**2.2 Reward-related information processing in the nucleus accumbens**

The nucleus accumbens plays a key role in the dopaminergic reward pathway and is implicated in overeating, leading to obesity. Two hypotheses have been proposed as the mechanisms that implicate the dopaminergic reward pathway in overeating and subsequently obesity [11]. The first hypothesis proposes that in obese subjects, responses to dopamine release during food intake (reward responses) are reduced. In line with this hypothesis, previous positron emission tomography (PET) studies have reported that dopamine D2 receptor utilization was reduced in obese subjects, which suggests a decrease in dopamine D2 receptor density. Thus, obese individuals may consume more food that causes dopamine release to compensate for the reduced reward response. A similar hypothesis (decreased dopamine D2 receptor utilization) has also been proposed as a mechanism for substance abuse. Furthermore, an animal study reported that when a chronic electrode was implanted in a specific brain region such as the medial forebrain bundle and an electric current (rewarding) was applied by a lever press, the animal preferred to press the lever [intracranial self-stimulation (ICSS) behavior]. However, in support of this first hypothesis, injection of a dopamine D2 receptor antagonist into the nucleus accumbens suppressed ICSS behavior due to electrical stimulation of the medial forebrain bundle. The second hypothesis proposes that obese people are more sensitive to cue stimuli (such as food smell and visual appearance, or conditioned stimuli) that predict food availability, rather than food ingestion itself. A higher sensitivity to food cues lead to an increase in food intake, which in turn, gradually reduces the reward response when food is consumed. Thus, food intake is

**27**

motivation for rewarding objects.

*Neural Mechanisms of Feeding Behavior and Its Disorders*

that there are multiple underlying mechanisms for obesity.

further increased by the individual to compensate for the reduced reward response. In support of the second hypothesis, in obese individuals as well as obesity-prone rats, responses to cues associated with food are enhanced in certain brain regions, including the nucleus accumbens [12, 13]. The nucleus accumbens is suggested to play a crucial role in incentive motivation: a process that translates expected reward derived from cues into behavioral manifestation for food acquisition [14]. A functional magnetic resonance imaging (fMRI) study on alcohol-dependent patients reported that activity in the ventral striatum, including the nucleus accumbens, increased in response to visual cues associated with alcohol [15]. It must be noted that obese individuals might have already experienced high-reward foods such as high fat or high calorie foods. Therefore, it is difficult to clarify which hypothetical mechanism causes a decrease in the reward response [11]. However, a human genetic study reported that obese subjects had a genetic polymorphism, called *Taq1A*, with decreased dopamine D2 receptor density [16]. These findings suggest

A study analyzed neuronal responses in the ventral striatum, including the nucleus accumbens during discrimination of foods from nonfoods in a lever-press feeding task, in monkeys (unpublished data). This task consisted of three phases: (1) a visual recognition phase during which various objects including food and nonfood were presented to the monkey by opening an opaque shutter in front of an object, (2) a lever-press phase during which the monkey pressed the lever by a predetermined number of times if food was presented, and (3) an ingestion phase during which the monkey could take food after the last lever press opened a transparent shutter in front of the object [17–19]. In this task, white and red cylinders were associated with drops of juice and water, respectively. A predetermined number of lever presses opened a valve to deliver a drop of juice and water, respectively. When a brown cylinder was presented, the monkey had to press the lever a predetermined number of times to avoid electric shock. An example of a ventral striatum neuron in the monkey that selectively responded to rewarding objects has been shown in **Figure 3A**. This neuron responded to the rewarding objects, including orange and white cylinders associated with juice and water, but not to the brown cylinder and aversive syringe associated with electric shock. It should be noted that the responses to the rewarding objects were not related to simple lever-pressing movements, since the neuron did not respond to the brown cylinder even though the monkey pressed the lever. The response magnitudes of this neuron to various objects have been shown in **Figure 3B**. The results showed that this neuron was highly responsive to the monkeys' favorite (highly rewarding) objects. The existence of this type of neuron in the ventral striatum suggests that the ventral striatum is involved in incentive

Previous studies have suggested that dopamine release in the ventral striatum is involved in the motivation for appetitive behaviors. Tonic (slow) increases in dopamine levels were shown to be involved in motivation [20, 21]. A positive correlation between reward-seeking behavior and dopamine levels was also reported [21]. Another study analyzed the activity of medium spiny neurons (MSNs), the major output neurons in the nucleus accumbens that receive dopaminergic projections [22] in the rat nucleus accumbens [23]. In this study, water-deprived rats were trained with an operant conditioning task in which the licking of a spout was associated with intragastric glucose (glucose group) or water (water group) infusion. After training, it was observed that rats in the glucose group were more vigorous in licking the spout in the absence of intragastric infusion. The inter-spike interval variability of MSNs, which reflects dopamine release in the striatum, was higher in the glucose group than in the water group [23]. These findings suggest that dopamine release in the nucleus accumbens plays a crucial role in motivation.

*DOI: http://dx.doi.org/10.5772/intechopen.92723*

#### *Neural Mechanisms of Feeding Behavior and Its Disorders DOI: http://dx.doi.org/10.5772/intechopen.92723*

*New Insights into Metabolic Syndrome*

and smell to the LHA (see below).

**Figure 2.**

prefrontal (orbital) cortex, which is a higher association area of olfaction and taste, sends not only cognitive information but also food information such as taste

*Interaction between the hypothalamic feeding control system and the emotion/reward system in the LHA. F: Fornix; LHA: Lateral hypothalamic area; V3: Third ventricle; PFC: Prefrontal cortex; DA: Dopamine; Acc:* 

The nucleus accumbens plays a key role in the dopaminergic reward pathway and is implicated in overeating, leading to obesity. Two hypotheses have been proposed as the mechanisms that implicate the dopaminergic reward pathway in overeating and subsequently obesity [11]. The first hypothesis proposes that in obese subjects, responses to dopamine release during food intake (reward responses) are reduced. In line with this hypothesis, previous positron emission tomography (PET) studies have reported that dopamine D2 receptor utilization was reduced in obese subjects, which suggests a decrease in dopamine D2 receptor density. Thus, obese individuals may consume more food that causes dopamine release to compensate for the reduced reward response. A similar hypothesis (decreased dopamine D2 receptor utilization) has also been proposed as a mechanism for substance abuse. Furthermore, an animal study reported that when a chronic electrode was implanted in a specific brain region such as the medial forebrain bundle and an electric current (rewarding) was applied by a lever press, the animal preferred to press the lever [intracranial self-stimulation (ICSS) behavior]. However, in support of this first hypothesis, injection of a dopamine D2 receptor antagonist into the nucleus accumbens suppressed ICSS behavior due to electrical stimulation of the medial forebrain bundle. The second hypothesis proposes that obese people are more sensitive to cue stimuli (such as food smell and visual appearance, or conditioned stimuli) that predict food availability, rather than food ingestion itself. A higher sensitivity to food cues lead to an increase in food intake, which in turn, gradually reduces the reward response when food is consumed. Thus, food intake is

**2.2 Reward-related information processing in the nucleus accumbens**

*Nucleus accumbens; VTA: Ventral tegmental area; PFC: Prefrontal cortex.*

**26**

further increased by the individual to compensate for the reduced reward response. In support of the second hypothesis, in obese individuals as well as obesity-prone rats, responses to cues associated with food are enhanced in certain brain regions, including the nucleus accumbens [12, 13]. The nucleus accumbens is suggested to play a crucial role in incentive motivation: a process that translates expected reward derived from cues into behavioral manifestation for food acquisition [14]. A functional magnetic resonance imaging (fMRI) study on alcohol-dependent patients reported that activity in the ventral striatum, including the nucleus accumbens, increased in response to visual cues associated with alcohol [15]. It must be noted that obese individuals might have already experienced high-reward foods such as high fat or high calorie foods. Therefore, it is difficult to clarify which hypothetical mechanism causes a decrease in the reward response [11]. However, a human genetic study reported that obese subjects had a genetic polymorphism, called *Taq1A*, with decreased dopamine D2 receptor density [16]. These findings suggest that there are multiple underlying mechanisms for obesity.

A study analyzed neuronal responses in the ventral striatum, including the nucleus accumbens during discrimination of foods from nonfoods in a lever-press feeding task, in monkeys (unpublished data). This task consisted of three phases: (1) a visual recognition phase during which various objects including food and nonfood were presented to the monkey by opening an opaque shutter in front of an object, (2) a lever-press phase during which the monkey pressed the lever by a predetermined number of times if food was presented, and (3) an ingestion phase during which the monkey could take food after the last lever press opened a transparent shutter in front of the object [17–19]. In this task, white and red cylinders were associated with drops of juice and water, respectively. A predetermined number of lever presses opened a valve to deliver a drop of juice and water, respectively. When a brown cylinder was presented, the monkey had to press the lever a predetermined number of times to avoid electric shock. An example of a ventral striatum neuron in the monkey that selectively responded to rewarding objects has been shown in **Figure 3A**. This neuron responded to the rewarding objects, including orange and white cylinders associated with juice and water, but not to the brown cylinder and aversive syringe associated with electric shock. It should be noted that the responses to the rewarding objects were not related to simple lever-pressing movements, since the neuron did not respond to the brown cylinder even though the monkey pressed the lever. The response magnitudes of this neuron to various objects have been shown in **Figure 3B**. The results showed that this neuron was highly responsive to the monkeys' favorite (highly rewarding) objects. The existence of this type of neuron in the ventral striatum suggests that the ventral striatum is involved in incentive motivation for rewarding objects.

Previous studies have suggested that dopamine release in the ventral striatum is involved in the motivation for appetitive behaviors. Tonic (slow) increases in dopamine levels were shown to be involved in motivation [20, 21]. A positive correlation between reward-seeking behavior and dopamine levels was also reported [21]. Another study analyzed the activity of medium spiny neurons (MSNs), the major output neurons in the nucleus accumbens that receive dopaminergic projections [22] in the rat nucleus accumbens [23]. In this study, water-deprived rats were trained with an operant conditioning task in which the licking of a spout was associated with intragastric glucose (glucose group) or water (water group) infusion. After training, it was observed that rats in the glucose group were more vigorous in licking the spout in the absence of intragastric infusion. The inter-spike interval variability of MSNs, which reflects dopamine release in the striatum, was higher in the glucose group than in the water group [23]. These findings suggest that dopamine release in the nucleus accumbens plays a crucial role in motivation.

#### **Figure 3.**

*Response of a monkey ventral striatal neuron to various objects. (A) Responses of a ventral striatal neuron to various objects. Upper histogram: Summed histogram of neuronal activity (bin width, 100 ms bin); lower histogram: Summed histograms of lever-press signals. Zero on time axis: Onset of object presentation. Calibration shown at the right of each histogram (4 spikes/bin). FR, number of lever presses; N, number of trials. (B) Average response magnitudes for various objects. Closed circle, response magnitude in each trial.*

However, the actual roles of dopamine and other neurotransmitters in motivation (wanting) and hedonic pleasure (liking) is still under debate [24–26].

#### **2.3 Food valuation in the amygdala**

Food is a rewarding object that induces pleasant emotions. The amygdala is involved in evaluating the biological (motivational) values of objects such as food. The activity of amygdala neurons has been reported to correlate with the biological value of sensory stimuli in monkeys as well as humans [17–19, 27, 28]. Therefore, it can be said that humans select an object with a high reward value based on value evaluation in the amygdala [27]. Thus, when the reward value of food decreases (devaluation), approaching behavior to the conditioned stimulus associated with that food also decreases. For example, the injection of lithium after eating causes discomfort, which reduces the reward value of the food to induce appetitive

**29**

significance of objects.

*Neural Mechanisms of Feeding Behavior and Its Disorders*

behaviors. However, it has been reported that basolateral amygdala lesions in rats abolished these changes in behaviors after devaluation [29]. Similarly, in another study, monkeys were trained to form an association between two pairs of specific objects with specific foods. They were then allowed to eat only one of the foods to satisfaction (i.e., devaluation of one of the foods) after which it was observed that the monkeys chose the specific object associated with the other food (food-specific satiety) when given the option. However, lesions to the amygdala or surgical disconnection of neural fibers between the amygdala and orbital cortex reduced such behavioral changes associated with devaluation [30, 31]. Also, in a human fMRI study, devaluation of a specific food after eating it reduced brain hemodynamic responses to the odor of that specific food in the amygdala and orbital cortex [32]. These findings suggest that the amygdala receives information about the internal state of the body from the hypothalamus and evaluates the expected reward. Since stimulation of the amygdala increases dopamine release in the nucleus accumbens [33], value information from the amygdala may modulate dopamine release in the nucleus accumbens. This further suggests that an interaction between the amygdala and the nucleus accumbens is crucial in behavioral alterations in devaluation or

Some previous studies have shown that recordings from the amygdala neurons of monkeys were made during performance of the same lever-press feeding task (as shown in **Figure 3**) [17–19]. About one-fourth of the recorded amygdala neurons responded differentially to various rewarding and aversive objects with biological value (differential neurons). An example of such differential neurons has been shown in **Figure 4A**. These neurons responded strongly to orange, which was highly preferred (**Figure 4Aa**) in the visual discrimination and ingestion phases, but had a weak response to raisin, which was less preferred (**Figure 4Af**). The neuron had no response to tape, which had no biological value (**Figure 4Ag**). The neurons also responded to an aversive spider model (**Figure 4Ad**) and a brown column associated with electric shock (**Figure 4Ae**). Furthermore, this neuron responded more strongly to the preferred white cylinder associated with juice compared with the less preferred red cylinder associated with water, in the visual discrimination and ingestion phases (**Figure 4Ab, c**). These results suggest that the activity of the amygdala neuron reflects the biological values of objects. Similar types of amygdala neurons have been reported in a human study [27]. In this study, recordings from the amygdala neurons of a patient were made, while the patient rated specific foods displayed on a monitor by bidding. It was observed that the activities of some amygdala neurons positively correlated with food evaluation (bid price).

In the same studies done by Nishijo et al. [17–19], the monkeys drank 80 mL of water after the initial recording, which reduced the values of the red cylinder. The same neuron in **Figure 4A** was again tested with the white and red cylinders. **Figure4B** shows altered neuronal responses in the amygdala after drinking. Although the neuron responded similarly to the white cylinder in the visual discrimination and ingestion phases, its responses to the red cylinder were attenuated (**Figure4B**). These findings suggest that neuronal activity in the amygdala changes online based on the biological

Delicious food is highly rewarding. Both humans and animals consume foods with high reward values. Several factors can affect the expected reward value of a specific food when cues associated with that food are presented. Several factors can also affect food reward when food is ingested. Factors that can affect the expected reward value and food reward includes the following: (1) sensory factors derived

**2.4 Comprehensive food valuation in the frontal orbital cortex**

*DOI: http://dx.doi.org/10.5772/intechopen.92723*

food-specific satiety.

#### *Neural Mechanisms of Feeding Behavior and Its Disorders DOI: http://dx.doi.org/10.5772/intechopen.92723*

*New Insights into Metabolic Syndrome*

**28**

**Figure 3.**

However, the actual roles of dopamine and other neurotransmitters in motivation

*Response of a monkey ventral striatal neuron to various objects. (A) Responses of a ventral striatal neuron to various objects. Upper histogram: Summed histogram of neuronal activity (bin width, 100 ms bin); lower histogram: Summed histograms of lever-press signals. Zero on time axis: Onset of object presentation. Calibration shown at the right of each histogram (4 spikes/bin). FR, number of lever presses; N, number of trials. (B) Average response magnitudes for various objects. Closed circle, response magnitude in each trial.*

Food is a rewarding object that induces pleasant emotions. The amygdala is involved in evaluating the biological (motivational) values of objects such as food. The activity of amygdala neurons has been reported to correlate with the biological value of sensory stimuli in monkeys as well as humans [17–19, 27, 28]. Therefore, it can be said that humans select an object with a high reward value based on value evaluation in the amygdala [27]. Thus, when the reward value of food decreases (devaluation), approaching behavior to the conditioned stimulus associated with that food also decreases. For example, the injection of lithium after eating causes discomfort, which reduces the reward value of the food to induce appetitive

(wanting) and hedonic pleasure (liking) is still under debate [24–26].

**2.3 Food valuation in the amygdala**

behaviors. However, it has been reported that basolateral amygdala lesions in rats abolished these changes in behaviors after devaluation [29]. Similarly, in another study, monkeys were trained to form an association between two pairs of specific objects with specific foods. They were then allowed to eat only one of the foods to satisfaction (i.e., devaluation of one of the foods) after which it was observed that the monkeys chose the specific object associated with the other food (food-specific satiety) when given the option. However, lesions to the amygdala or surgical disconnection of neural fibers between the amygdala and orbital cortex reduced such behavioral changes associated with devaluation [30, 31]. Also, in a human fMRI study, devaluation of a specific food after eating it reduced brain hemodynamic responses to the odor of that specific food in the amygdala and orbital cortex [32]. These findings suggest that the amygdala receives information about the internal state of the body from the hypothalamus and evaluates the expected reward. Since stimulation of the amygdala increases dopamine release in the nucleus accumbens [33], value information from the amygdala may modulate dopamine release in the nucleus accumbens. This further suggests that an interaction between the amygdala and the nucleus accumbens is crucial in behavioral alterations in devaluation or food-specific satiety.

Some previous studies have shown that recordings from the amygdala neurons of monkeys were made during performance of the same lever-press feeding task (as shown in **Figure 3**) [17–19]. About one-fourth of the recorded amygdala neurons responded differentially to various rewarding and aversive objects with biological value (differential neurons). An example of such differential neurons has been shown in **Figure 4A**. These neurons responded strongly to orange, which was highly preferred (**Figure 4Aa**) in the visual discrimination and ingestion phases, but had a weak response to raisin, which was less preferred (**Figure 4Af**). The neuron had no response to tape, which had no biological value (**Figure 4Ag**). The neurons also responded to an aversive spider model (**Figure 4Ad**) and a brown column associated with electric shock (**Figure 4Ae**). Furthermore, this neuron responded more strongly to the preferred white cylinder associated with juice compared with the less preferred red cylinder associated with water, in the visual discrimination and ingestion phases (**Figure 4Ab, c**). These results suggest that the activity of the amygdala neuron reflects the biological values of objects. Similar types of amygdala neurons have been reported in a human study [27]. In this study, recordings from the amygdala neurons of a patient were made, while the patient rated specific foods displayed on a monitor by bidding. It was observed that the activities of some amygdala neurons positively correlated with food evaluation (bid price).

In the same studies done by Nishijo et al. [17–19], the monkeys drank 80 mL of water after the initial recording, which reduced the values of the red cylinder. The same neuron in **Figure 4A** was again tested with the white and red cylinders. **Figure4B** shows altered neuronal responses in the amygdala after drinking. Although the neuron responded similarly to the white cylinder in the visual discrimination and ingestion phases, its responses to the red cylinder were attenuated (**Figure4B**). These findings suggest that neuronal activity in the amygdala changes online based on the biological significance of objects.

## **2.4 Comprehensive food valuation in the frontal orbital cortex**

Delicious food is highly rewarding. Both humans and animals consume foods with high reward values. Several factors can affect the expected reward value of a specific food when cues associated with that food are presented. Several factors can also affect food reward when food is ingested. Factors that can affect the expected reward value and food reward includes the following: (1) sensory factors derived

#### **Figure 4.**

*Changes in responsiveness of amygdala neurons in a monkey before and after drinking. (A) Responses of an amygdala neuron to various objects with and without biological significance. (B) Changes in neuronal responses to the white and red cylinders after drinking 80 mL of water. Upper and lower histograms indicate summed histograms of neuronal activity and lever-press signals (bin width, 200 ms), respectively. Calibration is shown on the right of each histogram (8 spikes/bin). FR, number of lever presses; N, number of trials. A zero on the time scale indicates stimulus onset.*

from foods such as taste and smell, (2) internal states such as hunger and satiety, (3) previous experience of ingesting foods, (4) cognitive factors, and (5) temporal factors such as immediate availability of foods [11, 34]. The orbital cortex receives multisensory inputs, such as visual and auditory inputs, and also functions as a secondary taste and olfactory area. This brain region also receives inputs from the somatosensory cortex and is involved in processing food texture in the oral cavity. Thus, the orbital cortex receives all food-related sensory inputs and is involved in the valuation of food reward. Second, the orbital cortex receives information on the internal states and other visceral information from the hypothalamus and amygdala and is involved in food-specific satiety (see above). It has been shown in monkeys that lesions of the orbital cortex as well as the amygdala impair

**31**

**Figure 5.**

*water in a two-dimensional space.*

*Neural Mechanisms of Feeding Behavior and Its Disorders*

food-specific satiety [35]. In humans, frontotemporal dementia with atrophy of the ventral frontal lobe leads to overeating even with awareness of satiety [36, 37]. Third, knowledge about specific foods or previous experience of eating certain foods may affect the reward values of it [34, 38]. Humans simulate the expected reward of foods using this cognitive information. For example, a famous food brand or food at a famous restaurant may have a high expected reward value. Fourth, the expected reward value of food decreases as the time for acquiring food (or food reward) increases (temporal discounting). For example, healthy foods that lead to longevity and slimming seem to be highly rewarding, but it takes a long time for the effects to be observed. Due to temporal discounting, the expected reward value of heathy foods may decrease. In contrast, junk food, which can provide an immediate reward when ingested, has a higher reward value than healthy

*Representation of taste solutions in the orbital cortex of rats. (A) An example of a response profile of a taste neuron to four basic and umami tastes. MSG, 0.1 M monosodium glutamate; GMP, 0.5 mM guanylic acid. Filled bars indicate significant taste responses (2.5 SD above or below the water response). The top bars indicate excitatory responses, and the down bars indicate inhibitory responses. (B) Distribution of taste solutions and* 

*DOI: http://dx.doi.org/10.5772/intechopen.92723*

*Neural Mechanisms of Feeding Behavior and Its Disorders DOI: http://dx.doi.org/10.5772/intechopen.92723*

*New Insights into Metabolic Syndrome*

**30**

**Figure 4.**

*on the time scale indicates stimulus onset.*

from foods such as taste and smell, (2) internal states such as hunger and satiety, (3) previous experience of ingesting foods, (4) cognitive factors, and (5) temporal factors such as immediate availability of foods [11, 34]. The orbital cortex receives multisensory inputs, such as visual and auditory inputs, and also functions as a secondary taste and olfactory area. This brain region also receives inputs from the somatosensory cortex and is involved in processing food texture in the oral cavity. Thus, the orbital cortex receives all food-related sensory inputs and is involved in the valuation of food reward. Second, the orbital cortex receives information on the internal states and other visceral information from the hypothalamus and amygdala and is involved in food-specific satiety (see above). It has been shown in monkeys that lesions of the orbital cortex as well as the amygdala impair

*Changes in responsiveness of amygdala neurons in a monkey before and after drinking. (A) Responses of an amygdala neuron to various objects with and without biological significance. (B) Changes in neuronal responses to the white and red cylinders after drinking 80 mL of water. Upper and lower histograms indicate summed histograms of neuronal activity and lever-press signals (bin width, 200 ms), respectively. Calibration is shown on the right of each histogram (8 spikes/bin). FR, number of lever presses; N, number of trials. A zero*  food-specific satiety [35]. In humans, frontotemporal dementia with atrophy of the ventral frontal lobe leads to overeating even with awareness of satiety [36, 37]. Third, knowledge about specific foods or previous experience of eating certain foods may affect the reward values of it [34, 38]. Humans simulate the expected reward of foods using this cognitive information. For example, a famous food brand or food at a famous restaurant may have a high expected reward value. Fourth, the expected reward value of food decreases as the time for acquiring food (or food reward) increases (temporal discounting). For example, healthy foods that lead to longevity and slimming seem to be highly rewarding, but it takes a long time for the effects to be observed. Due to temporal discounting, the expected reward value of heathy foods may decrease. In contrast, junk food, which can provide an immediate reward when ingested, has a higher reward value than healthy

#### **Figure 5.**

*Representation of taste solutions in the orbital cortex of rats. (A) An example of a response profile of a taste neuron to four basic and umami tastes. MSG, 0.1 M monosodium glutamate; GMP, 0.5 mM guanylic acid. Filled bars indicate significant taste responses (2.5 SD above or below the water response). The top bars indicate excitatory responses, and the down bars indicate inhibitory responses. (B) Distribution of taste solutions and water in a two-dimensional space.*

foods. The orbital cortex is suggested to comprehensively evaluate and integrate this food-related multiple sensory, visceral, and cognitive information, and is involved in the final decision of food selection based on these factors [34, 39].

Human fMRI studies have reported that activity in the orbital cortex correlates with the subjective pleasantness of liquid foods, including umami solutions [40, 41]. To investigate the representation of taste solutions in the orbital cortex, a study analyzed neuronal responses to various taste solutions in the rat orbital cortex after infusing the solutions into the oral cavity of awake rats through implanted chronic intraoral cannulae (unpublished data). **Figure 5A** shows an example of the neuronal activity recorded from the rat orbital cortex during ingestion of various taste solutions including monosodium glutamate (MSG). This neuron responded to 0.1 M MSG, but not to 0.1 M sodium chloride (NaCl). The neuron had an inhibitory response to sucrose solution. **Figure 5B** shows a representation of these taste solutions resulting from a multidimensional scaling analysis of response patterns of 21 orbital cortical taste neurons. Each taste solution is arranged almost in a straight line from the left, where the most aversive quinine and citric acid are located. On the right side, there are rewarding solutions including sucrose, GMP (0.5 mM guanylic acid, a kind of umami), and MSG + GMP (mixture of MSG and GMP). In addition, the MSG and water are located in the center of the space. These orders are fairly consistent with the orders of the rats' preference for taste solutions. These results suggest that the rat orbital cortical neurons represent reward values based on sensory information derived from food.
