Low Dopamine Striatal D2 Receptors are Associated With Prefrontal Metabolism in Obese Subjects: Possible Contributing Factors (2008)

PET imaging

PET scans were performed with a CTI-931 (Computer Technologies, Incorporated, Knoxville, Tenn.) tomograph (resolution 6×6×6.5 mm FWHM, 15 slices) with [¹¹C]raclopride and [¹⁸F]FDG. Details on procedures for positioning, arterial and venous catheterization, quantification of radiotracer and transmission and emission scans have been published for [¹¹C]raclopride (Volkow et al., 1993a), and for [¹⁸F]FDG (Wang et al., 1992). Briefly for [¹¹C]raclopride, dynamic scans were started immediately after iv injection of 4–10 mCi (specific activity>0.25 Ci/μmol at time of injection) for a total of 60 min. For [¹⁸F]FDG, one emission scan (20 min) was taken 35 min after an iv injection of 4–6 mCi of [¹⁸F]FDG. The scans were done on the same day; the [¹¹C]raclopride scan was done first and was followed by [¹⁸F]FDG, which was injected 2 h after [¹¹C]raclopride to allow for the decay of ¹¹C (half-life 20 min). During the study subjects were kept lying in the PET camera with their eyes open; the room was dimly lit and noise was kept at a minimum. A nurse remained with the subjects throughout the procedure to ensure that the subject did not fall asleep during the study.

Association between D2 receptors and prefrontal metabolism

The significant association between D2 receptors availability and metabolism in prefrontal regions is consistent with our prior findings in drug-addicted subjects (cocaine, methamphetamine and alcohol) in whom we showed that the reductions in D2 receptors were associated with decreased metabolism in prefrontal cortical regions (Volkow et al., 1993b; Volkow et al., 2001; Volkow et al., 2007).

Similarly in individuals at high familial risk for alcoholism we documented an association between D2 receptor availability and prefrontal metabolism (Volkow et al., 2006). Both obesity and addiction share in common the inability to restrain the behavior despite awareness of its negative effects. Inasmuch as prefrontal regions are implicated in various components of inhibitory control (Dalley et al., 2004) we postulate that the low D2 receptor availability in the striatum of obese subjects (Wang et al., 2001) and in rodent models of obesity (Hamdi et al., 1992; Huang et al., 2006; Thanos et al., 2008) may contribute to obesity in part via DA’s modulation of prefrontal regions that participate in inhibitory control.

The findings also suggest that dopaminergic regulation of prefrontal regions as it relates to the risk for obesity may be meditated through D2 receptors. This is consistent with genetic studies, which have specifically implicated the D2 receptor gene (TAQ-IA polymorphism), as one that is involved in vulnerability to obesity (Fang et al., 2005; Pohjalainen et al., 1998; Bowirrat and Oscar-Berman, 2005). Moreover, the TAQ-IA polymorphism, which appears to result in lower D2 receptor levels in brain (striatum) (Ritchie and Noble, 2003; Pohjalainen et al., 1998; Jonsson et al., 1999) was recently found to be associated with decreased ability to inhibit behaviors that result in negative consequences and with impaired activation of prefrontal regions (Klein et al., 2007). Similarly preclinical studies have shown that animals with low D2 receptor levels are more impulsive than their littermates with high D2 receptor levels (Dalley et al., 2007). Thus the findings from our study provide further evidence that the association of D2 receptors with inhibitory control and with impulsivity is mediated in part by their modulation of prefrontal regions. In this respect it is interesting to note that brain morphological studies have reported reduced gray matter volumes in prefrontal cortex in obese subjects when compared to lean individuals (Pannacciulli et al., 2006).

The association between D2 receptors and the DLPFC is particularly interesting since this region was recently implicated in endogenous inhibition of intentional action (Brass and Haggard, 2007). The evidence that neuronal activity precedes an individual’s conscious awareness of intention by 200–500 ms (Libet et al., 1983), has led some to question the concept of “free will” behind intentional actions and to propose that control reflects the ability to inhibit actions we do not want. Indeed, it was suggested that this veto power or “free won’t” may be the way we exert “free will” (Mirabella, 2007). In the case of obesity one could postulate that exposure to food or food conditioned cues will result in the non-volitional activation of neuronal systems involved in procuring and eating the food and that the control reflects the ability to inhibit these intentional actions to want to eat the food. One can conceive how improper function of DLPFC, which enables inhibition of actions that result in negative outcomes, such as eating when we are not hungry because we do not want to gain weight, could result in overeating. Imaging findings showing greater decreases in activation of the DLPFC after a meal in obese subjects than in lean individuals support this hypothesis (Le et al., 2006).

The association between D2 receptor availability and medial orbitofrontal cortex (OFC) and anterior CG is consistent with their involvement in appetite regulation (Pliquett et al., 2006). There are several ways that one can propose by which disrupted dopaminergic activation of the OFC and the CG could increase the risk for overeating. The medial OFC is involved with salience attribution including the value of food (Rolls and McCabe, 2007; Grabenhorst et al., 2007; Tremblay and Schultz, 1999) and thus its activation secondary to food-induced DA stimulation could result in an intense motivation to consume food with a concomitant inability to inhibit it. Moreover, because disruption in the activity of the OFC results in impairment in the reversal of learned associations when a reinforcer is devalued (Gallagher et al., 1999) this could result in continued eating when the value of food is devalued by satiety and could explain why damage of the OFC is associated with compulsive behaviors including overeating (Butter et al., 1963, Johnson, 1971). Also the OFC participates in learning stimulus-reinforcement associations and conditioning (Schoenbaum et al., 1998, Hugdahl et al., 1995) and could therefore participate in conditioned-cue elicited feeding (Weingarten, 1983). This is relevant because food-induced conditioned responses very likely contribute to overeating irrespective of hunger signals (Ogden and Wardle, 1990).

The dorsal CG (BA 32) is implicated in inhibitory control in situations that demand monitoring of activity and thus its disrupted activity along with that of DLPFC with which it interacts (Gehring and Knight 2000) is likely to further impair the ability of the obese individual to inhibit the tendency to overeat. The ventral CG (BA 25) is implicated in mediating the emotional responses to salient stimuli (rewarding as well as aversive) (Elliott et al., 2000) and imaging studies have shown that BA 25 is activated by natural and drug rewards (Breiter et al., 1997, Francis et al., 1999; Berns et al., 2001). Thus the negative association between D2 receptors and the tendency to eat when exposed to negative emotions we previously reported in healthy controls (Volkow et al., 2003) could be mediated by modulation of BA 25.

The association between metabolic activity in prefrontal regions and D2 receptors could reflect projections to the prefrontal cortex from ventral and dorsal striatum (Ray and Price, 1993), which are regions implicated in the reinforcing and motivational effects of food (Koob and Bloom, 1988) and/or from the ventral tegmental area (VTA) and substantia nigra (SN), which are the main DA projections to striatum (Oades and Halliday, 1987). However, the prefrontal cortex also sends projections to the striatum so the association could reflect the prefrontal regulation of DA striatal activity (Murase et al., 1993).

In non-obese controls the correlations between D2 receptor and prefrontal metabolism were not significant. In prior findings we had shown significant correlation between D2 receptor and prefrontal metabolism in addicted subjects with low D2 receptor availability but not in controls (Volkow et al., 2007). However, comparison of the correlations between the obese and the control groups were not significant, which suggests that it is unlikely that the association between D2 receptors and prefrontal metabolism is unique to obesity (or to addiction as per Volkow et al., 2007). It is more likely that that the stronger correlations seen in the obese individuals reflect the greater range of striatal D2 receptor measures in obese (Bmax/Kd range 2.1–3.7) than in control subjects (Bmax/Kd range 2.7–3.8).

In interpreting these findings it is also important to consider that [¹¹C]raclopride is a radiotracer whose binding to D2 receptors is sensitive to endogenous DA (Volkow et al., 1994) and thus the reductions of D2 receptor availability in obese subjects could reflect low receptor levels or increases in DA release. Preclinical studies in animal models of obesity have documented reduction in the concentration of D2 receptors (Thanos et al., 2008), which suggests that the reductions in obese subjects reflect decreases in D2 receptor levels.

Correlation between D2R and somatosensory cortex

We had not “a priori” hypothesized an association between D2 receptors and metabolism in somatosensory cortex. When compared with the frontal or temporal regions there is relatively little that is known about the influence of DA in the parietal cortex. In the human brain the concentration of D2 receptors and D2 mRNA in parietal cortex while much lower than in subcortical regions is equivalent to that reported in frontal cortex (Suhara et al., 1999; Mukherjee et al., 2002; Hurd et al., 2001). Though there is limited literature on the role of the somatosensory cortex in food intake and obesity. Imaging studies reported activation of the somatosensory cortex in normal weight subjects with exposure to visual images of low caloric foods (Killgore et al., 2003) and with satiety (Tataranni et al., 1999), and we had shown higher than normal baseline metabolism in the somatosensory cortex in obese subjects (Wang et al., 2002). Also a recent study reported that in obese individuals with leptin deficiency administration of leptin normalized their body weight and reduced brain activation in parietal cortex while viewing food-related stimuli (Baicy et al., 2007). The functional connectivity between the striatum and the somatosensory cortex was recently corroborated for the human brain by a meta-analysis study on 126 functional imaging studies, which documented co-activation of the somatosensory cortex with that of the dorsal striatum (Postuma and Dagher, 2006). However, from the correlations in our study we cannot ascertain the direction of the association; so we cannot determine if the association with D2 receptors reflects DA’s modulation of the somatosensory cortex and/or the influence of the somatosensory cortex on striatal D2 receptor availability. Indeed there is ample evidence that the somatosensory cortex influences brain DA activity including striatal DA release (Huttunen et al., 2003; Rossini et al., 1995; Chen et al., 2007). There is also evidence that DA modulates the somatosensory cortex in the human brain (Kuo et al., 2007). Inasmuch as DA stimulation signals saliency and facilitates conditioning (Zink et al., 2003, Kelley, 2004), DA’s modulation of the somatosensory cortex’s response to food is likely to play a role in the formation of conditioned association between food and food-related environmental cues and in the enhanced reinforcing value of food that occurs in obesity (Epstein et al., 2007).

Study limitations

A limitation for this study is that we did not obtain neuropsychological measures and thus we cannot assess if the activity in prefrontal regions is associated with behavioral measures of cognitive control in these obese subjects. Though neuropsychological studies on obesity are limited and the findings are confounded by the medical complications of obesity (i.e diabetes and hypertension), there is evidence that in obese subjects inhibitory control may be disrupted. Specifically, when compared with normal weight individuals, obese subjects make less advantageous choices, which is a finding consistent with impaired inhibitory control and with prefrontal dysfunction (Pignatti et al., 2006). Moreover rates of attention deficit hyperactivity disorder (ADHD), which involves disruption in impulsivity, are elevated in obese individuals (Altfas, 2002). Similarly impulsivity has been linked with high BMI in some populations (Fassino et al., 2003) and in healthy controls BMI has also been associated with performance in tasks of executive function that mediate impulsivity (Gunstad et al., 2007).

Also while in this paper we focus on the role that the prefrontal cortex has on inhibitory control and impulsivity we recognize that the prefrontal cortex is involved with a wide range of cognitive operations many of which are not disrupted in obese subjects (Kuo et al., 2006, Wolf et al., 2007). It is possible that the functions of the prefrontal cortex that contribute to obesity are the ones sensitive to DA modulation via striatal prefrontal pathways (Robbins, 2007; Zgaljardic et al., 2006).

Neither the dysregulation of prefrontal activity nor the impairment of executive function is specific for obesity. Indeed abnormalities in prefrontal metabolism and impairment in executive function have been documented in a wide range of disorders including those with dopaminergic involvement such as drug addiction, schizophrenia, Parkinson’s disease and ADHD (Volkow et al., 1993b; Gur et al., 2000; Robbins, 2007; Zgaljardic et al., 2006).

Another limitation was that the limited spatial resolution of the PET [₁₁C]raclopride method did not allow us to measure D2 receptor availability in small brain regions that are important in mediating food associated behaviors such as the hypothalamus.

Finally correlations do not imply causal associations and further studies are required to evaluate the consequences of disrupted DA brain activity in prefrontal function in obese subjects.

We thank David Schlyer, David Alexoff, Paul Vaska, Colleen Shea, Youwen Xu, Pauline Carter, Karen Apelskog, and Linda Thomas for their contributions. This research was supported by NIH’s Intramural Research Program (NIAAA) and by DOE (DE-AC01-76CH00016).