crisp’ coding [24]. Different combinations of these
components explained cell-specific time-courses, but there
was no attempt to identify the sources of the time-courses.
To explain the processes contributing to the genesis of
response dynamics, we performed a detailed analysis of
450
Sensory systems
Figure 2
(a)
Average firing rate
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
(b)
‘Portrait’ of NaCl firing rates as interpreted by AFP decoder
(c)
‘Portrait’ of NaCl firing rates as interpreted by LL decoder
A conceptualization of the static across-fiber
pattern and labeled-line models of gustatory
coding. (a) Hypothetical responses given as
average firing rate (over 5 s) minus the
background activity for 27 neurons to a given
a taste stimulus (e.g. 0.1M NaCl). The dark
bars represent the response of those neurons
that responded most strongly to NaCl. The
dashed line represents a hypothetical
threshold of firing rate. (b) A re-presentation
of (a) with color intensity coding the firing rate
for each of the 27 neurons (red, excitation;
blue, inhibition), representing how an AFP
‘decoder’ might read the responses.
(c) Another re-presentation of (a),
representing how a LL decoder might read
the responses. Black denotes that these
neurons’ responses are ‘on’ (as set by the
threshold) and white denotes ‘off’.
Current Opinion in Neurobiology
response dynamics in single-neuron GC ensembles in
awake rats [25••]. We found that GC responses exhibit
complex and neuron-specific time-varying responses to
gustatory stimuli (Figure 3a). Analysis of the spectral and
stimulus information content of these GC responses
revealed that the timings of firing rate changes were
driven, in part, by three distinct contributions. The earliest
was a somatosensory response to the tastant striking the
tongue, the second had a chemosensory origin and the
third was a combination of the two, related to the palatability of the tastant (Figure 3b). In summary, the most
parsimonious explanation of the GC responses is that neural
networks in different cortical areas contribute to the GC
responses at different times (it is worth noting that the
ideas advocated here are compatible with the recent
writings of Erickson, one of the pioneers of taste coding
theory, who has long since abandoned the version of AFP
presented above).
Temporal responses evoked by tastants have also been
investigated in humans using magnetoencephalography [26],
a noninvasive technique that combines the temporal sensitivity of elecroencephalography with the spatial resolution of
functional magnetic resonance imaging. These experiments
have revealed that, over the course of 1.5 s of processing
gustatory information, gustatory cortical regions may be activated repeatedly. The authors concluded that the necessarily
time-extensive process of gustation requires time-varying
neural activation, an hypothesis fully consistent with our
dynamic distributed model of gustatory coding.
Interactions in gustatory responses: I.
Within-area interactions
Over the past decade, interactive processing within neural
populations (typically measured as peaks in cross-correlograms or cross-spectra for pairs of neurons) has received
serious consideration as a potential mechanism of neural
coding. Much of the attention paid to interactive neural processing has been directed towards the search for functionally
significant patterns of ‘synchrony’, that is the near-simultaneous
(timescale of 1–10 ms) firing of individual action potentials
between neurons. Several recent studies have demonstrated
that GC or NST neuron pairs may respond synchronously to
stimulation with particular subsets of tastants [27–30]. These
interactions typically occur only between pairs of neurons
separated by <100 µm, and are most prevalent when the
neurons in the pair produce similar responses to the proffered
tastant [31•]. Such data may reflect the existence of direct
connections between the recorded pair, and have been
interpreted as evidence that nearby GC neurons form part of
columnar processing units [31•].
On a broader (hundreds of milliseconds) timescale, tastespecific cross-correlations of >100 ms duration were found
Gustatory processing is dynamic and distributed Katz, Nicolelis and Simon
451
Figure 3
(a)
sp/s
Gustatory cortical responses are dynamic and
multimodal. (a) Peri-stimulus time histograms
(PSTHs) of a single GC neurons’ responses
to NaCl, citric acid, sucrose and quinine
hydrochloride (Q-HCl). The PSTHs exhibit
time-varying responses both within and
between epochs (i.e. somatosensory contact
of the tastant, chemosensory processing of
the tastant and multi-sensory coding of
palatability). Periods of inhibitory firing rate
change are noted in blue/gray, and a period of
excitatory firing rate change is shown in
orange. Post-stimulus times are on the
abscissae and the response magnitude in
spikes per second (sp/s) is shown on the
ordinates. (b) A schematic of the ‘epochs’ of
tastant responses in GC neurons obtained
from analysis of the GC responses.
Somatosensory responses occur within
200 ms of tastant delivery, as the tastant
strikes the tongue. Chemosensory responses
occur thereafter. Approximately 1 s or more
following tastant delivery, a new
somatosensory contribution that is related to
the production of palatability-specific orofacial
behaviors comes into play, as the stimulus’
hedonic content is now coded. Reproduced
with permission from [25••].
NaCl
Citric acid
Sucrose
Q-HCl
40
20
0
-1
0
2
-1
0
2
-1
0
2
-1
0
2
Time post stimulus (s)
(b)
‘Epochs’
Contact
Chemosensation
Palatability
Input
Chemosensory
Somatosensory
between relatively large percentages of GC neurons in
awake rats [32••]. Figure 4a shows an example of such
cross-correlations. The neuron pair analyzed here crosscorrelated strongly (and negatively) only in the presence of
nicotine (orange line) and citric acid (green line). Peaked
cross correlations were not found for sucrose (black), NaCl
(maroon) or quinine (dark blue). When cross-correlations
are taken into account, 75% of the neurons in GC are
involved in gustation. In contrast, if only average firing
rates are considered, only 10–14% of the GC neurons
contribute to gustatory processing [25••,33].
These cross-correlations, which are found even in crosshemispheric pairs of GC neurons, may reflect state–rate
transitions in large neural networks associated with the
perceptual processing of tastants (Figure 1b; [15,16,34]).
Furthermore, these between-neuron interactions may
explain the previously described time-varying GC
response properties (Figure 3a; [25••]). Recent simulations
of the insect antennal lobe support the hypothesized
relationship between between-neuron interactions and
single-neuron rate changes on a timescale similar to those
observed in GC neurons [35••,36]. In a similar manner,
γ-amino butyric acid (GABA)ergic synapses have been
shown to play a role in shaping gustatory responses in the
GC and NST [37,38]. Applications of the GABAA antagonist bicuculline unmasked taste-related activity in GC and
changed the tastant causing the strongest (best) response
(Figure 4b). In summary, gustatory responses are formed of
the interactions between neurons within individual nuclei
in the CNS (Figure 1b).
0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4
Time post stimulus (s)
Interactions in gustatory responses: II.
Between-region interactions
In addition to interactions within single areas, interactions
between regions also modulate taste responses (see also [39]).
Indeed, recent evidence showed that feedback connections
between GC and NST modulate NST activity via both
excitatory pathways and GABAergic synapses [40•]. Figure 4c
shows an example of a GC–NST interaction. It is clear that
cortical activation (electrical stimulation) may excite or inhibit
NST activity, and that infusion of bicuculline in the NST
blocks the inhibitory impact of GC stimulation. These data
suggest that projections from the GC directly or indirectly
influence GABAergic interneurons in the NST.
Additional evidence for the interaction between different cortical areas comes from studies showing that activating the central
nucleus (CN) in the amygdala modulates gustatory responses in
the PbN [41•]. Figure 4d shows a single PbN neuron’s response
to 0.1 M NaCl in the absence (upper trace) and presence (lower
trace) of a periodic stimulation of the CN. Although amygdalar
stimulation usually inhibited responses to all tastants, in some
cases the impact was excitatory whereas in others it was taste
specific. In summary, these data suggest that theories of gustatory processing should incorporate the real-time modulation of
taste responses by both within-region neural networks and
between-region convergence of activation.
Interactions in gustatory responses: III. Input
from other systems
Early electrophysiological studies reported converging projections from gustatory and lingual somatosensory origins
452
Sensory systems
Figure 4
(a)
(b)
Pre
Bicuc
Post
0.1
Correlation (r)
N
0.0
S
-0.1
H
-0.2
Q
-1.0
-0.5
(c)
0
Lag (s)
Stim +
bicuc
1.0
Stimulus on times
(d)
sp/s
w/o stim
Stim
0.5
w/ stim
NaCl on:
Stim on:
Time
Evidence for intra-area (a,b) and between area (c,d) interactions in
gustatory coding. (a) Cross-correlations for two GC neurons in
response to six different tastants (lag [s] is on the abscissa, correlation
[r] is on the ordinate). Significant interactions can be seen in response
to nicotine (thick orange line) and citric acid (thick green line), but not to
water (light blue), NaCl (maroon), sucrose (black), or quinine (dark
blue). (b) PSTHs for a single GC neuron in response to NaCl (N),
sucrose (S), hydrochloric acid (H), and quinine (Q) depend on inhibition
in the cortical neural network. The first column shows the responses
under control conditions (Pre): note the lack of response to quinine. The
middle column shows the responses to the same stimuli during infusion
of GABA antagonist bicuculline (Bicuc): note that the time courses of
responses have changed, that the stimulus causing the largest
response has changed, and that a response to quinine has been
unmasked. The third column shows the response returning to baseline
following the removal of bicuculline (Post). Bar indicates duration of the
stimulus. (c) Neural activity in NST is affected by feedback from GC.
The responses of two NST neurons to electrical stimulation of GC are
shown (time is on the abscissa, response magnitude [spikes per
second, sp/s] is on the ordinate). In one neuron (top), GC stimulation
caused an excitatory response, whereas in the other (bottom) GC
stimulation caused an inhibitory response. The inhibitory response was
blocked by bicuculline, suggesting inhibitory feedback is caused by GC
activation of GABA-containing neurons. (d) Taste responses in the
pons (parabrachial nuclei) are affected by electrical stimulation of the
amygdala. In each row, vertical lines represent single action potentials.
Time progresses to the right. The top row shows this neuron’s response
to the application of NaCl without amygdala stimulation. The bottom
row shows the same neuron’s response to NaCl in the presence of
stimulation; each burst of stimulation lasted for one second. The
stimulation reduced the NaCl response, but did so in a temporally
complex manner. (a) reproduced with permission from [32••], (b)
reproduced with permission from [37], (c) reproduced with permission
from [40•] and (d) reproduced with permission from [41•].
in several relays of the gustatory pathway, especially at the
cortical level, where somatosensory and taste responses are
found in intermingled cells or even within the same cells
[2,42,43]. In this regard, it may not be so surprising to learn
that mechanically stimulating the tongue can produce
‘taste phantoms’ [44] and that temperature changes on the
tongue can induce or modulate specific tastes [45••]. In
addition, behavioral studies have shown that palatability
can be modulated by trigeminal input [46]. Nevertheless,
there is a paucity of studies showing how the somatosensory system may modulate electrophysiological responses
to gustatory stimuli. In one such study, chorda tympani
responses to tastants have been modulated by electrical
stimulation of the lingual branch of the trigeminal nerve
[47] and, in another study, the trigeminal stimulant
capsaicin has been shown to decrease chorda tympani
responses to NaCl [48].
The effects of visceral stimulation on gustatory responses
have been demonstrated in studies of conditioned taste
aversion [49,50], and in a study showing that NST responses
to tastants are modulated by gastric distension [51]. In
addition, single unit responses from neurons from the
caudolateral part of the orbital frontal cortex markedly
decrease in a taste-specific fashion as monkeys are fed to
satiety [52]. Although it is unclear where the convergences
Gustatory processing is dynamic and distributed Katz, Nicolelis and Simon
take place, what is clear is that gustation depends on
convergence between the taste system and the somatosensory/
visceral system.
Conclusions
We have attempted to justify the reasons for moving
beyond static and unimodal models of gustatory coding
towards models in which processing occurs in time, is
multimodal and involves interactions between neurons in
the same and in spatially separate gustatory regions. We
propose that the specifics of gustatory responses grow not
only out of information ascending from taste receptor
cells but rather from the cycling of information around a
massively interconnected system.
Acknowledgements
We are grateful to Robert Erickson for many fruitful discussions of data and
theory. We also acknowledge support from NIH grants DC-00403 (DBK),
DC-01065 (SAS), and DE-11121 (MALN), and by Philip Morris Incorporated.
References and recommended reading
Papers of particular interest, published within the annual period of review,
have been highlighted as:
• of special interest
•• of outstanding interest
1.
Dethier VG: A comparative role of taste in food intake: a
comparative view. In Mechanisms of Taste Transduction. Edited by
Simon SA, Roper SD. Boca Raton: CRC Press; 1993:3-25.
2.
Norgren R: Gustatory system. In The Human Nervous System.
Edited by Paxinos G. New York: Academic Press; 1990:845-861.
3.
Smith DV, St John SJ: Neural coding of gustatory information. Curr
Opin Neurobiol 1999, 9:427-435.
4.
••
Chandrashekar J, Mueller KL, Hoon MA, Adler E, Feng L, Guo W,
Zuker CS, Ryba NJ: T2Rs function as bitter taste receptors. Cell
2000, 100:703-711.
This is the first description of the G-protein coupling and function of some
T2Rs as bitter taste receptors. This paper also describes some of the polymorphisms found in the cycloheximide receptor of some mouse strains that
have a lesser sensitivity for this compound.
5.
••
Gilbertson TA, Boughter JD Jr, Zhang H, Smith DV: Distribution of
gustatory sensitivities in rat taste cells: whole-cell responses to
apical chemical stimulation. J Neurosci 2001, 21:4931-4941.
Using whole-cell patch clamp methods, these researchers provide impressive evidence that taste cells are broadly tuned and that, in fact, relatively few
taste receptor cells deliver information primarily about one taste.
453
13. Hellekant G, Ninomiya Y, Danilova V: Taste in chimpanzees. III:
Labeled-line coding in sweet taste. Physiol Behav 1998, 65:191-200.
14. Chang F-CT, Scott TR: Conditioned taste aversions modify neural
•
responses in the rat nucleus tractus solitarius. J Neurosci 1984,
4:1850-1862.
15. Salinas E, Sejnowski TJ: Correlated neuronal activity and the flow
of neural information. Nat Rev Neurosci 2001, 2:539-550.
16. Nowak LG, Bullier J: Cross-correlograms for neuronal spike trains.
Different types of temporal correlation in neocortex, their origin
and significance. In Time and the Brain. Edited by Miller R. Sydney,
Australia: Harwood Press; 2000:53-96.
17.
••
Friedrich RW, Laurent G: Dynamic optimization of odor
representations by slow temporal patterning of mitral cell activity.
Science 2001, 291:889-894.
In this paper, which focuses on the timescale of dynamic activity in the
hundreds of milliseconds, the authors outline a comprehensive description
of olfactory bulb function that emphasizes the process of perception, as
opposed to simply the identification of a code. As such, the interpretation of
the presented data integrates mechanism and function.
18. Sugase Y, Yamane S, Ueno S, Kawano K: Global and fine
information coded by single neurons in the temporal visual
cortex. Nature 1999, 400:869-873.
19. Ghazanfar AA, Nicolelis MA: Spatiotemporal properties of layer
V neurons of the rat primary somatosensory cortex. Cereb Cortex
1999, 9:348-361.
20. Mistretta CM: A quantitative analysis of rat chorda tympani fiber
discharge patterns. In Olfaction and Taste IV. Edited by Schneider D.
Stuttgart: Wissenschaftliche; 1971:294-300.
21. Ogawa H, Sato M, Yamashita S: Variability in impulse discharges in
rat chorda tympani fibers in response to repeated gustatory
stimulations. Physiol Behav 1973, 11:469-479.
22. Frank ME, Bieber SL, Smith DV: The organization of taste
sensibilities in hamster chorda tympani nerve fibers. J Gen
Physiol 1988, 91:861-896.
23. Di Lorenzo PM, Schwartzbaum JS: Coding of gustatory information
in the pontine parabrachial nuclei of the rabbit: temporal patterns
of neural response. Brain Res 1982, 251:245-257.
24. Erickson RP, Schiffman SS, Doetsch GS, Di Lorenzo PM,
Woodbury MA: A fuzzy set approach to the organization of the
gustatory system. Primary Sens Neuron 1995, 1:65-80.
25. Katz DB, Simon SA, Nicolelis MA: Dynamic and multimodal
•• responses of gustatory cortical neurons in awake rats. J Neurosci
2001, 21:4478-4489.
When time-varying response properties were taken into account, the
percentage of GC neurons identified as taste-specific jumped to 41%
(compared to 14% when only overall average responses were considered).
The observation of three temporal epochs map remarkably well onto those
discerned via mathematical analysis of NST data (see Figure 3 in [24]). The
authors analyzed activity in the primary somatosensory cortex to disassociate
chemosensory from somatosensory responses.
6.
Travers SP, Pfaffmann C, Norgren R: Convergence of lingual and
palatal gustatory neural activity in the nucleus of the solitary tract.
Brain Res 1986, 365:305-320.
26. Kobayakawa T, Ogawa H, Kaneda H, Ayabe-Kanamura S, Endo H,
Saito S: Spatio-temporal analysis of cortical activity evoked by
gustatory stimulation in humans. Chem Senses 1999, 24:201-209.
7.
Whitehead MC, Frank ME: Anatomy of the gustatory system in the
hamster: central projections of the chorda tympani and the lingual
nerve. J Comp Neurol 1983, 220:378-395.
27.
8.
Barnett EM, Evans GD, Sun N, Perlman S, Cassell MD: Anterograde
tracing of trigeminal afferent pathways from the murine tooth
pulp to cortex using herpes simplex virus type 1. J Neurosci 1995,
15:2972-2984.
28. Adachi M, Ohshima T, Yamada S, Satoh T: Cross-correlation
analysis of taste neuron pairs in rat solitary tract nucleus.
J Neurophysiol 1989, 62:501-509.
9.
Laurent G: A systems perspective on early olfactory coding.
Science 1999, 286:723-728.
Nakamura T, Ogawa H: Neural interaction between cortical taste
neurons in rats: a cross-correlation analysis. Chem Senses 1997,
22:517-528.
29. Yokota T, Eguchi K, Satoh T: Sensitivity of rat cortical neurons in
distinguishing taste qualities by individual and correlative
activities. Chem Senses 1997, 22:363-373.
10. Ghazanfar AA, Nicolelis MAL: The structure and function of
dynamic cortical and thalamic receptive fields. Cereb Cortex 2001,
11:183-193.
30. Yokota T, Eguchi K, Satoh T: Correlated discharges of two neurons
in rat gustatory cortex during gustatory stimulation. Neurosci Lett
1996, 209:204-206.
11. Ringach DL, Hawken MJ, Shapley R: Dynamics of orientation tuning
in macaque primary visual cortex. Nature 1997, 387:281-284.
31. Yokota T, Satoh T: Three-dimensional estimation of the distribution
•
and size of putative functional units in rat gustatory cortex as
assessed from the inter-neuronal distance between two neurons
with correlative activity. Brain Res Bull 2001, 54:575-584.
The authors of this paper measured the spatial and response relationships
between GC neuron pairs showing tight (i.e. probably monosynaptic)
connections between neurons. These careful recordings revealed that direct
12. Minear MM, Hammack SE, Lundy RF, Contreras RJ: Amiloride
inhibits taste nerve responses to NaCl and KCl in
Sprague-Dawley and Fischer 344 rats. Physiol Behav 1996,
60:507-516.
454
Sensory systems
connections between GC neurons are vanishingly rare beyond distances
commensurate with the size of cortical columns, but also that the efficacy of
even direct connections could be modulated in a taste-specific manner.
32. Katz DB, Simon SA, Nicolelis MA: Taste-specific neuronal
•• ensembles in the gustatory cortex of awake rats. J Neurosci 2002,
22:1850-1857.
The authors of this paper produced several simulations in order to test the
various possible sources of GC neuron–neuron interactions. The observed
interactions were not caused by between-trial magnitude coupling, onset
coupling, or spike-to-spike coupling. Instead, they depended on coupling
between single-trial rate changes, and to a significant extent on coupling in
sudden changes in rate. These cross-correlations provided additional
information as to tastant identity.
33. Hanamori T, Kunitake T, Kato K, Kannan H: Responses of neurons in
the insular cortex to gustatory, visceral, and nociceptive stimuli in
rats. J Neurophysiol 1998, 79:2535-2545.
34. Abeles M, Bergman H, Gat I, Meilijson I, Seidemann E, Tishby N,
Vaadia E: Cortical activity flips among quasi-stationary states. Proc
Natl Acad Sci USA 1995, 92:8616-8620.
35. Bazhenov M, Stopfer M, Rabinovich M, Abarbanel HD, Sejnowski TJ,
•• Laurent G: Model of cellular and network mechanisms for odorevoked temporal patterning in the locust antennal lobe. Neuron
2001, 30:569-581.
This paper proposes a unified neural network model that reproduces both
the observed oscillations in the locust olfactory bulb as well as slower
time-varying properties. Variations of the timescale at which time-varying firing
rates have been reported [17••,25••] are attributed to slow GABA synapses.
36. Bazhenov M, Stopfer M, Rabinovich M, Huerta R, Abarbanel HD,
Sejnowski TJ, Laurent G: Model of transient oscillatory synchronization
in the locust antennal lobe. Neuron 2001, 30:553-567.
37.
Ogawa H, Hasegawa K, Otawa S, Ikeda I: GABAergic inhibition and
modifications of taste responses in the cortical taste area in rats.
Neurosci Res 1998, 32:85-95.
38. Smith DV, Li CS: Tonic GABAergic inhibition of taste-responsive
neurons in the nucleus of the solitary tract. Chem Senses 1998,
23:159-169.
39. Di Lorenzo PM, Monroe S: Transfer of information about taste from
the nucleus of the solitary tract to the parabrachial nucleus of the
pons. Brain Res 1997, 763:167-181.
40. Smith DV, Li CS: GABA-mediated corticofugal inhibition of taste
•
responsive neurons in the nucleus of the solitary tract. Brain Res
2000, 858:408-415.
This is an excellent work, which continues from the basic findings reported
in [38], demonstrating that feedback from GC to NST probably contacts
both inhibitory and noninhibitory components of NST networks.
41. Lundy RF Jr, Norgren R: Pontine gustatory activity is altered by
•
electrical stimulation in the central nucleus of the amygdala.
J Neurophysiol 2001, 85:770-783.
This paper not only demonstrates amygdalar feedback onto PbN neurons,
but also examines how this impact affects responses to different tastes.
42. Yamamoto T, Yuyama N, Kawamura Y: Cortical neurons responding
to tactile, thermal and taste stimulations of the rat’s tongue. Brain
Res 1981, 221:202-206.
43. Yamamoto T, Matsuo R, Kiyomitsu Y, Kitamura R: Sensory inputs
from the oral region to the cerebral cortex in behaving rats: an
analysis of unit responses in cortical somatosensory and taste
areas during ingestive behavior. J Neurophysiol 1988,
60:1303-1321.
44. Todrank J, Bartoshuk LM: A taste illusion: taste sensation localized
by touch. Physiol Behav 1991, 50:1027-1031.
45. Cruz A, Green BG: Thermal stimulation of taste. Nature 2000,
•• 403:889-892.
Although the authors recognize the possibility of both peripheral and central
interpretations, this paper demonstrates that purely thermosensory stimuli
may provoke gustatory sensations, and offers the tantalizing possibility that
connections from the system processing temperature may cause activity in
the gustatory system.
46. Berridge KC, Fentress JC: Trigeminal-taste interaction in
palatability processing. Science 1985, 228:747-750.
47.
Wang Y, Erickson RP, Simon SA: Modulation of rat chorda tympani
nerve activity by lingual nerve stimulation. J Neurophysiol 1995,
73:1468-1483.
48. Osada K, Komai M, Bryant BP, Suzuki H, Goto A, Tsunoda K,
Kimura S, Furukawa Y: Capsaicin modifies responses of rat
chorda tympani nerve fibers to NaCl. Chem Senses 1997,
22:249-255.
49. Yasoshima Y, Yamamoto T: Short-term and long-term excitability
changes of the insular cortical neurons after the acquisition of
taste aversion learning in behaving rats. Neuroscience 1998,
84:1-5.
50. Shimura T, Tokita K, Yamamoto T: Parabrachial unit activities
after the acquisition of conditioned taste aversion to a nonpreferred HCl solution in rats. Chem Senses 2002,
27:153-158.
51. Glenn JF, Erickson RP: Gastric modulation of gustatory afferent
activity. Physiol Behav 1976, 16:561-568.
52. Rolls ET: Information processing in the taste system of primates.
J Exp Biol 1989, 146:141-164.