Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Mar 15.
Published in final edited form as: Brain Res. 2018 May 19;1731:145808. doi: 10.1016/j.brainres.2018.05.025

Lateral hypothalamic orexin glucose-inhibited neurons may regulate reward-based feeding by modulating glutamate transmission in the ventral tegmental area

Suraj B Teegala a,1, Zhenyu Sheng a,1,2, Miloni S Dalal a, Pamela R Hirschberg a, Kevin D Beck a,b, Vanessa H Routh a,*
PMCID: PMC6525648  NIHMSID: NIHMS1028379  PMID: 29787770

Abstract

Glucose inhibits ~60% of lateral hypothalamic (LH) orexin neurons. Fasting increases the activation of LH orexin glucose-inhibited (GI) neurons in low glucose. Increases in spontaneous glutamate excitatory postsynaptic currents (sEPSCs) onto putative VTA DA neurons in low glucose are orexin dependent (Sheng et al., 2014). VTA DA neurons modulate reward-based feeding. We tested the hypothesis that increased activation of LH orexin-GI neurons in low glucose increases glutamate signaling onto VTA DA neurons and contributes to reward-based feeding in food restricted animals. N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) currents on putative VTA DA neurons were measured using whole cell voltage clamp recording in horizontal brain slices containing the LH and VTA. Decreased glucose increased the NMDA receptor current for at least one hour after returning glucose to basal levels (P < 0.05; N = 8). The increased current was blocked by an orexin 1 receptor antagonist (P < 0.05; N = 5). Low glucose caused a similar persistent enhancement of AMPA receptor currents (P < 0.05; N = 7). An overnight fast increased the AMPA/NMDA receptor current ratio, an in vivo index of glutamate plasticity, on putative VTA DA neurons. Conditioned place preference (CPP) for palatable food was measured during LH dialysis with glucose. CPP score was negatively correlated with increasing LH glucose (P < 0.05; N = 20). These data suggest that increased activation of LH orexin-GI neurons in low glucose after weight loss, leads to enhanced glutamate signaling on VTA DA neurons, increases the drive to eat rewarding food, and may contribute to weight regain.

Keywords: Conditioned place preference, Electrophysiology, Microdialysis, NMDA, AMPA, Glucose

1. Introduction

Most individuals in developed countries are exposed to an almost unlimited food supply, largely consisting of highly palatable macronutrients (e.g., sugar and fat). Intake of these macronutrients can trigger ingestion beyond homeostatic needs (reward-based feeding). This increased intake of palatable foods is one factor in the modern obesity epidemic (Leigh et al., 2018). Obesity comorbidities include serious health concerns such as diabetes and cardiovascular disease (Kopelman, 2000; Krauss et al., 1998). However, maintenance of even a 10% body weight reduction significantly improves glycemia and cardiovascular risk factors (Wing et al., 2011). Unfortunately, <20% of individuals are able to achieve and maintain this weight loss (MacLean et al., 2011) in part because food restriction and weight loss enhance reward-based feeding (Fulton, 2010; Jewett et al., 1995). Thus, understanding how food restriction and weight loss enhance reward-based feeding will lead to better therapies for weight loss maintenance and thus improve the serious comorbidities associated with obesity.

The brain controls food intake. Within the brain, ventral tegmental area (VTA) dopamine (DA) projections to the nucleus accumbens drive reward-based feeding (Aston-Jones et al., 2010; Fields et al., 2007). The lateral hypothalamic (LH) orexin neurons activate the VTA DA neurons (Choi et al., 2010; Narita et al., 2006). In turn, metabolic status modulates LH orexin neurons. Fasting and hypoglycemia increase orexin expression (Cai et al., 1999; Yamanaka et al., 2003). Fasting also increases reward-based feeding (Fulton, 2010). Ghrelin, a hormone released from the stomach in the absence of food, activates orexin neurons and orexin mediates ghrelin-induced feeding (Toshinai et al., 2003). In contrast, the satiety hormone leptin decreases orexin expression and the activity of orexin neurons (Yamanaka et al., 2003). Approximately half of the LH orexin neurons are also glucose-inhibited (GI) neurons (Burdakov et al., 2005; Sheng et al., 2014; Yamanaka et al., 2003). We have found that the glucose sensitivity of orexin-GI neurons is a target of metabolic status. Fasting and ghrelin enhance, while leptin blunts, their activation in low glucose (Sheng et al., 2014). These data suggest that signals of peripheral energy status influence the neurocircuitry controlling reward-based feeding, in part, by regulating the glucose sensitivity of LH orexin-GI neurons.

LH orexin neurons mediate cocaine-based glutamate plasticity onto VTA DA neurons (Borgland et al., 2006). We have shown that activation of LH orexin-GI neurons in low glucose enhances spontaneous glutamate excitatory post-synaptic currents (sEPSCs) onto putative VTA DA neurons (Sheng et al., 2014). This suggests that increased activation of LH orexin-GI neurons during food restriction may induce glutamate plasticity onto VTA DA neurons.

Thus, we hypothesize that the changes in glucose sensitivity of LH orexin-GI neurons after food restriction and weight loss lead to persistent enhancement of glutamate signaling onto VTA DA neurons and a concomitant facilitation of reward-based feeding. Increased intake of palatable food may contribute to the difficulty maintaining weight loss after dieting. In order to test this hypothesis we determined whether low glucose lead to orexin-dependent increases in glutamate signaling onto putative VTA DA neurons and whether increased LH glucose levels reduce food preference in food restricted animals.

2. Results

Putative VTA DA neurons were identified by a strong hyperpolarization activated cation conductance (H current or Ih). We first determined whether N-methyl-D-aspartate (NMDA) receptor currents on putative VTA DA neurons were sensitive to glucose decreases and whether this effect persisted after return to basal glucose levels. Decreasing extracellular glucose from 2.5 mM to 0.25 mM for 60 min. increased the NMDA receptor current amplitude approximately twofold (P < 0.05; N = 8). This effect was maintained for at least one hour after return to 2.5 mM glucose (Fig. 1A). One hour was the minimum time that the recording was maintained in all of the 8 neurons tested and thus this time point was used for statistical analysis. However, in neurons in which the recording was stable for over an hour after return to 2.5 mM glucose, the persistent increase in NMDA receptor current amplitude was observed for the duration of the recording. One-way repeated measures ANOVA indicated a significant main effect of treatment on the frequency of NMDA currents in Hz (P = 0.039); however post hoc analysis did not show significant group differences (data not shown). In a separate experiment, neurons were first exposed to the orexin 1 receptor antagonist SB334287 (10 μM) in 2.5 mM glucose and then glucose levels were lowered to 0.25 mM for 60 min. In this case, NMDA receptor current amplitude did not increase. However, when the antagonist was washed out in the continued presence of 0.25 mM glucose a significant increase in NMDA current amplitude was observed (P < 0.05, N = 5; Fig. 1B). We next determined whether lowering glucose to levels seen in the hypothalamus after an overnight fast (~0.7 mM (De Vries et al., 2003)) was sufficient to alter α-amino-3-hydroxy-5-methy l-4-isoxazolepropionic acid (AMPA) receptor current amplitude and frequency. Decreased glucose to 0.7 mM increased both the frequency (Fig. 1C) and amplitude (Fig. 1D) of AMPA receptor currents and these changes persisted for at least 50 min upon return to 2.5 mM glucose (P < 0.05; N = 7).

Fig. 1.

Fig. 1

Open in a new tab

Increased NMDA and AMPA currents on putative VTA DA neurons in decreased glucose persist after return to basal glucose levels. A. Representative traces of sEPSCs recorded at a holding potential of +40 mV to release the Mg2+ block on the NMDA receptor and in the presence of 20 μM bicuculline to block GABA A receptors. In this condition, sEPSCs are seen as upward deflections. AMPA receptor currents decay rapidly, thus the current measured at 20 ms past the peak current predominantly reflects the NMDA receptor current amplitude. After 60 min in 0.25 mM glucose the NMDA receptor current amplitude has nearly doubled compared to that in 2.5 mM glucose. This effect is maintained for at least an hour after return to 2.5 mM glucose. B. Bar graphs representing NMDA receptor current amplitude in 2.5 mM glucose, 0.25 mM glucose and 60 min after return to 2.5 mM glucose. Amplitude was significantly greater in 0.25 mM glucose and one hour after return to 2.5 mM glucose compared to that observed initially in 2.5 mM glucose. Data analyzed by repeated measures ANOVA. Different letters represent statistical difference, P < 0.05, N = 8. W.O.; washout C. Bar graphs representing NMDA receptor current amplitude in 2.5 mM glucose, one hour in 0.25 mM glucose in the presence of the orexin 1 receptor antagonist, SB334867 (10 μM), and in 0.25 mM glucose one hour after washout of the orexin antagonist. Decreased glucose had no effect on NMDA receptor current amplitude in the presence of the orexin antagonist, however NMDA receptor current amplitude was nearly doubled in 0.25 mM glucose after washout of the drug. Data analyzed by repeated measures ANOVA. Different letters represent statistical difference, P < 0.05, N = 5. D. Representative traces of AMPA receptor currents recorded at 60 mV (when NMDA receptors are blocked by Mg2+) and in the presence of bicuculline. Under these conditions AMPA receptor currents are observed as downward deflections. AMPA receptor current frequency and amplitude was increased in 0.7 mM glucose and after 50 min return to 2.5 mM glucose compared to that in 2.5 mM glucose initially. E, F. Bar graphs representing AMPA receptor current frequency (E) and amplitude (F) in 2.5 mM glucose, 0.7 mM glucose and after 50 min return to 2.5 mM glucose. AMPA receptor current frequency and amplitude is significantly elevated in 0.7 mM glucose and after 50 min return to 2.5 mM glucose compared to that in 2.5 mM glucose initially. Data analyzed by repeated measures ANOVA. Different letters represent statistical difference, P < 0.05, N = 7.

We have previously shown that an overnight fast increased activation of orexin-GI neurons in 0.7 mM glucose (Sheng et al., 2014). In order to determine whether an overnight fast similarly enhanced sEPSCs on putative VTA DA neurons we compared sEPSC frequency and amplitude in 2.5 and 0.7 mM glucose in ad lib fed (N = 15) and overnight fasted (N = 14) mice. There were no differences in either frequency (Fig. 2A) or amplitude (Fig. 2B) of sEPSCs from fed animals in 2.5 and 0.7 mM glucose. Moreover, there were no differences in sEPSCs on putative VTA DA neurons from fed or fasted animals in 2.5 mM glucose. However, when glucose was lowered to 0.7 mM, there was a significant increase in both frequency (Fig. 2A) and amplitude (Fig. 2B) of sEPCSs from fasted mice compared to fed mice in either 2.5 or 0.7 mM glucose (P < 0.05). We also evaluated whether an overnight fast similarly enhanced the AMPA/NMDA receptor current amplitude ratio on putative VTA DA neurons from fed (N = 8) and overnight fasted (N = 6) mice in 2.5 mM glucose as an index of in vivo plasticity. The AMPA/ NMDA ratio was significantly greater in fasted vs fed mice (P < 0.05, unpaired Students t-test; Fig. 2C).

Fig. 2.

Fig. 2

Open in a new tab

Fasting increases sEPSC frequency in low glucose and the AMPA/NMDA ratio in putative VTA DA neurons. A. Bar graphs representing sEPSC frequency in 2.5 and 0.7 mM glucose in VTA neurons from ad lib fed (n = 15) and overnight fasted (n = 14) mice. There was a significant main effect of feeding state ((F (1,54) = 10.4, P = 0.002), two-way ANOVA. Tukey’s multiple comparisons test indicated that frequency of sEPSCs on VTA neurons from fasted mice in 0.7 mM glucose was significantly greater than those from fed mice in either 2.5 or 0.7 mM glucose. Different letters indicate statistical significance (P < 0.05). B. Bar graphs representing sEPSC amplitude in 2.5 and 0.7 mM glucose in VTA neurons from ad lib fed (n = 15) and overnight fasted (n = 14) mice. There were significant main effects of glucose concentration (F (1,54) = 5.10, P = 0.028) and feeding state ((F (1,54) = 4.91, P = 0.031), two-way ANOVA. Tukey’s multiple comparisons test indicated that frequency of sEPSCs on VTA neurons from fasted mice in 0.7 mM glucose was significantly greater than those from fed mice in 2.5 mM glucose. Different letters indicate statistical significance (P < 0.05). C. Bar graphs represent the AMPA/ NMDA receptor current ratio on VTA neurons from fed (n = 8) and fasted (n = 6) mice. The AMPA/NMDA receptor current ratio was significantly greater on VTA neurons from fed mice that that from fasted mice (P = 0.03; unpaired Students t-test).

In order to determine whether changes in LH glucose concentration affects feeding behavior in weight restricted rats we food restricted rats to 85% of their initial body weight and maintained that weight loss during conditioned place preference (CPP) training for chocolate (Fig. 3). Blood glucose levels were significantly reduced after the initial weight loss (94.9 ± 2.9 mg/dl) and on the day of CPP testing (87.5 ± 2.4 mg/dl) compared to that prior to food restriction and weight loss (124.2 ± 4.9 mg/dl; P < 0.0001, N = 23, repeated measures ANOVA followed by Tukey’s multiple comparison test). CPP training was verified by a significant increase in preference for the formerly non-preferred side of the CPP chamber after the rats had been trained to associate that side of the chamber with chocolate (Fig. 4a). Following CPP training we dialyzed the LH with glucose concentrations varying from that seen in the hypothalamus after an overnight fast (0.7 mM) to that reported when blood glucose levels are between 100 and 200 mg/dl as would be seen during postprandial excursions or fasting diabetic hyperglycemia (4.0 mM) (De Vries et al., 2003; Silver and Erecinska, 1994; Silver and Erecinska, 1998). Although variation was high (R2 = 0.282) there was a highly significant negative correlation between CPP score and LH glucose concentration (P = 0.016, N = 20; Fig. 4b, c). Interestingly, at LH glucose concentrations which exceeded 2.0 mM (that seen in a baseline satiated state (Silver and Erecinska, 1994)), the CPP was largely abolished (Fig. 4b, c). This effect was not due to osmotic differences in the dialysate because the CPP score obtained after matching the osmolality of the solution with 0.7 mM glucose to that of 4 mM glucose with mannitol was not different than 0.7 mM glucose without mannitol (N = 3).

Fig. 3.

Fig. 3

Open in a new tab

Time course of the behavioral study. On days 1–12 rats received 50% of their average daily intake in order to decrease their body weight to 85% of their initial weight. Food was adjusted to maintain this weight loss for the duration of the study. On day 13 they underwent surgery for implantation of the microdialysis guide cannula and were allowed 3 days to recover. After recovery they underwent 2 days of pre-testing to determine their initial preference for the striped or checkered chamber followed by 6 days of conditioning for chocolate on the initially non-preferred chamber. They were then evaluated to determine whether they had developed a preference for the chocolate associated chamber and re-tested the next day with glucose dialyzed into the LH.

Fig. 4.

Fig. 4

Open in a new tab

Increased LHA glucose concentration reduces CPP score in weight-restricted rats. A. The place conditioning score on the bar graph represents the time (min) spent in the initially non-preferred chamber of the CPP box minus that in the preferred chamber at the beginning of the study. Pre-test times were measured in the naïve rats. The rats were then conditioned to associate the non-preferred chamber with chocolate. Day 7/test day 1 is the time spent in the formerly non-preferred (now chocolate associated or conditioned) chamber minus that in the initially preferred chamber. A more positive score on day 7 relative to the pre-test indicates that the rat’s preference changed from the initially preferred side to the chocolate conditioned side and demonstrates establishment of CPP. ****p < 0.001, two-tailed Students t-test; Mean ± S.E.M (N = 23).B. Rats were tested for CPP for the chocolate preferred side of the chamber on day 7 in the absence of glucose dialysis and again on day 8 during LHA dialysis of glucose concentrations from 0.7 to 4.0 mM. The CPP score on the Y-axis represents the number of minutes on the conditioned side on day 7 subtracted from that on day 8. A positive score indicates that the CPP for the conditioned (chocolate) side persisted whereas decreasing CPP scores represent a decrease or reversal of preference for the chocolate conditioned side. There was a significant negative correlation between the level of glucose dialyzed into the LHA and CPP score (P = 0.016, N = 20) indicating that preference for the chocolate preferred side decreased with increasing LHA glucose. Circles represent CPP scores from rats dialyzed with varying glucose concentrations. Squares represent CPP scores from 3 rats in a separate experiment in which mannitol was added to 0.7 mM glucose to match the osmolality to that of the 4 mM glucose solution. Regression analysis was performed using Statistical Package for Social Sciences (SPSS, IBM) 23.0. C. Bar graph represents the amount of time (min) spent in the chocolate conditioned chamber on day 8 for rats dialyzed with <2 mM glucose compared to those dialyzed with >2 mM glucose. Those dialyzed with lower glucose levels spent significantly more time in the chocolate conditioned chamber whereas increasing LH glucose >2 mM reduced the amount of time spent in the chocolate conditioned chamber (P = 0.005, two-tailed Students t test, N = 20).

3. Discussion

The present results are consistent with our hypothesis that enhanced activation of LH orexin-GI neurons in low glucose may contribute to the difficulty maintaining weight loss after dieting by increasing the drive for palatable food. We show here that decreases in extracellular glucose seen in the hypothalamus during an overnight fast lead to an orexin-dependent increase in NMDA receptor current amplitude on putative VTA DA neurons that persists after glucose levels return to baseline. This is associated with a persistent increase in AMPA receptor current amplitude. Similarly, fasting increases the AMPA/NMDA receptor current ratio, an in vivo index of glutamate plasticity (Borgland et al., 2006). Interestingly, raising LH glucose levels in weight restricted animals decreased their preference for the chocolate associated chamber in a CPP test. These are the first data which directly implicate glucose sensing neurons in food seeking behavior. Moreover, they suggest that the glucose sensing mechanism of LH orexin-GI neurons might be a therapeutic target for weight loss maintenance. This would be of high clinical significance given the current obesity epidemic in developed countries (Dietz, 2015).

The VTA neurons that we recorded were not definitively identified as DA neurons (i.e., tyrosine hydroxylase positive). However, over 90% of VTA neurons exhibiting a strong H-current (Ih) are DA neurons (Korotkova et al., 2003). Thus, we refer to these neurons as ‘‘putative DA neurons” herein. We also did not demonstrate that the in vivo changes in the AMPA/NMDA receptor current ratio were orexin-dependent. However, the persistent increase in NMDA receptor current amplitude in low glucose was blocked by an orexin 1 receptor antagonist. Moreover, we have previously shown that increases in the frequency and amplitude of glutamate excitatory postsynaptic currents on putative VTA DA neurons in low glucose are orexin dependent (Sheng et al., 2014). In addition, we observed increased sEPSC frequency and amplitude on VTA DA neurons from fasted mice in 0.7 mM glucose compared to that in fed mice. Thus, it is reasonable to hypothesize that the increased AMPA/NMDA ratio on VTA DA neurons is due, in part, to enhanced activity of LH orexin-GI neurons in low glucose.

The present study focuses on the effects of decreased glucose during fasting. However, fasting also increases free fatty acid levels (Barrett, 1964). Fatty acids are sensed by hypothalamic neurons and could play a role in the fasting induced changes that we observed (Le Foll et al., 2013; Le Foll et al., 2014; Wang et al., 2005). We have recently shown that oleic acid directly decreases the activity of putative VTA DA neurons and operant feeding behavior (Hryhorczuk et al., 2017). Based on these data, it is unlikely that increased free fatty acids explain the increased excitatory signaling seen in the present study. Moreover, the change in food seeking behavior that we observed was dependent solely on changes in extracellular glucose levels in the LH. Our previous work indicated that changes in glucose sensitivity of LH orexin-GI neurons after fasting are consistent with the expected changes in leptin (decreased) and ghrelin (increased) levels (Yamanaka et al., 2003). That is, fasting and ghrelin enhance while leptin reduces activation of LH orexin-GI neurons in low glucose (Sheng et al., 2014). Moreover, ghrelin enhances the CPP for palatable food in mice in an orexin-dependent manner (Perello et al., 2010). Thus, it is reasonable to hypothesize that the effects of fasting are due, in part, to leptin/ghrelin induced changes in glucose sensitivity of LH orexin-GI neurons.

Although orexin-GI neurons are activated in low glucose, we did not lower LH glucose levels beyond that which would occur in the brain during food restriction using microdialysis. This is because dialyzing the brain with a lower than ambient glucose concentration would be compensated by glucose uptake from the blood. Thus, our baseline glucose concentration in the dialysate was 0.7 mM which has been reported in the hypothalamus after an over-night fast which began at dark onset (De Vries et al., 2003). The animals in the present study were not fasted overnight but rather food restricted. However, the amount of food needed to maintain body weight at 85% of initial weight was weighed and provided to the rats just before lights off. We observed that they consumed their food within an hour. The CPP test was done the following afternoon, thus even though the animals consumed a meal at the start of the dark cycle they had not eaten for nearly 24 h. Moreover, blood glucose level was 4.86 mM (87.5 mg/dl) on the day of testing. Hypothalamic glucose in overnight fasted animals is ~14% of blood levels (De Vries et al., 2003), thus in our study LH glucose concentration would have been ~0.7 mM. Increasing LH glucose level from 0.7 mM to levels which would be observed after a meal (4.0 mM; (Silver and Erecinska, 1994)) decreased preference for the CPP chamber associated with a chocolate reward. This strongly links LH glucose sensing pathways to the motivation to seek palatable food. We cannot distinguish between effects of glucose on LH orexin-GI neurons and melanin concentrating hormone neurons which are excited by glucose (Burdakov et al., 2005). Both orexin and melanin concentrating hormone increase food intake (Beck, 2000). Since increased glucose would activate the latter, the most reasonable conclusion is that increased LH glucose silenced the orexin-GI neurons and reduced food motivation. One possible issue with our dialysis study is that we did not osmotically match each glucose concentration. However, we did include a control group in which we used mannitol to osmotically match dialysate with 0.7 mM glucose to that with 4 mM glucose. As seen in Fig. 4 the CPP score for this group was identical to that of the 0.7 mM solution without mannitol. This suggests that the effect of increased LH glucose on CPP score was not due to changes in osmolality. There is also the possibility that the lack of reward in the chamber on the first test day diminished the CPP on the subsequent test day. However, we do not believe that this explains our results because we only saw a diminished response with dialysis of glucose concentrations > 2 mM. Rats were used for the dialysis study because the microdialysis probes are too heavy for conscious spontaneously active mice. On the other hand, mice were used for the electrophysiology studies because their brains were small enough to retain the LH, VTA and connections within one brain slice. While rats and mice are used for both types of studies there is always the possibility of species differences. That said, the data from both the behavioral and electrophysiological studies are consistent with our hypotheses. Finally, we only evaluated the glucose sensitivity of LH orexin-GI neurons (Sheng et al., 2014) and the AMPA/NMDA receptor current ratio after an overnight fast. Clearly, weight loss and maintenance for 4 weeks is a very different physiological state. However, decreased leptin levels and, in many but not all cases, increased ghrelin levels persist during weight loss maintenance in humans (Hill et al., 2013; Hinkle et al., 2013; Iepsen et al., 2016; Sumithran et al., 2011). Thus, these hormonal changes could contribute to long lasting changes in the glucose sensitivity of LH orexin-GI neurons and glutamate signaling onto VTA DA neurons which, in turn, drive intake of palatable food and promote weight regain.

In conclusion, we have shown that a brief exposure to decreased glucose (1 h) causes persistent increases in NMDA and AMPA current amplitudes on putative VTA DA neurons in vitro. In addition, an overnight fast which increases activation of LH orexin-GI neurons in low glucose (Sheng et al., 2014) increases the AMPA/ NMDA receptor current ratio onto putative VTA DA neurons suggesting glutamate plasticity in vivo. Finally, increasing LH glucose levels from those associated with an overnight fast to that seen after a meal reduces the CPP for a rewarding food. Together these data suggest that changes in the glucose sensitivity of LHA orexin-GI neurons after weight restriction could lead to enhanced reward-based feeding and contribute to the difficulty maintaining weight loss after dieting

4. Experimental procedures

4.1. Electrophysiology

Male 4–6 week old C57BL6 mice were housed on a 12:12 light:- dark cycle with water provided ad libitum. All procedures were in accordance with the Rutgers New Jersey Medical School Institutional Animal Care and Use Committee (IACUC). Most of the mice received standard rat chow (PicoLab Rodent Diet 20 5053) ad libitum; however some mice were fasted for ~24 h (food removed at 12 noon the day prior to sacrifice). On the day of experiment mice were anesthetized and transcardially perfused with ice-cold oxygenated (95%O2/5%CO2) perfusion solution in which NaCl was replaced with sucrose (composition in mM: 2.5 KCl, 7 MgCl2, 1.25 NaH2PO4, 28 NaHCO3, 0.5 CaCl2, 7 glucose, 1 ascorbate, and 3 pyruvate; osmolarity adjusted to ~300 mOsm with sucrose; pH 7.4). Brains were rapidly removed, placed in ice-cold (slushy) oxygenated perfusion solution and 320 μm slices containing the LH and VTA were made on a vibratome (Vibroslice, Camden Instruments, Camden, UK) as previously described (Sheng et al., 2014; Song et al., 2001; Wang et al., 2004). The brain slices were maintained in oxygenated artificial cerebrospinal fluid (aCSF; in mM: 126 NaCl, 1.9 KCl, 1.2 KH2PO4, 26 NaHCO3, 2.5 glucose, 1.3 MgCl2, and 2.4 CaCl2; osmolarity was adjusted to ~300 mOsm with sucrose; pH 7.4) for at least 1 h at room temperature. They were then transferred to the recording chamber for remainder of the day.

Whole cell voltage clamp recordings were made as previously described (Sheng et al., 2014). Borosilicate pipettes (4–6 MX; Sutter Instruments, Novato, CA) were filled with an intracellular solution containing (in mM): 128 K-gluconate, 10 KCl, 4 KOH, 10 HEPES, 4 MgCl2, 0.5 CaCl2, 5 EGTA, 2 Na2ATP, and 0.4 Na2GTP; pH 7.2. Osmolarity was adjusted to 300 mOsm with sucrose. Cells with membrane potentials more negative than −45 mV in 2.5 mM glucose and action potentials which overshoot 0 mV were considered viable for recording. Pipette access resistance under 30 MX with less than a 20% change during the time course of the experiment was considered acceptable. Voltage clamp recordings were low pass-filtered at 1 kHz and data were simultaneously digitized at 5 kHz. Putative VTA DA neurons from wild-type mice were identified by the presence of an Ih. Over 90% of VTA neurons exhibiting a strong Ih are DA neurons (Korotkova et al., 2003). A strong Ih was defined as a current sag greater than 150 pA at the end of a 500 ms voltage command to −120 mV (Ford et al., 2006; Sheng et al., 2014). All recordings were performed in the presence of the GABA-A receptor antagonist bicuculline (20 μM) and the GABA-B receptor antagonist saclofen (100 μM). Glucose concentrations were varied as described in the figures. Current frequency and amplitude was averaged over the last minute in each treatment. Neurons were clamped at −60 or +40 mV to record spontaneous AMPA or NMDA mediated currents, respectively. NMDA current amplitude was measured 20 ms after the peak current when AMPA currents have decayed. sEPSCs were measured at a holding potential of +40 mV and represented combined AMPA and NMDA currents. sEPSC amplitude was measured at the peak of the response. The AMPA/NMDA receptor current ratio in brain slices from fasted mice was measured in 2.5 mM glucose at a holding potential of +40 mV with and without the NMDA receptor antagonist (2R)-amino-5-phosphonovaleric acid (APV, 50 μM, Sigma). The peak current amplitude observed in the presence of APV (AMPA) was subtracted from that without APV (AMPA + NMDA) to give the NMDA current amplitude. The AMPA/NMDA receptor current ratio was then calculated by dividing the AMPA peak current by the NMDA peak current. Data were analyzed using repeated measures Analysis of Variance (ANOVA) followed by Dunnett’s multiple comparisons test, two-way ANOVA following by Tukeys multiple comparisons test or by Students t-test with P < 0.05 being considered statistically significant.

4.2. Conditioned place preference (CPP) behavioral test

Male Sprague Dawley rats (300–350 g) were housed on a 12:12 L:D cycle and provided with food and water ad libitum unless otherwise described. All procedures were IACUC approved. The time course for the experiment is shown in Fig. 3. After a 2 day acclimation to the facility daily food intake was measured for 2 days. Rats were then given 50% of their average daily intake until their body weight was 85% of their initial weight (approximately 2 weeks). Bilateral microdialysis guide cannulas (CMA11) were implanted to target 1 mm dorsal to the LH (AP: −2.5 or 2.86, ML: ±3.1, DV: −8.2; at a 12 degree angle). During recovery from surgery (3–5 days) rats were fed 75% of their initial average food intake until body weight was equal to that before surgery (85% of initial) and then weight was maintained at that level by feeding ~65% of their initial food intake. After recovery rats underwent 30 min pre-testing in the CPP box for 2 consecutive days with the barrier between the striped and checkered chambers removed to determine their initial preference. The rats were then conditioned for 30 min a day for the following 6 days. On days 1, 3, and 5 they were placed in the initially non-preferred chamber with chocolate (5 g, Nestle) broken up and placed evenly in the corners and on days 2, 4, and 6 they were placed in the initially preferred chamber without chocolate. On day 7 (test day 1) they were placed in the CPP box with the barrier between the chambers removed and the time spent in each chamber during a 30 min testing period was measured to verify that the rats developed a preference for the chocolate associated side of the chamber. On day 8 (test day 2), the guide cannulas were attached to a microdialysis pump (Harvard Apparatus CMA 402). Glucose (0.7 mM) in perfusion solution (composition in mM: NaCl 147, KCl 2.7, CaCl2 1.2, MgCl2 0.85) was dialyzed at a rate of 1–2 μl/min for 90 min in the home cage in order to stabilize LH glucose level at a uniform concentration between rats. The rats were again placed in the center of the CPP box with the barrier between the chambers removed for 30 min while the LH was continuously dialyzed with 0.7, 1.0, 2.0, 2.5, 3.0 or 4.0 mM glucose. All glucose concentrations were determined according to the standard efficiency of the dialysis probe for glucose (~8%) measured using a glucose oxidase kit (Thermofischer). A separate control group received 0.7 mM glucose with mannitol added to adjust osmolarity to that of the 4.0 mM glucose solution. The CPP score was calculated as the difference in time spent on the chocolate-conditioned chamber on test day 2 (day 8) subtracted from that on test day 1 (day 7). Regression analysis was performed using Statistical Package for Social Sciences (SPSS, IBM) 23.0.

Acknowledgements

Funding: This work was supported by the American Heart Association Grant-In-Aid 14GRNT20380639 (VHR) and by the National Institutes of Health Research Award 1R01DK10367 (VHR, KDB).

Abbreviations:

AMPA

α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

CPP

conditioned place preference

DA

dopamine

EPSCs

excitatory post-synaptic currents

GI

glucose-inhibited

HCN (Ih)

hyperpolarization-activated cation channel

LH

lateral hypothalamus

NMDA

N-methyl-D-aspartate

VTA

ventral tegmental area

References

  1. Aston-Jones G, Smith RJ, Sartor GC, Moorman DE, Massi L, Tahsili-Fahadan P, Richardson KA, 2010. Lateral hypothalamic orexin/hypocretin neurons: a role in reward-seeking and addiction. Brain Res 1314, 74–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barrett AM, 1964. Adventitious factors affecting the concentration of free fatty acids in the plasma of rats. Br. J. Pharmacol. Chemother 22, 577–584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beck B, 2000. Neuropeptides and obesity. [Review] [170 refs]. Nutrition 16, 916–923. [DOI] [PubMed] [Google Scholar]
  4. Borgland SL, Taha SA, Sarti F, Fields HL, Bonci A, 2006. Orexin A in the VTA is critical for the induction of synaptic plasticity and behavioral sensitization to cocaine. Neuron 49, 589–601. [DOI] [PubMed] [Google Scholar]
  5. Burdakov D, Gerasimenko O, Verkhratsky A, 2005. Physiological changes in glucose differentially modulate the excitability of hypothalamic melanin-concentrating hormone and orexin neurons in situ. J. Neurosci 25, 2429–2433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Cai XJ, Widdowson PS, Harrold J, Wilson S, Buckingham RE, Arch JR, Tadayyon M, Clapham JC, Wilding J, Williams G, 1999. Hypothalamic orexin expression: modulation by blood glucose and feeding. Diabetes 48, 2132–2137. [DOI] [PubMed] [Google Scholar]
  7. Choi DL, Davis JF, Fitzgerald ME, Benoit SC, 2010. The role of orexin-A in food motivation, reward-based feeding behavior and food-induced neuronal activation in rats. Neuroscience 167, 11–20. [DOI] [PubMed] [Google Scholar]
  8. De Vries MG, Arseneau LM, Lawson ME, Beverly JL, 2003. Extracellular glucose in rat ventromedial hypothalamus during acute and recurrent hypoglycemia. Diabetes 52, 2767–2773. [DOI] [PubMed] [Google Scholar]
  9. Dietz WH, 2015. The response of the US centers for disease control and prevention to the obesity epidemic. Annu. Rev. Public Health 36, 575–596. [DOI] [PubMed] [Google Scholar]
  10. Fields HL, Hjelmstad GO, Margolis EB, Nicola SM, 2007. Ventral tegmental area neurons in learned appetitive behavior and positive reinforcement. Annu. Rev. Neurosci 30, 289–316. [DOI] [PubMed] [Google Scholar]
  11. Ford CP, Mark GP, Williams JT, 2006. Properties and opioid inhibition of mesolimbic dopamine neurons vary according to target location. J. Neurosci 26, 2788–2797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fulton S, 2010. Appetite and reward. Front. Neuroendocrinol 31, 85–103. [DOI] [PubMed] [Google Scholar]
  13. Hill BR, Rolls BJ, Roe LS, De Souza MJ, Williams NI, 2013. Ghrelin and peptide YY increase with weight loss during a 12-month intervention to reduce dietary energy density in obese women. Peptides 49, 138–144. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hinkle W, Cordell M, Leibel R, Rosenbaum M, Hirsch J, 2013. Effects of reduced weight maintenance and leptin repletion on functional connectivity of the hypothalamus in obese humans. PLoS One 8, e59114. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hryhorczuk C, Sheng Z, Décarie-Spain L, Giguère N, Ducrot C, Trudeau L-É, Routh VH, Alquier T, Fulton S, 2017. Oleic acid in the ventral tegmental area inhibits feeding, food reward, and dopamine tone. Neuropsychopharmacology 43, 607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Iepsen EW, Lundgren J, Holst JJ, Madsbad S, Torekov SS, 2016. Successful weight loss maintenance includes long-term increased meal responses of GLP-1 and PYY3–36. Eur. J. Endocrinol 174, 775–784. [DOI] [PubMed] [Google Scholar]
  17. Jewett D, Cleary J, Levine A, Schaal D, Thompson T, 1995. Effects of neuropeptide Y, insulin, 2-deoxyglucose, and food deprivation on food-motivated behavior. Psychopharmacology 120, 267–271. [DOI] [PubMed] [Google Scholar]
  18. Kopelman PG, 2000. Obesity as a medical problem. [Review] [69 refs]. Nature 404, 635–643. [DOI] [PubMed] [Google Scholar]
  19. Korotkova TM, Sergeeva OA, Eriksson KS, Haas HL, Brown RE, 2003. Excitation of ventral tegmental area dopaminergic and nondopaminergic neurons by orexins/hypocretins. J. Neurosci 23, 7–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Krauss RM, Winston M, Fletcher BJ, Grundy SM, 1998. Obesity: impact on cardiovascular disease. Circulation 98, 1472–1476. [PubMed] [Google Scholar]
  21. Le Foll C, Dunn-Meynell A, Musatov S, Magnan C, Levin BE, 2013. FAT/CD36: a major regulator of neuronal fatty acid sensing and energy homeostasis in rats and mice. Diabetes 62, 2709–2716. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Le Foll C, Dunn-Meynell AA, Miziorko HM, Levin BE, 2014. Regulation of hypothalamic neuronal sensing and food intake by ketone bodies and fatty acids. Diabetes 63, 1259–1269. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Leigh S-J, Lee F, Morris MJ, 2018. Hyperpalatability and the generation of obesity: roles of environment, stress exposure and individual difference. Curr. Obesity Rep 7, 6–18. [DOI] [PubMed] [Google Scholar]
  24. MacLean PS, Bergouignan A, Cornier M-A, Jackman MR, 2011. Biology’s response to dieting: the impetus for weight regain. Am. J. Physiol. – Regul. Integr. Comp. Physiol 301, R581–R600. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Narita M, Nagumo Y, Hashimoto S, Narita M, Khotib J, Miyatake M, Sakurai T, Yanagisawa M, Nakamachi T, Shioda S, Suzuki T, 2006. Direct involvement of orexinergic systems in the activation of the mesolimbic dopamine pathway and related behaviors induced by morphine. J. Neurosci 26, 398–405. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Perello M, Sakata I, Birnbaum S, Chuang J-C, Osborne-Lawrence S, Rovinsky SA, Woloszyn J, Yanagisawa M, Lutter M, Zigman JM, 2010. Ghrelin increases the rewarding value of high-fat diet in an orexin-dependent manner. Biol. Psychiatry 67, 880–886. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sheng Z, Santiago AM, Thomas MP, Routh VH, 2014. Metabolic regulation of lateral hypothalamic glucose-inhibited orexin neurons may influence midbrain reward neurocircuitry. Mol. Cell. Neurosci 62, 30–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Silver IA, Erecinska M, 1994. Extracellular glucose concentration in mammalian brain: continuous monitoring of changes during increased neuronal activity and upon limitation in oxygen supply in normo-, hypo-, and hyperglycemic animals. J. Neurosci 14, 5068–5076. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Silver IA, Erecinska M, 1998. Glucose-induced intracellular ion changes in sugar-sensitive hypothalamic neurons. J. Neurophysiol 79, 1733–1745. [DOI] [PubMed] [Google Scholar]
  30. Song Z, Levin BE, McArdle JJ, Bakhos N, Routh VH, 2001. Convergence of pre- and postsynaptic influences on glucosensing neurons in the ventromedial hypothalamic nucleus. Diabetes 50, 2673–2681. [DOI] [PubMed] [Google Scholar]
  31. Sumithran P, Prendergast LA, Delbridge E, Purcell K, Shulkes A, Kriketos A, Proietto J, 2011. Long-term persistence of hormonal adaptations to weight loss. N. Engl. J. Med 365, 1597–1604. [DOI] [PubMed] [Google Scholar]
  32. Toshinai K, Date Y, Murakami N, Shimada M, Mondal MS, Shimbara T, Guan J-L, Wang Q-P, Funahashi H, Sakurai T, Shioda S, Matsukura S, Kangawa K, Nakazato M, 2003. Ghrelin-induced food intake is mediated via the orexin pathway. Endocrinology 144, 1506–1512. [DOI] [PubMed] [Google Scholar]
  33. Wang R, Liu X, Hentges ST, Dunn-Meynell AA, Levin BE, Wang W, Routh VH, 2004. The regulation of glucose-excited neurons in the hypothalamic arcuate nucleus by glucose and feeding-relevant peptides. Diabetes 53, 1959–1965. [DOI] [PubMed] [Google Scholar]
  34. Wang R, Cruciani-Guglielmacci C, Migrenne S, Magnan C, Cotero VE, Routh VH, 2005. The effects of oleic-acid (OA) on distinct populations of neurons in the hypothalamic arcuate nucleus (ARC) are dependent on extracellular glucose levels. J. Neurophysiol 55, 1491–1498. [DOI] [PubMed] [Google Scholar]
  35. Wing RR, Lang W, Wadden TA, Safford M, Knowler WC, Bertoni AG, Hill JO, Brancati FL, Peters A, Wagenknecht L, Group TLAR, 2011. Benefits of modest weight loss in improving cardiovascular risk factors in overweight and obese individuals with type 2 diabetes. Diabetes Care 34, 1481–1486. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Yamanaka A, Beuckmann CT, Willie JT, Hara J, Tsujino N, Mieda M, Tominaga M, Yagami K-I, Sugiyama F, Goto K, Yanagisawa M, Sakurai T, 2003. Hypothalamic orexin neurons regulate arousal according to energy balance in mice. Neuron 38, 701–713. [DOI] [PubMed] [Google Scholar]

RESOURCES