ABSTRACT
Dopamine transporter (DAT) mediates reuptake of extracellular dopamine into presynaptic neurons. We investigated the effects of glucose loading on striatal DAT in healthy young adults who underwent 18F-FP-CIT positron emission tomography (PET) scans, and sweet taste questionnaire (STQ). Thirty-five healthy subjects were enrolled in this study. Each subject visited the institution three times, for three brain PET scans (two 18F-FP-CIT PET scans after the infusion of glucose or placebo and one 18F-Fluorodeoxyglucose PET scan). All subjects underwent STQ, 12-item self-reporting to evaluate subjects’ reactions to eating sweets, craving for sweets and degree of control over eating sweet foods (STQ 1: sensitivity to mood altering effect of sweets, and STQ 2: impaired control over eating sweet foods). We created Bayesian models separately with STQ 1, and STQ 2 as predictors, with DAT availability and brain glucose uptake as a dependent variable. From caudate, glucose-loaded DAT availability was significantly higher than placebo-loaded DAT availability, and from putamen, glucose-loaded DAT availability showed the higher trend than placebo-loaded DAT availability. STQ was positively associated with glucose-loaded DAT availability. The effect of STQ markedly overlapped with zero on placebo-loaded DAT availability, and brain glucose uptake. In conclusion, the change of striatal DAT availability after glucose loading is associated with the vulnerability to sweet foods. This may indicate that individuals with higher DAT availability after glucose loading experience a rapid clearance of synaptic dopamine after consuming sweet foods, potentially leading to a desire for additional sweet foods.
- Dopamine transporter
- Striatum
- Sweet foods
- PET
INTRODUCTION
Obesity has nearly tripled worldwide since 1975 and has become one of the major public health threats. Obesity is a risk factor for malignancies of the colon [1], pancreas [2], thyroid [3], liver [4], and uterus [5] as well as for cardiovascular disease and diabetes mellitus [6].
The brain plays a critical role in controlling energy balance between energy intake and expenditure [7]. Food is a potent natural reinforcer, and fasting can further enhance its rewarding effects [8]. Dopamine is one of the neurotransmitters involved with eating behavior through modulation of the rewarding properties of food and the motivation and desire for food consumption [9]. Currently, there is no direct method to measure the synaptic dopamine concentration in the human brain. Therefore, positron emission tomography (PET) with dopamine receptor (DR) radiopharmaceuticals has been adopted to understand the role of dopaminergic system in obesity and eating behaviour. However, it is difficult to interpret the relationships between DR availability measured from PET scans and the synaptic dopamine level, which competes with DR radiopharmaceutical to bind DR [10,11]. In addition, brain DR availability is not different between lean and overweight/obese subjects from a recent meta-analysis [12].
Until recently, dopamine transporter (DAT), mediating reuptake of extracellular dopamine into presynaptic neurons [13], has not received as much attention in the area of eating behaviour, as DAT availability from brain PET scan was not correlated with body mass index (BMI) in previous studies [14-16]. However, the change of striatal DAT after glucose loading was reported by our group for the first time in humans [17]. Insulin has been shown to contribute to striatal DAT function by increasing DAT recruitment to the plasma membrane, and enhancing dopamine clearance via phosphatidylinositol 3-kinase signaling pathway, and downstream activation of Akt [13].
To address the effects of glucose loading on striatal DAT, we analyzed healthy young adults (mean 24.4 years) who underwent 18F-FP-CIT PET scans, and sweet taste questionnaire (STQ), which is a tool used to assess individual preferences and responses to sweet foods. In addition, 18F-FP-FDG PET scans were analyzed to investigate the association of STQ with striatal glucose metabolism. We used Bayesian hierarchical modeling to estimate the effects of 1) sensitivity to mood altering effect of sweet foods, and 2) impaired control over eating sweets on striatal DAT and glucose metabolism.
METHODS
Subjects
All participants signed an informed consent form prior to participation. Thirty-five healthy, male subjects were enrolled in this study. Subjects who had more than 10% change in weight over six months, were heavy smokers, or gave history of drug abuse, brain injury, neuropsychological disorders, or endocrine disorders were excluded. On the day of each visit, the subjects were instructed to fast overnight for at least 12 hours and abstain from smoking and alcohol consumption. The subjects visited the institution between 11 am and 12 pm to avoid the effect of diurnal variations in dopamine. During the visits, the height (m) and weight (kg) of the subject were measured. The body mass index (BMI) was calculated as follows: weight/height-2. The participants in this study were included in a previous study of striatal DAT changes after glucose loading [17,18]. This study was approved by the institutional review board of Pusan National University Hospital (PNUH-1707-019-057).
Image acquisition
Each subject visited the institution three times, on separate days, for three brain PET scans (two 18F-FP-CIT PET scans and one 18F-FDG PET scan). During the visits, the height (m) and weight (kg) of the subject were measured. For 18F-FP-CIT PET scans, bilateral antecubital veins were cannulated: one for blood sampling and for injection of 18F-FP-CIT, and the other for glucose or placebo infusions. The subjects were blinded and randomly assigned for either glucose or placebo infusions. Over 10 minutes, 300 mg/kg of glucose in a 50% solution, was administered. The placebo (normal saline) was also administered in the same speed and volume [19]. An intravenous bolus injection of 18F-FP-CIT was administered after the infusion of glucose or placebo. The emission data were acquired over 90 mins with 50 frames of progressively increasing durations (15 s × 8 frames, 30 s × 16 frames, 60 s × 10 frames, 240 s × 10 frames, and 300 s × 6 frames) using the Siemens Biograph 40 Truepoint PET/computed tomography (CT) (Siemens Healthcare, Knoxville, Tennessee, USA).
The dynamic PET data were collected in the 3-dimensional mode, with 148 slices with image sizes of 256 × 256 and pixel sizes of 1.3364 × 1.3364 mm2. These were reconstructed by filtered back projection using a Gaussian filter. The serum glucose level (mg/dL) and insulin level (μU/mL) were measured before and after the infusions of glucose and placebo. Serum glucose level was determined through an enzymatic reference method using hexokinase with the Glucose HK Gen.3 (Roche Diagnostics GmbH, Germany). Serum insulin level was determined through an electrochemiluminescence immunoassay method using Elecsys Insulin (Roche Diagnostics GmbH, Germany). For 18F-FDG PET scans, sixty minutes after injection of 18F-FDG, the emission data were acquired for 10 mins with Gemini TF PET/CT (Philips, Milpitas, CA, USA). The static PET data were collected in the 3-dimensional mode and reconstructed by the iterative ordered subset expectation maximization method.
Image Analysis
Oxford-GSK-Imanova striatal atlas from FMRIB Software Library v5.0 (https://fsl.fmrib.ox.ac.uk/fsl) was applied, which is an atlas involving sub-striatal regions of ventral striatum (VST), caudate, and putamen segmented according to the anatomical structure, and manually delineated on the non-linear MNI 152 template [20]. For 18F-FP-CIT PET, an averaged image (0-10 min after injection) was created from dynamic PET frames and spatially normalized to 15O-Water PET template in statistical parametric mapping 5 (Wellcome Trust Centre for Neuroimaging, United Kingdom). DAT availability, expressed in terms of binding potential (BPND), were measured by analyzing time-activity curve via the simplified reference tissue method [21] with a reference of cerebellum. For 18F-FDG PET, scans were normalized to 15O-Water PET template in statistical parametric mapping 5 (Wellcome Trust Centre for Neuroimaging, United Kingdom). The mean uptake of each striatal ROI was scaled to the mean of global cortical uptake of each individual, and defined as standardized uptake value ratio (SUVR). Image analysis was done with pmod version 3.6 (PMOD Technologies LLC, Zurich, Switzerland).
Sweet taste questionnaire (STQ)
STQ consists of 12-item self-reporting to evaluate subjects’ reactions to eating sweets, craving for sweets and degree of control over eating sweet foods [22]. Scores for each item range between 1 (strongly disagree) and 7 (strongly agree) with higher scores indicating elevated sensitivity to mood altering effect of sweet foods and impaired control over eating sweets. STQ identifies two factors in the individual’s attitude towards sweet foods; STQ 1: sensitivity to mood altering effect of sweets, and STQ 2: impaired control over eating sweet foods.
Statistical analysis
Normality was tested with Shapiro-Wilk test. Pearson correlation analysis was used to determine whether BMI, and serum glucose & insulin levels are associated with STQ 1, and STQ 2. Paired t-test was used to test the difference of serum glucose & insulin levels before, and after glucose loading, and the difference between glucose-loaded BPND, and placebo-loaded BPND of 18F-FP-CIT PET. After logarithmic transformation of BPND and SUVR, the effects of STQ on BPND and SUVR were investigated using Bayesian hierarchical modelling with brms [23-25] that applies the Markov-Chain Monte Carlo sampling tools of RStan [26]. We created models separately with STQ 1, and STQ 2 as predictors, with BPND and SUVR as a dependent variable, adjusting for age. We added subject and ROI as random intercepts to allow BPND and SUVR to vary between subjects and ROIs. Bayesian models were estimated using four Markov chains, each of which had 4,000 iterations including 1,000 warm-ups, thus totaling 12,000 post-warmup samples. The sampling parameters were slightly modified to facilitate convergence (max treedepth = 20). Statistical analysis was carried out in R Statistical Software (The R Foundation for Statistical Computing).
RESULTS
Thirty-five subjects (age 24.4±2.7 years, BMI 23.8±3.4 kg/m2) underwent two 18F-FP-CIT brain PET scans, and twenty-four underwent additional 18F-FDG brain PET scans. Mean 18F-FP-CIT and 18F-FDG brain PET scans are shown in Figure 1. STQ 1 (sensitivity to mood altering effect of sweets) ranged from 7 to 28 with the mean of 14.9±5.9. STQ 2 (impaired control over eating sweet foods) ranged from 5 to 19 with the mean of 9.1±4.3. Serum glucose level (before glucose loading: 85.2±9.4 mg/dL, after glucose loading: 117.6±48.3 mg/dL), and serum insulin level (before glucose loading: 7.0±4.4 μU/mL, after glucose loading: 19.7±21.3 μU/mL) were significantly increased after glucose loading (p<0.0001, p<0.0003). BMI was not correlated with either STQ 1 (r=-0.0868, p=0.6199) or STQ 2 (r=0.1027, p=0.5573). Serum glucose & insulin levels after glucose loading were not correlated with STQ 1 (r=-0.0066, p=0.97; r=-0.0063, p=0.9711), and STQ 2 (r=0.1276, p=0.4652; r=-0.0527, p=0.7637). From caudate, glucose-loaded DAT availability was significantly higher than placebo-loaded DAT availability (p=0.0495), and from putamen, glucose-loaded DAT availability showed the higher trend than placebo-loaded DAT availability (p=0.0761) (Figure 2).
From Bayesian analysis, glucose-loaded DAT availability from VST, caudate, and putamen was positively associated with both STQ 1 (sensitivity to mood altering effect of sweets), and STQ 2 (impaired control over eating sweet foods). However, the effects of STQ 1 and STQ 2 on placebo-loaded DAT availability, and brain glucose uptake markedly overlapped with zero (Figure 3).
DISCUSSION
In this study, sensitivity to mood altering effect of sweets, and impaired control over eating sweet foods were associated with glucose-loaded DAT availability, not with placebo-loaded DAT availability or brain glucose metabolism. However, neither BMI, nor serum glucose & insulin levels were correlated with sensitivity to mood altering effect of sweets, and impaired control over eating sweet foods.
The brain plays a crucial role in controlling energy balance [7,27]. Eating behaviour is controlled by both homeostatic regulatory brain circuits and those involved in reward and motivation [28]. Dopamine is one of the neurotransmitters involved in eating behaviour, modulating the rewarding properties of food and the motivation and desire to consume food [9]. There are two major hypotheses regarding the role of dopamine in obesity. The first hypothesis, dopamine hyperresponsiveness, proposes that there is a hypersensitivity to rewards and increased behavioral salience toward food, resulting in the excessive intake of palatable foods [29,30]. The second hypothesis, reward deficit model, states that subjects who are insensitive to rewards overeat to increase their endogenous dopamine levels to a normal amount of pleasure [29,30]. However, it is not possible to measure the synaptic dopamine levels directly in the human brain. Therefore, PET with dopamine receptor (DR) radiopharmaceuticals has been used to understand the role of dopaminergic system in obesity. However, the recent studies showed that brain DR availability is not different between lean and overweight/obese subjects both from a meta-analysis [12], and a large PET cohort study [31]. In addition, it is difficult to interpret the relationships between DR availability measured from PET scans and the synaptic dopamine level. Low DR binding potential could represent 1) low density of existing DR, 2) low affinity to bind DR or 3) greater amount of endogenous dopamine which competes with DR radiopharmaceutical to bind DR [10,11].
DAT has not received as much attention in the field of eating behaviour until recently. DAT actively translocates the synaptic dopamine from the extracellular space into the presynatptic neuron, and is a major target for addictive drugs including cocaine, amphetamine, and methamphetamine [14]. From animal studies, insulin can act on striatal insulin receptors to enhance dopamine uptake by DAT through phosphatidylinositol 3-kinase signaling pathway, and downstream activation of Akt, which increases DAT surface expression [13], leading to clearance of synaptic dopamine [32]. Previously, the change of striatal DAT after glucose loading was reported by our group for the first time in humans [17]. We adopted the intravenous injection of glucose to reflect the biological aspects of food intake. Unlike DR, DAT is not affected by endogenous dopamine concentrations. Depletion of dopamine production by □-Methyl-p-tyrosine did not affect 123I-β-CIT binding in non-human primates [33]. In addition, dopamine release by haloperidol did not affect 123I-FP-CIT binding in rats [34], implying nearly irreversible binding of 123I-FP-CIT to DAT [35]. In this study, the change of BPND from 18F-FP-CIT PET scan was visualized in dorsal striatum, not in VST, after glucose loading. Therefore, this change of DAT availability after glucose loading might represent the non-hedonic response of striatal DAT by insulin signaling, regardless of synaptic dopamine concentration.
Interestingly, glucose-loaded DAT availability was positively associated with sensitivity to mood altering effect of sweets, and impaired control over eating sweet foods from STQ, suggesting that the subject with higher DAT availability after glucose loading is the more vulnerable to sweet foods. This may indicate that individuals with higher DAT availability after glucose loading experience a rapid clearance of synaptic dopamine after consuming sweet foods, potentially leading to a desire for additional sweet foods. Therefore, insulin signaling at dopamine terminals might contribute to the rewarding properties of food consumption [36]. However, there was no significant correlation between STQ, and BMI, serum glucose & insulin levels. Individuals with the preference for the stronger sweet solution have a stronger mood altering effect of sweet foods and an impaired control over eating sweets [22], and are linked with alcohol use disorder [37]. Human brain utilizes glucose as its main source of energy, thus, brain glucose uptake, assessed by 18F-FDG PET can be utilized for quantifying neuronal activity in the human brain [38]. However, the effects of STQ 1 and STQ 2 on placebo-loaded DAT availability, and brain glucose uptake markedly overlapped with zero, implying that the baseline DAT availability, and neuronal activity of striatum does not reflect the vulnerability to sweet foods.
There are several limitations to consider in this study. First, although 35 healthy subjects were included, the sample size was small. Second, only males were included in this study, thus these results may not directly generalize to females. Third, the BPNDs that were measured did not distinguish between the DAT density and affinity.
In conclusion, the change of striatal DAT availability after glucose loading is associated with the vulnerability to sweet foods. This may indicate that individuals with higher DAT availability after glucose loading experience a rapid clearance of synaptic dopamine after consuming sweet foods, potentially leading to a desire for additional sweet foods.
Data Availability
All data produced in the present study are available upon reasonable request to the authors
FUNDING
No
DISCLOSURE
The authors had no conflict of interest.
Footnotes
Jihyun Kim march4900{at}hanmail.net
Keunyoung Kim nmpnuh{at}gmail.com
Seongho Seo dansoc{at}gmail.com
Myung Jun Lee mslayer9{at}gmail.com