Corticostriatal-hypothalamic circuitry and food motivation: integration of energy, reward, and action

Transcription

1 1 Corticostriatal-hypothalamic circuitry and food motivation: integration of energy, reward, and action Ann E. Kelley, Brian A. Baldo, Wayne E. Pratt and Matthew J. Will 1 Department of Psychiatry University of Wisconsin-Madison Medical School 6001 Research Park Blvd. Madison, WI This is a revised personal edition of a manuscript that has been published as noted below: Kelley, A. E., Baldo, B. A., Pratt, W. E., & Will, M. J. (2005) Corticostriatal-hypothalamic circuitry and food motivation: Integration of energy, reward, and action. Physiology and Behavior, 86(5), (doi: /j.physbeh ) This article may not exactly replicate the final version published in the journal. It is not the copy of record.

2 Abstract Striatal role in food motivation 2

3 3 1. Introduction One of the greatest threats to public health in the United States in the twenty-first century is obesity, which can be conceptualized as a disorder of appetitive motivation. The abundant availability of calorically dense foods such as fats and sweets in modern Western diets, as well as profound changes in the overall physical activity level of people compared with earlier times, is largely responsible for this epidemic [1, 2]. Neuroscience research can make critical contributions to the further understanding and treatment of this problem [3]. Traditionally, major focus has been directed to the hypothalamus, and rightly so given its crucial role in energy balance and food intake. However, much less is known about how the hypothalamus functions within its associated neural networks that integrate other factors involved in appetite, such as sensory factors, emotional processing, decision-making, and learning. These processes, in addition to the normal homeostatic mechanisms that drive the motivation for food, play a very significant role as determinants of when to eat or not eat. For example, emotional factors are clearly very important; ingestion of food provides a great deal of subjective pleasure, particularly if the food is rich in sugar or fat, and eating can be a source of comfort in depression or stressful states [4]. With regard to cognitive processing, sensory factors such as visual or olfactory stimuli can invoke food cravings even in the absence of energy deficit. Decision-making and inhibitory control are a key part of dieting and restraining from food intake. Our laboratory has been particularly interested in the role of neurotransmitter systems within the nucleus accumbens (Acb), a striatal region implicated in natural reward processes as well as addiction, in the control of food motivation and intake. Further, we are interested in the relationship that this region has with the hypothalamus, and in learning how integration between

4 4 energy balance sensing systems and reward or higher order motivational systems takes place at the neural level. In order to gain deeper knowledge about the interface between energy balance and cognitive control-emotional systems, we have begun to examine how cortical inputs from the amygdala and prefrontal cortex interact with striatal-hypothalamic circuitry. In this regard, the Acb is particularly well-positioned to serve as a node within circuits linking allocortical (amygdala, hippocampus) and neocortical feeding-related sensory processing to behavioral effector systems. Importantly, Acb efferent circuitry is organized to reach motor output systems in two distinct ways: by connections to pallidal structures that convey Acb processing to thalamocortical reentrant loops that eventually impinge on cortical (voluntary) motor output systems, and by direct projections to hypothalamic circuitry (arising exclusively within the Acb shell) that communicates to control centers for feeding-related motor patterns and autonomic arousal. A consideration of these two distinct output pathways provides a heuristic framework for interpreting some of the pharmacological and anatomical dissociations we have observed regarding the Acb control of feeding; this theme will be explored in the present review. First, however, it is useful to briefly review the specific pathways by which internal and external food- or appetite-related information gains access to the Acb, and how the Acb can, in turn, influence output effector pathways controlling feeding. The Acb receives brainstem information related to taste and visceral functions through a direct input from the nucleus of the solitary tract (NTS; to medial Acb shell), as well as an indirect input from gustatory cortex via parabrachial (PB) projections to gustatory (VPO) thalamus (to lateral shell and core) [5, 6]. Moreover, gustatory and viscerosensory regions within agranular insular cortex also project densely to the infralimbic and prelimbic regions of medial prefrontal cortex, which, in turn, project strongly to the Acb [7-10]. Taste and visceral information can also influence the via

5 5 two amygdala pathways: the NTS-PB-central nucleus of the amygdala-ventral tegmental area (VTA) connection, and the gustatory cortex-basolateral amygdala-accumbens pathway [11]. The central nucleus is particularly interesting as one of its major cortical inputs is from gustatory cortex [7, 12]. Pathways signaling internal homeostasis that eventually reach the Acb include projections from the lateral hypothalamus (LH, which has direct access to the arcuate nucleus, a critical command area for metabolic sensing) to the medial shell, either directly [13, 14] or through relays in midline thalamic nuclei [15-17]. With regard to behavioral effector routes, much of the output of the core reaches classic basal ganglia motor control circuits, while the shell s main effector systems appear to involve medial ventral pallidum and LH [18-20]. The downstream outputs from LH involve structures that directly control brainstem pattern generators for the motor actions of eating as well as autonomic structures [for general reviews, see 21, 22]. In summary, both the shell and core subregions of the Acb communicate extensively with circuitry that is well established to control taste perception, energy balance, and somatomotor effectors; these connections are summarized in Fig. 1. How, then, does intrinsic neurotransmitter signaling modulate this feeding-related information flow through the Acb to influence the manner in which behavior is organized in relation to ingestion? In this review, we describe work that has accrued in this laboratory regarding the roles of Acb GABAergic, dopaminergic and opioid peptide systems, and their interaction with other corticolimbic structures and the hypothalamus. Experiments have shown that these neurochemical systems play specific and dissociable roles in different aspects of food seeking, intake, and reward. Our work and that of other laboratories suggest four overarching hypotheses concerning these systems. First, stimulation of GABA A receptors or blockade of AMPA receptors in a circumscribed region of the (medial shell subregion) strongly

6 6 increases food intake in ad libitum-fed animals, via activation of the lateral hypothalamus as well as arcuate nucleus. We propose that an amino-acid coded subensemble of neurons in the shell communicates with energy balance-sensing systems within the hypothalamus (see Fig. 2), and reaches output pathways, via the lateral hypothalamus (LH) and central nucleus of amygdala, that primarily control feeding motor pattern generators and autonomic function. Second, opioid systems within the ventral striatum (and perhaps wider regions of striatum) recruit frontotemporal-lateral hypothalamic circuitry involved in affective regulation and palatability. Experiments indicate that opioid stimulation, in ad libitum-fed rats, results in marked and preferential increases in highly palatable food (fat, sucrose, salty solutions). We hypothesize that this response and in general, food intake evoked purely by palatability (i.e. not by energy deficit), engages basolateral amygdala/frontostriatal circuits, that exert control over the somatic voluntary motor systems. This system may be the substrate for affective and cognitive regulation of ingestive behavior. Third, recent work has indicated that ascending projections from the LH (perhaps orexin/hypocretin or MCH) may interact with midline thalamic-striatal pathways, that specifically engage striatal cholinergic interneurons, which in turn play a critical role in regulating striatal enkephalin gene expression [23]. Fourth, dopaminergic innervation of the striatum is also involved in food intake, but this system is concerned with motor activation and arousal, behavioral selection, and foraging strategies associated with changing motivational conditions, as well as initiating key intracellular plasticity mechanisms required for learning about food resources. In sum, these corticostriatalhypothalamic systems as a whole, via massive outputs to behavioral motor action systems, enable complex hierarchical control of adaptive ingestive behavior.

7 7 2. Control of feeding initiation by amino acid-coded Acb shell circuitry The control of feeding by GABAergic and glutamatergic substrates in the Acb shell was discovered quite serendipitously. In an experiment designed to explore the role of glutamatecoded inputs to Acb subregions in the control of spatial learning, the AMPA/kainate receptor antagonist DNQX was employed as a pharmacological tool to block glutamate transmission within discrete Acb territories. It was noted that animals receiving intra-acb shell DNQX ate voraciously when returned to their home cages after behavioral testing. Subsequent behavioral studies comfirmed that blocking AMPA/kainate (but not NMDA) glutamate receptor subtypes in the Acb shell produced a dramatic, short-latency feeding response [24, 25]. The magnitude of this effect was considerable; for example, intra-acb shell infusion of 750 ng of DNQX increased intake of a 10% sucrose solution by 65% [25], see Fig. 3B. A microinfusion mapping study revealed this to be a very anatomically circumscribed effect; as summarized in Fig. 3C, sites within the medial shell yielded the greatest response, whereas neighboring areas in Acb core, dorsal, or ventral striatum supported little or no effect [26]. Moreover, DNQX-associated behavioral changes were selective for food intake. As shown in Fig 3B, drinking was not increased, even in water-deprived rats; non-specific gnawing on wood chips was similarly unaffected [25]. Conversely, it was found that stimulating Acb shell glutamate receptors with local infusions of AMPA significantly decreased sucrose intake in food-deprived rats [25]. Taken together, these data were interpreted as reflecting an important and specific role of glutamate inputs to the medial Acb shell in negatively modulating feeding behavior. It should be noted that these experiments comprised one of the first demonstrations of dissociable behavioral functions of the Acb shell versus core.

8 8 Based on these findings, it was hypothesized that diminished glutamate transmission in the Acb shell, with a corresponding reduction in the activity of medium-spiny GABAergic Acb output neurons, was a critical neural event controlling feeding initiation. If this was the case, we reasoned that other manipulations producing an overall decrease in Acb neuronal activity would also augment feeding. Accordingly, it was found that excitotoxic lesions of the Acb shell, but not the core, resulted in a prolonged increase in the rate of weight gain [27]. Moreover, inhibiting local neuronal activity via GABA A receptor stimulation in the Acb shell (using the selective agonist, muscimol) produced a marked hyperphagia resembling the effect seen with intra-acb shell AMPA receptor blockade [28], see Fig. 3D. Intra-Acb shell muscimol-induced hyperphagia was even more extreme than that produced by DNQX; rats that were not food deprived were often observed to eat in excess of 8 g of rat chow in 30 min. In some instances, animals were observed to eat so voraciously that they could not swallow fast enough to keep up with their food intake. As with the DNQX effect, muscimol-induced hyperphagia was elicited most potently by infusions into the medial Acb shell; neighboring sites in the anterior Acb, where the core and shell are poorly differentiated, or posterior sites where the Acb shell merges into the bed nucleus of the stria terminalis, yielded little effect [28]. Also, similar to the experiments using DNQX, intra-acb shell muscimol infusions appeared to influence food intake specifically. As shown in Fig. 3F, ingestion of chow, sucrose, or high-carbohydrate or high-fat formulated diets was dramatically elevated, while intake of water or palatable but non-caloric saccharin or saline solutions was unaffected [29]. Additional experiments using a variety of GABA-selective agonists demonstrated that intra-acb shell stimulation of GABA B receptors using baclofen, or elevation of endogenous GABA levels using the GABA transaminase inhibitor, gamma-vinyl-gaba, also produced marked feeding effects [28]. Taken together,

9 9 these data supported the idea that functional inhibition of Acb shell medium spiny output neurons, whether by blockade of glutamatergic inputs (achieved either by postsynaptic AMPA receptor blockade or by stimulation of GABA B receptors, known to be localized on glutamate nerve terminals), or by direct stimulation of GABA A receptors (located on the medium-spiny neurons themselves), was sufficient to produce intense, behaviorally specific hyperphagia in adlibitum fed animals. The question then arose as to the circuitry mediating these effects. It was noted that, at least with regard to the magnitude and immediacy of the effect, hyperphagia induced by functional inhibition of the Acb shell strongly resembled stimulus-bound feeding seen in studies of lateral hypothalamic (LH) electrical stimulation [30] or in experiments using local infusions of glutamate receptor agonists into the LH [31]. Because the Acb shell represents the only striatal region sending direct projections to the LH [18, 32], it was hypothesized that inhibiting these outputs, which putatively contain GABA as a transmitter, could result in the disinhibition of feeding-related LH circuitry. Accordingly, inhibiting the LH with local infusions of muscimol was found to dose-dependently reverse the hyperphagia associated with intra-acb shell AMPA receptor blockade [33], see Fig. 4A. Also, as displayed in Fig. 4C & D, intra-acb shell infusions of muscimol induced a dramatic upregulation of Fos expression throughout the LH, including within neurons containing the feeding- and arousal-related peptide, orexin/hypocretin [34, 35]. It was also observed that intra-ach shell muscimol increased Fos expression in the arcuate nucleus, specifically within neurons containing the orexigenic peptide neuropeptide Y, while decreasing Fos expression in neurons containing the anorexic peptides CART and POMC [35, 36]. These effects were tentatively interpreted as reflecting the removal of an Acb-mediated tonic inhibitory influence upon specific peptide-coded hypothalamic feeding circuits.

10 10 Nevertheless, blockade of or receptors in the LH does not elicit feeding, suggesting that putative tonic inhibition exerted by Acb shell efferents occurs at a site interposed between the Acb shell and LH [28]. Such a relay could occur in the ventral pallidum. Blockade of receptors in the medial ventral pallidum, which receives direct GABA-ergic projections from the medial Acb shell and sends projections to the LH, produces hyperphagia on the order of that seen after intra-acb shell AMPA receptor blockade [37]. In addition, as shown in Fig. 4B, intra-acb shell muscimol-induced hyperphagia was blocked by LH of the NMDA receptor antagonist, AP5, suggesting the presence of an intervening glutamatergic synapse between the Acb shell projections and feeding-related LH neurons [34]. It could be, for example, that GABA-ergic afferents from the Acb shell or ventral pallidum synapse upon local intrahypothalamic glutamatergic circuits in the LH. Regardless of the precise route of control, however, these studies clearly establish that increased activity of LH substrates is necessary for the expression of Acb shell-mediated hyperphagia. We then sought to determine the behavioral mechanisms by which a putative Acb shell- LH circuit could mediate feeding initiation. As discussed above, the behavioral effects of intra- Acb shell AMPA receptor blockade or GABA receptor stimulation were found to be specific for food intake, raising the possibility that these manipulations elicit a state resembling a feeding central motivational state (CMS). We reasoned that if this was the case, it should be possible to mimic other behavioral aspects of the feeding CMS beyond a simple increase in food intake, such as the enhancement of the motivation to work for food reward, or to learn a new behavioral response to acquire food. As will be discussed in later sections, stimulating opioid or dopamine systems in the Acb augments responding on a progressive ratio schedule for sucrose pellets, an operant task used to assay the incentive-motivational properties of food [38]. It was expected

11 11 that intra-acb shell GABA receptor agonism would also enhance progressive-ratio responding, particularly considering the voracious nature of the feeding response seen with this manipulation. It was found, however, that intra-acb shell receptor stimulation (using the direct receptor agonist, muscimol) did not alter progressive ratio responding in rats trained on this task [38], see Fig. 3E. We then asked whether this neural manipulation would be sufficient to produce a state that supported food-reinforced operant learning in ad libitum-fed rats. Similar to the progressive ratio study, intra Acb shell muscimol failed to enable the acquisition of a foodreinforced operant response [39]. In a control experiment, it was found that the same dose of muscimol, injected at the same coordinates, increased intake of sugar pellets freely available in the operant chamber hoppers by 400% above vehicle control levels. Thus, GABA-mediated inhibition of the Acb shell produced neither an enhancement of the willingness to work for food reward (the progressive ratio experiment), nor a facilitation of the learning of a new response to obtain food, despite the fact that this manipulation increases the intake of freely available food far beyond the levels presumably dictated by homeostatic mechanisms in an ad libitium-fed animal. How then to characterize the motivational state elicited by the inhibition of neural output from the Acb shell? It is important to note, first, that this dissociation between effects on freefeeding versus working or learning to obtain food does not appear to be the consequence of a performance deficit. For example, in the progressive ratio study, intra-acb shell muscimol failed to enhance responding, but responding was not diminished and the discrimination between the correct and incorrect levers was maintained [38]. These results, taken together with the lack of effects on water intake [25], would also tend to argue against the possibility that GABAmediated inhibition of the Acb shell produces a non-specific augmentation of the incentive

12 12 salience of stimuli encountered in the environment. It should be noted that muscimol infusions into rostral Acb shell sites produces positive place conditioning, indicating that at these sites, inhibiting Acb shell output may produce general positive motivational effects beyond a restricted augmentation of food intake [40]. It is not understood, however, how these place conditioning effects relate to Acb shell muscimol-induced hyperphagia, as muscimol infusions into a subset of more posterior Acb shell sites yield strong feeding effects but are associated with place aversion and aversive-like taste reactivity responses to sucrose [40]. Thus, it is clear that, at least within certain Acb shell zones, there is a marked dissociation between muscimol-induced enhancement of feeding per se and other behavioral effects (augmented operant responding, enhanced learning of food-reinforced responses, positive place conditioning and affective responses to sucrose) typically associated with an appetitive CMS. These dissociations are quite interesting from the perspective of motivational theory. It is generally thought that organisms learn a variety of response strategies to obtain incentives in the environment, and that these varied behavioral strategies are utilized in a context-appropriate fashion as dictated by environmental contingencies and internal homeostatic states. In other words, numerous distinct goal-directed behavioral strategies become a part of an organism s overall behavioral repertoire; these response outputs are all available to be energized by the CMS, but are selected by the properties of the particular environmental circumstances at hand [for an excellent review, see 41]. In incentive-motivation models, this has been termed environmentally organized motor output [42]. It is curious, then, that inhibiting Acb shell neural output appears to selectively augment certain types of goal-directed behavior (ambulating to a food pellet and eating it) but not others (pressing a lever to obtain a food pellet). Also, strikingly, it would appear that in certain restricted zones of the Acb shell, food intake is

13 13 increased even though a taste reactivity test indicates that the food is disliked [40]. Moreover, intra-acb shell muscimol increases food intake regardless of macronutrient composition (Fig. 3F), and does not augment the intake of palatable but non-caloric tasteants. Therefore, it would seem that inhibiting the Acb shell produces a state favoring dramatic and selective increases in food-related ingestive behavior. However, this state is not manifested in observable behavioral changes under circumstances signaling the need for more complex food-seeking strategies (e.g., lever pressing), and moreover is apparently somewhat insensitive to the sensory/hedonic experience derived from eating the food. To account for this unusual motivational situation, we have made reference to a recent anatomical model proposed by Swanson, which holds that forebrain control over the initiation of motivated behaviors involves projections to a series of nuclei in the hypothalamus and midbrain, termed the behavioral control column [40]. The nuclei comprising this control column are proposed to represent integrative nodes within circuits that govern hindbrain and spinal cord motor pattern generators that produce the behavioral elements or fragments of specific classes of motivated behavior (i.e., eating, drinking, defensive reactions, etc.). We hypothesize that the medial Acb shell, with its idiosyncratic projections to the LH, has unique access among basal ganglia sites to feeding-selective nodes of the behavioral control column, and exerts negative modulation over these hypothalamic substrates. When this inhibition is artificially released by pharmacological inactivation of Ach shell output, the activation threshold for these feedingspecific nodes is decreased, biasing the animal towards consummatory behaviors directed at food even in an ad libitum-fed state. In agreement with this idea, a recent electrophysiological study showed that the onset of feeding consummatory responses is accompanied by the inhibition of a subpopulation of Acb neuronal units, providing general support for the idea that feeding onset

14 14 could be under negative modulatory control by Acb output [43]. Inherent in our hypothesis is the idea that removal of inhibitory input from these feeding-specific motor pattern controllers recapitulates or mimics only those behavioral characteristics of the feeding CMS mediated by the control column, i.e., responses most closely linked to procuring and ingesting food. This fragmented CMS would exclude more complex strategies and/or computations, such as operant responding to acquire food, which are represented in higher-order corticolimbic networks and therefore bypassed by the artificial, experimentally induced disinhibition of downstream feedingspecific motor controllers. This working hypothesis could account for the findings described above, including the observation that intra-acb shell muscimol fails to promote operant responding to obtain sugar pellets while dramatically increasing the intake of the same pellets when freely available in the food hoppers. What would be the physiological relevance and evolutionary benefit of such a mechanism? We have considered the possibility that the Acb shell acts as a sensory sentinel of sorts, performing a gating function over the feeding nodes of the control column (or the subcircuits in the LH that specifically reach the feeding control column). Thus, even under conditions such as hunger or starvation when the drive to feed is very strong, the animal must retain the ability to stop and switch away from feeding when necessitated by sudden threats in the environment. This could be achieved through a phasic glutamate signal reaching the Acb from corticolimbic sites such as the medial prefrontal cortex, hippocampus, or amygdala, which would immediately override the ongoing consummatory act and enable switching to other behavioral repertoires (e.g., escape or freezing). It is also likely, however, that glutamate signals to the Acb convey information that is important for the initiation of feeding. To resolve this discrepancy, we speculate that glutamatergic inputs related to feeding initiation reach broad areas

15 15 of the Acb and striatum, while feeding-interruption signals are more localized to the Acb shell. Thus, strong feeding drives may be accompanied by a widespread activation of medium-spiny neurons, which may initiate the development of a local intra-acb GABAergic signal that inhibits Acb shell output and serves to limit or gate the phasic control of incoming feeding-interruption signals. The overall effect of such a GABAergic mechanism would be to enable feeding to occur under a wider range of environmental situations, such as those containing somewhat unfamiliar or ambiguous cues, while the ability to rapidly interrupt feeding in response to a strong environmental stimulus, such as a clear, imminent threat would be preserved. This model would account for increases in feeding induced by either Acb shell glutamate receptor antagonism or GABA receptor agonism. An implication of this model is that it should be possible to enhance feeding via inactivation of the corticolimbic sites that convey putative glutamatergic feeding-interruption signals to the Acb shell. Conversely, the model implies that the influence of these glutamatergic signals could be overcome or short circuited by directly inhibiting the Acb shell via GABA receptor activation. In support of the first point, we have found that temporary inactivation of ventral medial prefrontal cortex, a region that projects strongly to the Acb shell, shifts the microstructure of feeding behavior toward longer feeding bouts while slightly reducing exploratory-like motor responses such as rearing [44]. This change in behavioral patterning is similar to what one would predict if there were a diminution in the ability of environmental stimuli to temporarily pull the animal away from feeding to engage in local exploration, foraging, or scanning the environment; however, other explanations, such as alterations in the hedonic impact of the consummatory act, are certainly possible and have yet to be tested.

16 16 With regard to the second point, we have observed that temporary inactivation of the basolateral amygdaloid complex, which decreases the increase in fat intake associated with opiate receptor agonism in the Acb [45], is completely ineffective in altering the hyperphagia associated with intra-acb shell GABA A receptor stimulation [46]. This dissociation has interesting implications for understanding the distinct ways in which Acb GABA and opioid systems regulate feeding-related information flow through the Acb. As will be discussed extensively in the next section, there is considerable evidence that Acb opioid transmisson enhances the positive hedonic aspects of palatable feeding, in contrast to the indiscriminate hyperphagia associated with intra-acb shell muscimol. Thus, the failure of basolateral amygdala inactivation to decrease muscimol-induced hyperphagia suggests that this feeding effect proceeds independently of information, arising in the basolateral amygdala, which is critical for Acbmediated processing of taste hedonics. These observations are consistent with the idea that artificially enhancing GABA tone specifically within the Acb shell bypasses the control over feeding behavior exerted by brain regions sending highly processed, glutamate-coded sensory and motivational information to the ventral striatum. 3. Opioids and food intake We now turn our attention to the evidence suggesting an important and specific role of striatal opioid peptides in governing the hedonic aspects of ingestion. The striatum contains the opioid peptides enkephalin and -endorphin [47, 48] and is rich in opioid receptors [49]. There has been a long history with regard to opiate administration and food intake; prior to the discovery of endogenous opioids, it was well known that systemic opiate agonists increased food intake while antagonists decreased feeding [50, 51]. The precise neural mechanisms of this

17 17 effect remain unknown, but current data has led to the dominant hypothesis that opioids specifically regulate palatability or hedonic evaluation of foods [52-56]. Simply stated, it is argued that opioid receptor agonism generally enhances food intake by increasing the positive hedonic valence of food, while opioid receptor antagonism reduces or blocks this affective response. Such a position gains support from human studies as well as the animal literature. Specifically, systemic infusions of the opioid antagonist naltrexone in humans reduces reported affective or pleasantness ratings of sweet and fatty foods while not affecting subjective reports of hunger or the ability to detect sweet or salty tastes [52, 57-60]. Similar data has been reported in rats from a variety of paradigms. In early reports, systemic naloxone administration was shown to abolish the preference for sweet solutions versus water in rats [61], and naloxone did not alter baseline caloric intake but selectively reduced the hyperphagia that results from presentation of a highly palatable diet [62]. Morphine has been shown to increase rat s preferences for sweetness, while naltrexone reduces sucrose preference as measured on preference-aversion curves [54, 63, 64]. Further validation has been reported utilizing taste reactivity measures in rats; morphine administration increases positive affective responses to intraoral administration of sucrose [65]. Opioid receptors are found throughout the neural pathways thought to relay taste information throughout the brain [49, 66-68], including the nucleus of the solitary tract, the pontine parabrachial nucleus, and the gustatory nuclei of the amygdala (including the central nucleus). They are furthermore present within the hypothalamus and the ventral tegmentum. While our research has focused on the role of opioids within the striatum, it should be noted that a substantial literature has accumulated that suggests an important role for opioid receptors in all of the aformentioned regions in food intake [69-72]. Furthermore, data from our laboratory (see below) and others [73, 74] suggest that interactions between these regions likely mediate food

18 18 intake and palatability. Thus, a distributed system of interconnected structures can be argued to be involved in an opioid-modulated axis of food motivation [56, 68, 75]. Our initial examinations of the effects of opioid manipulations of the striatum on food intake followed up on previous work in the literature showing that opioid receptor stimulation of the Acb increased food intake [76-78]. At that time, distinct lines of evidence were converging on the fact that the striatum was not a homogenous structure. Anatomical inquiry showed that connectivity of different striatal regions with regard to its afferent and efferent throughputs was not identical across its dorsal-ventral or medial-lateral extent [10, 79, 80]. It was suggested that the striatum contained a number of parallel circuits that might be involved in different cognitive and motor functions. Our microinfusion mapping experiments with GABA A agonists of the ventral striatum were consistent with such postulation [29], and indeed, additional behavioral research from many laboratories since that time has served to strengthen this hypothesis. With this framework in mind, the potential regional specificity of opioid receptor stimulation of the striatum on food intake in rats fed ad libitum was examined. In our initial report [81], it was demonstrated that intracranial morphine administration at high doses (up to 20 g/side) into five discrete regions of striatum (the Acb, ventromedial striatum, ventrolateral striatum, anterior dorsal striatum, and posterior dorsal striatum) resulted in significant increases in food intake (standard rat chow). The effect was most striking in the medial and ventral aspects of striatum. Lower doses of morphine (0.5 and 1.0 g bilaterally) that elicited robust and significant increases in these two regions did not cause similar increases when injected into ventrolateral or posterior dorsal regions of striatum. Thus, although morphine induced increases in food intake when administered throughout the striatum, the Acb and adjacent striatum were most sensitive to this effect.

19 19 Opioid mu, delta, and kappa receptors (and their mrna), are found throughout striatum, distributed in patches in the dorsal regions and more uniformly in the ventral striatum [49, 82]. Particularly dense levels of these receptors are observed in the Acb shell and adjoining ventrolateral caudate (of note, these are the regions of striatum that receive gustory cortical input). Furthermore, preproenkephalin (PPE) and preprodynorphin (PPE) mrna has been shown to be expressed in separate output neurons of the striatum (e.g., [83]). In experiments testing the specificity of each receptor subtype on the opioid-mediated feeding response, mu receptor stimulation (with DAMGO) resulted in the most robust increase of food intake (with milder increases observed with delta receptor agonism), while kappa receptor activation did not alter feeding [84]. Once again, the effect of treatment was most robust in the Acb, although significant stimulation of feeding was seen in the other striatal areas, particularly in the case of mu receptor stimulation. It was subsequently shown that mu receptor antagonism decreased chow intake [85]. These data corroborated a growing body of literature that suggested that activation of mu and delta receptors was more effective than kappa receptor agonism in enhancing food intake and in other reward paradigms [53, 86-88]. As mentioned previously, systemic pharmacological manipulations of opioids are thought to affect the intake of highly palatable foods. Moreover, it is believed that opioids do not impact the perception or recognition of taste per se, but rather the pleasure derived from tastes [59, 89, 90]. In particular, the literature suggests that fat is particularly influenced by opioid manipulation. Therefore, it was of interest to determine the nature of these effects with regards to palatable solutions and diets that varied in their macronutrient content. Consistent with the data reported with rat chow, stimulation of mu or delta receptors (with DAMGO or DPEN, respectively) of the Acb resulted in an increase in the consumption of 5% sucrose solution, in

20 20 animals fed ad libitum [91]. Neither kappa receptor stimulation with U50488H nor dynorphin affected intake measures. Furthermore, mu receptor agonism did not affect water intake. We have consistently failed to observe an increase in water consumption following Acb mu receptor activation, even in mildy water-deprived animals [91, 92], suggesting that the increased food intake is not merely due to general arousal- or activation-induced effects. The same anterior and ventral regions observed (above) to increase normal diet also cause robust increases in the intake of high fat diets [93]. Furthermore, in additional tests, Acb mu receptor stimulation preferentially stimulated the consumption of high fat diets [94]. Specifically, although DAMGO infusions resulted in increased intake of both fat- and carbohydrate-rich diets when either was presented alone, if rats were given a choice between both diets, drug treatment strongly increased their intake of the fat diet without affecting carbohydrate consumption. This change in fat consumption was completely abolished by systemic naltrexone administration. Notably, rats that have been food deprived for 24 hours were shown to increase their intake of both diets to an equal degree. However, deprived animals with intra-acb DAMGO treatment preferentially increased their fat intake relative to the highcarbohydrate diet (see Figure 5A). Intra-Acb injections of naltrexone also resulted in preferential reduction of the intake of high fat diet under food deprivation conditions (Figure 5B). The fact that rats tend to consume more calories of a high-fat diet under opioid influence may suggest that the diet is inherently more palatable, and that the effects of mu-receptor stimulation may be due to an increase in the pleasurable aspects of the food. However, it should be noted that all of the diets/solutions discussed thus far contained caloric content. It could therefore be argued that the treatment effect resulted from the rats shifting their intake preferences to diets highest in caloric

21 21 value, perhaps in an attempt to mediate some pharmacologically induced hunger state similar to but distinct from the fragmented CMS elicited by Acb shell inhibition (see previous section). In order to address whether striatal opioid manipulations specifically impacted hedonic aspects of feeding versus promoting caloric intake, we tested the effects of intra-acb DAMGO infusions upon the intake of palatable, but non-caloric, saccharin and salt solutions. Mu receptor activation resulted in a significant increase of the intake of both solutions, and left water intake unaltered [92] (Figure 5D). This result contrasts with the effects of Acb shell muscimol infusions (reviewed above), which did not increase saccharin intake. Thus, the CMS elicited by opioid receptor stimulation appears to be distinct from that produced by muscimol inactivation of the Acb shell, in that opioid-stimulation increases intake of palatable solutions that lack nutritive content. Additionally, mu receptor agonism within the Acb increases progressive ratio performance measures, which also contrasts with the effects seen under GABA A agonism [38] (figure 5C). However, similar to the effects seen with muscimol infusions into the Acb shell, it should be noted that mu receptor stimulation does not support learning of an instrumental response in rats fed ad libitum [39]. Moreover, Acb opioid agonism does not increase lever pressing for a stimulus previously associated with food delivery [95]. These data, consistent with other reports from the literature, suggest that the increases in intake following striatal opioid stimulation is the result of affecting the palatable and rewarding nature of preferred foods, a process which does not necessarily depend upon caloric content. However, stimulating this system in isolation does not produce a CMS capable of supporting new motor learning. Our next line of inquiry focused on identifying specific inputs and outputs to the striatum subserving intra-acb DAMGO-mediated augmentation of fat intake. As noted above, opioid receptor manipulation of several gustatory and hypothalamic regions have been shown to modify

22 22 food intake. There is considerable evidence that intercommunication between multiple brain regions is important for integrating need-based food intake, as well as for increases in consumption that is related to the palatable nature of foods (see above). In our initial attempts to determine what neural circuitry was involved in striatal DAMGO-elicited increase in food intake, we examined neural activation following Acb drug treatment. Rats were injected with either saline or DAMGO, and then were allowed to eat precisely 3 grams of the high-fat diet. The animals were sacrificed and their brains examined for immunocytochemical detection of cells containing Fos protein. Multiple brain regions increased the number cells expressing Foslike immunoreactivity following drug treatment. Notably, Fos-related increases were seen in lateral and dorsomedial hypothalamic areas (Figure 5E), in the midbrain (including the VTA), and in the nucleus of the solitary tract [93]. In short, mu receptor stimulation of the Acb appeared to recruit downstream regions that integrate motivational, metabolic, and autonomic aspects of ingestive behavior. We then examined the importance of many of these sites on DAMGO-induced feeding by selectively targeting individual regions and examining the effect of inactivating them (with the GABA A agonist muscimol) prior to intra-acb injections of DAMGO [75]. Inhibiting the lateral hypothalamus, dorsomedial hypothalamus, the ventral tegmental area, and the nucleus of the solitary tract all significantly reduced fat intake elicited by intra-acb mu receptor stimulation (dorsal hippocampal muscimol inactivation was without effect). In a more recent study, the effects of inactivation of the basolateral and central nuclei of the amygdala were both shown to decrease fat intake following DAMGO treatment of the Acb [45]. Notably, basolateral amygdala inactivation did not alter basal fat intake; infusions of muscimol into the central amygdala reduced both basal and DAMGO-elicited fat intake. Although much remains to be done with

23 23 regard to tracing the neural networks involved in the feeding response that ensues following mureceptor activation of the ventral striatum, these most recent data suggest that striatal opioid receptor activation has effects at multiple levels of feeding pathways, including regions implicated in relaying gustatory inputs (nucleus of the solitary tract), evaluating the rewarding value of the food (amygdala, ventral tegmentum), and modulating food intake relative to energy needs (hypothalamus). These data lend support to the theory that ventral striatal opioids contribute to the integration of reward-related aspects of feeding with those influencing energy balance. 4. A potential acetylcholine-opioid link in the modulation of food intake. A significant amount of research has gone into examining the striatal milieu with regards to the influences of individual neurotransmitters and their effects upon other chemical signals, as well as the firing patterns of individual neurons within the striatum [e.g., 96-99]. Given the interdependent nature of these signaling pathways, it is surprising that we see such diverse changes in food intake following different pharmacological interventions within the same region. For example, the voracious feeding seen following GABAergic receptor stimulation of the Acb shell is qualitatively different from the changes elicited by opioid receptor stimulation, in that the former elicits intake of any nutritive food source, while the latter preferentially promotes intake of palatable substances, even if they are non-caloric (i.e., saccharin, salt). Furthermore, as will be discussed below, dopaminergic signaling appears to modulate feeding only in relation to motivational arousal and incentive salience of food stimuli. Thus, despite the fact that these various manipulations have impacted the same anatomical locus (the ventral striatum), different

24 24 pharmacological treatments appear to affect different aspects of striatally-based feeding behaviors in fundamental ways. Although it is clear that the effects of mu receptor stimulation of the striatum is dependent upon numerous extra-striatal structures (see above section), little is known about how the effects are mediated intra-striatally. Opioid receptors within the striatum have their effects upon striatal interneurons and output neurons, and receptors are found upon axonal processes within the striatum as well [100, 101]. It remains a puzzle as to what precise mechanism opioid stimulation may use to promote palatable feeding, and how its influence may in turn be modulated by the effects of other intrastriatal signaling transmitters. As a means to begin to address the latter question, we have recently been examining the effect of inhibitory pretreatment of other neurotransmitter systems in the Acb upon DAMGO-induced feeding [102]. Toward this end, rats received various intra- Acb pretreatments prior to infusions of 0.25 g/side of DAMGO. They were then allowed 2-hour access to a high-fat diet. The individual pretreatments (tested on separate groups of animals) consisted of antagonists for opioid receptors (naltrexone), glutamate AMPA receptors (LY293558), dopamine D1 or D2 receptors (with SCH23390, or raclopride, respectively), or cholinergic muscarinic or nicotinic receptors (with scopolamine or mecamylamine, respectively). As would be expected, intra-acb naltrexone completely blocked DAMGO s effect upon fat intake, without changing baseline feeding. In contrast, inhibition of AMPA receptors (at doses that did not stimulate feeding), dopaminergic D1 or D2 receptors, or nicotinic receptors had no effect on either baseline intake or DAMGOelicited feeding. However, blockade of acetylcholine muscarinic receptors reduced fat intake, in rats treated with DAMGO or saline. Thus, although the feeding effect elicited by intrastriatal mu

25 25 receptor activation is resistant to the inactivation of AMPA, dopaminergic, and nicotinic receptor activity, both it and baseline feeding is reduced by acetylcholine muscarinic receptor blockade. Unlike the cortex, which receives acetycholine input predominantly from projection neurons of the basal forebrain, the cholinergic signal within the Acb is derived from its intrinsic pool of interneurons. Although comprising only 1-2% of all striatal neurons, the large aspiny cholinergic interneuron extend its dendritic and axonal processes up to a millimeter 3 of tissue, allowing each to integrate input from and influence a broad expanse of the striatal milieu [103]. They modulate striatal function via their effects on muscarinic receptors, located on the dendritic shafts and soma of medium spiny neurons, as well as incoming dendrites from cortical and midbrain efferent regions [104, 105]. Cholinergic output from the interneurons also stimulates nicotinic receptors, which are predominantly found on incoming axons from extrastriatal sites [103]. The neurons display firing rate changes in relation to salient events, including that of unexpected food reward [ ]. Furthermore, as with dopaminergic neurons of the midbrain [110], over the course of learning about stimuli consistently paired with reward, these neurons shift their responsiveness from an unexpected reward to stimuli that come to predict it. Consistent with the plastic nature of their firing pattern, lesions of acetylcholine interneurons impair reward-related learning [111]. Intrastriatal scopolamine impairs instrumental learning as well as reversal learning in the t-maze [ ]. Furthermore, striatal acetylcholine levels increase as animals reverse place learning [116], as well as during spontaneous shifts from spatial- to response-guided strategies while solving a t-maze task [117]. Compared to the work on learning-related effects of intrastriatal Acb manipulations, to date there has been relatively little research on the potential role of striatal acetylcholine on feeding behavior, with some notable exceptions. Research by Hoebel and his colleagues has

26 26 implicated Acb acetylcholine output in mediating satiety mechanisms. Using microdialysis techniques, it has been shown that acetylcholine output increases over the duration of feeding [118]. The authors suggest that this rise in acetylcholine assists in inhibiting feeding, because manipulations of hypothalamus that induce food intake lowers Acb acetylcholine, and pharmacological manipulations that reduce feeding result in increased Acb cholinergic output [ ]. Additionally, Hajnal et al. [121] have shown that lesions of the cholinergic interneurons of the striatum result in decreased body weight and alterations in food intake, in animals fed ad libitum as well as in response to a 24 hour food deprivation challenge. Such data provide intriguing insight into the possible role of striatal acetylcholine in modulating feeding behaviors, in addition to its proven role in learning paradigms. Given the presumed role for cholinergic signaling in mediating satiety mechanism proposed by Hoebel and colleagues, it may seem paradoxical that blocking muscarinic receptors within the Acb results in decreased food intake. However, the experiment testing the effect of scopolamine on DAMGO-induced fat intake (reviewed above) is the second experiment in which we have seen a reduction in short-term intake of food following muscarinic receptor blockade of the Acb. Previously, we have reported that inhibition of Acb muscarinic, but not nicotinic, receptors blocks learning and performance of an instrumental response for food (in this case lever pressing). We also found that drug-treated rats were as fast as controls in performing the behaviors required in the instrumental chamber, but that they ate less sucrose when it was made freely available for 15 minutes and that progressive ratio performance was reduced under the influence of Acb scopolamine [112]. These data suggest to us that muscarinic receptor blockade reduced the rewarding value of the reinforcer, in this case sucrose, rather than having a specific effect upon learning per se. Considering our data together with that of Hoebel and colleagues, it

27 27 seems likely that in addition to whatever effects Acb acetylcholine may play on mediating satiety, normal cholinergic tone serves additional roles in food seeking and/or consumption. Thus, converging evidence suggests that striatal acetylcholine is, at the very least, involved in striatal mechanisms underlying motor learning, and may also serve an important role in striatal mechanisms of food intake. We also noticed, serendipitously, that in the instrumental learning study the animals that had undergone the scopolamine treatment lost weight relative to the control animals over the course of five treatments. Animals treated with the nicotinic receptor antagonist mecamylamine showed no such effect. This led us to examine the effect of intrastriatal scopolamine upon 24-hour food intake in animals fed ad libitum. Following bilateral infusions of 5 or 10 g scopolamine into the Acb or anterior dorsal striatum, rats reduced their subsequent daily intake of chow by as much as 50% [122] (Figure 6). We observed an initial locomotor increase immediately following the drug infusion that lasted approximately 30 minutes; otherwise, no gross locomotor deficits were observed. Animals recovered their normal feeding in the following 24-hour period. Clearly, temporary disruption of local striatal muscarinic signaling receptor results in a long-lasting, anorexic effect. Based on our earlier finding that muscarinic receptor blockade reduced intra-acb DAMGO-mediated augmentation of fat intake, we wondered whether the 24 h decrease in food intake induced by scopolamine was mediated by a long-lasting decrease in striatal opioid function. There is a significant literature suggesting that muscarinic receptors mediate the expression of opioid mrna. In particular, it has been shown that systemic stimulation of muscarinic receptors increases striatal mrna for preproenkephalin, and both systemic and intrastriatal muscarinic antagonism block the upregulation of striatal opioid mrna that follows systemic amphetamine treatment [ ]. Other studies have shown that opioids modulate

28 28 acetylcholine activity, suggesting that the two neurotransmitter systems may functionally interact [127, 128]. Thus, we speculated that the long term effect of scopolamine treatment upon striatal food intake systems might be linked to the expression of opioid peptides. Therefore, we examined opioid preproenkephalin and preprodynorphin mrna expression in the same animals showing reduced food intake 24 hours after scopolamine infusion. Relative to vehicle-injected rats, animals with scopolamine infusions showed downregulation of preproenkephalin that was apparent 24 hours following the drug [129] (Figure 6). Striatal preprodynorphin was either not altered (following Acb injections), or increased near the infusion site (anterior dorsal injections). As reviewed above, it is opioid mu receptor stimulation that most potently increases food intake. Preproenkephalin mrna translates the peptide precursor to met- and leu-enkephalins, the endogenous ligands that preferentially bind to the mu (and delta) receptor. In our hands, we have consistently shown that manipulations of diet or food consumption alter striatal preproenkephalin mrna expression preferentially to that of preprodynorphin, which codes the peptide precursors to dynorphin (the endogenous kappa receptor ligand). For instance, giving rats limited access to chocolate Ensure over the course of three weeks has been shown to reduce preproenkephalin levels in the striatum, in a similar fashion to the effects observed in rats with repeated opiate drug exposure [130]. Additionally, we have recently examined striatial preproenkephalin and preprodynorphin gene expression in different food motivational conditions [131]. In these experiments, there were two main groups, rats that were chronically food restricted and those that were fed ad libitum. On the treatment day, half of the animals in each group received their daily food ration (before lights out); the remaining rats did not. Behavior and gene expression was measured at the beginning of the dark cycle, when rats are normally active and feeding. The results indicated that preproenkephalin mrna activity was correlated,

29 29 not with deprivation condition, but rather with the acute ingestion of food (Figure 5F). Animals that had been given their food ration, and had the opportunity to eat, had low expression of preproenkephalin; animals that had not received their food ration had relatively high expression levels. Preprodynorphin showed no such change across conditions. As a positive control, neuropeptide Y expression in the arcuate nucleus of the hypothalamus was also measured; in this case, neuropeptide Y was increased in all animals in the food restricted group, whether or not the animals had recently eaten their daily food ration. These intriguing data indicate a double dissociation between striatal and hypothalamic peptide systems that modulate food intake, with striatal enkephalin influencing palatability and consumption, and NPY modulating long-term energy balance. What role, then, could intrastriatal acetylcholine play in mediating these changes in enkephalin gene expression? We suggest that the cholinergic interneurons serve to modulate opioid expression (and, in relation to feeding pathways, particularly preproenkephalin signaling) based on their inputs from extrinsic sources. Although it has been shown that cortical stimulation results in post-synaptic currents within the striatal cholinergic interneuron [29], anatomical examination suggests that their predominant inputs to these neurons arise from the intralaminar thalamic nuclei [ ]. Mapping studies have also shown that the thalamic regions that project to striatum, including to the cholinergic interneurons, are heavily innervated by lateral hypothalamic regions, particularly by neurons containing the feeding- and arousalrelated peptide orexin/hypocretin [13, 14]. Thalamic regions known to receive orexin/hypocretin input functionally impact striatal cholinergic systems; in the monkey, cooling of the parafascicular nucleus of the thalamus reduces the responsiveness of striatal tonically-active (cholinergic) neurons to salient rewarded [108]. In rats, lesions to the same thalamic regions

30 30 reduce striatal acetylcholine output in response to dopaminergic receptor agonists [135]. Thus, the thalamus is well-positioned to relay hypothalamic energy-state signals, as putatively mediated by orexin/hypocretin, to the giant aspiny interneurons of the striatum. We have recently proposed a functional role for this pathway in modulating striatal enkephalin expression with regard to food intake [23]. In particular, modulation of enkephalin availability may be one means by which acetylcholine might regulate food intake. As it has been generally shown (see above) that muscarinic stimulation increases striatal preproenkephalin mrna, while muscarinic antagonism decreases it, it is possible that our scopolamine manipulations are having their long term effect by directly or indirectly reducing enkephalin availability within striatum. Such a downregulation may result in a general reduction of the hedonic aspects of all food, thus inhibiting food intake. In the normal, untreated animal, energy state signaling from the hypothalamic-thalamic-striatal axis may result in an increase of enkephalin availability during times when feeding would normally occur. This upregulation may serve to stimulate feeding (particularly on palatable substances) beyond that which would otherwise be required to maintain energy homeostasis. Indirect support for this hypothesis can be found in the literature. Stimulating the lateral hypothalamus with orexin A agonists results in feeding in rats; this effect can be reversed with intra-acb infusions of naltrexone [136]. Within an evolutionary context, an opioid-driven motivation to overeat, particularly when presented with a highly palatable and caloric food source, would serve to increase fat stores and assist in promoting the survival of the individual in the event of future famine.

31 31 5. Dopamine transmission in the Acb and feeding Although an exhaustive review of the role of dopamine in appetitive motivation is beyond the scope of the present discussion, and can be found in other excellent papers [ ], it is useful to consider several key ideas that are relevant to the topic of the dopaminergic control of feeding. A highly influential early hypothesis, the anhedonia hypothesis promulgated mainly by R. A. Wise and colleagues, held that dopamine transmission mediates the rewarding or hedonic properties of stimuli or events that served as reinforcers. A major impetus for this idea was the finding that some behavioral effects of dopamine receptor antagonism on food- or electrical brain stimulation-reinforced operant responding were reminiscent of the effects of extinction, in that animals would exhibit normal or even increased lever-pressing before eventually ceasing to respond within the testing sessions [142, 143]. This similarity between dopamine antagonist effects and extinction represented an important observation, because of the great difficulties in dissociating purported motivational effects of dopamine receptor antagonism from the profound motoric impairment produced by these drugs. Based in part on the reasoning that the extinction-like pattern in lever pressing could not be easily explained as a motor impairment, and because this pattern resembled the behavioral consequences of removing response-contingent reinforcement, it was hypothesized that dopamine receptor antagonism similarly removed the rewarding or hedonic properties of reinforcement delivery [144]. The idea that central dopamine systems mediated reward also seemed consistent with the well-known euphorigenic effects of cocaine and amphetamine, both of which exert potent indirect agonist effects on the dopamine system.

32 32 The anhedonia hypothesis had a profound influence on the study of the behavioral role of dopamine systems; a great deal of the foundational work in this area has made implicit or explicit reference to this theory, either to support or refute it [for a small sampling, see 145, ]. Nevertheless, even in the early days of the theory there were strong hints that it provided an incomplete account of dopamine s role in appetitive motivation. Much of this evidence came from studies of the dopaminergic control of feeding. It was observed that restricted dopaminergic lesions in the Acb or local intra-acb microinfusions of dopamine antagonists, while strongly decreasing spontaneous locomotor activity, had no effect on or slightly increased food intake [148, 154]. We have recently replicated this effect, and shown that a similar pattern is seen in both the Acb core and shell [155]. These effects can be interpreted as reflecting a more selective role for dopamine transmission in the anticipatory/approach phase versus the consummatory phase of feeding, an idea developed in early studies indicating that orienting responses toward a conditioned stimulus (CS) signaling food delivery were blocked by doses of dopamine receptor antagonists that spared food intake [150, 156]. Similarly, postmortem neurochemical measures of dopamine metabolism in rats indicated enhanced turnover associated with orienting responses to the CS but not with ingestive behaviors [157]. Subsequent tests of this hypothesis utilizing techniques that assay dopamine efflux (e.g., microdialysis and in vivo voltammetry) have yielded mixed results, some supporting [ ] and others contradicting [ ] a selective role for dopamine transmission in the anticipatory phase of feeding. An additional complexity is introduced by important series of microdialysis studies by Di Chiara and colleagues showing that dopamine dynamics differ substantially between the Acb core and Acb shell in relation to distinct aspects of appetitive and aversive motivational states. These investigators demonstrated that dopamine efflux in both the Acb core and shell is closely linked

33 33 to exposure and consumption of novel, palatable foodstuffs; however, food-related conditioned stimuli (which elicit strong orienting behaviors), familiar palatable tastes, and unfamiliar aversive tastes augment extracellular dopamine levels only in the Acb core [164, 165]. These functional dissociations raise the possibility that some of the conflicting findings mentioned above may have resulted from a failure to resolve distinct, localized dopamine dynamics within Acb subregions. Nevertheless, such discrepancies do not remove the difficulty to the anhedonia hypothesis posed by the failure of dopamine receptor antagonism, or restricted intra-acb 6- OHDA lesions, to decrease feeding, even though these manipulations greatly diminish locomotor activity (see Fig. 7). To address this problem, Salamone and colleagues carried out an important series of studies examining in great detail the behavioral sequelae of moderate Acb dopamine depletons. It was found that these depletions reduced the motor effort an animal was willing to expend in obtaining food reward, but did not diminish approach or intake when food was easily available [145, ]. For example, hungry rats with 6-OHDA-induced Acb dopamine depletions chose to avoid climbing over an obstacle to obtain a large number of food pellets, in favor of ambulating down an open runway to obtain a smaller number of pellets, even though these dopamine-compromised rats ate all the available pellets [145]. Similarly, Acb dopamine depletions markedly reduced operant responding on high ratio schedules, while sparing performance and ingestion of earned food pellets on low ratio schedules [168]. These results seemed difficult to reconcile with either the idea that dopamine mediated the rewarding impact of food, or with the hypothesis that dopaminergic transmission was necessary for appetitive approach and goal-seeking per se. Instead, it was hypothesized that dopamine transmission in the Acb mediates a complex process related to evaluating the effort to be expended in obtaining a

34 34 goal (although it remains unknown whether effort in this sense reflects exertion, caloric cost, expenditure of time, etc.), thereby accounting for the selective effects of dopamine-depleting lesions on more demanding behavioral tasks. Around the same time, a conceptually related line of research emerged focusing on the unconditioned stereotyped behavioral responses to pleasant versus aversive tastes (e.g., orofacial movements, facial gestures, certain types of forelimb movements) [55, 149]. These reactions are very consistent, reliable, cross-species responses to pleasant (such as sucrose) or unpleasant (such as quinine) tastes [169], and are present in humans from the time of birth. Although these unconditioned motor acts are present even in decerebrate rats [170], there is evidence that they can be modulated by higher-order forebrain influences, and have thus been proposed to represent a window into the internal hedonic experience of taste [for review, see 141]. It was found that massive chemical lesions of the dopamine system, sufficient to render the animals completely incapable of initiating behaviors aimed at obtaining food, did not affect the hedonic-like reactions to sucrose placed in the mouth [141]. This finding was interpreted as indicating that the neural systems mediating the hedonic impact of preferred tastes was distinct from that mediating the ability of incentives to elicit goal-directed behavior; in colloquial language, these two processes were termed liking and wanting, respectively. With regard to neurochemical mediation, it was proposed that liking depends upon opiatergic transmission (as discussed in section 3 of this review), and wanting is governed by the mesolimbic dopamine system. The specific role of dopamine was proposed to involve the tagging of neural representations of stimuli in the environment as motivationally significant, or possessing incentive salience, thus explaining the lack of goal-directed actions in animals with functional impairments in dopamine transmission. As will be discussed in detail later in this review, the liking/wanting model has

35 35 proved to be an extremely useful heuristic framework with which to understand the dissociable effects of opiatergic and dopaminergic manipulations on feeding. In a general way, the incentive salience model is quite compatible with the anergia hypothesis described above, in the sense that they both ascribe to the dopamine system a role in mediating those aspects of appetitive motivation most closely connected to the production or augmentation of goal-directed behavior, rather than those components related to the generation of an internal state with hedonic valence. Indeed, it has been suggested that overcoming response costs can be viewed as a specific subtype of wanting. Our own work supports this general theme. As described above, we have observed that intra-acb infusions of dopamine receptor antagonists, regardless of the subtype selectivity of the antagonists or the Acb subregion into which they are injected, dramatically reduces the motor output associated with a hungry state without reducing (or in some cases, actually increasing) food intake [155, 171], see Fig. 7 A, C. It has also been noted that increasing intra-acb dopamine transmission with local amphetamine injections does not increase food intake per se [148] see Fig. 7F, but markedly augments performance in tasks designed to evaluate the ability of primary or secondary reinforcers to elicit goal-seeking behavior. Specifically, as shown in Fig. 7D-E, intra-acb amphetamine infusions, in either the Acb core or the shell, markedly increase performance on a progressive ratio task for food reward [38], as well as on an operant second-order schedule for a food-associated conditioned stimulus [172, 173]. Interestingly, however, intra-acb amphetamine infusions do not enable the acquisition of food-reinforced operant responding in ad libitum-fed rats [39]. We interpret these results as reflecting the specificity of dopamine-dependent processes in energizing the anticipatory/approach phase of the feeding CMS versus the hedonic aspects of the consummatory act; in this manner, selective pharmacological augmentation of Acb dopamine

36 36 transmission can be seen as artificially fragmenting the coordinated activation of the multiple neurochemical systems by which the CMS is instantiated. This hypothesized fragmentation accounts for the relative ineffectiveness of dopamine manipulations to alter feeding behavior in conditions requiring only minimal motoric output (i.e., emphasizing the consummatory phase), or where coordinated enhancement of both the anticipatory/approach phase and the hedonic impact of the consummatory act are critical (such as in eliciting a state that supports new operant learning in ad libitum-fed rats). It is important to note that in addition to its role in motivational processes, dopamine transmission in the striatum may also play a related role in mediating local circuit-based and intracellular dynamics related to learning [137, 140, 165], although see [174, 175]. A substantial body of electrophysiological and neurochemical work has accrued demonstrating that the ascending dopamine systems are critically involved in reward-related learning. An influential line of research on the electrophysiological responses of ventral tegmental area neurons has indicated that these cells may convey to the Acb a type of training signal related to the relevance of stimuli in predicting reward [ ], although there is debate as to whether dopamine transmission specifically marks stimuli with positive motivational valence versus motivationally relevant stimuli in general [see discussions in 137, 179]. Dopamine, in conjunction with glutamatergic systems in the striatum, is also thought to promote important postsynaptic plasticity-related events in association with learning [ ]. Accordingly, it has been has shown that coincident activation of postsynaptic D1 dopamine and NMDA glutamate receptors in the Acb core is required for the consolidation of a new food-reinforced operant response [185, 186], and for the acqusition and consolidation of appetitive Pavlovian associations [187].

37 37 6. Synthesis and conclusions To briefly review, we have observed multiple dissociations with regard to amino acid, opioid, and dopaminergic modulation of food intake and food-seeking behaviors. Thus, blockade of glutamate or stimulation of GABA systems in the Acb shell mediates large increases in food intake measured in free-feeding tests, but does not enhance operant responding to obtain food. In contrast, augmenting Acb dopamine transmission increases operant responding for food (progressive ratio task) or stimuli associated with food (conditioned reinforcement paradigm), but not food intake under free-feeding conditions. Finally, stimulating striatal mu-opioid receptors augments both free-feeding and progressive ratio performance, but does not influence responding in the conditioned reinforcement paradigm. None of these manipulations is sufficient to produce a state that supports new food-reinforced operant learning in ad libitum-fed rats. Moreover, although manipulations of both Acb shell amino acid systems and striatal opioid

38 38 systems enhance food intake, the amino acid-coded effect is indiscriminate with regard to the macronutrient or sensory properties of the food, while the opioid effect favors intake of palatable, energy-dense high sugar/fat foodstuffs. Finally, the feeding-related effects of these three systems exhibit a great deal of heterogeneity with regard to their localization within the striatum; GABA and glutamate-associated effects demonstrate a strikingly distinct localization to the Acb shell, while opioid-mediated hyperphagia can be elicited from a number of sites broadly distributed throughout the striatum, although more ventrally localized sites are associated with larger effects. These pharmacological, behavioral, and anatomical dissociations have interesting implications for a basic understanding of how appetitive motivational states are instantiated in the brain. In this regard, it is useful to briefly review two elements of current motivational theory that are germane to an interpretation of our results. First, there is a long history of viewing appetitively motivated behaviors as occurring in two distinct phases: an anticipatory/approach phase characterized by high arousal, exploratory behavior, and instrumental strategies designed to obtain a goal object, and a consummatory phase which involves commerce with the goal object ( consummatory refers to the consummation of the anticipatory/approach phase, rather than to consumption in the sense of ingestion) [see 41]. In contrast to behavioral output associated with the anticipatory/approach phase, interactions with the goal object are characterized by very stereotyped, often repetitive motor sequences (e.g., chewing, swallowing) and a degree of disengagement from the environment and decreased vigilance. Second, models of incentive motivation postulate, in a general sense, that any particular CMS (i.e., for food, water, mating, etc.) is actualized though distinct but coordinated motor and autonomic output structures that together enable motivated behaviors to be carried out

39 39 successfully. For example, Bindra wrote that the behavioral expression of the CMS ( observed responses ) represented the coordinated contributions of systems mediating autonomic discharge, postural adjustments, and environmentally organized motor output [42]. At the time, however, the anatomical substrates mediating these processes were only beginning to be characterized. An influential contribution to this topic was made by Mogenson [188], who noted that the Acb is anatomically well-positioned to serve as an interface between allocortical and neocortical regions traditionally associated with the limbic system, and pallido-thalamic circuits that eventually reached areas of cortex concerned with voluntary motor control (thus the characterization of the Acb as a structure translating motivation to action ). This idea had a profound impact on the development of subsequent theories of Acb function, including the now well-accepted idea that mesocortical dopamine projections to the Acb influence behavior by modulating incoming corticolimbic glutamate-coded signals. Mogenson s theories have also influenced the development of ideas regarding the discrete neurochemical coding of distinct aspects of appetitive motivational states. Consonant with this general idea, the argument has been developed in the current paper that the numerous behavioral dissociations we have observed with regard to Acb control of food motivation can be best understood within the heuristic framework of distinct subsystems in the Acb mediating discrete aspects of appetitive motivation. To briefly summarize, we propose that neurchemical systems in the Acb mediate the following three processes: (1) Enabling and augmenting the goal-seeking instrumental strategies associated with the anticipatory/approach phase of the feeding CMS, and participating in plasticity related to stamping in successful new motor strategies, as mediated by dopamine transmission throughout the Acb. This can account for the failure of dopamine

40 40 antagonists to block, and amphatamine to enhance, food intake, despite the fact that amphetamine dramatically augments progressive ratio performance and responding for food-associated conditioned reinforcers (CRs). (2) Augmenting the hedonic experience of the feeding consummatory act, particularly with regard to palatable foods, as mediated by opioid transmission throughout the Acb and striatum under the regulatory control of cholinergic interneurons. This could explain the preferential enhancement of sugar/fat intake and augmented progressive ratio performance, and, if (as we hypothesize) increased opioid tone enhances specifically the hedonic impact of palatable food currently being tasted, the failure of intra-acb opioid stimulation to enhance CR-maintained responding. (3) Maintaining a sentinel -like control function over motor/autonomic circuits mediating the patterned behavioral activities of the consummatory act, as mediated by a subensemble of GABA output cells in the Acb shell. We posit that removing this negative control over feeding pattern generators results in hyperphagia. This mechanism could explain why the hyperphagia elicited by GABA stimulation or glutamate receptor blockade in the Acb shell is somewhat insensitive to the sensory properties of food (palatability and macronutrient content), and why these manipulations do not enhance complex instrumental responses to obtain food. Further, based on our observation that none of these manipulations alone is sufficient to enable the acquisition of a food-reinforced operant response to acquire food in ad libitum-fed animals, we hypothesize that the coordinated activity of all these systems, in which anticipatory/approach- and consummatory response-related substrates are not artificially pulled apart from one another, is critical for the instantiation of a state that supports learning new goal-

41 41 directed strategies to obtain food. As discussed previously, these ideas find excellent support and agreement from other data and theories in the literature. For example, similarly to the framework outlined above, the incentive salience theory discussed previously indicates that rewarding/hedonic aspects of the consummatory act ( liking ) can be separated from the identification of stimuli as motivationally relevant ( wanting ), with opioid systems mediating liking and dopamine systems more strongly connected to wanting [55]. The question arises, however, as to the physiological basis for these dissociations in motivational functions, particularly considering that these neurotransmitter systems strongly interact with each other within the basal ganglia. One clue is provided by the anatomical specificity of some of our effects, particularly the amino-acid associated hyperphagia that is obtained exclusively in the Acb shell. Thus, while outputs from the Acb core and overlying striatum are characterized by the classic cortico-striatal-pallido-thalamic circuitry that enables basal ganglia processing to access voluntary motor systems, efferents from the shell project in a much more restricted way to subregions of the ventral pallidum and, unlike rest of the striatum, to the lateral hypothalamus [18, 19, 20]. We propose that this arrangement enables opioid and dopamine systems to broadly engage frontotemporal cortical circuits that contribute to the positive emotion associated with certain foods (opioids) and to assist in directing behavior toward foods and the stimuli that predict them (dopamine). In contrast, ventral striatal neurons bearing GABA receptors and AMPA receptors, in a restricted medial ventral striatal region, are able to directly control brainstem feeding motor pattern generators, but have much more limited access to the voluntary motor system. These anatomical distinctions can explain why, in contrast to dopamine and opioid stimulation, GABA or glutamate manipulations in the Acb shell are

42 42 unable to influence complex instrumental strategies to obtain food despite the fact that they produce profound hyperphagia. Finally, although it seems clear from an evolutionary perspective why we need systems to seek after goal objects (dopamine) and to interrupt feeding in order to flee threats (Acb shell amino acids), it is somewhat puzzling to discern what the benefit of striatal opioid processing might be. In this regard, it is interesting to consider that throughout most of our evolutionary history, food was a scarce commodity and it was difficult to foresee impending periods of famine. Hence, there was a distinct evolutionary advantage in developing a reserve of stored energy through the intake of calorically dense foods. As discussed previously, we speculate that the striatal opioid mediation of palatability may have developed as a way to prolong the intake of energy-rich foods beyond immediate homeostatic needs, by increasing the incentive value of these foods in a manner based entirely upon their sensory properties (which would be predictive of their energy content). Nevertheless, because there would still be an advantage to maximizing energy-dense food intake under conditions of energy deficit, we posit that the striatal opioid system exhibits responsivity to (but not dependence upon) hypothalamic energy-sensing systems; as stated previously, the cholinergic interneurons of the striatum represent an interesting potential substrate to act as such an interface, because they are positioned to receive strong input from thalamic regions that themselves have access to information regarding energy balance and feeding-relevant circadian patterns arising in the hypothalamus. Although still in its early stages, this theory may provide interesting new directions for the study of striatal control of appetitive motivation.

43 43 Acknowledgements. The research described within has been supported by National Institute on Drug Abuse grants DA09311 and DA04788 to A.E.K. B.A.B was supported by a NARSAD young investigator award, as well as by a National Institute of Mental Health National Research Service Reward MH W.E.P. and M.J.W. were supported by National Research Service Awards DA and MH-68981, respectively. We would like to thank Carol Dizack for her assistance with the figures.

44 44 References 1. Drewnowski, A.; Levine, A. S. Sugar and fat--from genes to culture. J Nutr, (3): 829S-830S. 2. Hill, J. O.; Wyatt, H. R.; Reed, G. W.; Peters, J. C. Obesity and the environment: where do we go from here? Science, (5608): Saper, C. B.; Chou, T. C.; Elmquist, J. K. The need to feed: homeostatic and hedonic control of eating. Neuron, (2): Dallman, M. F.; Pecoraro, N.; Akana, S. F.; La Fleur, S. E.; Gomez, F.; Houshyar, H.; Bell, M. E.; Bhatnagar, S.; Laugero, K. D.; Manalo, S. Chronic stress and obesity: a new view of "comfort food". Proc Natl Acad Sci U S A, (20): Saper, C. B. Convergence of autonomic and limbic connections in the insular cortex of the rat. J Comp Neurol, (2): Ricardo, J. A.; Koh, E. T. Anatomical evidence of direct projections from the nucleus of the solitary tract to the hypothalamus, amygdala, and other forebrain structures in the rat. Brain Res, (1): Shi, C. J.; Cassell, M. D. Cortical, thalamic, and amygdaloid connections of the anterior and posterior insular cortices. J Comp Neurol, (4): Vertes, R. P. Differential projections of the infralimbic and prelimbic cortex in the rat. Synapse, (1): Hurley, K. M.; Herbert, H.; Moga, M. M.; Saper, C. B. Efferent projections of the infralimbic cortex of the rat. J Comp Neurol, (2):

45 Berendse, H. W.; Galis-de Graaf, Y.; Groenewegen, H. J. Topographical organization and relationship with ventral striatal compartments of prefrontal corticostriatal projections in the rat. J Comp Neurol, (3): McDonald, A. J.; Jackson, T. R. Amygdaloid connections with posterior insular and temporal cortical areas in the rat. J Comp Neurol, (1): McDonald, A. J.; Shammah-Lagnado, S. J.; Shi, C.; Davis, M. Cortical afferents to the extended amygdala. Ann N Y Acad Sci, : Peyron, C.; Tighe, D. K.; van den Pol, A. N.; de Lecea, L.; Heller, H. C.; Sutcliffe, J. G.; Kilduff, T. S. Neurons containing hypocretin (orexin) project to multiple neuronal systems. J Neurosci, (23): Baldo, B. A.; Daniel, R. A.; Berridge, C. W.; Kelley, A. E. Overlapping distributions of orexin/hypocretin- and dopamine-beta-hydroxylase immunoreactive fibers in rat brain regions mediating arousal, motivation, and stress. J Comp Neurol, (2): Berendse, H. W.; Voorn, P.; te Kortschot, A.; Groenewegen, H. J. Nuclear origin of thalamic afferents of the ventral striatum determines their relation to patch/matrix configurations in enkephalin-immunoreactivity in the rat. J Chem Neuroanat, (1): Berendse, H. W.; Groenewegen, H. J. Organization of the thalamostriatal projections in the rat, with special emphasis on the ventral striatum. J Comp Neurol, (2): Wright, C. I.; Groenewegen, H. J. Patterns of overlap and segregation between insular cortical, intermediodorsal thalamic and basal amygdaloid afferents in the nucleus accumbens of the rat. Neuroscience, (2):

46 Heimer, L.; Zahm, D. S.; Churchill, L.; Kalivas, P. W.; Wohltmann, C. Specificity in the projection patterns of accumbal core and shell in the rat. Neuroscience, (1): Zahm, D. S.; Heimer, L. Specificity in the efferent projections of the nucleus accumbens in the rat: comparison of the rostral pole projection patterns with those of the core and shell. J Comp Neurol, (2): Mogenson, G. J.; Swanson, L. W.; Wu, M. Neural projections from nucleus accumbens to globus pallidus, substantia innominata, and lateral preoptic-lateral hypothalamic area: an anatomical and electrophysiological investigation in the rat. J Neurosci, (1): Saper, C. B. The central autonomic nervous system: conscious visceral perception and autonomic pattern generation. Annu Rev Neurosci, : Swanson, L. W. Cerebral hemisphere regulation of motivated behavior. Brain Res, (1-2): Kelley, A. E.; Baldo, B. A.; Pratt, W. E. A proposed hypothalamic-thalamic-striatal axis for the integration of energy balance, arousal, and food reward. J Comp Neurol, in press. 24. Maldonado-Irizarry, C. S.; Kelley, A. E. Excitatory amino acid receptors within nucleus accumbens subregions differentially mediate spatial learning in the rat. Behav Pharmacol, (5 And 6): Stratford, T. R.; Swanson, C. J.; Kelley, A. Specific changes in food intake elicited by blockade or activation of glutamate receptors in the nucleus accumbens shell. Behav Brain Res, (1-2):

47 Kelley, A. E.; Swanson, C. J. Feeding induced by blockade of AMPA and kainate receptors within the ventral striatum: a microinfusion mapping study. Behav Brain Res, (1-2): Maldonado-Irizarry, C. S.; Kelley, A. E. Excitotoxic lesions of the core and shell subregions of the nucleus accumbens differentially disrupt body weight regulation and motor activity in rat. Brain Res Bull, (6): Stratford, T. R.; Kelley, A. E. GABA in the nucleus accumbens shell participates in the central regulation of feeding behavior. J Neurosci, (11): Basso, A. M.; Kelley, A. E. Feeding induced by GABA(A) receptor stimulation within the nucleus accumbens shell: regional mapping and characterization of macronutrient and taste preference. Behav Neurosci, (2): Berridge, K. C.; Valenstein, E. S. What psychological process mediates feeding evoked by electrical stimulation of the lateral hypothalamus? Behav Neurosci, (1): Duva, M. A.; Tomkins, E. M.; Moranda, L. M.; Kaplan, R.; Sukhaseum, A.; Jimenez, A.; Stanley, B. G. Reverse microdialysis of N-methyl-D-aspartic acid into the lateral hypothalamus of rats: effects on feeding and other behaviors. Brain Res, (1-2): Groenewegen, H. J.; Russchen, F. T. Organization of the efferent projections of the nucleus accumbens to pallidal, hypothalamic, and mesencephalic structures: a tracing and immunohistochemical study in the cat. J Comp Neurol, (3):

48 Maldonado-Irizarry, C. S.; Swanson, C. J.; Kelley, A. E. Glutamate receptors in the nucleus accumbens shell control feeding behavior via the lateral hypothalamus. J Neurosci, (10): Stratford, T. R.; Kelley, A. E. Evidence of a functional relationship between the nucleus accumbens shell and lateral hypothalamus subserving the control of feeding behavior. J Neurosci, (24): Baldo, B. A.; Gual-Bonilla, L.; Sijapati, K.; Daniel, R. A.; Landry, C. F.; Kelley, A. E. Activation of a subpopulation of orexin/hypocretin-containing hypothalamic neurons by GABAA receptor-mediated inhibition of the nucleus accumbens shell, but not by exposure to a novel environment. Eur J Neurosci, (2): Zheng, H.; Corkern, M.; Stoyanova, I.; Patterson, L. M.; Tian, R.; Berthoud, H. R. Peptides that regulate food intake: appetite-inducing accumbens manipulation activates hypothalamic orexin neurons and inhibits POMC neurons. Am J Physiol Regul Integr Comp Physiol, (6): R Stratford, T. R.; Kelley, A. E.; Simansky, K. J. Blockade of GABAA receptors in the medial ventral pallidum elicits feeding in satiated rats. Brain Res, (1-2): Zhang, M.; Balmadrid, C.; Kelley, A. E. Nucleus accumbens opioid, GABaergic, and dopaminergic modulation of palatable food motivation: contrasting effects revealed by a progressive ratio study in the rat. Behav Neurosci, (2): Hanlon, E. C.; Baldo, B. A.; Sadeghian, K.; Kelley, A. E. Increases in food intake or food-seeking behavior induced by GABAergic, opioid, or dopaminergic stimulation of the nucleus accumbens: is it hunger? Psychopharmacology (Berl), (3):

49 Reynolds, S. M.; Berridge, K. C. Positive and negative motivation in nucleus accumbens shell: bivalent rostrocaudal gradients for GABA-elicited eating, taste "liking"/"disliking" reactions, place preference/avoidance, and fear. J Neurosci, (16): Berridge, K. C. Motivation concepts in behavioral neuroscience. Physiol Behav, (2): Bindra, D. A unfied interpretation of emotion and motivation. Ann N Y Acad Sci, (3): Taha, S. A.; Fields, H. L. Encoding of palatability and appetitive behaviors by distinct neuronal populations in the nucleus accumbens. J Neurosci, (5): Baldo, B. A.;Kelley, A. E. Baldo BA, Kelley AE. (2004) Effects of GABA-A mediated inhibition of infralimbic or agranular insular cortex on the microstructure of deprivationinduced and palatability-driven feeding. Society for Neuroscience Abstracts, Will, M. J.; Franzblau, E. B.;Kelley, A. E. The amygdala is critical for opioid-mediated binge eating of fat. Neuroreport, (12): Baldo, B. A.; Sadeghian, K.; Kelley, A. E. Temporary pharmacological inactivation of ventral orbital cortex alters feeding microstructure in rats. Society for Neuroscience Abstracts, Sweep, C. G.; Wiegant, V. M.; De Vry, J.; Van Ree, J. M. Beta-endorphin in brain limbic structures as neurochemical correlate of psychic dependence on drugs. Life Sci, (16): Pickel, V. M.; Sumal, K. K.; Beckley, S. C.; Miller, R. J.; Reis, D. J. Immunocytochemical localization of enkephalin in the neostriatum of rat brain: a light and electron microscopic study. J Comp Neurol, (4):

50 Mansour, A.; Khachaturian, H.; Lewis, M. E.; Akil, H.; Watson, S. J. Autoradiographic differentiation of mu, delta, and kappa opioid receptors in the rat forebrain and midbrain. J Neurosci, (8): Martin, W. R.; Wikler, A.; Eades, C. G.; Pescor, F. T. Tolerance to and Physical Dependence on Morphine in Rats. Psychopharmacologia, : Herman, B. H.; Holtzman, S. G. Repeated administration of naltrexone and diprenorphine decreases food intake and body weight in squirrel monkeys. Life Sci, (1): Yeomans, M. R.; Gray, R. W.; Mitchell, C. J.; True, S. Independent effects of palatability and within-meal pauses on intake and appetite ratings in human volunteers. Appetite, (1): Cooper, S. J.; Kirkham, T. C. Opioid mechanisms in the control of food consumption and taste preferences. In: Herz, A. ed Handbook of Experimental Pharmacology. Opioids II. Berlin: Springer-Verlagl; 1993; Vol. 104/II. 54. Calcagnetti, D. J.; Reid, L. D. Morphine and acceptability of putative reinforcers. Pharmacol Biochem Behav, (4): Berridge, K. C. Food reward: brain substrates of wanting and liking. Neurosci Biobehav Rev, (1): Levine, A. S.; Billington, C. J. Opioids as agents of reward-related feeding: a consideration of the evidence. Physiol Behav, (1): Arbisi, P. A.; Billington, C. J.; Levine, A. S. The effect of naltrexone on taste detection and recognition threshold. Appetite, (2): Fantino, M.; Hosotte, J.; Apfelbaum, M. An opioid antagonist, naltrexone, reduces preference for sucrose in humans. Am J Physiol, (1 Pt 2): R91-6.

51 Drewnowski, A.; Krahn, D. D.; Demitrack, M. A.; Nairn, K.; Gosnell, B. A. Taste responses and preferences for sweet high-fat foods: evidence for opioid involvement. Physiol Behav, (2): Yeomans, M. R.; Gray, R. W. Selective effects of naltrexone on food pleasantness and intake. Physiol Behav, (2): Le Magnen, J.; Marfaing-Jallat, P.; Miceli, D.; Devos, M. Pain modulating and reward systems: a single brain mechanism? Pharmacol Biochem Behav, (5): Apfelbaum, M.; Mandenoff, A. Naltrexone suppresses hyperphagia induced in the rat by a highly palatable diet. Pharmacol Biochem Behav, (1): Siviy, S. M.; Reid, L. D. Endophinergic modulation of acceptability of putative reinforcers. Appetite, (4): Czirr, S. A.; Reid, L. D. Demonstrating morphine's potentiating effects on sucrose-intake. Brain Res Bull, (5): Doyle, T. G.; Berridge, K. C.; Gosnell, B. A. Morphine enhances hedonic taste palatability in rats. Pharmacol Biochem Behav, (3): Aicher, S. A.; Goldberg, A.; Sharma, S.; Pickel, V. M. mu-opioid receptors are present in vagal afferents and their dendritic targets in the medial nucleus tractus solitarius. J Comp Neurol, (2): Xia, Y.; Haddad, G. G. Ontogeny and distribution of opioid receptors in the rat brainstem. Brain Res, (2): Glass, M. J.; Billington, C. J.; Levine, A. S. Opioids and food intake: distributed functional neural pathways? Neuropeptides, (5):

52 Echo, J. A.; Lamonte, N.; Ackerman, T. F.; Bodnar, R. J. Alterations in food intake elicited by GABA and opioid agonists and antagonists administered into the ventral tegmental area region of rats. Physiol Behav, (1): Levine, A. S.; Olszewski, P. K.; Mullett, M. A.; Pomonis, J. D.; Grace, M. K.; Kotz, C. M.; Billington, C. J. Intra-amygdalar injection of DAMGO: effects on c-fos levels in brain sites associated with feeding behavior. Brain Res, (1-2): Kotz, C. M.; Billington, C. J.; Levine, A. S. Opioids in the nucleus of the solitary tract are involved in feeding in the rat. Am J Physiol, (4 Pt 2): R Wilson, J. D.; Nicklous, D. M.; Aloyo, V. J.; Simansky, K. J. An orexigenic role for muopioid receptors in the lateral parabrachial nucleus. Am J Physiol Regul Integr Comp Physiol, (5): R Bodnar, R. J.; Lamonte, N.; Israel, Y.; Kandov, Y.; Ackerman, T. F.; Khaimova, E. Reciprocal opioid-opioid interactions between the ventral tegmental area and nucleus accumbens regions in mediating mu agonist-induced feeding in rats. Peptides, (4): Glass, M. J.; Billington, C. J.; Levine, A. S. Naltrexone administered to central nucleus of amygdala or PVN: neural dissociation of diet and energy. Am J Physiol Regul Integr Comp Physiol, (1): R Will, M. J.; Franzblau, E. B.; Kelley, A. E. Nucleus accumbens mu-opioids regulate intake of a high-fat diet via activation of a distributed brain network. J Neurosci, (7): Mucha, R. F.; Iversen, S. D. Increased food intake after opioid microinjections into nucleus accumbens and ventral tegmental area of rat. Brain Res, (2):

53 Majeed, N. H.; Przewlocka, B.; Wedzony, K.; Przewlocki, R. Stimulation of food intake following opioid microinjection into the nucleus accumbens septi in rats. Peptides, (5): Evans, K. R.; Vaccarino, F. J. Amphetamine- and morphine-induced feeding: evidence for involvement of reward mechanisms. Neurosci Biobehav Rev, (1): Alexander, G. E.; Crutcher, M. D.; DeLong, M. R. Basal ganglia-thalamocortical circuits: parallel substrates for motor, oculomotor, "prefrontal" and "limbic" functions. Prog Brain Res, : Kelley, A. E.; Domesick, V. B.; Nauta, W. J. The amygdalostriatal projection in the rat-- an anatomical study by anterograde and retrograde tracing methods. Neuroscience, (3): Bakshi, V. P.; Kelley, A. E. Striatal regulation of morphine-induced hyperphagia: an anatomical mapping study. Psychopharmacology (Berl), (2): Delfs, J. M.; Kong, H.; Mestek, A.; Chen, Y.; Yu, L.; Reisine, T.; Chesselet, M. F. Expression of mu opioid receptor mrna in rat brain: an in situ hybridization study at the single cell level. J Comp Neurol, (1): Steiner, H.; Gerfen, C. R. Role of dynorphin and enkephalin in the regulation of striatal output pathways and behavior. Exp Brain Res, (1-2): Bakshi, V. P.; Kelley, A. E. Feeding induced by opioid stimulation of the ventral striatum: role of opiate receptor subtypes. J Pharmacol Exp Ther, (3): Kelley, A. E.; Bless, E. P.; Swanson, C. J. Investigation of the effects of opiate antagonists infused into the nucleus accumbens on feeding and sucrose drinking in rats. J Pharmacol Exp Ther, (3):

54 Shippenberg, T. S.; Elmer, G. I. The neurobiology of opiate reinforcement. Crit Rev Neurobiol, (4): Gosnell, B. A.; Levine, A. S.; Morley, J. E. The stimulation of food intake by selective agonists of mu, kappa and delta opioid receptors. Life Sci, (12): Bodnar, R. J.; Glass, M. J.; Ragnauth, A.;C ooper, M. L. General, mu and kappa opioid antagonists in the nucleus accumbens alter food intake under deprivation, glucoprivic and palatable conditions. Brain Res, (1-2): Ruegg, H.; Yu, W. Z.; Bodnar, R. J. Opioid-receptor subtype agonist-induced enhancements of sucrose intake are dependent upon sucrose concentration. Physiol Behav, (1): Levine, A. S.; Weldon, D. T.; Grace, M.; Cleary, J. P.; Billington, C. J. Naloxone blocks that portion of feeding driven by sweet taste in food-restricted rats. Am J Physiol, (1 Pt 2): R Zhang, M.; Kelley, A. E. Opiate agonists microinjected into the nucleus accumbens enhance sucrose drinking in rats. Psychopharmacology (Berl), (4): Zhang, M.; Kelley, A. E. Intake of saccharin, salt, and ethanol solutions is increased by infusion of a mu opioid agonist into the nucleus accumbens. Psychopharmacology (Berl), (4): Zhang, M.; Kelley, A. E. Enhanced intake of high-fat food following striatal mu-opioid stimulation: microinjection mapping and fos expression. Neuroscience, (2):

55 Zhang, M.; Gosnell, B. A.; Kelley, A. E. Intake of high-fat food is selectively enhanced by mu opioid receptor stimulation within the nucleus accumbens. J Pharmacol Exp Ther, (2): Cunningham, S. T.; Kelley, A. E. Opiate infusion into nucleus accumbens: contrasting effects on motor activity and responding for conditioned reward. Brain Res, (1): Blanchet, F.; Gauchy, C.; Perez, S.; Soubrie, P.; Glowinski, J.; Kemel, M. L. Control by GABA and tachykinins of the evoked release of acetylcholine in striatal compartments under different modalities of NMDA receptor stimulation. Brain Res, (1): Blanchet, F.; Kemel, M. L.; Gauchy, C.; Desban, M.; Perez, S.; Glowinski, J. N-methyl- D-aspartate-evoked release of [3H]acetylcholine in striatal compartments of the rat: regulatory roles of dopamine and GABA. Neuroscience, (1): DeBoer, P.; Heeringa, M. J.; Abercrombie, E. D. Spontaneous release of acetylcholine in striatum is preferentially regulated by inhibitory dopamine D2 receptors. Eur J Pharmacol, (2-3): Di Chiara, G.; Morelli, M.; Consolo, S. Modulatory functions of neurotransmitters in the striatum: ACh/dopamine/NMDA interactions. Trends Neurosci, (6): Wang, H.; Moriwaki, A.; Wang, J. B.; Uhl, G. R.; Pickel, V. M. Ultrastructural immunocytochemical localization of mu opioid receptors and Leu5-enkephalin in the patch compartment of the rat caudate-putamen nucleus. J Comp Neurol, (4):

56 Wang, H.; Moriwaki, A.; Wang, J. B.; Uhl, G. R.; Pickel, V. M. Ultrastructural immunocytochemical localization of mu-opioid receptors in dendritic targets of dopaminergic terminals in the rat caudate-putamen nucleus. Neuroscience, (3): Will, M. J.; Kelley, A. E. Role of amygdala, lateral hypothalamus, and glutamate in the enhancement of fat intake induced by intra-accumbens mu-opioid stimulation. Society for Neuroscience Abstracts, Zhou, F. M.; Wilson, C. J.; Dani, J. A. Cholinergic interneuron characteristics and nicotinic properties in the striatum. J Neurobiol, (4): Izzo, P. N.; Bolam, J. P. Cholinergic synaptic input to different parts of spiny striatonigral neurons in the rat. J Comp Neurol, (2): Hersch, S. M.; Gutekunst, C. A.; Rees, H. D.; Heilman, C. J.; Levey, A. I. Distribution of m1-m4 muscarinic receptor proteins in the rat striatum: light and electron microscopic immunocytochemistry using subtype-specific antibodies. J Neurosci, (5 Pt 2): Sardo, P.; Ravel, S.; Legallet, E.; Apicella, P. Influence of the predicted time of stimuli eliciting movements on responses of tonically active neurons in the monkey striatum. Eur J Neurosci, (5): Ravel, S.; Legallet, E.; Apicella, P. Responses of tonically active neurons in the monkey striatum discriminate between motivationally opposing stimuli. J Neurosci, (24):

57 Matsumoto, N.; Minamimoto, T.; Graybiel, A. M.; Kimura, M. Neurons in the thalamic CM-Pf complex supply striatal neurons with information about behaviorally significant sensory events. J Neurophysiol, (2): Apicella, P.; Ravel, S.; Sardo, P.; Legallet, E. Influence of predictive information on responses of tonically active neurons in the monkey striatum. J Neurophysiol, (6): Schultz, W. Predictive reward signal of dopamine neurons. J Neurophysiol, (1): Kitabatake, Y.; Hikida, T.; Watanabe, D.; Pastan, I.; Nakanishi, S. Impairment of rewardrelated learning by cholinergic cell ablation in the striatum. Proc Natl Acad Sci U S A, (13): Pratt, W. E.; Kelley, A. E. Nucleus accumbens acetylcholine regulates appetitive learning and motivation for food via activation of muscarinic receptors. Behav Neurosci, (4): Ragozzino, M. E.; Jih, J.; Tzavos, A. Involvement of dorsomedial striatum in behavioral flexibility: Role of muscarinic cholinergic receptors. Brain Res, : Tikhonravov, D. L.; Shapovalova, K. B.; Dyubkacheva, T. A. Effects of microinjection of scopolamine into the neostriatum of rats on performance of a food conditioned reflex at different levels of fixation. Neurosci Behav Physiol, (3): Prado-Alcala, R. A. Is cholinergic activity of the caudate nucleus involved in memory? Life Sci, (23): Ragozzino, M. E.; Choi, D. Dynamic changes in acetylcholine output in the medial striatum during place reversal learning. Learn Mem, (1): 70-7.

58 Chang, Q.; Gold, P. E. Switching memory systems during learning: changes in patterns of brain acetylcholine release in the hippocampus and striatum in rats. J Neurosci, (7): Mark, G. P.; Rada, P.; Pothos, E.; Hoebel, B. G. Effects of feeding and drinking on acetylcholine release in the nucleus accumbens, striatum, and hippocampus of freely behaving rats. J Neurochem, (6): Mark, G. P.; Weinberg, J. B.; Rada, P. V.; Hoebel, B. G. Extracellular acetylcholine is increased in the nucleus accumbens following the presentation of an aversively conditioned taste stimulus. Brain Res, (1-2): Rada, P.; Mark, G. P.; Hoebel, B. G. Galanin in the hypothalamus raises dopamine and lowers acetylcholine release in the nucleus accumbens: a possible mechanism for hypothalamic initiation of feeding behavior. Brain Res, (1-2): Hajnal, A.; Szekely, M.; Galosi, R.; Lenard, L. Accumbens cholinergic interneurons play a role in the regulation of body weight and metabolism. Physiol Behav, (1-2): Pratt, W. E.; Kelley, A. E. Twenty-four hour food intake is reduced following muscarinic receptor antagonism of the nucleus accumbens or anterior dorsal striatum. In: Zuberbeuhler, C. A. ed. Proceedings of the Society for the Study of Ingestive Behavior. Appetite, (3): Weisinger, G.; DeCristofaro, J. D.; LaGamma, E. F. Tissue- and treatment-specific usage of multiple preproenkephalin transcriptional start sites. J Biol Chem, (7):

59 Weisinger, G.; Zinder, O.; DeCristofaro, J. D.; LaGamma, E. F. Novel transcriptional mechanisms are involved in regulating preproenkephalin gene expression in vivo. Biochem Biophys Res Commun, (2): Wang, J. Q.; McGinty, J. F. Muscarinic receptors regulate striatal neuropeptide gene expression in normal and amphetamine-treated rats. Neuroscience, (1): Wang, J. Q.; McGinty, J. F. Intrastriatal injection of a muscarinic receptor agonist and antagonist regulates striatal neuropeptide mrna expression in normal and amphetaminetreated rats. Brain Res, (1-2): Jabourian, M.; Bourgoin, S.; Perez, S.; Godeheu, G.; Glowinski, J.; Kemel, M. L. Mu opioid control of the N-methyl-D-aspartate-evoked release of [3H]-acetylcholine in the limbic territory of the rat striatum in vitro: diurnal variations and implication of a dopamine link. Neuroscience, (3): Jabourian, M.; Venance, L.; Bourgoin, S.; Ozon, S.; Perez, S.; Godeheu, G.; Glowinski, J.; Kemel, M. L. Functional mu opioid receptors are expressed in cholinergic interneurons of the rat dorsal striatum: territorial specificity and diurnal variation. Eur J Neurosci, (12): Pratt, W. E.; Kelley, A. E. Striatal muscarinic receptor blockade reduces 24-hour food intake and striatal preproenkephalin gene expression. Society for Neuroscience Abstracts, Kelley, A. E.; Will, M. J.; Steininger, T. L.; Zhang, M.; Haber, S. N. Restricted daily consumption of a highly palatable food (chocolate Ensure(R)) alters striatal enkephalin gene expression. Eur J Neurosci, (9):

60 Will, M. J.; Vander Heyden, W.; LaVaute, T.; Kelley, A. E. Motivational state (foodrestriction) and acute food intake effects on striatal preproenkephalin and hypothalamic neuropeptide Y gene expression profiles. Society for Neuroscience Abstracts, Meredith, G. E.; Wouterlood, F. G. Hippocampal and midline thalamic fibers and terminals in relation to the choline acetyltransferase-immunoreactive neurons in nucleus accumbens of the rat: a light and electron microscopic study. J Comp Neurol, (2): Dube, L.; Smith, A. D.; Bolam, J. P. Identification of synaptic terminals of thalamic or cortical origin in contact with distinct medium-size spiny neurons in the rat neostriatum. J Comp Neurol, (4): Lapper, S. R.; Bolam, J. P. Input from the frontal cortex and the parafascicular nucleus to cholinergic interneurons in the dorsal striatum of the rat. Neuroscience, (3): Consolo, S.; Baronio, P.; Guidi, G.; Di Chiara, G. Role of the parafascicular thalamic nucleus and N-methyl-D-aspartate transmission in the D1-dependent control of in vivo acetylcholine release in rat striatum. Neuroscience, (1): Sweet, D. C.; Levine, A. S.; Kotz, C. M. Functional opioid pathways are necessary for hypocretin-1 (orexin-a)-induced feeding. Peptides, (2): Horvitz, J. C. Dopamine gating of glutamatergic sensorimotor and incentive motivational input signals to the striatum. Behav Brain Res, (1-2): Wise, R. A. Dopamine, learning and motivation. Nat Rev Neurosci, (6):

61 Salamone, J. D.; Correa, M. Motivational views of reinforcement: implications for understanding the behavioral functions of nucleus accumbens dopamine. Behav Brain Res, (1-2): Beninger, R. J.; Miller, R. Dopamine D1-like receptors and reward-related incentive learning. Neurosci Biobehav Rev, (2): Berridge, K. C.; Robinson, T. E. What is the role of dopamine in reward: hedonic impact, reward learning, or incentive salience? Brain Res Brain Res Rev, (3): Wise, R. A. Neuroleptic attenuation of intracranial self-stimulation: reward or performance deficits? Life Sci, (7): Wise, R. A.; Spindler, J.; dewit, H.; Gerberg, G. J. Neuroleptic-induced "anhedonia" in rats: pimozide blocks reward quality of food. Science, (4352): Wise, R. A. Neuroleptics and operant behavior: the anhedonia hypothesis. Behav Brain Sci, : Salamone, J. D.; Cousins, M. S.; Bucher, S. Anhedonia or anergia? Effects of haloperidol and nucleus accumbens dopamine depletion on instrumental response selection in a T- maze cost/benefit procedure. Behav Brain Res, (2): White, N. M. Reward or reinforcement: what's the difference? Neurosci Biobehav Rev, (2-3): Ettenberg, A.; Camp, C. H. Haloperidol induces a partial reinforcement extinction effect in rats: implications for a dopamine involvement in food reward. Pharmacol Biochem Behav, (4):

62 Bakshi, V. P.; Kelley, A. E. Dopaminergic regulation of feeding behavior: II. Differential effects of amphetamine microinfusion into three striatal subregions. Psychobiology, (3): Berridge, K. C.; Venier, I. L.; Robinson, T. E. Taste reactivity analysis of 6- hydroxydopamine-induced aphagia: implications for arousal and anhedonia hypotheses of dopamine function. Behav Neurosci, (1): Blackburn, J. R.; Phillips, A. G.; Fibiger, H. C. Dopamine and preparatory behavior: I. Effects of pimozide. Behav Neurosci, (3): Hajnal, A.; Smith, G. P.; Norgren, R. Oral sucrose stimulation increases accumbens dopamine in the rat. Am J Physiol Regul Integr Comp Physiol, (1): R Kornetsky, C.; Duvauchelle, C. Dopamine, a common substrate for the rewarding effects of brain stimulation reward, cocaine, and morphine. NIDA Res Monogr, : Hoebel, B. G.; Monaco, A. P.; Hernandez, L.; Aulisi, E. F.; Stanley, B. G.; Lenard, L. Self-injection of amphetamine directly into the brain. Psychopharmacology (Berl), (2): Koob, G. F.; Riley, S. J.; Smith, S. C.; Robbins, T. W. Effects of 6-hydroxydopamine lesions of the nucleus accumbens septi and olfactory tubercle on feeding, locomotor activity, and amphetamine anorexia in the rat. J Comp Physiol Psychol, (5): Baldo, B. A.; Sadeghian, K.; Basso, A. M.; Kelley, A. E. Effects of selective dopamine D1 or D2 receptor blockade within nucleus accumbens subregions on ingestive behavior and associated motor activity. Behav Brain Res, (1-2):

63 Blackburn, J. R.; Phillips, A. G.; Fibiger, H. C. Dopamine and preparatory behavior: III. Effects of metoclopramide and thioridazine. Behav Neurosci, (4): Blackburn, J. R.; Phillips, A. G.; Jakubovic, A.; Fibiger, H. C. Dopamine and preparatory behavior: II. A neurochemical analysis. Behav Neurosci, (1): Kiyatkin, E. A.; Gratton, A. Electrochemical monitoring of extracellular dopamine in nucleus accumbens of rats lever-pressing for food. Brain Res, (2): Phillips, A. G.; Atkinson, L. J.; Blackburn, J. R.; Blaha, C. D. Increased extracellular dopamine in the nucleus accumbens of the rat elicited by a conditional stimulus for food: an electrochemical study. Can J Physiol Pharmacol, (5-6): Roitman, M. F.; Stuber, G. D.; Phillips, P. E.; Wightman, R. M.; Carelli, R. M. Dopamine operates as a subsecond modulator of food seeking. J Neurosci, (6): Radhakishun, F. S.; van Ree, J. M.; Westerink, B. H. Scheduled eating increases dopamine release in the nucleus accumbens of food-deprived rats as assessed with on-line brain dialysis. Neurosci Lett, (3): Hernandez, L.; Hoebel, B. G. Feeding and hypothalamic stimulation increase dopamine turnover in the accumbens. Physiol Behav, (4-5): Wilson, C.; Nomikos, G. G.; Collu, M.; Fibiger, H. C. Dopaminergic correlates of motivated behavior: importance of drive. J Neurosci, (7 Pt 2): Bassareo, V.; Di Chiara, G. Differential responsiveness of dopamine transmission to food-stimuli in nucleus accumbens shell/core compartments. Neuroscience, (3):

64 Bassareo, V.; De Luca, M. A.; Di Chiara, G. Differential Expression of Motivational Stimulus Properties by Dopamine in Nucleus Accumbens Shell versus Core and Prefrontal Cortex. J Neurosci, (11): Salamone, J. D.; Arizzi, M. N.; Sandoval, M. D.; Cervone, K. M.; Aberman, J. E. Dopamine antagonists alter response allocation but do not suppress appetite for food in rats: contrast between the effects of SKF 83566, raclopride, and fenfluramine on a concurrent choice task. Psychopharmacology (Berl), (4): Cousins, M. S.; Atherton, A.; Turner, L.; Salamone, J. D. Nucleus accumbens dopamine depletions alter relative response allocation in a T-maze cost/benefit task. Behav Brain Res, (1-2): Aberman, J. E.; Salamone, J. D. Nucleus accumbens dopamine depletions make rats more sensitive to high ratio requirements but do not impair primary food reinforcement. Neuroscience, (2): Grill, H. J.; Norgren, R. The taste reactivity test. I. Mimetic responses to gustatory stimuli in neurologically normal rats. Brain Res, (2): Grill, H. J.; Norgren, R. The taste reactivity test. II. Mimetic responses to gustatory stimuli in chronic thalamic and chronic decerebrate rats. Brain Res, (2): Bakshi, V. P.; Kelley, A. E. Dopaminergic regulation of feeding behavior: I. Differential effects of haloperidol infusion into three striatal subregions. Psychobiology, (3): Kelley, A. E.; Delfs, J. M. Dopamine and conditioned reinforcement. I. Differential effects of amphetamine microinjections into striatal subregions. Psychopharmacology (Berl), (2):

65 Cador, M.; Taylor, J. R.; Robbins, T. W. Potentiation of the effects of reward-related stimuli by dopaminergic-dependent mechanisms in the nucleus accumbens. Psychopharmacology (Berl), (3): Berridge, K. C. Espresso reward learning, hold the dopamine: theoretical comment on Robinson et al. (2005). Behav Neurosci, (1): Robinson, S.; Sandstrom, S. M.; Denenberg, V. H.; Palmiter, R. D. Distinguishing whether dopamine regulates liking, wanting, and/or learning about rewards. Behav Neurosci, (1): Schultz, W.; Dayan, P.; Montague, P. R. A neural substrate of prediction and reward. Science, (5306): Tobler, P. N.; Fiorillo, C. D.; Schultz, W. Adaptive coding of reward value by dopamine neurons. Science, (5715): Ljungberg, T.; Apicella, P.; Schultz, W. Responses of monkey dopamine neurons during learning of behavioral reactions. J Neurophysiol, (1): Di Chiara, G. Nucleus accumbens shell and core dopamine: differential role in behavior and addiction. Behav Brain Res, (1-2): Cepeda, C.; Buchwald, N. A.; Levine, M. S. Neuromodulatory actions of dopamine in the neostriatum are dependent upon the excitatory amino acid receptor subtypes activated. Proc Natl Acad Sci U S A, (20): Cepeda, C.; Levine, M. S. Dopamine and N-methyl-D-aspartate receptor interactions in the neostriatum. Dev Neurosci, (1): 1-18.

66 Floresco, S. B.; Blaha, C. D.; Yang, C. R.; Phillips, A. G. Dopamine D1 and NMDA receptors mediate potentiation of basolateral amygdala-evoked firing of nucleus accumbens neurons. J Neurosci, (16): Thomas, M. J.; Malenka, R. C. Synaptic plasticity in the mesolimbic dopamine system. Philos Trans R Soc Lond B Biol Sci, (1432): Wickens, J. R.; Reynolds, J. N.; Hyland, B. I. Neural mechanisms of reward-related motor learning. Curr Opin Neurobiol, (6): Smith-Roe, S. L.; Kelley, A. E. Coincident activation of NMDA and dopamine D1 receptors within the nucleus accumbens core is required for appetitive instrumental learning. J Neurosci, (20): Hernandez, P. J.; Andrzejewski, M. E.; Sadeghian, K.; Panksepp, J. B.; Kelley, A. E. AMPA/kainate, NMDA, and dopamine D1 receptor function in the nucleus accumbens core: a context-limited role in the encoding and consolidation of instrumental memory. Learn Mem, (3): Dalley, J. W.; Laane, K.; Theobald, D. E.; Armstrong, H. C.; Corlett, P. R.; Chudasama, Y.; Robbins, T. W. Time-limited modulation of appetitive Pavlovian memory by D1 and NMDA receptors in the nucleus accumbens. Proc Natl Acad Sci U S A, (17): Mogenson, G. J.; Jones, D. L.; Yim, C. Y. From motivation to action: functional interface between the limbic system and the motor system. Prog Neurobiol, (2-3):

67 67 Figure Captions Figure 1. A schematic view of brain circuitry involved in food intake and motivation. Pathways coded by glutamate as the main neurotransmitter are shown in blue, while dopamine pathways are shown in red. Tan lines arising from the lateral hypothalamus (LH) indicate widespread direct and indirect projections from hypothalamus to neocortex and forebrain limbic structures, as discussed in Swanson (2000). Arrows in black indicate projections that are coded by GABA (from striatal regions and the central nucleus of the amygdala), or that are unknown. Primary gustatory inputs arise from the nucleus of the solitary tract (NTS) and the parabrachial nucleus (PB), prior to converging on cortical and thalamic regions involved in affect, memory, and behavioral inhibition. Although not shown in this figure, the influences of striatal opioids and acetylcholine on food motivation arise from axon collaterals of the medium spiny output neurons of the striatum and acetylcholine-containing interneurons, respectively. Figure 2. A schematic view of hypothalamic circuitry involved in food intake. Circulating factors in the blood modulate the activity of energy sensing neurons in the arcuate nucleus, that serve to modulate food-directed behaviors via their activation of outputs from lateral hypothalamic regions to thalamo-cortical systems, central autonomic effectors, and motor pattern generators. Of particular interest is the convergence of inputs to the hypothalamus from the amygdala, prefrontal cortex, and nucleus accumbens shell. These inputs putatively allow direct modulation of feeding behaviors based upon cognitive and affective signaling. Figure 3. Feeding induced by pharmacological manipulations of amino acid signaling in the nucleus accumbens core.

68 68 Infusions into the Acb shell (A) of the AMPA/kainate receptor antagonist DNQX specifically increase the consumption of food, but not water, compared to vehicle infusions (B). This effect is largely limited to infusions within the anterior-posterior extent of the Acb shell; treatment of other striatal regions, including the adjacent Acb core, was without effect on food intake (C). Infusions of the GABA A agonist muscimol into the Acb shell also increased feeding (D), suggesting that inhibition of the Acb shell might promote consummatory behaviors. However, doses of muscimol that were effective in stimulating food intake did not increase the amount of work rats would exert to earn a food reward in a progressive ratio paradigm (E). It is suggested that Acb shell inhibition activates downstream motor pattern generators that induce food intake when food is freely available, but does not recruit forebrain cortical circuitry involved in organizing behavioral strategies toward the procurement of food. GABA A agonism, furthermore, increases the intake of caloric foods irrespective of their content. Consumption of carbohydrate- or fat- based diets is equivalently increased in rats treated with intra Acb shell muscimol treatment (F). Figure 4. Food intake resulting from GABA A agonism of the Acb depends upon neurotransmission within the lateral hypothalamus. Food intake elicited by Acb shell infusions of the AMPA/kainite receptor antagonist DNQX (A), or the GABA A agonist muscimol (B) was dose-dependently decreased by inhibition of the lateral hypothalamus. Furthermore, Fos expression in the lateral hypothalamus, as well as the arcuate nucleus of the hypothalamus, was dramatically increased following muscimol infusions into the Acb shell (compared C with D). Arc, arcuate nucleus; VL PeF, ventrolateral perifornical area. Figure 5. Opioid modulation of food intake.

69 69 Stimulation of mu opioid receptors with DAMGO favors the intake of highly palatable high-fat diets. Animals that have been food deprived for 24 hours feed upon both highcarbohydrate diet as well as fat diet, but preferentially shift their consumption to the high fat diet under the effects of nucleus accumbens mu receptor stimulation (A). Similarly, naltrexone antagonism of opioid receptors of the nucleus accumbens preferentially reduces the intake of fat under deprivation conditions (B). Unlike muscimol inactivation of the accumbens shell, mu receptor stimulation of the nucleus accumbens increases breakpoint within a progressive ratio paradigm, suggesting an important role for opioids within the striatum on the reinforcing properties of food reward. DAMGO treatment also increases intake of palatable, but non-caloric, saccharin and saline solutions, without affecting water intake (even in mildly water-deprived animals, D). Mu receptor stimulation increases Fos expression in a network of brain regions involved in the modulation of food intake, including the dorsomedial hypothalamic region (DMH) shown here (E). Finally, expression of preproenkephalin mrna varies as a function of motivational state. Both food deprived animals (shown here) and rats fed ad libitum show downregulated preproenkephalin levels following a meal; if food is withheld at lights out, preproenkephalin expression remain high. Figure 6. Muscarinic receptor antagonism of the nucleus accumbens reduces 24-hour food intake and striatal preproenkephalin mrna expression.

70 70 Figure 7. The effects of striatal dopaminergic manipulations on food intake and appetitive behaviors. Although striatal dopamine is involved in behavioral activation and increasing learned appetitive behaviors toward food, it does not appear to directly modulate food intake. Doses of SCH (a D1 antagonist) that markedly decreased locomotor activity (A) increased the length of feeding bouts in hungry rats (B). Thus, although treated rats reduced the total number of feeding bouts within a 30 min session compared to vehicle infusion (not shown), the amount of food consumed remained unaltered (C). Furthermore, dopamine agonism increases locomotor activity, lever pressing under a progressive ratio paradigm, and lever pressing to achieve conditioned reinforcement (D-F). However, amphetamine infusion into the Acb did not increase food intake, and in fact reduced it at the highest dose tested (F). The effects of Acb amphetamine infusions on conditioned reinforcement were not duplicated by treatments with DALA, an opioid agonist.

71 Figure 1 Striatal role in food motivation 71

72 Figure 2 Striatal role in food motivation 72

73 Figure 3 Striatal role in food motivation 73

74 Figure 4 Striatal role in food motivation 74

75 Figure 5 Striatal role in food motivation 75