. Author manuscript; available in PMC: 2026 Oct 15.

Published in final edited form as: Cell Rep. 2026 Sep 15;45(9):117990. doi: 10.1016/j.celrep.2026.117990

Locus coeruleus activation transforms cortical taste representations

Will Fan ^1,², Paula To ^1,^2,³, Natale R Sciolino ^1,^2,^3,^4,^5,^6,^7,^*

PMCID: PMC13663908 NIHMSID: NIHMS2212249 PMID: 42747960

SUMMARY

Noradrenergic neurons in the locus coeruleus (LC) shape sensory processing, yet how LC activity influences population taste coding remains unclear. Using optogenetic LC activation with miniscope imaging in the gustatory cortex (GC), we examined LC modulation of multiple taste attributes. Phasic LC activation strengthens correlations between neuronal responses and palatability and expands the dynamic range of stimulus representations along a palatability axis. This expansion is driven by an aversive shift in the representations of all tastants except sucrose, the most palatable stimulus. For mixture ratio and concentration, phasic activation expands and rotates attribute axes, potentially reflecting dependencies between these attributes and palatability. These transformations likely arise from multiplicative gain modulation and more flexible tuning changes. Tonic LC activation affects fewer neurons and does not expand attribute axes. Together, the findings show that LC activation reorganizes GC population geometry in a pattern-dependent manner, linking neuromodulation with feeding behavior and affective processing.

Graphical Abstract

graphic file with name nihms-2212249-f0001.webp

In brief

Fan et al. showed that phasic, but not tonic, locus coeruleus activation strengthens the relationship between cortical activity and palatability while reorganizing taste population geometry to increase separation between highly palatable and less palatable tastes. The findings reveal how neuromodulation rapidly reshapes cortical codes to emphasize behaviorally relevant information.

INTRODUCTION

Sensory processing is adaptively shaped by an animal’s internal state, including fluctuations in arousal, attention, motivation, and stress.^1–6 One of the best studied sources of modulatory input to sensory circuits is the noradrenergic locus coeruleus (LC).^7–10 LC activity tracks behavioral state and is commonly described in phasic and tonic firing modes¹¹: phasic activation is linked to vigilance, attentional shifts, and motivated actions,^12–14 whereas elevated tonic firing predominates during stress and hyperarousal.^15,16 Through its widespread projections to nearly all sensory networks,¹⁷ the LC can broadly influence circuit dynamics through norepinephrine (NE) release. LC stimulation and pharmacological NE application produce diverse effects across sensory regions, including bidirectional modulation of responses,^18–20 changes in tuning selectivity,^21–23 improvement in signal-to-noise ratio,^24,25 and alterations in response timing.^18,19,26 These effects depend on brain region^18,27 and LC activity pattern.^26,28

Although LC’s modulatory effects have been well-characterized at the single neuron level, it remains unclear how LC activity causally reshapes neural population codes, the level at which perception and behavior emerge.^29,30 Population responses can be represented as points in high-dimensional activity space, where each axis corresponds to the activity of one neuron. Across stimuli, these points form a low-dimensional manifold whose geometry captures the relationships among stimuli as encoded by the population.³¹ Neuromodulation can scale, shift, rotate, or warp this geometry, with distinct implications for feature encoding and downstream readout. Despite the promise of this approach, only a few studies have applied population-based analyses to LC modulation,^27,28,32 and how LC activity transforms sensory coding geometry remains poorly understood. Defining these transformations could reveal general principles of how neuromodulators bias sensory processing toward behaviorally relevant stimulus features.

In the present study, we investigated how distinct patterns of LC activation influence taste representations in the primary gustatory cortex (GC) at both the single-neuron and population levels. GC is a strong model system for several reasons. First, compared to other sensory modalities, gustation remains underrepresented in the neuromodulation literature, with no prior work directly testing the causal impact of LC activity on GC coding. Second, GC population activity encodes not only taste identity and intensity, but also palatability^33–36—a behaviorally relevant dimension linking sensory coding to affective processing and feeding behavior. Third, GC palatability coding is flexibly regulated by behavioral states such as attention³⁷ and sickness,³⁸ but the pathways mediating these effects remain unknown. Given LC’s role in adaptive sensory-guided decisions,^27,39–41 we hypothesized that LC activation would reorganize GC population geometry by enhancing palatability representations, among other impacts.

To test this hypothesis, we combined miniscope calcium imaging in GC with optogenetic LC activation in awake mice receiving passively delivered taste stimuli. We examined phasic and tonic LC activation across three taste attributes: palatability, mixture ratio, and concentration. Phasic LC activation enhanced the correlation between neuronal responses and palatability and expanded the dynamic range of taste representations along a palatability-relevant axis. Similar stretching occurred for mixture ratio and intensity; however, the corresponding population axes also rotated, potentially due to their correlation with palatability. These population-level changes likely arise from a combination of gain modulation and flexible tuning changes. In contrast, tonic LC stimulation modulated fewer GC neurons and did not produce stretching along any attribute axis, despite inducing greater NE release. Together, these findings show that LC activity reshapes GC population coding in a pattern-dependent manner and provide causal insights into how neuromodulation reorganizes sensory representations of behaviorally relevant stimulus attributes.

RESULTS

Probing the effects of phasic LC activation on taste-evoked GC activity

To determine how phasic LC activation influences taste-evoked responses in GC, we expressed GCaMP8m (CaMKIIα-driven) in GC excitatory neurons and the excitatory opsin ChrimsonR (Flp dependent) in NE-producing LC neurons of Dbh^Flpo mice¹⁷ (Figure 1A). A relay lens was implanted over GC for miniscope imaging, and an optical fiber was positioned above the ipsilateral LC for optogenetic stimulation. ChrimsonR-expressing LC axons densely innervated GC and intermingled with GCaMP8m-expressing neurons (Figures 1B and 1C).

To control stimulus delivery, mice were implanted with an intraoral cannula (IOC) (Figure 1D). Tastants were delivered through four independent microtubes within the IOC, preventing cross-stimulus mixing. Each session consisted of 96 trials with repeated presentations of four stimuli (Figure 1D). On half of the trials, tastant delivery was preceded by a 0.5-s phasic LC stimulation train (0.5 s; 20 Hz, 5-ms pulses), ending 0.5 s before taste delivery. This stimulation timing to mimic LC activation that precedes feeding⁴² and may prime taste processing. To test whether LC-dependent modulation varied across sensory dimensions, we examined three stimulus sets that differed along defined attributes: palatability (sucrose > NaCl > citric acid > quinine), mixture composition (four sucrose-NaCl ratios), and intensity (four sucrose concentrations) (Figure 1E).

To confirm that this stimulation protocol increased LC output to GC, we performed validation experiments in separate Dbh^Flpo mice expressing the NE sensor GRAB_NE2h⁴³ in GC and Flp-dependent ChrimsonR in the ipsilateral LC (Figures 1F and 1G). Using spectrally resolved fiber photometry,^42,44,45 we measured NE dynamics in GC with the same stimulation timing as in miniscope experiments, but without tastant delivery (Figure 1D). Phasic LC stimulation robustly increased GC NE release across mice (Figures 1H and 1I). Although NE fluorescence did not fully return to baseline within the 25–35 s inter-trial interval, this prolonged decay was likely exaggerated by GRAB_NE2h off-kinetics (τ_off = 1.93 s⁴³), as NE transients evoked by brief LC or NE pathway stimulation typically return to baseline within 10 s in fast-scan cyclic voltammetry studies.^46–50 Thus, it is unlikely that endogenous NE signals persist across successive trials. Notably, NE fluorescence signals during stimulation (stim) trials were clearly distinguishable from the no-stimulation (no-stim) trials (Figure 1J), confirming the efficacy of our stimulation protocol and supporting comparisons of GC activity across conditions.

PHASIC LC ACTIVATION BIDIRECTIONALLY MODULATES TASTE-EVOKED GC RESPONSES

We first examined the effects of phasic LC activation on GC responses to four basic taste stimuli that vary in palatability. We identified neurons exhibiting tastant-dependent responses in stim and/or no-stim trials (hereafter termed taste-discriminative neurons). Among these, 20.8% showed differential responses in stim versus no-stim trials for at least one tastant; these were classified as LC-modulated neurons (Figure S1A). In tdTomato control mice, only 2.88% of taste-discriminative neurons met this criterion (Figure S1A), suggesting that the effect largely reflect LC activation rather than photo-stimulation artifacts.

Consistent with findings in other LC target regions,^18–20 phasic LC activation produced both enhancement and suppression of GC taste responses (Figure 2A). In some cases, LC modulation recruited previously non-responsive neurons (cell 2529) or fully suppressed excitatory responses, occasionally producing sign reversals (cell 1655) (Figure 2A). Across all significant cell-tastant pairs, individual neurons could be modulated by multiple tastants, but the direction of modulation within each neuron was consistently enhancing or suppressive (Figure 2B). The proportions of LC-enhanced and LC-suppressed neurons were roughly balanced.

Figure 2. — (A) Mean taste-evoked responses from six representative GC neurons with and without LC stimulation. Tastants were delivered at 0 s; shading indicates ±SEM.

(B) *Top*: modulation direction of all LC-modulated GC neurons. Each tick shows the effect of LC activation on responses to sucrose (S), NaCl (N), citric acid (C), or quinine (Q). Gold and teal indicate LC-enhanced and LC-suppressed cell-taste pairs, respectively. *Bottom*: same data sorted by modulation direction.

(C) Decoding accuracy for LC-modulated, taste-discriminative GC neurons across population sizes. Data are mean ± SD of bootstrapped means. Non-significant (n.s.) permutation tests.

(D) Correlations between response magnitude and palatability rank across LC-modulated neurons and time bins. Two-way RM ANOVA: time × stimulation (F_{5.766, 1418}= 6.259, p < 0.0001); stimulation (F_{1.000, 246.0}= 13.41, p = 0.0001); time (F_{5.112, 1258}= 33.98, p < 0.0001). Bonferroni post hoc: ***p < 0.001, **p < 0.01, *p < 0.05. Mean ± SEM (247 cells).

(E) PCA of LC-modulated GC population responses to tastants (12 trials/condition). Arrows indicate palatability axes.

(F) Projection of responses onto the corresponding palatability axis.

(G) Permutation tests for dynamic-range ratio (left) and palatability-axis angle (right) in LC-modulated neurons. Violin plots show null distributions (median, solid line; quartiles, dashed); magenta lines indicate observed values.

(H) Responses projected onto a common palatability axis defined from no-stimulation trials. two-way ANOVA: tastant (F_3,88= 482.7, p < 0.0001); stimulation (F_1,88= 40.22, p < 0.0001); tastant × stimulation (F_3,88= 2.824, p = 0.0433). Bonferroni’s post hoc: ***p < 0.001 Black lines indicate means; dots represent individual trials (12 trials/condition).

Phasic LC activation enhances GC palatability correlation without altering tastant separability

To determine how LC activation alters GC taste encoding, we analyzed a pseudo-population of LC-modulated, taste-discriminative neurons pooled across mice. To assess population-level taste discriminability, we trained linear decoders to classify tastants using population dynamics. Decoding accuracy increased monotonically with population size; however, accuracy did not differ significantly between stim and no-stim trials (Figure 2C), suggesting that phasic LC activation does not alter overall tastant discriminability in GC.

We next examined palatability encoding using a standard four-tastant panel.^36,51–54 These tastants differ in hedonic value to mice, with the expected ranking of sucrose > NaCl > citric acid > quinine, which we confirmed using a brief-access taste test (Figure S2). To quantify palatability-related information at the single-neuron level, we calculated the absolute rank correlation (Kendall’s tau) between each neuron’s mean response to the four tastants and their palatability ranking^34,55 (Figure 2D). Palatability correlation increased gradually following tastant delivery and peaked around 3 s. This time course is delayed relative to prior electrophysiology studies,^33,36 in which palatability coding peaks ~1–2.5 s after taste delivery, likely reflecting the slower kinetics of calcium imaging. Among LC-modulated neurons, mean palatability correlation was significantly higher in stim than no-stim trials, demonstrating that phasic LC activation enhances palatability coding at the single-neuron level.

Phasic LC activation reshapes GC population coding along a palatability axis

To examine how LC activation reshapes GC palatability coding at the population level, we performed principal component analysis (PCA) on evoked responses of LC-modulated neurons. Tastant responses formed a low-dimensional manifold organized primarily along a palatability-related axis (Figure 2E). We approximated the palatability axis separately for stim and no-stim conditions as vectors connecting the centroids of palatable (sucrose, NaCl) and aversive (citric acid, quinine) tastants. The most prominent effect of phasic LC activation was a coherent shift in the response manifold that was largely orthogonal to the palatability axes (Figure 2E), which may reflect changes in other GC coding directions (see Discussion).

To isolate LC-induced changes related to palatability coding, we projected responses onto their respective palatability axes (Figure 2F). Phasic LC activation increased the separation between sucrose and quinine, i.e., the dynamic range of the palatability axis, by 1.55-fold. A permutation test confirmed that this increase significantly exceeded that expected from trial-to-trial variability (Figure 2G). We next tested whether the palatability axis itself shifted in orientation and found no change between stim and no-stim conditions (Figure 2G). Both the dynamic range expansion and axis stability persisted when the analysis was extended to the broader population of taste-discriminative neurons (Figures S3A and S3B).

Pairwise tastant distances revealed that this expansion was non-uniform and was driven primarily by increased separation between sucrose and the remaining tastants (Figure S4A). Because palatability axis orientation was stable across conditions, we projected responses onto a common (no stim derived) axis and found that phasic LC modulation shifted NaCl, citric acid, and quinine responses toward an aversive direction, while leaving sucrose unchanged (Figure 2H). Thus, the increased separation between sucrose and other tastants reflects a reorganization of palatability representation rather than a sucrose-specific effect.

To determine whether LC modulation expanded the stimulus manifold beyond the palatability dimension, we quantified dispersion, the mean perpendicular distance from the palatability axis. Phasic LC activation did not significantly alter dispersion, though trended toward expansion (p = 0.074; Figure S5A). Together with the non-uniform pairwise tastant distance changes, this indicates that LC modulation does not simply scale the population representation isotropically. Procrustes analysis further showed that LC modulation reorganized tastant centroids into a configuration more closely aligned with palatability order (Figures S5B and S5C).

Phasic LC activation stretches and rotates GC representations of mixture ratios

Given GC’s role in encoding mixture composition,^56,57 we examined how phasic LC activation affects responses to sucrose-NaCl mixtures (90:10, 65:35, 35:65, and 10:90). Because sucrose and NaCl differ in palatability (Figure S2), these mixtures likely vary in both mixture-ratio and palatability. Consistent with the palatability experiment, 20.6% of mixture-ratio-discriminative neurons were LC-modulated (Figure S1A), including both LC-enhanced and LC-suppressed subtypes (Figure S6A). Among LC-modulated neurons, mixture-ratio decoding accuracy did not significantly differ between stim and no-stim trials (Figure 3A), indicating that phasic LC activation did not alter overall mixture separability. In contrast, LC activation significantly increased the correlation between neuronal responses and mixture-ratio ranking (Figure 3B), suggesting enhanced encoding of graded mixture differences.

To characterize population-level effects, we performed PCA on mixture responses (Figure 3C). We defined a mixture-ratio axis by connecting the 90:10 and 10:90 centroids and projected responses onto the respective axis. Phasic LC activation increased the dynamic range along the mixture axis (Figures 3C–3E). Pairwise distance analysis showed that this expansion was non-uniform and driven primarily by larger separations between mixtures with greater compositional differences (Figure S4B). LC activation rotated the mixture axis (Figure 3E), possibly reflecting differential stretching of latent dimensions encoding pure mixture composition and palatability (see Discussion). Both stretching and axis rotation persisted when analyses were extended to all mixture-ratio-discriminative neurons (Figures S3C and S3D), indicating that the effect was robust across the broader population. Dispersion showed a trend toward reduced separation from the mixture axis (Figure S5D). Together with increased dynamic range, this supports selective elongation of the stimulus manifold along a mixture-relevant axis, as visualized by Procrustes analysis (Figures S5E and S5F).

Phasic LC activation modestly expands and rotates GC representations of sucrose concentration

We next examined LC’s effect on GC encoding of tastant intensity using four sucrose concentrations (50, 100, 250, and 500 mM). Because sucrose palatability increases monotonically with concentration,^58,59 these stimuli likely vary in both intensity and palatability. Consistent with prior results, 19.3% of sucrose-concentration-discriminative neurons were LC modulated (Figure S1A), comprising both LC-enhanced and LC-suppressed subtypes (Figure S6B).

Among LC-modulated neurons, phasic LC activation did not significantly alter decoding accuracy or the correlation between neuronal responses and concentration (Figures 3F and 3G), indicating limited modulation of intensity coding. Nevertheless, PCA revealed a stimulation dependent increase in dynamic range along the sucrose-concentration axis (Figures 3H–3J). Pairwise distance analysis showed only modest changes, with only the 250 versus 50 mM comparison surviving correction (Figure S4C), consistent with relatively weak dynamic-range expansion. LC modulation also induced a significant rotation of the concentration axis (Figure 3J), possibly reflecting differential stretching of latent dimensions encoding concentration and palatability. When analyses were extended to all sucrose-concentration-discriminative neurons, axis rotation persisted, whereas dynamic-range stretching was no longer significant (Figures S3E and S3F), suggesting that axis rotation is the more robust population-level effect. Dispersion showed a significant LC-induced expansion orthogonal to the concentration axis (Figures S5G–S5I), possibly reflecting non-linear concentration coding not captured by the approximated axis.

Parsing tuning modulations underlying population-level effects of phasic LC activation

The preceding analyses showed that phasic LC activation induces distinct geometric transformations of GC population activity, including (but not limited to) dynamic-range expansion and axis rotation (Figure 4A). Because representational geometry is determined by neuronal tuning,³¹ we next examined LC-induced tuning changes underlying these transformations. We defined each neuron’s tuning profile as its mean response across a tastant set, and, using a model comparison approach, classified LC-induced changes as scale-shift modulation, flexible modulation, or no modulation for each experiment (Figure 4B). Scale-shift modulation reflects a uniform transformation of the tuning profile, captured by a common multiplicative factor and additive offset across tastants. In contrast, flexible modulation reflects tastant-specific changes that cannot be explained by uniform scaling or shifting. Example scale-shift and flexible neurons from the palatability experiment are shown in Figures 4C and 4D, respectively. Scale-shift modulation preserved the relative response pattern across tastants, whereas flexible modulation could reorganize it, including changes in rank order. Across experiments, basic-taste-discriminative neurons showed more scale-shift modulation and less flexible modulation than mixture-ratio- and sucrose-concentration-discriminative neurons (Figures 4E, 4H, 4K, and S7A).

Figure 4. — (A) Summary of LC-induced transformations across stimulus-discriminative GC neurons.

(B) Schematic of scale, shift, and flexible modulation. Gray indicates tuning without LC stimulation.

(C and D) Example GC neurons showing scale-shift (C) or flexible modulation (D) during no-stimulation (left) and stimulation (right) trials. Curves show mean responses to sucrose (S), NaCl (N), citric acid (C), and quinine (Q).

(E, H, and K) Proportion of modulation types among basic taste (E), mixture ratio (H), and sucrose-concentration discriminative (K) GC neurons.

(F, I, and L) Linear regressions of absolute modulation weight versus log_2.

LC-mediated gain modulation is structured and attribute specific

We next examined the additive and multiplicative factors associated with scale-shift-modulated neurons, given their potential contribution to population-level transformations. Additive factors shift the population manifold without changing its shape or size, whereas multiplicative factors can expand, compress, and/or rotate it. Both additive and log-transformed multiplicative factors spanned positive and negative values (Figure S7B). Median log-multiplicative factors were positive across all experiments, indicating an overall bias toward expansion (Figure S7C). Multiplicative effects were less expansive in the sucrose-concentration experiment than in the mixture experiment, consistent with the more modest expansion of sucrose dynamic range.

Given the relevance of multiplicative effects (gain modulation) to manifold expansion, we next examined whether multiplicative factors related to each neuron’s contribution to the coding axis, measured by its absolute axis weight. No correlation was observed in the palatability experiment (Figure 4F), whereas significant negative correlations emerged in the mixture-ratio and sucrose-concentration experiments (Figures 4I and 4L). Thus, under phasic LC modulation, neurons contributing less strongly to these coding axes were preferentially expanded, whereas stronger contributors were relatively compressed.

Distinct contributions of scale-shift and flexible modulation to LC-induced transformations

We next applied PCA separately to scale-shift and flexibly modulated neurons to determine their contributions to coding-space transformations. Scale-shift modulation increased dynamic range in the palatability and mixture-ratio experiments (Figures 4G and 4J), but not in the sucrose-concentration experiment (Figure 4M), consistent with the stronger expansion bias of multiplicative factors in the former two datasets (Figure S7C). Despite its potential to reweight neuronal contributions, scale-shift modulation did not induce axis rotation in any experiment. Thus, scale-shift modulation primarily contributed to LC-induced dynamic-range expansion.

In contrast, flexible modulation induced significant axis rotation across all experiments (Figures 4G, 4J, and 4M), consistent with its ability to reshape neuronal tuning. Flexible modulation also increased dynamic range in the sucrose-concentration experiment, but not in the palatability or mixture-ratio experiments. Analysis of rank correlations revealed that scale-shift modulation increased response-stimulus correlation in the palatability and mixture experiments (Figures S7D, S7F, and S7H), whereas flexible modulation increased these correlations across all attributes (Figures S7E, S7G, and S7I). Together, these findings indicate that scale-shift and flexible modulation make distinct contributions to LC-induced population transformations, with the former primarily supporting dynamic-range expansion and the latter driving axis rotation.

Tonic LC activation has limited impact on GC taste encoding

Previous work suggests that phasic and tonic LC activation can produce qualitatively different effects on downstream circuits.^26,28,60 To examine the impact of tonic LC activation on GC taste encoding, we conducted three additional experiments using tastants varying in palatability, mixture ratio, and sucrose concentration, paired with an alternative stimulation protocol. Tastant trials were divided into eight blocks, half of which included 5-Hz tonic LC stimulation (Figure 5A). Validation experiments confirmed that this protocol elicited robust NE release that persisted throughout each stimulation block (Figures 5B and 5C).

Figure 5. — (A) Tastant deliveries were divided into eight blocks (12 deliveries per block). Tonic LC photostimulation (5 Hz) occurred in four blocks. Each stimulation block is followed by a gap period (145–155 s) before the onset of the next block.

(B) Example fiber-photometry trace showing NE dynamics evoked by tonic LC stimulation; vertical lines indicate tastant timing (no tastant was delivered in this validation experiment).

(C) NE fluorescence averaged over 1-s epochs for tonic-stimulation and no-stimulation trials. two-way ANOVA: stimulation (F_2,282 = 5392, p < 0.0001); stimulation × subjects (F_2,282 = 97.60, p < 0.0001); subjects (F_2,282 = 14.00, p < 0.0001). Šídák’s post hoc: ***p <0.001. Dots show n = 48 trials/condition (3 mice); lines show means.

(D, F, and H) Population responses of LC-modulated neurons projected onto the palatability (D), mixture-ratio (F) and sucrose-concentration (H) axes for stimulation (right) and no-stimulation (left) trials (12 trials/condition).

(E, G, and I) Permutation tests for dynamic-range ratios (left) and axis angles (right) for tonic LC-modulated neurons in the palatability (E), mixture-ratio (G), and sucrose-concentration (I) experiments. Violin plots show permuted distributions (median, solid black; quartiles, dashed), with observed values in magenta.

Despite robust NE release, tonic LC activation modulated only ~10%–12% of stimulus-discriminative neurons across attributes (Figure S1B), approximately half the proportion observed under phasic stimulation (Figure S1A), though still higher than in tdTomato controls. Tonic activation did not affect palatability or mixture-ratio decoding but reduced decodability for sucrose concentration (Figure S8). Tonic activation also did not alter response-stimulus correlations for any taste attribute (Figure S8). PCA showed that tonic LC activation produced axis rotation across all attributes without changing dynamic range (Figures 5D–5I). When all stimulus discriminative neurons were included, rotation persisted for mixture ratio and sucrose concentration, but not palatability (Figure S9). Together, these findings suggest that, unlike phasic activation, tonic LC activation produced modest transformations that are not clearly aligned with behaviorally relevant stimulus dimensions.

DISCUSSION

Here, we examined how LC activation shapes cortical taste encoding at single-neuron and population levels. A central advance of this study is that it links LC activation patterns to specific geometric transformations of gustatory cortical population activity, revealing how a major neuromodulatory system can reorganize taste representations across multiple stimulus dimensions. Phasic LC activation (1) bidirectionally modulated taste-evoked responses, with each neuron showing a consistent direction of modulation, and (2) enhanced correlations between neuronal responses and graded stimulus properties in the palatability and mixture-ratio experiments. At the population level, phasic LC activation (1) stretched the dynamic range of GC activity along attribute-relevant axes and (2) rotated the mixture-ratio and sucrose-concentration axes, without affecting the palatability axis. Dynamic-range stretching was driven by structured multiplicative modulation, whereas axis rotation reflected flexible reshaping of neuronal tuning. In contrast to phasic LC activation, tonic activation influenced fewer GC neurons and produced axis rotation without dynamic-range stretching. Together, these results indicate that brief LC activation reorganizes GC population coding in a pattern-dependent manner, providing understanding of how neuromodulation biases behaviorally relevant taste representations, particularly along the palatability dimension.

Population-level effects of LC modulation

The geometry of population activity determines how sensory information is represented. Expansion along an attribute-relevant axis is linked to increased perceptual separation along that dimension, and selective attention can stretch neural representations along task-relevant axes.^61,62 Although LC activity is more commonly associated with global arousal than with top-down control, it may nonetheless bias information processing toward specific sensory attributes. Consistent with this idea, phasic LC activation primarily increased the separation between sucrose and other basic tastes along the palatability dimension without expanding orthogonal dimensions, suggesting selective reorganization of palatability coding. The absence of expansion along other taste-differentiating dimensions may help explain the unchanged decoding performance under LC modulation. These findings are consistent with prior work showing that behavioral state differentially influences taste identity and palatability representations in GC.^37,38 Notably, the palatability shift was driven mainly by aversive shifts in NaCl, citric acid, and quinine responses, whereas sucrose responses remained largely unchanged. This effect differs from other forms of palatability reorganization previously observed in GC. For instance, task disengagement increases the separation between appetitive and aversive tastants,³⁷ while familiarization clusters appetitive tastants.⁶³ In our study, the aversive shift in all tastants except the most palatable one could bias animals toward selecting highly palatable foods, a potentially adaptive strategy when navigating unfamiliar foraging environments. The relative stability of sucrose representations may relate to its caloric value, a possibility that could be tested by comparing LC modulation of sucrose and non-caloric sweeteners.

Axis rotation provides an additional way to alter population geometry. The rotation of the mixture and sucrose-concentration axes admits several interpretations that do not require alterations in downstream readout. One interpretation is that these axes reflect blending of two latent dimensions: objective stimulus properties and palatability. If LC modulation scales these dimensions unequally, the resulting attribute axis would rotate toward the dimension undergoing greater expansion. An intriguing possibility is that the latent palatability dimension underwent preferential expansion, which could explain the population-level transformation observed across all three phasic stimulation experiments. This interpretation is also consistent with our observation that multiplicative modulation expands low-contribution neurons while compressing high-contribution neurons. If these groups preferentially represent distinct latent dimensions, such opposing gain modulation would reweight those dimensions. Alternatively, LC activation may recruit previously non-discriminative neurons into stimulus encoding, effectively pulling the attribute axis toward the subspace they span. These neurons would likely be classified as flexibly modulated, consistent with their role in both driving axis rotation and increasing response-stimulus correlations. Importantly, axis estimates could be sensitive to noise, particularly when the defining centroids are close together, and thus the precise angle should not be interpreted literally. This variability is addressed here by assessing significance through permutation testing.

Beyond stretching and rotation, phasic LC activation produced a sizable coherent shift of the response manifold largely orthogonal to the palatability axis, which is likely driven by the additive (shifting) components of scale-shift modulation. Although this shift could reflect another taste-related coding dimension, we did not observe large displacements aligned with the mixture-ratio or sucrose-concentration axes. Alternatively, the LC-induced shift orthogonal to the palatability axis may reflect an internal-state dimension, such as arousal, that is linked to LC output.^64,65 This shift may also represent causal or compensatory changes in the network that enable or accompany expansion and rotation in taste-relevant coding dimensions.

Mechanisms of LC-induced neuronal effects

LC modulation can shape cortical taste responses through adrenergic receptor-mediated effects on GC neurons, local recurrent GC circuits, and/or other brain regions that influence GC. Both glutamatergic and GABAergic GC neurons express transcripts for several adrenergic receptors, including the α1-, α2-, and β1-receptor subtypes.⁶⁶ Activation of these receptors can influence intrinsic and synaptic properties of cortical neurons,^67–71 potentially contributing to the enhancement, suppression, and diverse tuning effects observed here. LC output may also influence neuronal responses indirectly through its effects on other cortical microcircuit components, including inhibitory neurons, which were not examined in this study. Additionally, changes in GC activity may reflect LC’s impact on other nodes of the taste pathway, such as the gustatory thalamus and basolateral amygdala, which primarily convey chemosensory^72–74 and hedonic information,^52,75,76 respectively. Thus, our results may reflect both local adrenergic modulation within GC and LC-driven changes across the broader gustatory network. How LC influences GC neurophysiology and inter-regional interactions remains an important open question.

Characterizing whether LC enhances or suppresses GC responses to different tastants alone is not sufficient to explain population reorganization. Population-level effects such as palatability-axis expansion and coding-axis rotation depend on how tuning profiles are reshaped. To examine LC-mediated tuning changes, we examined scale-shift and more flexible modulatory effects. Decomposing modulation into additive and multiplicative components (scale-shift) is a common framework for characterizing modulatory effects on neuronal tuning.^77–79 Multiplicative (gain) modulation, in particular, is a well-established mechanism by which the LC-NE system can facilitate cognition and information processing.^11,80–82 Although the cellular mechanisms underlying additive and multiplicative modulation have been extensively studied,^79,83,84 the basis of flexible modulation remains less clear, given its heterogeneous nature. Recruitment of coding neurons, a form of flexible modulation described above, may reflect unmasking of attribute-relevant inputs. Alternatively, flexible modulation may arise from differential scaling of narrowly tuned, labeled-line-like inputs, effectively altering the relative contribution of distinct sensory streams. Importantly, our modulation classes are descriptive rather than mechanistic. This framework provides a way to relate diverse single-neuron tuning changes to the reorganization of GC population coding induced by LC activation.

Links to state-dependent modulation

Prior work has shown enhanced GC palatability coding during task disengagement³⁷ and LiCl-induced sickness.³⁸ Both behavioral conditions are accompanied by shifts in cortical state reflected in local field potential patterns. Because LC-NE activity is tightly coupled to cortical state^85,86 and can rapidly modulate it,^87–90 LC signaling may provide a route through which internal state influences sensory coding by altering cortical state. Whereas these prior studies examined behavioral states lasting tens of minutes, our results show that gustatory representations can be reorganized on the timescale of transient LC activation, contributing to moment-to-moment variability in sensory responses.⁴

Understanding LC function requires moving beyond general labels such as “arousal” to consider specific LC dynamics in sensory-guided behaviors. In the context of taste, prior work shows that LC phasic activation occurs immediately before consumption⁴² and during brief gaps between lick bouts⁴⁵; these dynamics were mimicked by our phasic stimulation pattern and may transiently prime GC before each influx of gustatory input. LC activity is also suppressed during consumption bouts,^12,42,45 potentially stabilizing sensory representations by reducing LC-driven fluctuations. More broadly, many learning paradigms deliver gustatory rewards immediately after reward-seeking actions, so phasic LC activation accompanying these actions^13,14,91 may extend into reward processing and influence learning.

The distinct effects of phasic and tonic LC activation highlight the importance of activation pattern. A classic framework proposes that moderately aroused states, associated with phasic LC activation, optimize information processing, whereas hyper-aroused states, associated with high tonic LC activation, do not.^2,11,92 Although prior studies have shown that phasic and tonic LC activation can differentially affect sensory processing,^26,28 our results show that the sensory-enhancing effects of phasic LC activation on GC neurons were not evident under high tonic activation. However, these differences may also reflect stimulation timing, because tonic stimulation occurred throughout the experiment whereas phasic stimulation occurred selectively before tastant delivery. The timing of LC activity relative to sensory stimuli has been proposed to partially account for the differential effects of phasic and tonic LC stimulation on sensory processing.⁹³ Future experiments comparing the same LC stimulation pattern before versus during tastant delivery will be needed to isolate timing effects relative to sensory input. Tonic LC activation reduced decodability only for sucrose-concentration-coding neurons, suggesting a selective diminishment in intensity representation. A proposed mechanism for divergent effects of LC activity patterns is that phasic activation preferentially engages high-affinity α1 receptors through moderate NE release, whereas sustained tonic activity recruits lower-affinity β receptors with distinct cellular consequences.⁹⁴ Consistent with our findings, recent fMRI work suggests that phasic LC activation produces brain-wide interaction patterns that may favor sensory processing relative to tonic activation.⁶⁰Together, these findings suggest that LC’s influence on sensory coding depends on its activation pattern.

Open questions

Our work also provides a foundation for probing more advanced aspects of population coding, including response dynamics and noise correlations. GC taste responses unfold as sequences of ensemble states that encode presence, identity, and palatability^33,34,95; how LC activity shapes these transitions remains unknown and will likely require higher-temporal-resolution-recordings, such as electrophysiology. Another open question is how LC affects noise correlations—the shared trial-to-trial variability between neurons—which influence population information capacity^96–98 and are regulated by internal state⁹⁹ and neuromodulation.¹⁰⁰ Addressing this will require simultaneously recorded populations and larger numbers of trials than used here. In addition to encoding basic properties of taste stimuli, GC dynamics contribute to taste expectation,^54,101 multisensory integration,^102–105 and taste-guided decision making.^57,106,107 By revealing how LC reshapes GC taste representations, our findings provide a basis for future studies to understand how state-dependent neuromodulation influences the neural computations underlying natural feeding behavior.

Limitations of the study

Several limitations should be considered when interpreting our findings. First, whether LC-induced shifts in GC population responses reflect arousal remains to be tested. Future experiments could examine taste-evoked responses during natural arousal fluctuations monitored with EEG or pupil measures to test whether this orthogonal shift is reproduced. Second, the behavioral consequences of LC-induced coding reorganization remain untested. A full assessment will require brief-access,⁵⁸ two-bottle choice,¹⁰⁸ and taste-reactivity¹⁰⁹ assays to measure consummatory behavior, taste preference, and affective orofacial reactions across sex-balanced cohorts, as these behaviors are known to exhibit sex differences.^110–112 Third, our interpretation of axis rotation as reflecting unequal stretching of covarying taste attributes remains speculative. A stronger test would present the mixture-ratio or concentration stimulus sets together with the standard four-taste panel in the same session, allowing the latter to provide a stable palatability reference axis. Fourth, the effects observed in each experiment may depend on the tastant battery, including stimulus selection and concentration range. Finally, because our population-level analyses rely on distinct attribute axes defined in separate PCA spaces for each stimulus set, direct statistical comparisons across experiments are not available. Therefore, differences among the palatability, mixture-ratio, and sucrose-concentration results should be interpreted cautiously.

STAR★METHODS

Detailed methods are provided in the online version of this paper and include the following:

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Adult male and female mice older than postnatal day 60 (>P60) were used in all experiments. Dbh^Flpo mice¹⁷ were maintained as heterozygotes on a C57BL/6J background. Mice were housed under a reversed 12-h light/dark cycle, with lights off at 9 A.M. (zeitgeber time 0, ZT0), and all experiments were conducted during the dark period. Mice had ad libitum access to food and water except during experiments requiring water restriction. Prior to surgery, mice were group housed. Both male and female mice were included, but the study was not designed or powered to assess sex-dependent effects, which therefore remain a limitation. All animal procedures were approved by the University of Connecticut Institutional Animal Care and Use Committee (IACUC) and were conducted in accordance with the Guide for the Care and Use of Laboratory Animals.

METHOD DETAILS

Survival surgical procedures

All survival surgical procedures were performed under 1–4% isoflurane anesthesia using aseptic techniques. Mice were positioned in a stereotaxic frame (Model 942, Kopf Instruments) atop a warming pad (RT-0514, Kent Scientific) to maintain body temperature. Anesthesia was delivered via a nose cone and continuously monitored throughout the procedure to ensure proper depth. For analgesia, ketoprofen (5 mg/kg i.p., Covetrus) was administered during surgery and again at 24- and 48-h post-operatively for all procedures.

Intracranial viral injections

Viral constructs were packaged in the indicated serotypes, and titers are reported as genome copies per milliliter (GC/mL) following dilution in physiological saline. Intracranial injections were performed after a midline scalp incision, followed by skull exposure and leveling. Unilateral craniotomies were made using a surgical drill (OmniDrill 35, WPI). Viral solutions were loaded into glass micropipettes pulled from capillary tubes (Fisher Scientific, Part 21171), backfilled with mineral oil (Fisher Scientific, Part O1211), and connected to a Nanoliter 2020 injector (WPI).

For miniscope imaging with LC activation, Dbh^Flpo mice received unilateral LC injections of AAVs encoding either the Flp-dependent optogenetic activator ChrimsonR^42,113 (AAV-Ef1a-fDIO-ChrimsonR-tdTomato; serotype 5; 5 × 10¹² GC/mL; Lot #: 09012021; n = 13 mice) or a fluorescent control⁴² (AAV-Ef1a-fDIO-tdTomato; serotype 5; 2.9 × 10¹² GC/mL; Lot #: 09012021; n = 3 mice). In the same surgery, a CaMKIIa-driven calcium indicator (GCaMP8m) was injected in the ipsilateral GC using AAV-CaMKIIa-jGCaMP8m-WPRE (serotype 9; 1.15 × 10¹³ GC/mL; Lot #: V197511) to target pyramidal neurons. For LC injections, 500 nL was delivered at 100 nL/min. Coordinates, relative to bregma were: anterior-posterior (AP) −5.45 mm, medial-lateral (ML) −1.25 mm, and dorsal-ventral (DV) −3.85. We used unilateral LC manipulation because LC projections to cortex are predominantly ipsilateral (~95%).¹¹⁴ For GC injection, 500 nL was delivered at 50 nL/min. Coordinates, relative to bregma: AP +1.2 mm, ML −3.3 mm, DV −1.95 (from brain surface).

To monitor NE dynamics in the GC during LC activation, a separate cohort of Dbh^Flpo mice (n = 3 mice) received unilateral AAV injections of Flp-dependent ChrimsonR^42,113 into the LC and a 4:1 cocktail of AAVs expressing the constitutive NE biosensor GRAB_NE2h⁴³ (AAV-hSyn-GRAB-NE2h; serotype 9; titer: 8 × 10¹² GC/mL; Lot #: 20240529) and tdTomato (AAV-CAG-tdTomato; serotype 9; titer: 4.0 × 10¹⁰ GC/mL; Lot #: V197427) into the ipsilateral GC. For this GC injection, 500 nL of the viral cocktail was infused at 100 nL/min. Coordinates relative to bregma: AP +1.2 mm, ML −3.3 mm, DV −2.0 (from brain surface).

After each injection, the micropipette was left in place for 5 min to minimize backflow. Incisions were sutured, and mice recovered on a heating pad for at least 24 h.

Intracranial optical fiber and grin lens implantation

All procedures preceding viral injection, including pre-operative care, were conducted as previously described. This section outlines the implantation of custom fiber-optic probes and gradient refractive index (GRIN) lenses. Fiber-optic probes were constructed using 200 μm, 0.39 NA multimode fiber (FT200EMT, Thorlabs) housed in ceramic ferrules (MM-CON2007–2300, Precision Fiber Products) and were used for optogenetic stimulation in the LC or NE fluorescence detection in the GC. GRIN lenses (ProView Integrated Lens, 1 mm Ø × 4 mm length, with baseplate; Inscopix) were used for GC imaging.

For mice expressing either ChrimsonR or tdTomato in the LC, along with GCaMP8m in GC, a second surgery was performed ≥ three weeks after viral injection to implant the optical components. A small craniotomy was made above the LC and a larger one above the GC using a 1.4 mm inverted-cone bur (Panther). An optical fiber was unilaterally implanted above the LC at coordinates AP: −5.45 mm, ML: −1.25 mm, DV: −3.85 mm relative to bregma, and secured to the skull using Metabond (S371, S398, and S396, Parkell), while leaving the GC craniotomy open.

To guide GRIN lens placement above the GC, a flattened needle was lowered to 1.65 mm below the brain surface and then withdrawn. The GRIN lens was then implanted at 1.80 mm depth and secured with Metabond. For mice expressing ChrimsonR in LC and GRAB_NE2h in GC, a second surgery was performed three weeks post-injection to implant optical fibers above both regions. The LC fiber was implanted at AP −5.45 mm, ML −1.25 mm, DV −3.85 mm (bregma-referenced), and GC at AP +1.2 mm, ML −3.3 mm, DV: −2.0 mm (relative to the brain surface). Incisions were sutured, and post-operative care followed the protocol described above.

Intraoral cannulae (IOC) implantation

A third surgery was performed to implant IOC in mice for miniscope recordings, at least two weeks after optical implantations. Each IOC consisted of a 1.5 cm segment of plastic tubing (1883T3, McMaster-Carr), with one end flared to secure a trimmed plastic washer (93785A812, McMaster-Carr). Mice were anesthetized via subcutaneous (S.C.) injection of ketamine (100 mg/kg) and xylazine (12.5 mg/kg) mixture (Covetrus).

A scalp incision was made to expose the lateral edge of the skull. A needle was attached to the non-flared end of the IOC, which was inserted lateral to the second/third maxillary molar, passed through the masseter muscle and beneath the zygomatic arch, and exited at the scalp edge behind the eye. After removing the needle, the exposed end of the IOC was connected to a push-in fitting (QSM-M3–3, Festo). Mice were placed in a stereotaxic frame equipped with a heating pad and maintained on 1–4% isoflurane anesthesia. The IOC was secured to the skull using Metabond and dental acrylic, and the incision was sutured closed. A fishing line plug was inserted into the IOC to prevent obstruction. To prevent infection, Penicillin G procaine (3,000 units/lb., Norbrook; 0.01 to 0.02 mL; intramuscular) was administered immediately post-surgery and again 48 h later. All other post-operative care followed the protocol described above.

Water restriction protocol

Following IOC surgery, mice were initially provided with water ad libitum. Starting at least two weeks post-surgery, they were gradually transitioned to a 23-h water restriction protocol to stabilize fluid intake prior to neural recordings. Acclimation occurred over three days, with restriction periods progressively increasing from 12 to 23 h. Once acclimated, mice received a minimum of one hour of water access daily, after passive tastant delivery (described below), to account for potential post-experiment satiety. Body weight and general health were monitored and recorded throughout restriction, which was limited to a maximum of two weeks.

Passive tastant delivery

Taste stimuli were delivered using a gravity-driven, valve-controlled perfusion system (VC3–4PG, ALA Scientific Instrument). This system included four fluid reservoirs, each connected via separate 1/16″ PVC tubing (MMF-4, ALA Scientific Instruments) that converged at a custom-made manifold composed of four jointed miniature polyimide tubes (085-III, MicroLumen). Before each experiment, the IOC plug was removed, and the cannulae was cleaned using a fishing line to dislodge debris, followed by rinsing with deionized water-filled syringe, and flushing with compressed air to prevent obstruction.

To ensure consistent tastant delivery, mice were briefly anesthetized with 4% isoflurane, and any food debris or scar tissue near the IOC opening was removed with forceps. Mice were then placed in an empty arena and allowed to recover for at least 20 min. Immediately before each experimental session, the manifold was inserted into the IOC and secured using the push-in fitting. The manifold was designed so that its tip aligned flush with the IOC opening.

Three sets of taste stimuli were tested in separate experiments, including: 1) Basic tastes: 200 mM sucrose, 50 mM NaCl, 20 mM citric acid, and 1 mM quinine, 2) Sucrose concentrations: 50 mM, 100 mM, 250 mM, and 500 mM, and 3) Sucrose-NaCl mixtures: 100 mM sucrose and 100 mM NaCl mixed at ratios of 90:10, 65:35, 35:65, and 10:90. Each experiment consisted of 96 trials with tastants presented in pseudo-random order with intertrial intervals of 25–35 s. During each trial, a solenoid valve opened for 0.08 s to deliver 8 μL of tastant. The perfusion system was rinsed and calibrated daily to ensure consistent fluid delivery. We did not include water rinses between tastant deliveries, consistent with prior studies^36,38,52,76 because awake mice rapidly clear tastants through salivation. Any residual cross-trial tastant interactions would be expected to increase trial-to-trial variability rather than introduce systematic bias across conditions.

In vivo miniscope calcium imaging

To monitor large-scale, single-cell Ca²⁺ activity in the GC of freely moving mice, we used the nVoke 2.0 miniature microscope system (Inscopix). The miniscope was secured to the baseplate by tightening a small set screw and connected to a commutator (Inscopix), which enabled high-speed video and signal transmission while minimizing cable entanglement. The commutator was linked to a data acquisition (DAQ) box (Inscopix).

On the day of the experiment, mice were placed in an operant chamber (ENV-370W, Med Associates) within a sound-attenuating cubicle (ENV-017M, Med Associates) and given 15 min to habituate. The miniscope data were collected through the Inscopix Data Acquisition Software (IDAS). The miniscope has an excitation blue LED (455 ± 8 nm) that was set to 0.2 arbitrary unit power. The imaging gain (ranging from 2.0 to 3.0) was adjusted based on the fluorescence intensity histogram to ensure that the highest possible dynamic range was achieved without signal saturation. Data were collected at 20 frames per second (fps) with ~50 ms exposure time. Synchronization of calcium imaging with tastant delivery and optogenetic stimulation was achieved using MedPC scripts (Med Associates) and transistor–transistor logic (TTL) signals delivered at 20 Hz.

Optogenetic stimulation procedures

To activate LC neurons expressing ChrimsonR, a 532 nm green laser (Shanghai Laser & Optics Century Co., Ltd., China) was coupled to an optical patch cable (M72L02, M83L01, Thorlabs) and connected to the implantable probe via a mating sleeve (ADAL4–5, Thorlabs), which was wrapped in tinfoil to prevent light leakage. Laser power was adjusted to 8 mW at the fiber tip.

Two photostimulation paradigms were employed. In the phasic stimulation paradigm, half of the tastant delivery events were preceded by a 0.5-s train of laser pulses (20 Hz, 5 ms pulse duration). Tastants were administered 0.5 s after the photostimulation train ended. In the tonic stimulation paradigm, a total of 96 tastant delivery events were organized into eight blocks. Photostimulation at 5 Hz was applied during four of these blocks in the following sequence: on, off, off, on, on, off, off, on. Within each stimulation block, the tonic pulse train began 45–55 s before the first tastant delivery and continued until 10 s after the final delivery. The subsequent block began 145–155 s after the preceding stimulation ended.

Experimental timeline

Two weeks following IOC implantation, mice were subjected to 23-h water restriction to enhance motivation for tastant sampling. Prior to recording sessions, mice were habituated to wearing the miniscope during passive taste delivery in the testing chamber for three days. Each habituation session consisted of 96 tastant trials, including 200 mM sucrose, 50 mM NaCl, 20 mM citric acid, and 1 mM quinine, allowing mice to become familiar with the stimuli.

The experimental design combined three sets of taste stimuli with two optogenetic stimulation paradigms, resulting in six distinct conditions: 1) Basic tastes–phasic stimulation, 2) Sucrose concentrations–phasic stimulation, 3) Mixtures–phasic stimulation, 4) Basic tastes–tonic stimulation, 5) Sucrose concentrations–tonic stimulation, and 6) Mixtures–tonic stimulation. These conditions were tested in the order listed above. Experimental sessions were occasionally separated by rest days, during which mice received 1 h of water access without undergoing testing.

Fiber photometry recordings of photostimulation-evoked NE release

Five weeks post-optical fiber implantations, in vivo fiber photometry recordings were conducted to measure NE release in the GC following optogenetic stimulation of LC neurons. Recordings were conducted using a custom-built spectrally resolved fiber photometry system.^42,44,45 Excitation light at 488 nm (OBIS 488LS-20, Coherent) was adjusted to ~60 μW power and directed into a fluorescence cube (DFM1, Thorlabs), reflected by a dichroic mirror (ZT488/561rpc-UF1, Chroma), and focused via an achromatic fiber port (PAFA-X-4-A, Thorlabs) onto a patch cable (M83L01, Thorlabs). The distal end was connected via a quick-release interconnect (ADAL3, Thorlabs) to an optical probe mounted on the mouse skull.

Fluorescence emitted from the tissue traveled back through the same optical path, passed through the dichroic mirror, and was filtered (ZET 488/561m, Chroma). The filtered signal was routed through a fiber port (PAF2S-11A, Thorlabs) and transmitted to a spectrometer (Ocean FX, Ocean Optics) via patch cable (M200L02-A, Thorlabs). Time-resolved fluorescence emission spectra of GRAB_NE2h and TdTomato signals were acquired using OceanView software (version 2.0.8) at 25 frames per second, with a 31 ms integration time. Synchronization of calcium imaging and optogenetic stimulation was achieved using MedPC (Med Associates) scripts and TTL signals delivered at 25 Hz.

Two recording sessions were conducted to test distinct photostimulation paradigms for activating ChrimsonR-expressing LC neurons and their effects on NE release in GC. In the first session, mice received phasic stimulation, consisting of 0.5-s trains of light pulses (5-ms pulse duration; 532 nm) delivered at 20 Hz. In the second session, mice received tonic stimulation at 5 Hz pulses. Both stimulation conditions used the same trial structure as the phasic and tonic miniscope experiments, respectively, except without tastant delivery.

Brief-access taste testing

Palatability of the four tastants used in the palatability experiment was assessed using a brief-access taste test. Eight Dbh^Flpo mice (separate from the above experimental subjects) were water restricted for 23 h prior to testing. During testing, mice licked a metal spout fitted with four microtubes, each connected to a reservoir containing one tastant solution. Licks were detected with a lickometer (Med Associates ENV-250), and each lick triggered delivery of ~1 μL of tastant via the opening of a solenoid valve.

A trial began with the mouse’s first lick on the spout; the spout remained accessible for 10 s before being retracted. After a 10-s inter-trial interval, the spout was extended again to allow initiation of the next trial. Each session lasted 30 min, during which tastants were delivered in pseudo-random order and lick timestamps were recorded. The tastants tested—200 mM sucrose, 50 mM NaCl, 20 mM citric acid, and 1 mM quinine—were identical to those used in the palatability experiment.

Tissue collection and processing

At the conclusion of the experiments, brain tissue was collected from all mice to verify viral expression and optical implant placement. Mice were deeply anesthetized with isoflurane and perfused transcardially with 1% phosphate-buffered saline (PBS) containing heparin (1:1000), followed by 4% paraformaldehyde in PBS (PFA/PBS), also supplemented with heparin. Brains were post-fixed overnight in 4% PFA/PBS at 4°C on a shaker, then cryoprotected in 30% sucrose in PBS for 48 h. Following cryoprotection, brains were embedded in optimal cutting temperature (O.C.T.) compound (Thermo Fisher), frozen, and stored at −80°C. Coronal sections (40-μm) were cut using a cryostat (Leica CM 3050S) and transferred into a cryoprotectant storage solution containing 30% sucrose, 30% ethylene glycol, and 40% 0.1 M phosphate buffer. Sections were stored at −80°C until further processing.

Immunohistochemistry

Brain sections were rinsed in 0.1 M PBS and blocked for 1 h at room temperature on a shaker in 5% normal goat serum made in 0.1% Triton PBS. Sections were then incubated overnight at 4°C in primary antibodies, including chicken anti-GFP (1:10,000; Abcam; Lot # GR3361041–14), rabbit anti-DsRed (1:1000; Takara; Lot # 2103116), and mouse anti-TH (1:500; GeneTex; Lot # 822104279). The next day, tissues were washed in 0.1% Triton X-100 dissolved in 0.1 M PBS and incubated for 2 h at room temperature on a shaker with secondary antibodies (1:1000; Fisher), including goat anti-chicken Alexa 488 (Lot # 2304258), goat anti-rabbit Alexa 568 (Lot # 2273773), and goat anti-mouse Alexa 633 (Lot # 2304276). After final rinses in PBS, brain sections were mounted onto slides and cover slipped using Prolong Diamond Anti-Fade mounting medium with DAPI (4^′,6-diamidino-2-phenylindole; P36971, Thermo Fisher).

Digital image processing

Immunofluorescent-labeled brain sections were imaged at 10× magnification using a Leica DM6 epifluorescent microscope (Leica Microsystems, Wetzlar, Germany). To enhance image clarity and reduce out-of-focus blur, all DM6 images were deconvolved using the Leica Instant Computational Clearing algorithm (LAS X 3.8.1). High-magnification images of the regions of interest were acquired at 40x using a Lecia SP8 spectral confocal microscope (Leica Microsystems). Digital images were exported from LASX software (v3.7.5) as uncompressed .lif or .tif files and opened in FIJI (v2.0) for further analysis. z-stacks were converted to maximum-intensity projections, and image adjustments were limited to uniform brightness and contrast changes to enhance fluorescent signal visibility.

Data presentation and visualization

All graphical illustrations were made using Adobe Illustrator, Photoshop, GraphPad Prism, and MATLAB.

QUANTIFICATION AND STATISTICAL ANALYSIS

Data analysis was performed using custom scripts written in MATLAB (MathWorks) and Prism 10 (GraphPad). Unless otherwise indicated, statistical significance in figures is denoted as *p < 0.05, **p < 0.01, and ***p < 0.001; n.s. indicates p ≥ 0.05.

Miniscope movie pre-processing and cell-identification

Miniscope recordings were pre-processed using Inscopix Data Processing Software (IDPS), including spatial and temporal down-sampling, bandpass filtering, and motion correction. Cell bodies were identified using constrained non-negative matrix factorization (CNMFe),¹¹⁵ producing fluorescence traces in units of dF/noise. Extracted cells and traces were visually inspected to remove poorly identified cells. The traces were down-sampled to 10 Hz.

Identification of taste-discriminative and LC-modulated neurons

All analyses were performed in MATLAB unless noted otherwise. Neuronal activity was aligned to tastant delivery and baseline-corrected by subtracting the mean response during the 1–4 s window preceding delivery. Aligned responses were pooled across subjects to construct pseudo-populations. Analyses focused on the 5 s post-tastant delivery period, divided into ten 500-ms time bins.

To identify stimulus-discriminative neurons, repeated-measures (RM) two-way ANOVAs (tastant × time bin) were performed on each neuron. Neurons were classified as stimulus-discriminative if either the tastant main effect or the tastant × time interaction was significant (p <0.05, two-tailed). To ensure robust evoked responses, we applied two additional criteria: (i) paired t-tests comparing each post-delivery time bin with baseline, retaining neurons with at least one significant time bin at p <0.001; and (ii) mean evoked response that differed from baseline by ≥1 dF. This procedure was applied separately to stim and no-stim trials, and neurons identified in either condition were combined. Thus, neurons were included in the stimulus-discriminative category if they met criteria in either condition, including cases near the classification boundary.

LC-modulated neurons were identified using RM two-way ANOVAs (stim-condition × time bin) separately for each tastant. Neurons were classified as LC-modulated if either the stim-condition main effect or interaction reached significance (p <0.01) for at least one tastant. The proportion of LC-modulated neurons in experimental and tdTomato control mice was compared using exact binomial tests (performed in R).

Population decoding

To decode tastant identity, linear discriminant classifiers (MATLAB fitcdiscr function) were trained on time-bin-resolved population responses. Decoding accuracy was evaluated via 6-fold cross-validation. To estimate performance for a given population size, 100 subpopulations were constructed by sampling neurons with replacement, and mean accuracy was computed. To test differences between stim and no-stim conditions, trial labels were permuted 1,000 times for each population size to generate a null distribution. The observed decoding difference was compared to the null distribution (two-tailed p-value). Significance across population sizes required at least three sizes with p <0.05. If this criterion was not met, no significance (n.s.) was noted on graphs.

Response–stimulus correlation

Attribute-relevant information at the single-neuron level was quantified by computing Kendall’s rank correlation coefficient (Kendall’s tau) between each neuron’s mean responses to tastants and the ranked attribute values. LC stimulation effects were assessed by comparing the absolute values of tau across stim and no-stim conditions using RM two-way ANOVA (stimulation × time) followed by Bonferroni post hoc tests (using GraphPad Prism).

Principal component analysis (PCA) and LC-induced transformations

Because PCA is sensitive to response magnitude, each neuron was normalized by its response to its preferred tastant. PCA was applied to mean population responses for each trial type (4 tastants × 2 stimulation conditions), using all baseline and 10-s post-delivery time points. This procedure stabilizes the PCA axes by fitting them to stimulus-averaged responses rather than noisy single-trial responses. Single-trial responses were averaged within the high response–stimulus correlation window (1.5–4.5 s post-delivery) and projected onto the PCA axes derived from trial-type means. Quantifications used the first six principal components (~75% variance).

The palatability axis was defined as the vector connecting palatable (sucrose, NaCl) and aversive (citric acid, quinine) centroids, computed separately for stim and no-stim trials. Responses were projected onto this axis and normalized such that palatable centroid = 1 and aversive centroid = −1. Mixture-ratio and sucrose-concentration axes were defined similarly, except end-range tastants were used. Dynamic range was the difference between projection scores of the end-range tastants. The angle between stim- and no-stim-derived axes were computed using their dot product. LC-induced changes in tastant-pair separation were normalized by the no-stim dynamic range. We used normalized distance differences rather than distance ratios because ratios can become unstable when baseline distances are small.

Dynamic range changes, axis rotations, and pairwise distance changes were assessed by permuting trial labels between stim and no-stim condtions 1,000 times. Permutation-derived null distributions yielded two-tailed p-values for dynamic range ratios and one-tailed p-values for axis rotation. The latter choice assumed that LC-induced rotation would increase, rather than decrease, the angle that arose from variability in axis estimation. Pairwise distance changes were Bonferroni corrected for six comparisons. Dispersion orthogonal to the attribute axis was calculated as the mean perpendicular distance of points from the axis; LC-induced changes were assessed using the same permutation procedure. Procrustes transformations were used to align centroids between stim and no-stim conditions for visualization in the first three PCs.

For palatability, responses from both trial types were projected onto the no-stim-derived axis, and scores were compared using RM two-way ANOVA (tastant × stim condition) with Bonferroni-corrected post hoc tests (performed in GraphPad Prism).

Modulation type analysis

Neuronal tuning was defined as mean responses to each tastant during 1.5–4.5 s post-delivery (same window as PCA). For stimulus-discriminative neurons, stim vs. no-stim responses were compared using three models: (i) Null: main effect of tastant only; (ii) Scale–shift: stimulation modifies tastant tuning via multiplicative and additive effects:

r_stim = mr_no-stim+ a, where r denotes the tastant-evoked response, and m and a represent the multiplicative and additive factors, respectively; (iii) Flexible: tastant-specific changes are allowed by including interaction terms between tastant identity and stimulation condition, enabling arbitrary reshaping of the tuning profile. Null and flexible models were fit with linear least squares regression (MATLAB fitlm function); scale–shift was fit with nonlinear least squares (fitnlm). Models were compared using Akaike Information Criterion (AIC) with hierarchical selection: scale–shift had to outperform null (ΔAIC > 2) to indicate modulation; flexible was selected only if it outperformed both. To prevent unstable nonlinear fits, scale–shift fits were constrained (m: 0–10; a: −10–10). Fits exceeding these bounds typically yielded extreme values and were thus excluded from the scale–shift category and evaluated only against the null and flexible models.

Proportions of scale–shift vs. flexible neurons were compared across experiments using logistic regression (binary response variable: modulation type; predictor: attribute), with pairwise Wald tests and Tukey correction (performed in R).

To relate multiplicative modulation to neuronal contributions to stimulus coding, we quantified each neuron’s contribution to attribute axes from the no-stim condition. Attribute axes, defined in PCA space, were projected back into neural activity space by multiplying each axis vector with the PCA coefficient matrix, yielding a vector of neuronal weights. Each neuron’s contribution was defined as the absolute value of its weight. For scale–shift modulated neurons, multiplicative factors were log-transformed to symmetrize expansive and compressive effects, and their relationship to neuronal contributions was assessed using Pearson’s correlation.

Fiber photometry data analysis

Raw photometry signals were preprocessed using a custom R script to linearly unmix the overlapping GRAB_NE2h and tdTomato signals, following established protocols.^44,116 To mitigate artifacts in GRAB_NE2h fluorescence signal (e.g., due to tissue displacement, cable bending during animal movement), the ratio of unmixed GRAB_NE2h to tdTomato signals (fluorescence ratio) was computed and used for all subsequent analyses, as performed previously.^42,44,45

Fluorescence ratio traces were analyzed using custom MATLAB scripts. The percent change in fluorescence (%ΔF/F) was calculated by normalizing fluorescence ratios to their baseline mean values.

Brief-access taste test analysis

Licking data from the brief-access test were first processed in MATLAB to calculate the mean lick counts for each 10-s trial. Lick counts were then analyzed using an RM one-way ANOVA, followed by Šídák’s post hoc comparisons (performed in GraphPad Prism).

Supplementary Material

NIHMS2212249-supplement-1.pdf^{(902.6KB, pdf)}

SUPPLEMENTAL INFORMATION

Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2026.117990.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER

Antibodies

Anti-GFP (chicken)	Abcam	Abcam Cat # ab13970; RRID: AB_300798
Anti-dsRed (rabbit)	Takara	Takara Cat # 632496; RRID: AB_10013483
Anti-TH (mouse)	GeneTex	GeneTex Cat # GTX10372; RRID: AB_373127
Goat anti-Chicken Alexa Fluor 488	Thermo Fisher Scientific	Thermo Fisher Scientific Cat # A-11039; RRID: AB_2534096
Goat anti-Rabbit Alexa Fluor 568	Thermo Fisher Scientific	Thermo Fisher Scientific Cat # A-11036; RRID: AB_10563566
Goat anti-Mouse Alexa Fluor 633	Thermo Fisher Scientific	Thermo Fisher Scientific Cat # A-21052; RRID: AB_2535719

Bacterial and Virus Strains

AAV-Ef1a-fDIO-ChrimsonR-tdTomato	Gift from Patricia Jensen	Addgene plasmid # 171027; RRID: Addgene_171027; Viral prep: NIEHS Viral Vector Core
AAV-Ef1a-fDIO-tdTomato	Gift from Patricia Jensen	Addgene plasmid # 128434; RRID: Addgene_128434; Viral prep: NIEHS Viral Vector Core
AAV-CaMKIIa-jGCaMP8m-WPRE	Gift from Loren Looger	Addgene viral prep # 176751-AAV9; RRID: Addgene_176751
AAV-hSyn-GRAB-NE2h	Gift from Yulong Li	WZ Biosciences; viral prep # L003011-AAV9
AAV-CAG-tdTomato	Gift from Edward Boyden	Addgene viral prep # 59462-AAV9; RRID: Addgene_59462

Deposited Data

Dataset	This paper	Zenodo: doi:10.5281/zenodo.21826182
Analysis code	This paper	Zenodo: doi:10.5281/zenodo.21826182

Experimental Models: Organisms/Strains

Mouse; Dbh^Flpo	The Jackson Laboratory	JAX Stock # 033952; RRID: IMSR_JAX:033952

Software and Algorithms

Fiji (ImageJ2)	NIH	https://imagej.net/software/imagej2/; RRID: SCR_002285
Illustrator	Adobe	https://www.adobe.com/products/illustrator.html; RRID: SCR_010279
Inscopix Data Acquisition Software (IDAS)	Inscopix	https://inscopix.com/software-analysis-miniscope-imaging/
Inscopix Data Processing Software (IDPS)	Inscopix	https://inscopix.com/software-analysis-miniscope-imaging/
Leica Application Suite X (LASX)	Leica Microsystems	https://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/; RRID: SCR_013673
MATLAB	The MathWorks	https://www.mathworks.com/products/matlab.html; RRID: SCR_001622
OceanView 2.0	OceanOptics	https://www.oceanoptics.com/software/oceanview/
Prism 10	GraphPad	https://www.graphpad.com/; RRID: SCR_002798

Other

nVoke 2.0 miniscope system	Inscopix	https://inscopix.com/nvoke-system/; RRID: SCR_023028
Coherent OBIS 488 nM Laser System, 20 mW	Edmund Optics	https://www.edmundoptics.com/p/obis-488nm-ls-20mw-laser/29838/; Edmund Optics Cat #: 87–459
Green 532 nM DPSS Laser	Shanghai Laser & Optics Century Co.	https://www.lasercentury.com/product.asp?id=609; Shanghai Laser Cat #: GL532T3–200FC
MedPC Behavioral Control	Med Associates Inc.	https://med-associates.com/product/med-pc/; RRID: SCR_012156
OCEAN FX Miniature Spectrometer	Ocean Optics	https://www.oceanoptics.com/spectrometer/ocean-fx/; Ocean Optics Cat #: OCEAN-FX-VIS-NIR-ES
VC-Plus-4 Channel Gravity System	ALA Scientific Instruments	https://alascience.com/products/vcplus-4-channel-gravity-systems/; ALA Cat #: ALAVCPLUS-4CLG

Open in a new tab

Highlights.

Phasic locus coeruleus (LC) activation reorganizes gustatory cortex coding
LC activation expands representational distance along a palatability axis
Gain modulation and flexible tuning changes underlie LC-induced population effects
Tonic LC activation produces limited effects on gustatory coding

ACKNOWLEDGMENTS

We thank Dr. Alfredo Fontanini for guidance on the intraoral delivery method. We also thank Leon Novak, Julia Chernyak, Anthony Esposito, and Abigail McFarland for assistance with genotyping, and Nikolas Supczak for histology support. We acknowledge the University of Connecticut’s Animal Care Services for animal husbandry support, the Advanced Light Microscopy Facility (RRID:SCR_027547) for imaging guidance, and the Mechanical/Glass Facility for custom intraoral connectors. This work was supported by the National Institutes of Health (R01 DC023564 and R00 DK119586), the Brain Research Foundation Seed Grant (BRFSG 2023-09), and UConn CLAS Research Equipment Fund and Startup Funds (to N.R.S.).

Footnotes

DECLARATION OF INTERESTS

The authors declare no competing interests.

DECLARATION OF GENERATIVE AI AND AI-ASSISTED TECHNOLOGIES IN THE WRITING PROCESS

During the preparation of this work, the author(s) used ChatGPT for proofreading. The author(s) reviewed and edited the output as needed and take full responsibility for the content of the published article.

RESOURCE AVAILABILITY

Lead contact

For questions or additional information, please contact the lead contact, Natale R. Sciolino (natale.sciolino@uconn.edu). All reasonable requests will be accommodated.

Materials availability

No new or unique reagents were generated.

Data and code availability

Data: All neural and behavioral datasets generated here have been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Code: Original analysis code used has been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

REFERENCES

1.Harris KD, and Thiele A (2011). Cortical state and attention. Nat. Rev. Neurosci. 12, 509–523. 10.1038/nrn3084. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.McGinley MJ, David SV, and McCormick DA (2015). Cortical Membrane Potential Signature of Optimal States for Sensory Signal Detection. Neuron 87, 179–192. 10.1016/j.neuron.2015.05.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Polack PO, Friedman J, and Golshani P (2013). Cellular mechanisms of brain state-dependent gain modulation in visual cortex. Nat. Neurosci. 16, 1331–1339. 10.1038/nn.3464. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Fontanini A, and Katz DB (2008). Behavioral states, network states, and sensory response variability. J. Neurophysiol. 100, 1160–1168. 10.1152/jn.90592.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Matteucci G, Guyoton M, Mayrhofer JM, Auffret M, Foustoukos G, Petersen CCH, and El-Boustani S (2022). Cortical sensory processing across motivational states during goal-directed behavior. Neuron 110, 4176–4193.e10. 10.1016/j.neuron.2022.09.032. [DOI] [PubMed] [Google Scholar]
6.McGirr A, LeDue J, Chan AW, Boyd JD, Metzak PD, and Murphy TH (2020). Stress impacts sensory variability through cortical sensory activity motifs. Transl. Psychiatry 10, 20. 10.1038/s41398-020-0713-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Waterhouse BD, and Navarra RL (2019). The locus coeruleus-norepinephrine system and sensory signal processing: A historical review and current perspectives. Brain Res. 1709, 1–15. 10.1016/j.brainres.2018.08.032. [DOI] [PubMed] [Google Scholar]
8.McBurney-Lin J, Lu J, Zuo Y, and Yang H (2019). Locus coeruleus-norepinephrine modulation of sensory processing and perception: A focused review. Neurosci. Biobehav. Rev. 105, 190–199. 10.1016/j.neubiorev.2019.06.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Edeline JM (2012). Beyond traditional approaches to understanding the functional role of neuromodulators in sensory cortices. Front. Behav. Neurosci. 6, 45. 10.3389/fnbeh.2012.00045. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Jacob SN, and Nienborg H (2018). Monoaminergic Neuromodulation of Sensory Processing. Front. Neural Circ. 12, 51. 10.3389/fncir.2018.00051. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Aston-Jones G, and Cohen JD (2005). An integrative theory of locus coeruleus-norepinephrine function: adaptive gain and optimal performance. Annu. Rev. Neurosci. 28, 403–450. 10.1146/annurev.neuro.28.061604.135709. [DOI] [PubMed] [Google Scholar]
12.Aston-Jones G, and Bloom FE (1981). Norepinephrine-containing locus coeruleus neurons in behaving rats exhibit pronounced responses to non-noxious environmental stimuli. J. Neurosci. 1, 887–900. 10.1523/JNEUROSCI.01-08-00887.1981. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Bouret S, and Sara SJ (2004). Reward expectation, orientation of attention and locus coeruleus-medial frontal cortex interplay during learning. Eur. J. Neurosci. 20, 791–802. 10.1111/j.1460-9568.2004.03526.x. [DOI] [PubMed] [Google Scholar]
14.Breton-Provencher V, Drummond GT, Feng J, Li Y, and Sur M (2022). Spatiotemporal dynamics of noradrenaline during learned behaviour. Nature 606, 732–738. 10.1038/s41586-022-04782-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Page ME, and Valentino RJ (1994). Locus coeruleus activation by physiological challenges. Brain Res. Bull. 35, 557–560. 10.1016/0361-9230(94)90169-4. [DOI] [PubMed] [Google Scholar]
16.Valentino RJ, and Van Bockstaele E (2008). Convergent regulation of locus coeruleus activity as an adaptive response to stress. Eur. J. Pharmacol. 583, 194–203. 10.1016/j.ejphar.2007.11.062. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Robertson SD, Plummer NW, de Marchena J, and Jensen P (2013). Developmental origins of central norepinephrine neuron diversity. Nat. Neurosci. 16, 1016–1023. 10.1038/nn.3458. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Devilbiss DM, and Waterhouse BD (2004). The effects of tonic locus ceruleus output on sensory-evoked responses of ventral posterior medial thalamic and barrel field cortical neurons in the awake rat. J. Neurosci. 24, 10773–10785. 10.1523/JNEUROSCI.1573-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Bouret S, and Sara SJ (2002). Locus coeruleus activation modulates firing rate and temporal organization of odour-induced single-cell responses in rat piriform cortex. Eur. J. Neurosci. 16, 2371–2382. 10.1046/j.1460-9568.2002.02413.x. [DOI] [PubMed] [Google Scholar]
20.Sato H, Fox K, and Daw NW (1989). Effect of electrical stimulation of locus coeruleus on the activity of neurons in the cat visual cortex. J. Neurophysiol. 62, 946–958. 10.1152/jn.1989.62.4.946. [DOI] [PubMed] [Google Scholar]
21.Manunta Y, and Edeline JM (1999). Effects of noradrenaline on frequency tuning of auditory cortex neurons during wakefulness and slow-wave sleep. Eur. J. Neurosci. 11, 2134–2150. 10.1046/j.1460-9568.1999.00633.x. [DOI] [PubMed] [Google Scholar]
22.McLean J, and Waterhouse BD (1994). Noradrenergic modulation of cat area 17 neuronal responses to moving visual stimuli. Brain Res. 667, 83–97. 10.1016/0006-8993(94)91716-7. [DOI] [PubMed] [Google Scholar]
23.Ego-Stengel V, Bringuier V, and Shulz DE (2002). Noradrenergic modulation of functional selectivity in the cat visual cortex: an in vivo extracellular and intracellular study. Neuroscience 111, 275–289. 10.1016/s0306-4522(02)00011-8. [DOI] [PubMed] [Google Scholar]
24.Manella LC, Petersen N, and Linster C (2017). Stimulation of the Locus Ceruleus Modulates Signal-to-Noise Ratio in the Olfactory Bulb. J. Neurosci. 37, 11605–11615. 10.1523/JNEUROSCI.2026-17.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Hirata A, Aguilar J, and Castro-Alamancos MA (2006). Noradrenergic activation amplifies bottom-up and top-down signal-to-noise ratios in sensory thalamus. J. Neurosci. 26, 4426–4436. 10.1523/JNEUROSCI.5298-05.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Vazey EM, Moorman DE, and Aston-Jones G (2018). Phasic locus coeruleus activity regulates cortical encoding of salience information. Proc. Natl. Acad. Sci. USA 115, E9439–E9448. 10.1073/pnas.1803716115. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Geng C, Li R, Li S, Liu P, Peng Y, Liu C, Wang Z, Zhang H, and Li A (2025). Noradrenergic inputs from the locus coeruleus to anterior piriform cortex and the olfactory bulb modulate olfactory outputs. Nat. Commun. 16, 260. 10.1038/s41467-024-55609-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Devilbiss DM, and Waterhouse BD (2011). Phasic and tonic patterns of locus coeruleus output differentially modulate sensory network function in the awake rat. J. Neurophysiol. 105, 69–87. 10.1152/jn.00445.2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Vyas S, Golub MD, Sussillo D, and Shenoy KV (2020). Computation Through Neural Population Dynamics. Annu. Rev. Neurosci. 43, 249–275. 10.1146/annurev-neuro-092619-094115. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Barack DL, and Krakauer JW (2021). Two views on the cognitive brain. Nat. Rev. Neurosci. 22, 359–371. 10.1038/s41583-021-00448-6. [DOI] [PubMed] [Google Scholar]
31.Kriegeskorte N, and Wei XX (2021). Neural tuning and representational geometry. Nat. Rev. Neurosci. 22, 703–718. 10.1038/s41583-021-00502-3. [DOI] [PubMed] [Google Scholar]
32.Devilbiss DM, Page ME, and Waterhouse BD (2006). Locus ceruleus regulates sensory encoding by neurons and networks in waking animals. J. Neurosci. 26, 9860–9872. 10.1523/JNEUROSCI.1776-06.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Katz DB, Simon SA, and Nicolelis MA (2001). Dynamic and multimodal responses of gustatory cortical neurons in awake rats. J. Neurosci. 21, 4478–4489. 10.1523/JNEUROSCI.21-12-04478.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Sadacca BF, Mukherjee N, Vladusich T, Li JX, Katz DB, and Miller P (2016). The Behavioral Relevance of Cortical Neural Ensemble Responses Emerges Suddenly. J. Neurosci. 36, 655–669. 10.1523/JNEUROSCI.2265-15.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Mukherjee N, Wachutka J, and Katz DB (2019). Impact of precisely-timed inhibition of gustatory cortex on taste behavior depends on single-trial ensemble dynamics. Elife 8, e45968. 10.7554/eLife.45968. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Levitan D, Lin JY, Wachutka J, Mukherjee N, Nelson SB, and Katz DB (2019). Single and population coding of taste in the gustatory cortex of awake mice. J. Neurophysiol. 122, 1342–1356. 10.1152/jn.00357.2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Fontanini A, and Katz DB (2006). State-dependent modulation of time-varying gustatory responses. J. Neurophysiol. 96, 3183–3193. 10.1152/jn.00804.2006. [DOI] [PubMed] [Google Scholar]
38.Stone BT, Lin JY, Mahmood A, Sanford AJ, and Katz DB (2022). LiCl-induced sickness modulates rat gustatory cortical responses. PLoS Biol. 20, e3001537. 10.1371/journal.pbio.3001537. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Ghosh S, and Maunsell JHR (2024). Locus coeruleus norepinephrine contributes to visual-spatial attention by selectively enhancing perceptual sensitivity. Neuron 112, 2231–2240.e5. 10.1016/j.neuron.2024.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Escanilla O, Arrellanos A, Karnow A, Ennis M, and Linster C (2010). Noradrenergic modulation of behavioral odor detection and discrimination thresholds in the olfactory bulb. Eur. J. Neurosci. 32, 458–468. 10.1111/j.1460-9568.2010.07297.x. [DOI] [PubMed] [Google Scholar]
41.Rodenkirch C, Liu Y, Schriver BJ, and Wang Q (2019). Locus coeruleus activation enhances thalamic feature selectivity via norepinephrine regulation of intrathalamic circuit dynamics. Nat. Neurosci. 22, 120–133. 10.1038/s41593-018-0283-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Sciolino NR, Hsiang M, Mazzone CM, Wilson LR, Plummer NW, Amin J, Smith KG, McGee CA, Fry SA, Yang CX, et al. (2022). Natural locus coeruleus dynamics during feeding. Sci. Adv. 8, eabn9134. 10.1126/sciadv.abn9134. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Feng J, Dong H, Lischinsky JE, Zhou J, Deng F, Zhuang C, Miao X, Wang H, Li G, Cai R, et al. (2024). Monitoring norepinephrine release in vivo using next-generation GRAB. Neuron 112, 1930–1942.e6. 10.1016/j.neuron.2024.03.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Meng C, Zhou J, Papaneri A, Peddada T, Xu K, and Cui G (2018). Spectrally Resolved Fiber Photometry for Multi-component Analysis of Brain Circuits. Neuron 98, 707–717.e4. 10.1016/j.neuron.2018.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Fan W, Engborg CB, and Sciolino NR (2024). Locus Ceruleus Dynamics Are Suppressed during Licking and Enhanced Postlicking Independent of Taste Novelty. eNeuro 11. 10.1523/ENEURO.0535-23.2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Deal AL, Bass CE, Grinevich VP, Delbono O, Bonin KD, Weiner JL, and Budygin EA (2020). Bidirectional Control of Alcohol-drinking Behaviors Through Locus Coeruleus Optoactivation. Neuroscience 443, 84–92. 10.1016/j.neuroscience.2020.07.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.McCall JG, Siuda ER, Bhatti DL, Lawson LA, McElligott ZA, Stuber GD, and Bruchas MR (2017). Locus coeruleus to basolateral amygdala noradrenergic projections promote anxiety-like behavior. Elife 6, e18247. 10.7554/eLife.18247. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Budygin E, Grinevich V, Wang ZM, Messi ML, Meeker WR, Zhang J, Stewart WM, Milligan C, and Delbono O (2025). Aging disrupts locus coeruleus-driven norepinephrine transmission in the prefrontal cortex: Implications for cognitive and motor decline. Aging Cell 24, e14342. 10.1111/acel.14342. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Fox ME, Bucher ES, Johnson JA, and Wightman RM (2016). Medullary Norepinephrine Projections Release Norepinephrine into the Contralateral Bed Nucleus of the Stria Terminalis. ACS Chem. Neurosci. 7, 1681–1689. 10.1021/acschemneuro.6b00210. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Park J, Takmakov P, and Wightman RM (2011). In vivo comparison of norepinephrine and dopamine release in rat brain by simultaneous measurements with fast-scan cyclic voltammetry. J. Neurochem. 119, 932–944. 10.1111/j.1471-4159.2011.07494.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Neese C, Bouaichi CG, Needham T, Bauer M, Bertram R, and Vincis R (2022). Active Licking Shapes Cortical Taste Coding. J. Neurosci. 42, 8658–8669. 10.1523/JNEUROSCI.0942-22.2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Lin JY, Mukherjee N, Bernstein MJ, and Katz DB (2021). Perturbation of amygdala-cortical projections reduces ensemble coherence of palatability coding in gustatory cortex. Elife 10, e65766. 10.7554/eLife.65766. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Jezzini A, Mazzucato L, La Camera G, and Fontanini A (2013). Processing of hedonic and chemosensory features of taste in medial prefrontal and insular networks. J. Neurosci. 33, 18966–18978. 10.1523/JNEUROSCI.2974-13.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Samuelsen CL, Gardner MPH, and Fontanini A (2012). Effects of cue-triggered expectation on cortical processing of taste. Neuron 74, 410–422. 10.1016/j.neuron.2012.02.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Sadacca BF, Rothwax JT, and Katz DB (2012). Sodium concentration coding gives way to evaluative coding in cortex and amygdala. J. Neurosci. 32, 9999–10011. 10.1523/JNEUROSCI.6059-11.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Maier JX, and Katz DB (2013). Neural dynamics in response to binary taste mixtures. J. Neurophysiol. 109, 2108–2117. 10.1152/jn.00917.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Kogan JF, and Fontanini A (2024). Learning enhances representations of taste-guided decisions in the mouse gustatory insular cortex. Curr. Biol. 34, 1880–1892.e5. 10.1016/j.cub.2024.03.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Glendinning JI, Gresack J, and Spector AC (2002). A high-throughput screening procedure for identifying mice with aberrant taste and oromotor function. Chem. Senses 27, 461–474. 10.1093/chemse/27.5.461. [DOI] [PubMed] [Google Scholar]
59.Dotson CD, and Spector AC (2004). The relative affective potency of glycine, L-serine and sucrose as assessed by a brief-access taste test in inbred strains of mice. Chem. Senses 29, 489–498. 10.1093/chemse/bjh051. [DOI] [PubMed] [Google Scholar]
60.Grimm C, Duss SN, Privitera M, Munn BR, Karalis N, Frässle S, Wilhelm M, Patriarchi T, Razansky D, Wenderoth N, et al. (2024). Tonic and burst-like locus coeruleus stimulation distinctly shift network activity across the cortical hierarchy. Nat. Neurosci. 27, 2167–2177. 10.1038/s41593-024-01755-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Zhang XY, Bobadilla-Suarez S, Luo X, Lemonari M, Brincat SL, Siegel M, Miller EK, and Love BC (2025). Adaptive stretching of representations across brain regions and deep learning model layers. Nat. Commun. 16, 10302. 10.1038/s41467-025-65231-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Sereno AB, and Lehky SR (2018). Attention Effects on Neural Population Representations for Shape and Location Are Stronger in the Ventral than Dorsal Stream. eNeuro 5. 10.1523/ENEURO.0371-17.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Staszko SM, Boughter JD, and Fletcher ML (2022). The impact of familiarity on cortical taste coding. Curr. Biol. 32, 4914–4924.e4. 10.1016/j.cub.2022.09.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Carter ME, Yizhar O, Chikahisa S, Nguyen H, Adamantidis A, Nishino S, Deisseroth K, and de Lecea L (2010). Tuning arousal with optogenetic modulation of locus coeruleus neurons. Nat. Neurosci. 13, 1526–1533. 10.1038/nn.2682. [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Vazey EM, and Aston-Jones G (2014). Designer receptor manipulations reveal a role of the locus coeruleus noradrenergic system in isoflurane general anesthesia. Proc. Natl. Acad. Sci. USA 111, 3859–3864. 10.1073/pnas.1310025111. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Yao Z, van Velthoven CTJ, Kunst M, Zhang M, McMillen D, Lee C, Jung W, Goldy J, Abdelhak A, Aitken M, et al. (2023). A high-resolution transcriptomic and spatial atlas of cell types in the whole mouse brain. Nature 624, 317–332. 10.1038/s41586-023-06812-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Salgado H, Treviño M, and Atzori M (2016). Layer- and area-specific actions of norepinephrine on cortical synaptic transmission. Brain Res. 1641, 163–176. 10.1016/j.brainres.2016.01.033. [DOI] [PubMed] [Google Scholar]
68.McCormick DA, Pape HC, and Williamson A (1991). Actions of norepinephrine in the cerebral cortex and thalamus: implications for function of the central noradrenergic system. Prog. Brain Res. 88, 293–305. 10.1016/s0079-6123(08)63817-0. [DOI] [PubMed] [Google Scholar]
69.Devilbiss DM, and Waterhouse BD (2000). Norepinephrine exhibits two distinct profiles of action on sensory cortical neuron responses to excitatory synaptic stimuli. Synapse 37, 273–282. [DOI] [PubMed] [Google Scholar]
70.Labarrera C, Deitcher Y, Dudai A, Weiner B, Kaduri Amichai A, Zylbermann N, and London M (2018). Adrenergic Modulation Regulates the Dendritic Excitability of Layer 5 Pyramidal Neurons In Vivo. Cell Rep. 23, 1034–1044. 10.1016/j.celrep.2018.03.103. [DOI] [PubMed] [Google Scholar]
71.Zhang Z, Cordeiro Matos S, Jego S, Adamantidis A, and Séguéla P (2013). Norepinephrine drives persistent activity in prefrontal cortex via synergistic alpha1 and alpha2 adrenoceptors. PLoS One 8, e66122. 10.1371/journal.pone.0066122. [DOI] [PMC free article] [PubMed] [Google Scholar]
72.Pritchard TC, Hamilton RB, and Norgren R (1989). Neural coding of gustatory information in the thalamus of Macaca mulatta. J. Neurophysiol. 61, 1–14. 10.1152/jn.1989.61.1.1. [DOI] [PubMed] [Google Scholar]
73.Verhagen JV, Giza BK, and Scott TR (2003). Responses to taste stimulation in the ventroposteromedial nucleus of the thalamus in rats. J. Neurophysiol. 89, 265–275. 10.1152/jn.00870.2001. [DOI] [PubMed] [Google Scholar]
74.Samuelsen CL, Gardner MPH, and Fontanini A (2013). Thalamic contribution to cortical processing of taste and expectation. J. Neurosci. 33, 1815–1827. 10.1523/JNEUROSCI.4026-12.2013. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Piette CE, Baez-Santiago MA, Reid EE, Katz DB, and Moran A (2012). Inactivation of basolateral amygdala specifically eliminates palatability-related information in cortical sensory responses. J. Neurosci. 32, 9981–9991. 10.1523/JNEUROSCI.0669-12.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Mahmood A, Steindler JR, and Katz DB (2026). Sensory and palatability coding of taste stimuli in cortex involves dynamic and asymmetric cortico-amygdalar interactions. J. Neurophysiol. 135, 613–629. 10.1152/jn.00503.2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
77.Arandia-Romero I, Tanabe S, Drugowitsch J, Kohn A, and Moreno-Bote R (2016). Multiplicative and Additive Modulation of Neuronal Tuning with Population Activity Affects Encoded Information. Neuron 89, 1305–1316. 10.1016/j.neuron.2016.01.044. [DOI] [PMC free article] [PubMed] [Google Scholar]
78.Lin IC, Okun M, Carandini M, and Harris KD (2015). The Nature of Shared Cortical Variability. Neuron 87, 644–656. 10.1016/j.neuron.2015.06.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Sripati AP, and Johnson KO (2006). Dynamic gain changes during attentional modulation. Neural Comput. 18, 1847–1867. 10.1162/neco.2006.18.8.1847. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Eldar E, Cohen JD, and Niv Y (2013). The effects of neural gain on attention and learning. Nat. Neurosci. 16, 1146–1153. 10.1038/nn.3428. [DOI] [PMC free article] [PubMed] [Google Scholar]
81.Shine JM, Aburn MJ, Breakspear M, and Poldrack RA (2018). The modulation of neural gain facilitates a transition between functional segregation and integration in the brain. Elife 7, e31130. 10.7554/eLife.31130. [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Guedj C, Meunier D, Meunier M, and Hadj-Bouziane F (2017). Could LC-NE-Dependent Adjustment of Neural Gain Drive Functional Brain Network Reorganization? Neural Plast. 2017, 4328015. 10.1155/2017/4328015. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Chance FS, Abbott LF, and Reyes AD (2002). Gain modulation from background synaptic input. Neuron 35, 773–782. 10.1016/s0896-6273(02)00820-6. [DOI] [PubMed] [Google Scholar]
84.Ferguson KA, and Cardin JA (2020). Mechanisms underlying gain modulation in the cortex. Nat. Rev. Neurosci. 21, 80–92. 10.1038/s41583-019-0253-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Noei S, Zouridis IS, Logothetis NK, Panzeri S, and Totah NK (2022). Distinct ensembles in the noradrenergic locus coeruleus are associated with diverse cortical states. Proc. Natl. Acad. Sci. USA 119, e2116507119. 10.1073/pnas.2116507119. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.Fazlali Z, Ranjbar-Slamloo Y, Adibi M, and Arabzadeh E (2016). Correlation between Cortical State and Locus Coeruleus Activity: Implications for Sensory Coding in Rat Barrel Cortex. Front. Neural Circ. 10, 14. 10.3389/fncir.2016.00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
87.Marzo A, Totah NK, Neves RM, Logothetis NK, and Eschenko O (2014). Unilateral electrical stimulation of rat locus coeruleus elicits bilateral response of norepinephrine neurons and sustained activation of medial prefrontal cortex. J. Neurophysiol. 111, 2570–2588. 10.1152/jn.00920.2013. [DOI] [PubMed] [Google Scholar]
88.Durán E, Yang M, Neves R, Logothetis NK, and Eschenko O (2021). Modulation of Prefrontal Cortex Slow Oscillations by Phasic Activation of the Locus Coeruleus. Neuroscience 453, 268–279. 10.1016/j.neuroscience.2020.11.028. [DOI] [PubMed] [Google Scholar]
89.Neves RM, van Keulen S, Yang M, Logothetis NK, and Eschenko O (2018). Locus coeruleus phasic discharge is essential for stimulus-induced gamma oscillations in the prefrontal cortex. J. Neurophysiol. 119, 904–920. 10.1152/jn.00552.2017. [DOI] [PubMed] [Google Scholar]
90.Wang X, Li Z, Wang X, Chen J, Guo Z, Qiao B, and Qin L (2024). Effects of phasic activation of locus ceruleus on cortical neural activity and auditory discrimination behavior. J. Neurosci. 44, e1296232024. 10.1523/JNEUROSCI.1296-23.2024. [DOI] [PMC free article] [PubMed] [Google Scholar]
91.Clayton EC, Rajkowski J, Cohen JD, and Aston-Jones G (2004). Phasic activation of monkey locus ceruleus neurons by simple decisions in a forced-choice task. J. Neurosci. 24, 9914–9920. 10.1523/JNEUROSCI.2446-04.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Yerkes RM, and Dodson JD (1908). The relation of strength of stimulus to rapidity of habit-formation. Journal of Comparative Neurology & Psychology 18, 459–482. 10.1002/cne.920180503. [DOI] [Google Scholar]
93.Devilbiss DM (2019). Consequences of tuning network function by tonic and phasic locus coeruleus output and stress: Regulating detection and discrimination of peripheral stimuli. Brain Res. 1709, 16–27. 10.1016/j.brainres.2018.06.015. [DOI] [PubMed] [Google Scholar]
94.Ramos BP, and Arnsten AFT (2007). Adrenergic pharmacology and cognition: focus on the prefrontal cortex. Pharmacol. Ther. 113, 523–536. 10.1016/j.pharmthera.2006.11.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Jones LM, Fontanini A, Sadacca BF, Miller P, and Katz DB (2007). Natural stimuli evoke dynamic sequences of states in sensory cortical ensembles. Proc. Natl. Acad. Sci. USA 104, 18772–18777. 10.1073/pnas.0705546104. [DOI] [PMC free article] [PubMed] [Google Scholar]
96.Azeredo da Silveira R, and Rieke F (2021). The Geometry of Information Coding in Correlated Neural Populations. Annu. Rev. Neurosci. 44, 403–424. 10.1146/annurev-neuro-120320-082744. [DOI] [PubMed] [Google Scholar]
97.Kohn A, Coen-Cagli R, Kanitscheider I, and Pouget A (2016). Correlations and Neuronal Population Information. Annu. Rev. Neurosci. 39, 237–256. 10.1146/annurev-neuro-070815-013851. [DOI] [PMC free article] [PubMed] [Google Scholar]
98.Panzeri S, Moroni M, Safaai H, and Harvey CD (2022). The structures and functions of correlations in neural population codes. Nat. Rev. Neurosci. 23, 551–567. 10.1038/s41583-022-00606-4. [DOI] [PubMed] [Google Scholar]
99.Kobak D, Pardo-Vazquez JL, Valente M, Machens CK, and Renart A (2019). State-dependent geometry of population activity in rat auditory cortex. Elife 8, e44526. 10.7554/eLife.44526. [DOI] [PMC free article] [PubMed] [Google Scholar]
100.Minces V, Pinto L, Dan Y, and Chiba AA (2017). Cholinergic shaping of neural correlations. Proc. Natl. Acad. Sci. USA 114, 5725–5730. 10.1073/pnas.1621493114. [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Gardner MPH, and Fontanini A (2014). Encoding and tracking of outcome-specific expectancy in the gustatory cortex of alert rats. J. Neurosci. 34, 13000–13017. 10.1523/JNEUROSCI.1820-14.2014. [DOI] [PMC free article] [PubMed] [Google Scholar]
102.de Araujo IE, and Simon SA (2009). The gustatory cortex and multisensory integration. Int. J. Obes. 33, S34–43. 10.1038/ijo.2009.70. [DOI] [PMC free article] [PubMed] [Google Scholar]
103.Samuelsen CL, and Fontanini A (2017). Processing of Intraoral Olfactory and Gustatory Signals in the Gustatory Cortex of Awake Rats. J. Neurosci. 37, 244–257. 10.1523/JNEUROSCI.1926-16.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
104.Maier JX (2017). Single-neuron responses to intraoral delivery of odor solutions in primary olfactory and gustatory cortex. J. Neurophysiol. 117, 1293–1304. 10.1152/jn.00802.2016. [DOI] [PMC free article] [PubMed] [Google Scholar]
105.Nash AN, Shakeshaft M, Bouaichi CG, Odegaard KE, Needham T, Bauer M, Bertram R, and Vincis R (2025). Cortical coding of gustatory and thermal signals in active licking mice. J. Physiol. 603, 909–928. 10.1113/JP287499. [DOI] [PMC free article] [PubMed] [Google Scholar]
106.Fonseca E, de Lafuente V, Simon SA, and Gutierrez R (2018). Sucrose intensity coding and decision-making in rat gustatory cortices. Elife 7, e41152. 10.7554/eLife.41152. [DOI] [PMC free article] [PubMed] [Google Scholar]
107.Vincis R, Chen K, Czarnecki L, Chen J, and Fontanini A (2020). Dynamic Representation of Taste-Related Decisions in the Gustatory Insular Cortex of Mice. Curr. Biol. 30, 1834–1844.e5. 10.1016/j.cub.2020.03.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
108.Peng Y, Gillis-Smith S, Jin H, Tränkner D, Ryba NJP, and Zuker CS (2015). Sweet and bitter taste in the brain of awake behaving animals. Nature 527, 512–515. 10.1038/nature15763. [DOI] [PMC free article] [PubMed] [Google Scholar]
109.Grill HJ, and Norgren R (1978). The taste reactivity test. I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Res. 143, 263–279. 10.1016/0006-8993(78)90568-1. [DOI] [PubMed] [Google Scholar]
110.Lemon CH, Li J, Ali MSS, Ngum NM, Zumpano KT, and Roberts CJ (2025). Parabrachial Calca Neurons Influence Aversive and Appetitive Taste Function. eNeuro 12. 10.1523/ENEURO.0191-25.2025. [DOI] [PMC free article] [PubMed] [Google Scholar]
111.Clarke SN, and Ossenkopp KP (1998). Taste reactivity responses in rats: influence of sex and the estrous cycle. Am. J. Physiol. 274, R718–724. 10.1152/ajpregu.1998.274.3.R718. [DOI] [PubMed] [Google Scholar]
112.Lin JY, Stone BT, Rahamim OM, Craddock AE, and Katz DB (2026). Changes in Palatability Processing across the Estrous Cycle Are Modulated by Hypothalamic Estradiol Signaling. eNeuro 13. 10.1523/ENEURO.0225-25.2026. [DOI] [PMC free article] [PubMed] [Google Scholar]
113.Klapoetke NC, Murata Y, Kim SS, Pulver SR, Birdsey-Benson A, Cho YK, Morimoto TK, Chuong AS, Carpenter EJ, Tian Z, et al. (2014). Independent optical excitation of distinct neural populations. Nat. Methods 11, 338–346. 10.1038/nmeth.2836. [DOI] [PMC free article] [PubMed] [Google Scholar]
114.Caestecker S, Lescrauwaet E, Boon P, Carrette E, Raedt R, and Vonck K (2025). The Locus Coeruleus-Noradrenergic System in the Healthy and Diseased Brain: A Narrative Review. Eur. J. Neurol. 32, e70337. 10.1111/ene.70337. [DOI] [PMC free article] [PubMed] [Google Scholar]
115.Zhou P, Resendez SL, Rodriguez-Romaguera J, Jimenez JC, Neufeld SQ, Giovannucci A, Friedrich J, Pnevmatikakis EA, Stuber GD, Hen R, et al. (2018). Efficient and accurate extraction of in vivo calcium signals from microendoscopic video data. Elife 7, e28728. 10.7554/eLife.28728. [DOI] [PMC free article] [PubMed] [Google Scholar]
116.Wilson LR, Plummer NW, Evsyukova IY, Patino D, Stewart CL, Smith KG, Konrad KS, Fry SA, Deal AL, Kilonzo VW, et al. (2024). Partial or Complete Loss of Norepinephrine Differentially Alters Contextual Fear and Catecholamine Release Dynamics in Hippocampal CA1. Biol. Psychiatry Glob. Open Sci. 4, 51–60. 10.1016/j.bpsgos.2023.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS2212249-supplement-1.pdf^{(902.6KB, pdf)}

Data Availability Statement

Lead contact

For questions or additional information, please contact the lead contact, Natale R. Sciolino (natale.sciolino@uconn.edu). All reasonable requests will be accommodated.

Materials availability

No new or unique reagents were generated.

Data and code availability

Data: All neural and behavioral datasets generated here have been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Code: Original analysis code used has been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Data: All neural and behavioral datasets generated here have been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Code: Original analysis code used has been deposited at Zenodo (doi:10.5281/zenodo.21826182).
Other items: Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

PERMALINK

Locus coeruleus activation transforms cortical taste representations

Will Fan

Paula To

Natale R Sciolino

SUMMARY

Graphical Abstract

In brief

INTRODUCTION

RESULTS

Probing the effects of phasic LC activation on taste-evoked GC activity

Figure 1. Approaches for assessing phasic LC modulation of taste-evoked GC responses.

PHASIC LC ACTIVATION BIDIRECTIONALLY MODULATES TASTE-EVOKED GC RESPONSES

Figure 2. Phasic LC activation bidirectionally modulates taste-evoked responses and reorganizes GC palatability representation.

Phasic LC activation enhances GC palatability correlation without altering tastant separability

Phasic LC activation reshapes GC population coding along a palatability axis

Phasic LC activation stretches and rotates GC representations of mixture ratios

Figure 3. Phasic LC activation reshapes GC representations of mixture composition and sucrose intensity.

Phasic LC activation modestly expands and rotates GC representations of sucrose concentration

Parsing tuning modulations underlying population-level effects of phasic LC activation

Figure 4. Phasic LC activation reshapes neuronal tuning in an attribute-dependent manner.

LC-mediated gain modulation is structured and attribute specific

Distinct contributions of scale-shift and flexible modulation to LC-induced transformations

Tonic LC activation has limited impact on GC taste encoding

Figure 5. Effects of prolonged tonic LC activation on GC representation of taste attributes.

DISCUSSION

Population-level effects of LC modulation

Mechanisms of LC-induced neuronal effects

Links to state-dependent modulation

Open questions

Limitations of the study

STAR★METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

METHOD DETAILS

Survival surgical procedures

Intracranial viral injections

Intracranial optical fiber and grin lens implantation

Intraoral cannulae (IOC) implantation

Water restriction protocol

Passive tastant delivery

In vivo miniscope calcium imaging

Optogenetic stimulation procedures

Experimental timeline

Fiber photometry recordings of photostimulation-evoked NE release

Brief-access taste testing

Tissue collection and processing

Immunohistochemistry

Digital image processing

Data presentation and visualization

QUANTIFICATION AND STATISTICAL ANALYSIS

Miniscope movie pre-processing and cell-identification

Identification of taste-discriminative and LC-modulated neurons

Population decoding

Response–stimulus correlation

Principal component analysis (PCA) and LC-induced transformations

Modulation type analysis

Fiber photometry data analysis

Brief-access taste test analysis

Supplementary Material

KEY RESOURCES TABLE.

Highlights.

ACKNOWLEDGMENTS

Footnotes

RESOURCE AVAILABILITY

Lead contact

Materials availability

Data and code availability

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

Lead contact

Materials availability

Data and code availability

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases