Search

The overall goal of this thesis was to investigate how the ventral pallidum tracks experimentally-induced changes in the hedonic value of tastes. Taste reactions allows us to measure how ‘liked’ or ‘disliked’ a taste is by a rat. These reactions are not purely reflexive as they are altered when circumstances change: hunger can increase and satiety can decrease the tastiness of foods (Berridge 1996) and a salt appetite can make a normally ‘disliked’ salt solution ‘liked’ (Berridge et al. 1984, Schulkin 1991). Therefore taste reactions reflect the palatability of tastes. Understanding how the reward pathway tracks changes in the hedonic value of tastes may elucidate how neural coding is affected in disorders such as drug addiction and obesity. The first project investigated how neurons in the ventral pallidum (VP) track changes in hedonic value of a sweet taste by a conditioned taste aversion. When a taste changed from ‘liked’ to ‘disliked’, the predominant VP response also changed: taste responsive VP cells tastes typically increase firing rate in response to ‘liked’ tastes whereas the hedonically devalued taste typically triggered a decrease in firing rate. The second project compared the effects of dopamine and opioid modulation of ‘liked’, relatively neutral, and ‘disliked’ taste stimuli. We found that opioid modulation enhanced the hedonic value of quinine (which is normally ‘disliked’), but that only morphine resulted in a VP population response that mirrored the hedonic value of the tastes. The third project tested the effects of dopamine and opioid modulation on the hedonic value of a pair of cues (water infusions) and reward (a sucrose infusion). Overall, the cues did increase in hedonic value via association with reward, and the VP reflected the similarity in hedonic value between the cues and reward by responding similarly to cues and reward, under most drug conditions. Amphetamine decreased the proportion of units that responded to the first cue with an increase in firing rate, relative to vehicle control. In summary, the VP tracked changes in hedonic value under a variety of experimental manipulations, supporting evidence of its role in the neural coding of hedonia.

Deep brain stimulation (DBS) has been shown to be an effective treatment for Parkinson’s disease and other movement disorders including essential tremor and dystonia. Given its success, DBS is also being investigated as a potential treatment for psychiatric disorders including depression, obsessive compulsive disorder, eating disorders, and addiction. Although the specific therapeutic mechanisms of DBS are not known, studies suggest that this type of electrical stimulation may be causing an entrainment or regularization of firing patterns in neurons, interfering with the information being processed in the underlying neural circuit. The overall goal of this thesis is to evaluate the potential neural interference effects of DBS-like stimulation on motivation and reward consumption. Two structures were targeted: the ventral pallidum (VP), which is thought to be an area of convergence for processing reward and reward-related information, and the central nucleus of the amygdala (CeA), which is thought to be especially involved in focusing motivation for particular cues and rewards. The effects on DBS in the VP and CeA on reward-seeking behaviors, food consumption, and hedonic value of tastes were assessed. Results showed that DBS in the VP produced complex patterns of neuronal firing; however, it did not disrupt neural coding of reward and had only minimal effects on food consumption and motivation. DBS in the CeA also resulted in similar complex firing patterns, and additionally (in contrast to the VP) disrupted neural coding of reward. This disruption was reflected in altered behavior. DBS in the CeA invoked an immediate and profound decrease in the consumption of (p < 0.001) and motivation to work for (p < 0.001) sucrose pellets by more than a factor of four. DBS in the CeA also decreased the hedonic value of and increased aversive reactions to sucrose (p = 0.003). Overall results suggest the DBS is modulating neural activity in the underlying target structure, but the location of electrode is very important and this DBS-induced change in neuronal firing may or may not disrupt coding for reward. Data suggests that CeA may be an effective target for blocking food consumption and motivation.

Endogenous opioids play an important role in regulating reinstatement of extinguished drug seeking. Ten experiments investigated the neuroanatomical locus for opioid regulation of renewal, alcohol-primed reinstatement, and re-acquisition of extinguished beer seeking response. Rats were trained to nose poke for alcoholic beer in a particular context. This response was extinguished in either an alternate context or in the same context. Recovery of responding was produced in one of three ways. In renewal, rats were returned to the acquisition context. In primed reinstatement, rats were allowed to consume a small amount of beer at the commencement of test. In reacquisition, rats were allowed to respond once more for alcoholic beer. Within each of these experiments three different challenges to opioid neurotransmission were applied: a systemic injection of the general opioid receptor antagonist, naloxone, an infusion of the µ-opioid receptor antagonist CTAP into the accumbens shell, or a similar infusion into the ventral pallidum. Systemic naloxone attenuated renewal, primed reinstatement, and reacquisition. Intra-pallidal infusions of CTAP also attenuated all three forms of relapse; however intra-accumbal infusions reduced context-induced reinstatement only. In order to characterise the role of the ventral pallidum in reinstatement, active projections to this region during renewal were examined. The retrograde neuronal tracer Cholera Toxin-b subunit was injected into the pallidum of rats prior to training and testing for renewal. Retrograde-labelled neurons and c-Fos protein immunoreactivity within the accumbens shell and elsewhere in the brain were examined. Active projections specific to renewal were not found in the accumbens shell, however they were seen in the accumbens core, rostral basolateral amygdala, and the paraventricular thalamus. These findings are consistent with the hypothesis that opioid receptors mediate the hedonic impact of ethanol and ethanol associated cues. While µ-opioid receptors in the accumbens shell play a more specific role in reinstatement for context, the ventral pallidum may represent a final common locus for production of ethanol-seeking responding. The pathways recruited during context-induced reinstatement support and extend the current understanding the role of the ventral pallidum in reinstatement, showing that it is a site of convergence of striatal, thalamic and amygdala inputs.

There are three questions that I propose to answer in this dissertation in order to elucidate the role of the ventral pallidum in hedonic ‘liking’ processes, or pleasure generation. The first question is where in the ventral pallidum (VP) or adjacent region does neural dysfunction eliminate positive ‘liking’ and instead cause aversive ‘disliking’ and loss of reward even for sucrose taste. The goal is to identify the specific ‘disliking’ site in VP, where loss of function causes aversion to sweet reward. I found a 0.8 mm3 aversive site in the caudal VP where GABA inhibition or lesion suppressed hedonic ‘liking’ reactions to a sweet taste and replace them with aversive ‘disliking’ reactions (along with loss of food appetite). The second question is to compare and contrast the VP and the NAc shell (NAcSh) regarding whether their hedonic hotspots are necessary for normal positive ‘liking’ reactions. This question asks whether each hotspot is both necessary (for normal ‘liking’) and sufficient (able to enhance ‘liking’ above normal). I found that the caudal VP is the only site that is both necessary (for normal ‘liking’) and sufficient (able to enhance ‘liking’ above normal) for ‘liking’ of a sweet taste. In comparison, the NAcSh, though widely recognized for its role in reward, is only sufficient for ‘liking’ enhancement but not necessary for the maintenance of normal ‘liking’. The third goal is to identify other neurotransmitters in the VP that can amplify ‘liking’ reactions. Before this dissertation, only microinjections of DAMGO, u-opioid agonist, in the posterior VP cause increased of hedonic ‘liking’ reactions. I identify two other neurotransmitters that intensely enhanced ‘liking’: orexin and anandamide. Both the orexin hedonic hotspot and the anandamide hotspot are located in the posterior VP at a site that highly overlaps with the opioid hedonic hotspot. Overall, the findings of the dissertation suggest that VP is actively involved in pleasure generation in the brain, and the malfunction of VP leads to loss of pleasure. The results have broader implications on human reward processes and related clinical disorders, such as depression, addictions, and obesity.

A fundamental question in the neuroscience of reward is how the brain takes sensory input and adds hedonic or incentive value to it. My dissertation focuses on neural mechanisms for 'liking' and 'wanting' in two limbic brain structures, the ventral pallidum (VP) and nucleus accumbens (NAc). The VP is a chief output site for the NAc, and together they form a circuit linking limbic areas to cognitive and motor systems. My goal is to identify neurochemical, neuroanatomical and electrophysiological mechanisms for 'liking' and 'wanting' in VP and NAc, and in circuit features of how the two structures are functionally connected together. In my first project, I use a microinjection and immediate early gene (Fos) mapping technique to identify a hedonic hotspot in the caudal VP where mu opioids can function to elevate 'liking' for taste rewards and 'wanting' for food. The NAc has also been discovered recently to contain its own opioid hotspot for 'liking' and 'wanting'. Thus, in my second project, I seek to characterize if, and how, these two hotspots interact. I first observe that opioid stimulation in one hotspot recruits neurobiological Fos activity in the other, suggesting they communicate bidirectionally. To explore functional connections I next use simultaneous NAc and VP microinjections, and find that both hotpots work together as obligatory stages of a single circuit, rather than independently, to elevate reward 'liking'. However, I discover that the NAc can elevate eating independently of the VP, revealing overlapping but distinct circuitry for 'liking' versus 'wanting'. My third project asks the question of how neuronal activity within a hotspot might encode increases in 'liking' or 'wanting'. Using simultaneous NAc microinjection and VP electrophysiological recording, I find that VP neurons can encode elevations in 'liking' and 'wanting' that arise from opioid or dopamine stimulation in the NAc. These results reveal a causal role for the VP in increasing reward 'liking' and 'wanting', reveal a code for those functions within VP, and reveal how the VP functions within a larger reward circuit. The VP may therefore be an important target for future research on normal and pathological goal-directed behavior.

The ventral pallidum (VP) is a brain structure essential for reward learning and motivation. However, little is known about how VP neurons actually process reward information. In these studies, neural activity in the VP was recorded in freely moving rats as they learned a Pavlovian task in which specific stimuli predicted a reward. The first study demonstrated that VP neuronal firing encodes learning and reward by increasing the number of neurons that respond to reward-predicting stimuli, and by varying the speed at these neurons fire. In subsequent studies, I asked whether VP firing to stimuli during the task reflect purely predictive information or also motivational reward (incentive salience) properties of the stimuli. I induced physiological shifts in rats' motivational state after learning by sensitizing rats to amphetamine (which enhances incentive motivation for rewards) or depleting rats of sodium (which selectively enhances salt taste reward). Changes in VP firing following these motivational shifts allowed us to compare directly prediction models versus motivational incentive salience models. After amphetamine sensitization, VP neurons immediately increased their response to the predictive stimulus associated with a relevant reward, indicating coding of motivational incentive salience. Our novel computational Profile Analysis technique showed a shift in VP firing toward the stimulus with highest motivational incentive salience. These shifts in coding occurred without new learning about the predictive tones, again indicating motivational value influenced by physiological states. My last experiments showed that VP firing codes hedonic impact of rewards themselves, in addition to incentive motivational value of learned stimuli that predict reward. After physiological sodium depletion caused the taste of concentrated triple-seawater salt, which normally is 'disliked', to become strongly 'liked', VP neurons increased firing to the taste in a manner that coded hedonic reward value. My results demonstrate that VP neurons code the motivational value of cues and rewards, tracking that value through shifts in homeostasis, and extracting information about learned stimuli that reflects the current motivational value of the rewards they predict. These data emphasize the ability of VP neurons to multiplex information about reward prediction and motivation by integrating learned predictive associations and dynamic physiological states. My studies reveal important roles for VP neurons in coding normal reward learning and motivation, and suggest that VP dysfunction may be implicated in human motivational disorders including drug addiction and eating disorders.

The nucleus accumbens is a key limbic region implicated in mechanisms underlying reward and addiction. Evidence for involvement of the Nacc in responses to drugs of abuse (including cocaine and morphine) has partly come from studies using Fos as a marker for neuronal activation. While previous reports have consistently shown that Nacc neurons are activated following acute administration of cocaine, studies of morphine-induced activation have produced variable results. In the present study, we used two markers of neural activation, phosphorylation of the MAP-kinase, ERK, and expression of Fos, to investigate Nacc activation in response to either cocaine or morphine. In addition, we used a combination of retrograde tract tracing and immunocytochemistry to investigate whether activated Nacc neurons send projections to the ventral pallidum (VP), a major efferent target of the Nacc. Whereas cocaine induced neural activation at ten minutes, one hour, and two hours following drug administration, morphine only induced activation two hours following injection. Furthermore, while cocaine-activated neurons were present at all rostral-caudal Nacc levels, morphine-activated neurons were restricted to the rostral Nacc. Finally, cocaine induced activation of a subset of VP-projecting neurons in all rostral-caudal levels of the Nacc, while morphine did not. Thus, cocaine and morphine induce different temporal and regional patterns of activation in the Nacc, with cocaine producing more rapid and widespread activation than morphine. The only area of overlap in the patterns of activation produced by either drug was in the rostral Nacc, suggesting that this subdivision may be particularly important in reward and addiction.