Beyond Dopamine: Broader Neuromodulatory Contributions in Sign- and Goal-Tracking

Abstract

Sign- and goal-tracking behaviors provide a powerful framework for understanding individual differences in cue-motivated responding and vulnerability to addiction. While dopamine has long been recognized as central to the attribution of incentive salience, growing evidence indicates that multiple neuromodulatory systems contribute to these behaviors through distinct mechanisms. This review synthesizes findings on the cholinergic, serotonergic, adrenergic, opioid, and stimulant (nicotine, caffeine) systems in modulating sign- and goal-tracking in rodents. A better understanding of these mechanisms may guide individualized interventions to reduce cue-triggered relapse and improve addiction treatment outcomes.

Why Study Individual Differences

Understanding individual differences is critical for uncovering the neurobiological mechanisms that drive vulnerability to substance abuse disorders. While many individuals experiment with drugs, only a subset develops compulsive use, highlighting the need to investigate why some brains are more susceptible than others (Deroche-Gamonet et al., 2004; George & Koob, 2017). Animal models play a key role in identifying both shared and divergent neurobiological pathways underlying substance use. Studying these individual differences not only reveals novel targets for intervention but also opens the door to more personalized therapeutic strategies that address the specific molecular or circuit-level alterations contributing to addiction risk (Ragia & Manolopoulos, 2018; Yadid et al., 2013).

Sign Tracking as a Model to Understand Individual Differences in Addiction

One well-established model of individual differences relevant to addiction is the distinction between sign-tracking and goal-tracking behavior (Gheidi et al., 2023; Robinson & Flagel, 2009). In Pavlovian conditioning paradigms, some animals (sign-trackers, STs) (Hearst & Jenkins, 1974) are drawn to cues that predict reward delivery, while others (goal-trackers, GTs) (Boakes, 1977) focus on the location of reward availability. A third phenotype (intermediate trackers, ITs) exists along this continuum, which does not show dominant responding to one or the other, demonstrating a mixture of ST and GT with no clear preference for either. This suggests that there are fundamental differences in how incentive salience is attributed (Saunders & Robinson, 2013). For instance, STs and GTs differ in how they respond to both cues predicting reward and the reward itself. Given the choice between cocaine and food, STs prefer cocaine more often than GTs (Tunstall & Kearns, 2015), and ST behavior has been linked to impulsivity (Flagel et al., 2010; Lovic et al., 2011).

Pavlovian conditioned approach (PCA) behaviors provide a powerful and well-validated framework for understanding individual differences in addiction vulnerability. These behavioral phenotypes are remarkably stable across time and heritable across generations, and they robustly predict a range of addictive-relevant outcomes, including the rapid acquisition of drug-self administration, escalation of drug seeking, resistance to extinction, and heightened cue-induced reinstatement after abstinence (Morrow et al., 2011; Saunders & Robinson, 2013; Tomie et al., 2008; Ahrens et al., 2015). ST reflects the attribution of salience to reward-predictive cues, a motivational process that parallels core features of human addiction such as elevated cue-reactivity, attentional capture by drug-associated stimuli, cue-evoked craving, and relapse triggered by environmental reminders of drug use. Conversely, GT is associated with greater flexibility and reduced cue-sensitivity, aligning with lower relapse risk. Thus, individual variation in the ST / GT behavior models trait-like, cross-species differences in cue-motivated responding making this paradigm a strong translational tool for studying addiction liability.

This translational relevance underscores the value of the ST / GT framework dissecting the neurobiological mechanisms that confer vulnerability to substance use disorders. While dopamine has long been recognized as central to incentive salience attribution, growing evidence demonstrates that cholinergic, serotonergic, adrenergic, opioid and adenosine signalling also shape these behavioral tendencies. Identifying how these neuromodulators bias individuals toward ST or GT has important clinical implications: it can clarify which neurotransmitter systems mediate susceptibility to addiction, and it may illuminate why certain individuals are more prone to relapse when exposed to drug-associated cues. Relapse, a defining feature of addiction, can be triggered by environmental stimuli months or even years after abstinence (Frawley & Smith, 1992; Lee et al., 2022; Moos & Moos, 2006; Zhu et al., 2018). Understanding the mechanisms that drive heightened cue-reactivity in STs may therefore reveal predictive markers of addiction risk that emerge well before drug exposure and identify novel targets for relapse prevention.

The Use of the Pavlovian Conditioned Approach Index (PCA)

To contextualize the neurobiological differences between STs and GTs, it is first necessary to discuss the Pavlovian Conditioned Approach Index (PCA) for several reasons, a widely used measure for quantifying individual variation in cue-directed behavior (Meyer et al., 2012). This index is a composite score that considers multiple dependent variables. It is typically calculated using three behavioral measures during the cue (conditioned stimulus) period:

Probability Difference Score

$$P=(\text{Probability of lever contact})-(\text{Probability of food cup entry})$$

Response Bias Score

$$\begin{gathered} R=\underset{¯}{\text{Number of lever contacts}-\text{Number of food cup entries}} \\ \text{Number of lever contacts}+\text{Number of food cup entries} \end{gathered}$$

Latency Difference Score

$$L=\underset{¯}{\text{Latencty to food cup}-\text{Latency to lever contact}}$$

The PCA Index is computed as the average of these three normalized scores $(\text{PCA Index}=\underset{¯}{P+R+L}∕3)$. Each score ranges from 1 to +1. Positive values (+0.5 to +1.0) indicate ST bias. Negative values (−0.5 to −1.0) indicate GT bias. Intermediate values (−0.5 to +0.5) indicate IT phenotype. It's worth noting that some studies don't generate PCA scores, but we've still included them in this review (Angelyn et al., 2021; Palmatier et al., 2013; Stringfield et al., 2018; Stringfield et al., 2017; Winstanley et al., 2004)

Neurobiology of Sign and Goal Tracking

The dopamine system is one of the most extensively studied neurotransmitter systems in relation to ST and GT (Gyawali et al., 2023; Iglesias et al., 2023; Lopez et al., 2015; Roughley & Killcross, 2019). Multiple laboratories using pharmacological (Bacharach et al., 2023), optogenetic (Iglesias et al., 2023), and electrochemical approaches (Flagel et al., 2011) have consistently linked increased dopamine in the nucleus accumbens (NAc) to ST, particularly when the cue is present (Gillis & Morrison, 2019). Lesioning the NAc core region has been shown to result in deficits in ST behavior (Cardinal et al., 2002; Chang et al., 2012). Likewise, depleting dopamine (Parkinson et al., 2002) or blocking dopamine receptors (Saunders & Robinson, 2012) has a similar result. While both STs and GTs exhibit phasic dopamine release in the NAc in response to reward-predictive cues, only STs' cue-evoked responses increase while reward-evoked responses diminish across training. This shift, absent in GTs, suggests the acquisition of ST uniquely relies on mesolimbic dopamine-encoded reward prediction errors (Flagel et al., 2011). While the role of dopamine is important, its role has been comprehensively reviewed elsewhere (Tomie et al., 2008) and will not be the focus of this article. Readers are referred to Tomie et al., (2008) for a detailed discussion of dopaminergic contribution to ST and GT phenotypes. Here, we aim to survey the literature in an assessment of other neurotransmitter systems that may also play a role in PCA and ST / GT behavior.

The Role of the Endocannabinoid System in Sign and Goal Tracking

Given the well-established role that cue-evoked dopamine release in the NAc plays in ST behavior, the endocannabinoid system is a strong candidate neuromodulatory system underlying individual differences in cue-motivated responding. Endocannabinoid signalling in the ventral tegmental area (VTA) exerts powerful control over mesolimbic dopamine release transmission (Cheer et al., 2000; Cheer et al., 2004; Cheer et al., 2007; Lupica & Riegel, 2005; Lupica et al., 2004; Oleson & Cheer, 2012; Wenzel et al., 2018), and multiple studies demonstrate that endocannabinoids directly modulate the motivational value of reward-predictive cues. For example, pharmacological blockade of endocannabinoid CB1 receptors in the VTA reduces lever contacts and lever approach probability in STs and suppresses cue-evoked NAc dopamine release (Bacharach et al., 2018; Oleson & Cheer, 2012; Bacharach et al., 2023). Similarly, administration the non-selective cannabinoid agonist CP-55,940 during PCA training in rats produced a dose-dependent decrease in ST, reflected in fewer lever contacts and lower PCA indices, without affecting GT behavior (Gheidi et al., 2020).

Critically, Kruse et al., (2019) provide compelling evidence that the endocannabinoid system can shape long-term vulnerability to ST phenotypes. Using a model of adolescent voluntary THC ([−]-trans-Δ⁹-tetrahydrocannabinol) consumption, the authors showed that adolescent males, but not females, displayed enhanced ST in adulthood, along with selective downregulation of CB1 receptor-expressing glutamatergic terminals in the VTA. These long-lasting, sex-specific alterations in CB1-mediated control of VTA excitatory inputs suggest that developmental perturbations of endocannabinoid signaling shift dopamine-dependent cue-learning mechanisms toward a heightened incentive-salience profile. Kruse et al. (2019) thus directly link adolescent cannabinoid exposure with an adult ST bias via persistent endocannabinoid-driven changes in the VTA circuitry, supporting the model that endocannabinoid function contributes causally to individual variation in cue-motivated behavior.

Radioligand autoradiography and site-specific manipulations (e.g., fatty acid amide hydrolase (FAAH), the enzyme that degrades endogenous cannabinoids) further converge on the VTA as a key locus where CB1 signalling shapes ST behavior. STs exhibited elevated CB1 expression in the infralimbic cortex compared to GTs (Gheidi et al., 2020) and intra-VTA CB1 antagonism selectively reduces lever-directed responding in STs (Bacharach et al., 2023). Fiber photometry and optogenetic work confirm that CB1 receptors modulate VTA-NAc dopamine dynamics in a manner than tracks with cue-evoked approach, indicating that endocannabinoid activity gates the expression of incentive salience. Inhibition of VTA neurons with halorhodopsin while recording in the NAc reduced lever contacts selectively in STs. Similarly, intra-VTA CB1 receptor blockade with rimonabant decreased ST in STs but not intermediates (Bacharach et al., 2023).

Taken together, these findings indicate that endocannabinoids, particularly CB1 receptor signaling in the VTA → NAc circuit (Bacharach et al., 2023) and infralimbic cortex (Gheidi et al., 2020), plays a selective and mechanistically coherent role in modulating mesolimbic dopamine transmission that promotes ST behavior. Across studies, manipulating CB1 activity, whether by systemic antagonism, site-specific VTA blockade, or cannabinoid agonist administration, consistently alters cue-driven responding in STs, but not GTs, suggesting that the endocannabinoid system specifically modulates the attribution of incentive salience rather than general reward learning or goal-directed processes. Moreover, adolescence is a critical period for endocannabinoid-dopamine circuit maturation (Meyer et al., 2018; Tseng & Molla, 2025). Given that ST and cue-driven drug seeking rely on intact CB1 receptor-mediated facilitation of VTA dopamine release, adolescent cannabis exposure, which directly targets CB1 receptors, could bias individuals toward altered emotional or cue salience processing, potentially enhancing vulnerability to drug-related cues or impairing adaptive control over reward seeking (Draycott et al., 2014; Ferland et al., 2023; Tseng & Molla, 2025). Developmental perturbations (Kruse et al., 2019), systemic pharmacology, and region-specific CB1 manipulations converge on the conclusion that endocannabinoid tone biases individuals towards or away from high cue-reactivity phenotypes. This identifies CB1-modulated VTA dopamine plasticity as a critical pathway through which endocannabinoid signalling contributes to addiction-relevant susceptibility to cue-driven drug-seeking.

The Role of the Cholinergic System in Sign and Goal Tracking Behavior

Nicotine has long been recognized for its reinforcement-enhancing properties, particularly its ability to increase the motivational value of reward-predictive cues. Several studies have demonstrated that nicotine reliably increases responding for otherwise weak or neutral stimuli, enhancing both primary reinforcement and the incentive value of associated cues (Donny et al., 2003; Chaudhri et al., 2007; Olausson et al., 2004). Moreover, researchers have also shown that nicotine functions not only as a primary reinforcer, but also as a potent conditioned reinforcer. More specifically, nicotine-paired cues acquire heightened incentive salience, drive approach behavior, and promote cue-controlled responding (Palmatier et al., 2007; Murray & Bevins, 2007). Collectively, this literature demonstrates that activation of nicotinic acetylcholine receptors (nAChRs) amplifies the motivational significance of reward-predictive cues, directly aligning nicotine’s behavioral profile with incentive-salience processes central to ST.

More broadly, the cholinergic system influences attentional control and reward processing through widespread projections from the basal forebrain to cortical and limbic regions (Ballinger et al., 2016). Acetylcholine acts at two main receptor classes: fast-acting nAChRs, ligand-gated ion channels that modulate excitability and neurotransmitter release, and slower muscarinic acetylcholine receptors (mAChRs), G-protein coupled receptors, which regulate longer-timescale signalling (Akaike & Izumi, 2018). Nicotinic receptors in the VTA can directly modulate dopamine neuron activity (Liu et al., 2012), linking cholinergic signaling to the attribution of incentive salience, suggesting a mechanistic route through cholinergic tone may bias individuals toward either ST or GT phenotypes. Consistent with this framework, manipulations of the cholinergic system can shift PCA behavior. In the context of developing non-invasive pharmacological approaches for substance use disorders, we examined the effects of muscarinic (Scopolamine) and nicotinic (Mecamylamine) receptor antagonists on PCA behavior. Our results showed that pretreatment with scopolamine dose-dependently decreased ST (reducing lever contacts, lever approach probability, and PCA index) and increased GT, whereas pretreatment with mecamylamine selectively reduced lever-directed responding without significantly changing magazine entries or reclassifying animals into different phenotypic categories (Gheidi et al., 2023). These findings reinforce the view that cholinergic signaling is deeply embedded in the neurobiology of cue-driven approach behavior and may contribute to individual differences in addiction-relevant cue responsivity.

Systemic nicotine administration has produced mixed effects on PCA behavior. In some studies, nicotine biased CPA measures (approach, probability, and latency) toward GT (increasing magazine entries and food cup approach probability) without altering ST parameters (Angelyn et al., 2021), while in others, acute nicotine administration increased ST behavior, even in nicotine-naïve rats, suggesting that nicotine can enhance the incentive salience of cues, independent of reinforcement history (Overby et al., 2018). Additional reports have indicated that nicotine increases both ST and GT (Palmatier et al., 2013; Stringfield et al., 2017), with sex differences in ST reported (Stringfield et al., 2019). Moreover, the tendency to ST predicts cue-induced reinstatement during nicotine self-administration, and is enhanced by nicotine, but not alcohol (Versaggi et al., 2016).

One study also showed that independent of nicotine exposure, drug-naïve female rats displayed increased cue-directed behavior compared to males and increased ST citing that brain-derived neurotrophic factor did not play a role (Stringfield et al., 2019). Females will take more nicotine in extended-access models (Chellian et al., 2024). Female rodents have a greater proportion of dopamine projecting neurons compared to males (Kritzer & Creutz, 2008), and VTA-dopamine neurons are sensitive to circulating levels of estradiol (E2) (Kokane & Perrotti, 2020). Sex differences in nicotine receptor density also exist. For instance, in female rodents, nicotine does not lead to an upregulation of β2 nAChRs, whereas it does in males (Koylu et al., 1997). Evidence in humans indicates that men tend to smoke for nicotine’s reinforcing effects, while women are more influenced by non-nicotine factors such as negative affective regulation (related to stress), or conditioning (likely due to the fact that they show less nicotine-evoked dopamine release in the ventral striatum) (Verplaetse et al., 2025). Women also face greater difficulty quitting and respond less to nicotine replacement therapies (Perkins et al., 1999; Verplaetse et al., 2025), highlighting the need to understand sex-specific neurochemical mechanisms of nicotine use. Nicotine occupies a distinctive position among drugs of abuse. For individuals without prior exposure, it often produces strong aversive effects (Fowler & Kenny, 2014), and its modest euphoric properties are subject to rapid tolerance (Benowitz, 2010). Yet, despite these features and well-known negative health consequences, nicotine remains highly addictive, with cessation proving difficult for many users. One mechanism that helps reconcile this paradox is nicotine’s ability to enhance reinforcement: as a reinforcement enhancer, it can amplify the incentive value of other stimuli, including cues associated with alcohol and other drugs of abuse Perkins et al., 2017).

Interestingly, Pitchers et al. (2017) examined how a discriminative stimulus (a cue signaling cocaine availability) influenced relapse in ST and GT rats, and looked at the role of basal forebrain cholinergic activity in this process. The discriminative stimulus triggered stronger reinstatement in GTs compared to STs, and cholinergic disruption reduced cue-induced cocaine seeking in GTs. In STs, reinstatement was not affected, indicating distinct neural mechanisms underlying relapse in the two phenotypes. Cholinergic neurons likely influence relapse by modulating attentional and cognitive control processes, making GTs particularly sensitive to discriminative stimuli. Nicotine exposure has also been shown to promote behavioral inflexibility (Stringfield et al., 2018): nicotine-treated rats persisted in responding to outdated cue locations during reward-omission trials and in cue–reward reassignment tasks, suggesting impaired reward updating. One study examining the how acetylcholine receptors in the NAc core influenced the flexibility of cue-motivated behavior described nAChRs and mAChRs as having opposing roles in regulating the persistence and restructuring of cue-directed motivation (Townsend et al., 2023). Using an omission procedure (where lever deflection during cue presentation cancelled reward delivery), the authors tested how blocking nAChRs or mAChRs receptors affected behavioral adaptation. They found that nAChRs blockade promoted flexibility later but initially heightened cue engagement, whereas mAChR blockade showed more flexibility from the outset. Thus, modulating accumbal acetylcholine signaling may offer a means to reshape maladaptive cue-driven behaviors, providing a potential therapeutic avenue for conditions characterized by rigid, compulsive responses to reward-associated cues.

Mechanistic studies have probed the cholinergic contributions to attentional processes in STs and GTs. Using chemogenetic hM4Di (human M4 muscarinic inhibitory) DREADDs (Designer Receptors Exclusively Activated by Designer Drugs) inhibition of cholinergic neurons in the horizontal limb of the diagonal band (HBD), it was found that suppression of these neurons impaired sustained attention task performance in GTs but not STs, reducing goal-hit rates in the former (Kucinski et al., 2022). These findings indicate that superior attentional performance in GTs depends on intact cholinergic function in the basal forebrain. Cortical acetylcholine availability also appears to shape tracking phenotype. Pharmacological inhibition of presynaptic choline transporters (CHT) with VU6001221 in GT rats induced a phenotypic shift towards ST. Reflected in increased lever contacts, decreased lever latency, reduced magazine entries, and increased magazine entry latency. Furthermore, basal forebrain electrostimulation (BF-ES) was performed in rats anesthetized with urethane and ultimately enhanced CHT expression in GT, but not ST rats, thus indicating that GT rats possess a greater capacity to mobilize CHT to maintain GT behavior during attentional demand (Koshy Cherian et al., 2017).

In sum, cholinergic signaling influences ST and GT in a receptor-specific and context-dependent manner. Muscarinic antagonism reduces ST and promotes GT, while nicotinic antagonism selectively suppresses ST, without affecting GT. Nicotine itself can bias behavior toward either phenotype depending on dose, sex, and task parameters, consistent with its reinforcement-enhancing effects and complex modulation of dopamine signaling. Mechanistically, basal forebrain cholinergic function and cortical acetylcholine availability appear critical for sustaining GT behavior, whereas disruptions in these systems, whether via pharmacological blockade, attentional impairment, or altered receptor regulation, tend to bias animals toward ST.

The Role of the Serotonergic System in Sign and Goal Tracking

Compared to other neurotransmitter systems, relatively few studies have examined serotonin’s role in ST & GT behavior. Early work, primarily in the context of impulsivity, showed that intracerebroventricular infusions of the 5-HT neurotoxin 5,7-dihydroxytryptamine (5,7-DHT). dihydroxytryptamine increased approaches to both CS− and CS+ with lesioned animals making more approaches to the CS+ during omission trials (Winstanley et al., 2004) suggesting impaired discrimination and inhibitory control. Similarly, food-restricted serotonin knockout rats also did not differ from wildtype controls in either ST or GT responses (Nonkes et al., 2014) indicating that global serotoninergic disruptions do not uniformly shift PCA phenotypes.

More recent pharmacological studies provide clearer evidence of serotonergic modulation. Systemic selective serotonin reuptake inhibitors (SSRIs) administration reliably decreases ST. Escitalopram produced a dose-dependent decrease in ST, reducing lever contacts and lever-approach probability and increasing latency to approach the lever, but only in animals that exhibit a ST phenotype, without comparable effects in GTs (Holden, 2023). Likewise, citalopram, escitalopram, and fluoxetine reduced lever-directed behavior in male Sprague-Dawley rats, although effects on GT varied across compounds. These findings suggest that increasing extracellular serotonin attenuate the motivational pull of conditioned cues, raising the possibility that SSRIs may reduce cue-triggered relapse by dampening ST tendencies (Holden, 2023).

Campus et al. (2016) provide the most detailed developmental analysis of serotonergic contributions to PCA. In this work, DBA/2J and C57BL/6J mice displayed robust ST during adolescence.However, only DBA/2J mice maintained a relaible ST phenotype into adulthood. Overall, PCA scores in mice were generally positive, consistent with a bias toward ST behavior. To test whether serotonergic signaling in the medial prefrontal cortex (mPFC) regulates this developmental trajectory, Campus et al. (2016) selectively depleted mPFC serotonin in adolescent C57BL/6J mice using the neurotoxin 5,7-dihydroxytryptamine (5,7-DHT). This manipulation caused C57BL/6J mice, to develop persistent adult ST behavior, mimicking the DBA/2J phenotype. These findings demonstrate that serotonergic tone in the mPFC during adolescence plays a critical role in constraining the emergence and persistence of ST behavior.

Taken together, the limited literature indicates that serotonin modulates ST and GT through multiple routes: (1) serotonergic disruption can impair cue discrimination and inhibitory control, increasing maladaptive cue engagement; (2) SSRIs reduce ST, suggesting that elevated serotonin dampens incentive-salience attribution; and (3) developmental serotonin depletion in the mPFC promotes persistent adult ST. Although sparse, these converging findings support serotonin as a key regulator of the balance between ST and GT, with alterations in serotonergic signaling, whether pharmacological or neurodevelopmental, biasing individuals toward or away from cue-directed behavior.

The Role of the Adrenergic System in Sign and Goal Tracking

Compared to other neuromodulators, the adrenergic system has received even less attention than the serotonergic system in the context of ST and GT. However, two recent studies provide important insight. Holden (2022, 2023) tested atomoxetine (a norepinephrine reuptake inhibitor) and guanfacine an alpha-2-adrenergic agonist) in rats performing the PCA task. Atomoxetine produced a dose-dependent suppression of both ST and GT, but the pattern differed by behavioral phenotype. ST was robustly reduced at all doses. Importantly, atomoxetine primarily reduced the intensity of lever and magazine-directed responding (PCA component scores) rather than fully converting ST animals into GT; many rats instead shifted toward intermediate phenotype scores. This included fewer lever contacts, shorter lever contact durations, reduced probability of approaching the lever during the conditioned stimulus, and increased latency to first elver contact. GT was only reduced at the highest dose tested. This included reduced magazine entries, lower probability of approaching the food cup, and increases latency to approach the magazine. This suggests that modest elevations in prefrontal norepinephrine preferentially disrupts cue-directed incentive behaviors (ST), whereas GT, which relies more on predictive value than incentive salience, only becomes disrupted when NE levels are sufficiently high to interfere with goal-directed approach or attentional selection. In contrast, guanfacine failed to alter either ST or GT across a wide dose range, indicating that not all adrenergic manipulations influence conditioned approach in the same way.

The divergent patterns map onto their pharmacological profiles. Both drugs result in increased extracellular norepinephrine (in the PFC specifically) and are used clinically to treat attention deficit and hyperactivity disorder (ADHD) (Arnsten & Pliszka, 2011), yet their divergent outcomes highlight a critical point: the behavioral effects of norepinephrine depend on receptor subtype engagement (α1, α2, β) and circuit-level action. Atomoxetine increases extracellular norepinephrine broadly by blocking the norepinephrine transporter, thereby engaging multiple adrenergic receptor subtypes (Garnock-Jones & Keating, 2009) across cortical and subcortical circuits. Guanfacine, by contrast, selectively stimulates presynaptic α2A autoreceptors, which reduces norepinephrine release through negative feedback in some pathways (Arnsten, 2020). Atomoxetine enhances norepinephrine signaling across multiple receptor systems, while guanfacine dampens norepinephrine output in certain pathways. In ST and GT, the outcome seems to depend not only on the amount of norepinephrine but on where and how it is acting. Thus, despite both raising extracellular norepinephrine in the PFC, their net circuit effects differ dramatically, highlighting that adrenergic modulation of ST and GT depends on receptor specificity, circuit engagement, and the degree of noradrenergic elevation. Accordingly, the adrenergic system’s role in cue-motivated behavior is mechanism-dependent because the same neurotransmitter produces opposite behavioral outcomes depending on dose, receptor subtype, and circuit-level action.

The timing of adrenergic modulation is equally critical. Pasquariello et al. (2018) demonstrated that reducing norepinephrine synthesis with the dopamine β-hydroxylase inhibitor disulfiram suppressed ST only when administered before Pavlovian training, with no impact when given after ST was already acquired. Blockade of α1-adrenergic receptors (prazosin) or β-adrenergic receptors (propranolol) also selectively reduced ST without affecting GT. Together, these studies show that ST acquisition and expression depend on intact norepinephrine transmission, whereas GT is comparatively insensitive to adrenergic manipulation, again supporting a more norepinephrine-dependent motivational mechanism in ST.

Within a broader framework, these pharmacological effects align with the known role of the locus coeruleus (LC), the brain’s primary source of norepinephrine (Dahlström & Fuxe, 1965). LC neurons exhibit phasic firing to predictive and salient cues (Aston-Jones et al., 1997) and support attentional orienting / shifts, behavioral flexibility, and adaptive cue-outcome updating (when an organism changes what it expects a cue to predict) (Aston-Jones et al., 2000; Grella & Donaldson, 2024; McBurney-Lin et al., 2022). Importantly, LC norepinephrine activity increases when contingencies change, facilitating updating during extinction, reversal learning, and prediction-error conditions (Grella et al., 2021; Janitzky et al., 2015; Rorabaugh et al., 2017). Here, phasic signaling specifically refers to rapid, event-locked bursts of LC activity that enhance learning about conditioned stimuli. This helps adjust cue-driven behavior when the cue’s meaning shifts. In contrast, dysregulated or exaggerated adrenergic signaling may exaggerate the motivational value of reward-predictive cues, contributing to the persistence of ST even when contingencies change.

Overall, the literature indicates that ST, but not GT, is strongly dependent on adrenergic signaling, and that the behavioral consequences of adrenergic manipulation depend on when norepinephrine is altered (during learning vs after), where in the circuit it is modulated (PFC vs subcortical targets), and which receptors are engaged. This unified framework clarifies why adrenergic drugs produce heterogeneous effects across studies and underscores norepinephrine’s selective contribution to the motivational component of cue-directed behavior.

The Role of the Opioid System in Sign and Goal Tracking

Given that ST reflects the attribution of incentive salience to reward-predictive cues, whereas GT reflects the reward itself, the opioid system may regulate both behaviors by modulating motivational value. One study examined whether direct stimulation of μ-opioid receptors in the central amygdala (CeA) could influence ST & GT in rats. Site-specific microinjections of DAMGO, a selective μ-opioid receptor agonist (D-Ala2,NMePhe4,Glyol5enkephalin) enhanced approach behaviors toward each phenotype’s preferred target. In GTs, DAMGO increased contacts with the food magazine, while in STs, it increased lever interactions (DiFeliceantonio & Berridge, 2012). These effects indicate that CeA μ-opioid activation amplifies the motivational salience of reward-related cues and outcomes, intensifying existing behavioral biases rather than shifting animals between phenotypes. In the same study, DAMGO increased lever presses and food cup entries only in response to reward-predictive cues, with no effect on cues unassociated with reward. This demonstrates that the observed increases were linked to reward expectation, rather than general hyperactivity.

In addition to μ-opioid receptors, the brain also expresses δ-and κ-opioid receptors (KORs) each with distinct functional profiles and anatomical distributions (Erbs et al., 2015; Le Merrer et al., 2009). Although direct evidence linking these receptor systems to ST & GT is limited, research in reward-related circuits points to their potential roles. δ-opioid agonists like DPDPE (D-Pen 2, D-Pen 5] enkephalin) and deltorphin II, similar to DAMGO and morphine, induce neuronal hyperpolarization by activating inhibitory Gi/o G-proteins, reducing the influx of calcium ions and opening G-protein-dependent inward rectifier potassium channels (GIRKS) (Reeves et al., 2022). This suggests that δ-opioid receptors may be able to modulate excitability in brain regions involved in reward processing and may potentially act in similar ways to μ-opioid modulation. In fact, stimulating opioid receptors within the NAc shell has been shown to enhance hedonic “liking” responses (Castro & Berridge, 2016). This highlights the potential for these receptors to influence incentive salience and cue attraction.

In the above study, they found that the κ-opioid receptor agonist, U50488H, enhanced hedonic “liking” responses (Castro & Berridge, 2016). KOR activation produces a well-established aversive state, largely via dynorphin-mediated suppression of mesolimbic dopamine (Bals-Kubik et al., 1993; Grella et al., 2014; Land et al., 2009); thus, its role in reward processing is not well-understood. With respect to conditioned place preference, low doses of the κ-opioid receptor agonist salvinorin A induced preference but high doses induced aversion (Braida et al., 2008). Similarly, rats will self-administer low doses but not high doses (Braida et al., 2008). These findings raise the possibility that mild κ-opioid receptor activation could, like μ-opioid stimulation, enhance the motivational salience of cues or outcomes thereby amplifying pre-existing ST or GT tendencies. In contrast, high dose κ-opioid receptor activation would potentially suppress ST (especially dopamine-dependent effects) consistent with the dopaminergic suppression (Gray et al., 1999) and dysphoric states associated with robust κ-opioid receptor stimulation (Carlezon et al., 2006). Moreover, this aversion competes with appetitive motivation and can reduce cue attraction independently of incentive-salience mechanisms. In the context of ST/GT, such aversive motivational signaling would be expected to blunt ST, which depends on dopamine-mediated incentive salience, while more moderately affecting GT behaviors that rely on cognitive expectancy. Importantly, the aversive component of KOR activation may mask or override typical cue-driven approach responses. Because ST is particularly sensitive to dopamine suppression and negative affect, KOR-induced aversion could selectively diminish CS-directed approach, although this has yet to be empirically tested in ST/GT assays.

The Role of the Adenosine System in Sign and Goal Tracking

As a psychostimulant, caffeine modulates reinforcement sensitivity primarily through adenosine receptor antagonism, leading to downstream increases in dopaminergic neurotransmission (Hsu et al., 2010). These effects have promoted investigation into caffeine’s influence on ST and GT behavior given its potential to enhance motivational salience and cognitive performance. In one study, rats received caffeine or saline prior to each session of a PCA procedure in which a lever predicted food delivery (Holden et al., 2022). ST and GT were quantified via lever contacts, food magazine entries, and associated latencies and probabilities. Caffeine administration produced a significant, dose-dependent increase in ST (lever contacts, lever-approach probability, and PCA index) behavior without affecting GT. This pattern indicates that caffeine selectively enhances the motivational value attributed to reward-predictive cues, rather than the reward location itself. The authors suggest that this effect may be mediated by increased dopamine release in the NAc and elevated acetylcholine levels in the medial PFC.

Importantly, these effects are consistent with a broader literature on adenosine A₂A receptor antagonists, which can facilitate dopaminergic transmission in mesolimbic reward pathways (specifically at D2 receptors), thereby enhancing the incentive salience of conditioned cues (Prasad et al., 2021; Prediger et al., 2005). Together, these results suggest that caffeine may enhance sensitivity to conditioned cues without altering general reward approach, raising the possibility that common stimulants could exacerbate cue-triggered relapse risk in individuals predisposed to ST.

Discussion

The studies reviewed here demonstrate that ST and GT behavior in rodents are shaped by multiple neurotransmitter systems, each contributing distinct but highly interconnected pathways that converge on dopaminergic function. Although dopamine remains central to attributing incentive salience, its activity is modulated by cholinergic, serotonergic, adrenergic, opioid, and adenosine systems in ways that differentially bias cue-motivated behavior. Cholinergic manipulations often bias GTs toward ST, likely by disrupting prefrontal and striatal cholinergic inputs that normally regulate attention and thereby constrain dopamine-driven cue attraction. Serotonin generally reduces ST, and several studies suggest this occurs through serotonergic inhibition of mesolimbic dopamine neuron firing, though effects vary by strain and drug-specific pharmacology. Adrenergic manipulations primarily reduce ST while generally not affecting GT except at higher doses, consistent with noradrenergic modulation of dopamine release in the NAc and its role in amplifying cue-triggered motivation. This indicates a dependence on noradrenergic signaling for cue-directed behavior. Opioid receptor stimulation amplifies pre-existing phenotype tendencies reflecting μ- and κ-opioid modulation of dopamine tone that enhances incentive salience without necessarily shifting animals between ST and GT categories. Finally, nicotine and caffeine increase ST behavior in a mechanistic and dose-dependent manner, which aligns with their well-established capacity to potentiate dopaminergic responses to reward-predictive cues. This also underscores the potential for common stimulants to heighten vulnerability to cue-driven relapse. Notably, while some sex-differences have been reported, the majority of these studies were done solely in males, and thus the neuromodulatory mechanisms described here may not generalize to females, and sex-specific contributions of these neuromodulators may differ in important ways.

Across studies, these neuromodulators influence ST and GT behavior through partially distinct but highly interacting circuit nodes, particularly the NAc, PFC, and LC. Dopamine-dependent incentive salience attribution is centered in the NAc core (Cardinal et al., 2002; Chang et al., 2012; Gillis & Morrison, 2019; Bacharach et al., 2023; Iglesias et al., 2023; Flagel et al., 2011; Saunders & Robinson, 2012; Tomie et al., 2008), where cholinergic (Townsend et al., 2023), endocannabinoid (Bacharach et al., 2018; Oleson & Cheer, 2012; Bacharach et al., 2023), adenosingergic (Hsu et al., 2010; Prasad et al., 2021; Prediger et al., 2005) and opioid (Castro & Berridge, 2016) signaling modulate dopamine release dynamics to bias cue-directed behavior. Serotonergic and adrenergic systems additionally shape prefrontal functions, including attentional control and behavioral flexibility (Arnsten, 2020; Arnsten & Pliszka, 2011; Puig & Gulledge, 2011). Additionally, the LC provides noradrenergic input to both the PFC and NAc, regulating arousal-based gain and amplifying the motivational significance of cues (Aston-Jones et al., 1997). Thus, the effects of these neurotransmitter system manipulations on ST/GT phenotypes are best understood not in isolation, but as neuromodulatory influences on a broader LC–PFC–NAc circuit that determines cue reactivity, behavioral regulation, and susceptibility to ST behavior.

A clearer understanding of how these neuromodulatory systems bias individuals toward ST or GT also highlights several potential pharmacological targets relevant for reducing cue-driven relapse vulnerability. Cholinergic modulation of attentional control and top-down suppression suggests that α7 nicotinic and M1 muscarinic receptors may serve as targets for dampening maladaptive cue attraction (Gheidi et al., 2023). Serotonergic influences on behavioral inhibition point to 5-HT2A-C, 5-HT1B receptors, already implicated in impulsivity and drug seeking, as promising mechanisms for reducing ST tendencies (Bevilacqua et al., 2011; Higgins et al., 2011; Anastasio et al., 2011; Nautiyal et al., 2015). Noradrenergic regulation of cue-evoked arousal through α2A and β-adrenergic receptors may offer opportunities to modulate excessive cue reactivity and enhance cognitive control over drug-associated stimuli (Arnsten, 2020). Likewise, opioid receptor manipulations, particularly κ-opioid receptor antagonists, which reduce dysphoric and stress-enhanced cue responsivity, could mitigate vulnerabilities in individuals who strongly attribute incentive salience to cues (Gray et al., 1999; Carlezon et al., 2006). Finally, adenosinergic modulation of dopamine signaling through A1 and A2A receptors provides another route for altering motivational value assignment without directly targeting dopamine transmission (Quarta et al., 2004). Together, these neuromodulatory pathways expand the set of therapeutic leverage points beyond dopamine alone and underscore the practical value of ST/GT research for identifying treatments aimed at reducing cue-triggered relapse risk in substance use disorders.

Collectively, these findings reinforce that dopamine is not an isolated driver of ST and GT but instead, sits at the center of a broader neuromodulatory network. Disentangling how each system shapes dopaminergic firing, release, and receptor activation offers a more mechanistic understanding of why individuals differ in their susceptibility to cue-driven relapse. Future research should therefore move beyond single-system manipulations and examine how these neuromodulators interact under conditions of stress, genetic variation, and chronic drug exposure. Such work will be essential for developing individualized interventions that target system-level dysfunction rather than dopamine alone.

Data Availability Statement

The raw data is not available and was not used