NTS and VTA oxytocin reduces food motivation and food seeking

Abstract

Oxytocin (OT) is a neuropeptide whose central receptor-mediated actions include reducing food intake. One mechanism of its behavioral action is the amplification of the feeding inhibitory effects of gastrointestinal (GI) satiation signals processed by hindbrain neurons. OT treatment also reduces carbohydrate intake in humans and rodents, and correspondingly, deficits in central OT receptor (OT-R) signaling increase sucrose self-administration. This suggests that additional processes contribute to central OT effects on feeding. This study investigated the hypothesis that central OT reduces food intake by decreasing food seeking and food motivation. As central OT-Rs are expressed widely, a related focus was to assess the role of one or more OT-R-expressing nuclei in food motivation and food-seeking behavior. OT was delivered to the lateral ventricle (LV), nucleus tractus solitarius (NTS), or ventral tegmental area (VTA), and a progressive ratio (PR) schedule of operant reinforcement and an operant reinstatement paradigm were used to measure motivated feeding behavior and food-seeking behavior, respectively. OT delivered to the LV, NTS, or VTA reduced 1) motivation to work for food and 2) reinstatement of food-seeking behavior. Results provide a novel and additional interpretation for central OT-driven food intake inhibition to include the reduction of food motivation and food seeking.

INTRODUCTION

Oxytocin (OT) is a neuropeptide produced by neurons of the hypothalamic paraventricular (PVH) and supraoptic nuclei (SON) whose central receptor-mediated actions in rodents include reduced food intake (6, 7, 9, 49, 57). Consistent with its agonist effects on feeding, deficits in endogenous central OT receptor (OT-R) signaling increase food intake (6–8, 12, 14, 38, 53, 68, 71, 72). Given its contribution to feeding control, there is great interest in the OT system for its potential in obesity treatment development, highlighting the need to better understand the neurobehavioral mechanisms responsible for OT’s intake inhibitory actions (40, 41).

In considering the mechanistic bases of central OT’s anorectic effect, it is worth noting that OT treatment decreases sucrose self-administration and sucrose-seeking behavior in rodents (11, 37, 47, 73), whereas OT-knockout mice show an increase in self-administration and motivation for sucrose compared with wild-type controls (28, 47, 50, 61). These data collectively suggest that OT-R signaling would reduce aspects of appetitive control. Given that the intake-inhibitory effects of peripheral OT are mediated at least in part by central mechanisms (30–32), and central OT reduces high-sugar and high-fat food intake in humans (55) and animals (13, 58), central OT-Rs likely govern these behaviors. However, whether central OT administration and central OT-R activation reduce motivated feeding behaviors is yet to be investigated.

This study tested the hypothesis that central (lateral ventricle, LV) delivery of OT, providing ligands to multiple central OT-R-expressing sites, reduces the motivation to work for and to seek palatable food. This study also sought to determine whether the effects of central OT signaling on food motivation and food seeking can be attributed to one or more OT-R-expressing nuclei. OT-Rs are expressed widely throughout the mammalian brain (22, 33, 66), and reduced food intake has been observed in response to activation of OT-Rs expressed in several nuclei (42, 64), including but not limited to the midbrain ventral tegmental area (VTA) and hindbrain nucleus tractus solitarius (NTS) (47, 52). The role of OT signaling in the NTS and VTA in the control of motivated feeding behavior is of particular interest, given that both nuclei receive direct synaptic connections from PVH OT-producing neurons (56, 70), and that reducing OT-R signaling in both sites increases food intake (47, 53), suggesting a role for endogenous OT-R signaling in feeding inhibitory control in both sites.

Collectively, the data presented here support the hypothesis that central OT, delivered to the LV, NTS, or VTA, reduces food motivation and food seeking. The data overall identify reductions in food motivation and food-seeking behavior as a novel mechanism of central OT-induced intake inhibition and highlight that this mechanism contributes to OT-induced intake inhibition in at least two OT-R-expressing sites, the NTS and VTA.

METHODS

Subjects

Adult male Sprague–Dawley rats (250–265 g on arrival, Charles River Laboratories, Wilmington, MA) were individually housed in wire hanging cages under a 12:12-h light/dark cycle (lights off at 10:30 AM). Rats had ad libitum access to pelleted chow (Laboratory Rodent Diet 5001; LabDiet, St. Louis, MO) and water, unless otherwise stated. All procedures conformed to the institutional standards of the University of Pennsylvania Institutional Animal Care and Use Committee and were consistent with the NIH Guide for the Care and Use of Laboratory Animals.

Surgery

Rats received intramuscular ketamine (90 mg/kg; Midwest Veterinary Supply, Norristown, PA), xylazine (2.7 mg/kg; Anased, Shenadoah, IA), and acepromazine (0.64 mg/kg Midwest Veterinary Supply, Norristown, PA) followed by subcutaneous analgesia (2.0 mg/kg Loxicom; Midwest Veterinary Supply) postsurgery. For ventricular infusion of OT, rats were implanted with a unilateral guide cannula (26 gauge; Plastics One, Roanoke, VA) with the tip positioned 2 mm above the LV (coordinates: 0.9 mm posterior to the bregma, 1.6 mm lateral to the midline, and 2.8 mm ventral from the skull, at a 90°angle). For VTA and NTS infusion of OT, rats were implanted with a bilateral guide cannula (26 gauge; Plastics One) positioned 2.5 mm above the VTA (coordinates: 5.8 mm posterior to the bregma, 6.2 mm ventral from the skull, on the midline with 0.5 mm cannula to cannula) or NTS [coordinates: 1.7 mm anterior to the occipital, 6.55 ventral from the skull, and on the midline with 0.5 mm c-c at a 15° angle (anterior-posterior)].

LV cannula placements were verified by assessing the sympathoadrenal-mediated hyperglycemic response to 5-thio-d-glucose [210 µg in 2 µL of artificial cerebral spinal fluid (aCSF)]. A postinjection increase in blood glucose level of 100% or greater from the baseline was required for subject inclusion. VTA and NTS cannula placements were histologically evaluated postmortem with a 200-nL injection of 2% Pontamine Sky Blue. Representative images of the injection sites are depicted in Fig. 1 for the NTS and in Fig. 2 for the VTA.

Fig. 1.

Representative photomicrograph and schematic depicting injection sites in the NTS. The white arrow designates an injection site. Coordinates are relative to the bregma. NTS, nucleus tractus solitaries. Brain maps are derived from Swanson (64a).

Fig. 2.

Representative photomicrograph and schematic depicting injection sites in the VTA. The white arrow designates an injection site. Coordinates are relative to the bregma. VTA, ventral tegmental area. Brain maps are derived from Swanson (64a).

General Experimental Procedures

All experiments were conducted using a within-subject, counterbalanced design with at least 48-h between experimental conditions. OT (Bachem Americas Inc., H-2510) was dissolved in sterile water (Fisher BioReagents). For LV injection, OT or vehicle was administered with a microsyringe at a volume of 1 μL manually. For NTS and VTA injections, OT or vehicle was injected using a Harvard Apparatus PHD 2000 infusion pump at a delivery rate of 2.6 μL/min at a volume of 100 nL. NTS and VTA injections were made unilaterally, confining injection volumes to injection targets and allowing for multiple tests without compromising tissue integrity. After all injections, injectors were left in place for 30 s to allow for drug diffusion before withdrawal.

Behavioral Procedures

Chow intake.

Short-term food intake was measured to evaluate whether a given dose of centrally delivered OT suppressed food intake. Food was removed 2 h before dark onset, and rats received an injection of vehicle or OT immediately before dark, when food was replaced. Food intake was measured 0.5 h after dark onset, accounting for spillage. Only 0.5 h was measured, as previous studies showed that the effects of OT are quite short (30, 47, 52).

Motivation to work for palatable food: progressive ratio schedule of reinforcement.

Rats were habituated to 45-mg sucrose pellets (Bio-Serv, Frenchtown, NJ) in their home cages and were trained to lever-press for sucrose, as previously described (1, 2, 34). Two levers were presented in the operant chamber (Med Associates, San Diego, CA), an active lever and an inactive lever. Active-lever presses resulted in sucrose delivery, whereas inactive-lever presses had no programmed consequences and served as a control for nonconditioned rates of operant responding. Briefly, rats were trained to press a lever to obtain pellets first on a fixed ratio (FR)1, then FR3 and FR5 schedules of reinforcement (1, 3, and 5 lever presses required to receive one pellet, respectively). Rats were then moved to a progressive ratio (PR) schedule of reinforcement, where the effort to obtain each pellet increased exponentially throughout the session, using the formula F(i) = 5e^0.2i − 5, where F(i) is the number of lever presses to obtain the next pellet at i = pellet number. The PR session ended when a 20-min period elapsed without the rat earning a pellet. Testing commenced once a stable PR baseline was established (active-lever presses within 15% of average responding over 3 days). Animals were injected with OT or its vehicle immediately before test sessions, and responses on the active and inactive levers as well as the number of pellets earned were recorded 30 min from the start of the session.

Palatable food seeking: pellet and cue-primed reinstatement.

A reinstatement paradigm was used to evaluate food seeking. The reinstatement paradigm had three phases: self-administration training, extinction, and reinstatement testing. These procedures were adapted from previously published work in other laboratories (20, 62) and our own (54).

self-administration training.

Rats were trained to self-administer 45-mg chocolate pellets (12.7% fat and 66.7% carbohydrate; TestDiet, Richmond, IN). Chocolate pellets were used because rats had already formed learned associations with 45-mg sucrose pellets in the PR paradigm. In addition, prior work has described that mixed-meal pellets provide more robust reinstatement than sucrose pellets (15). Similar to the PR paradigm, active-lever presses resulted in chocolate pellet delivery, whereas inactive-lever presses did not. During training, rats earned a chocolate pellet paired with a discrete light-tone cue under a fixed ratio-1 20-s timeout reinforcement schedule. Active-lever presses during the timeout period did not yield any chocolate pellets or cues. Once there was a stable baseline on FR1 20-s timeout (active-lever presses within 15% of average responding over 3 days), rats proceeded to the extinction phase. Responses on the active and inactive levers as well as pellets earned were recorded.

extinction.

During extinction, responses on the active lever led to neither presentation of the light-tone cue nor delivery of a chocolate pellet. Behavior was considered to be extinguished when rats met the extinction criterion (≤25 responses on the active lever for 2 consecutive days). Responses on the active and inactive levers were recorded at 30 min from the start of the session.

pellet and cue-primed reinstatement testing.

Animals were injected with OT or its vehicle immediately before reinstatement testing. At the start of reinstatement testing, three chocolate pellets were provided in the food cup to serve as a prime, and responding on the active lever led to contingent presentations of the discrete light-tone cue; however, additional pellets were not delivered. Responses on the active and inactive levers were recorded. Reinstatement was evaluated using a within-subject design, so that after each drug condition, rats were switched back to extinction and were only evaluated under the reinstatement paradigm for another condition after reaching the extinction criterion.

Experiment 1: LV

Experiment 1a: to determine the effects of LV OT delivery on dark cycle chow intake.

Rats (n = 9) received a 1-µL injection of 0, 0.25, or 1 μg OT immediately before the onset of the dark cycle using a within-subject counterbalanced design such that all rats received all treatments. Chow intake was analyzed at 0.5 h after injection/dark onset. Based on this experiment and those published previously (6, 7, 9, 49, 57), 1 µg was used in all subsequent LV experiments.

Experiment 1b: to assess whether LV OT delivery affects motivation to work for palatable food.

A cohort of rats (n = 13) received a 1-µL injection of 0 or 1 μg OT to the LV immediately before the PR test session using a within-subject counterbalanced design such that all rats received all treatments.

Experiment 1c: to determine whether LV OT delivery affects food-seeking behavior.

Rats from experiment 1b (n = 13) were also used in this experiment. Rats received a 1-µL injection of 0 or 1 μg OT to the LV immediately before the reinstatement test session using a within-subject counterbalanced design such that all rats received all treatments.

Experiment 2: NTS

Experiment 2a: to determine effects of NTS OT delivery on dark-cycle chow intake.

Rats (n = 11) received a 100-nL injection of OT (0, 0.25, or 1μg) immediately before onset of the dark cycle, and chow intake was measured after 0.5 h using a within-subject counterbalanced design such that all rats received all treatments. The doses were chosen based on experiment 1a and published work showing that unilateral NTS injection of 0.3 μg did not significantly reduce chow intake, whereas unilateral NTS injection of 1 μg OT did (52).

Experiment 2b: to evaluate whether NTS OT delivery affects motivation to work for palatable food.

One group of rats (n = 11) received a 100-nL unilateral injection of 0 or 0.25 μg OT to the NTS immediately before the PR test session using a within-subject counterbalanced design such that all rats received all treatments. Another group of rats (n = 11) received a unilateral injection of 0 or 1 µg OT to the NTS using a within-subject counterbalanced design such that all rats received all treatments.

Experiment 2c: to examine whether NTS OT delivery affects food seeking.

The same group of rats from experiment 2a (n = 11) received a 100-nL unilateral NTS-directed injection of 0, 0.25, or 1 μg OT immediately before the reinstatement test session using a within-subject counterbalanced design such that all rats received all treatments.

Experiment 3: VTA

Experiment 3a: to determine effects of VTA OT delivery on dark-cycle chow intake.

Rats (n = 13) received a 100-nL unilateral injection of OT (0, 0.25, or 1 μg) immediately before onset of the dark cycle, and chow intake was measured after 0.5 h using a within-subject counterbalanced design such that all rats received all treatments. The doses were chosen based on experiment 1a and published work showing that unilateral VTA injection of 0.3 μg did not significantly reduce chow intake, whereas unilateral NTS injection of 1 μg OT did (47).

Experiment 3b: to determine whether VTA OT delivery affects motivation to work for palatable food.

The same cohort of rats from experiment 3a (n = 13) received a 100-nL unilateral injection of 0, 0.25, or 1 μg OT to the VTA immediately before the PR test session using a within-subject counterbalanced design such that all rats received all treatments.

Experiment 3c: to evaluate whether VTA OT delivery affects food-seeking behavior.

A separate cohort of rats (n = 11) received a 100-nL unilateral VTA-directed injection of 0 or 1 μg OT immediately before the reinstatement test session using a within-subject counterbalanced design such that all rats received all treatments.

Statistical Analyses

Data were analyzed using Prism 8 for macOS (GraphPad). PR and reinstatement lever presses were analyzed by two-way repeated-measures ANOVA for main effects and post hoc comparisons, either Tukey’s or Sidak’s tests of multiple comparisons, when applicable. Chow intake and pellets earned in PR (VTA test) were analyzed by one-way ANOVA for main effects, and post hoc comparisons were used when applicable. For LV and NTS tests, pellets earned in PR were analyzed using t tests, as effects were only analyzed at one dose and at one time point (0.5 h). α levels were set to 0.05 for all analyses.

RESULTS

Experiment 1a: LV OT Delivery Reduced Chow Intake

There was a significant effect of OT dose on chow intake [F(2,16) = 6.877, P = 0.007]. Pairwise comparisons show that 1 µg/µL OT significantly reduced chow intake versus vehicle (P = 0.005) and 0.25 µg/µL OT was without effect on chow intake (Fig. 3A).

Fig. 3.

LV OT reduces food intake, food motivation, and food-seeking behavior. LV OT reduced chow intake (A) and reduced active-lever presses (B) and pellets earned (C) under a progressive ratio schedule of reinforcement. Pellet plus cue priming increased active-lever presses during reinstatement testing (D), which was reduced by LV OT delivery (E). LV, lateral ventricle; OT, oxytocin. *P < 0.05 and error bars represent SE.

Experiment 1b: LV OT Delivery Reduced Motivation to Work for Palatable Food

There was a significant effect of LV OT treatment on operant responding (dose × lever interaction [F(1,12) = 6.707, P = 0.0237]). Figure 3B shows that LV OT decreased active-lever presses at 1 µg/µL (P = 0.004) when compared with the vehicle, and there was no significant difference in inactive-lever presses between the vehicle- and OT-treated animals. Figure 3C shows that rats treated with 1 µg/µL OT also earned significantly fewer pellets than those treated with the vehicle [t(12) = 3.092, P = 0.009].

Experiment 1c: LV OT Delivery Reduced Palatable Food-Seeking Behavior

The reinstatement paradigm was first validated by comparing active-lever presses under the extinction phase versus vehicle treatment during reinstatement testing. There was a significant effect of phase on operant responding (phase × lever interaction [F(1,12) = 26.54, P = 0.0002]). As shown in Fig. 3D, rats treated with the vehicle during reinstatement testing significantly increased active-lever presses (P < 0.0001) compared with lever presses under the last 2 days of extinction, indicating reinstatement of food-seeking behavior. There was no significant difference in inactive-lever pressing between extinction and reinstatement testing. Analysis of LV OT on reinstatement of palatable food seeking revealed a significant effect of treatment on operant responding (dose × lever interaction: [F(1,12) = 7.348, P = 0.0189]). Figure 3E shows that 1 µg/µL OT reduced active-lever presses during reinstatement testing (P = 0.001) when compared with the vehicle, and there was no significant difference in inactive-lever presses between vehicle and OT treatment.

Experiment 2: NTS

Experiment 2a: NTS OT delivery reduced chow intake.

There was a main effect of OT treatment on chow intake [F(2,14) = 6.267, P = 0.0114]. Pairwise comparisons revealed that 1 µg/100 nL OT (P = 0.0065) but not 0.25 µg/100 nL OT significantly reduced chow intake versus vehicle (Fig. 4A). Out of the 11 animals that underwent all experimental conditions, eight animals that had cannula targeted at the NTS were included in the final analyses for this experiment and experiment 2c.

Fig. 4.

NTS OT delivery reduces food intake, food motivation, and food seeking. NTS OT delivery reduced chow intake (A). NTS OT reduced active-lever presses and pellets earned in a progressive ratio schedule of reinforcement with 0.25 µg OT (B and C) and active-lever presses and pellets earned under a progressive ratio schedule of reinforcement with 1 µg OT (D and E). Pellet plus cue priming increased active-lever presses during reinstatement testing (F), which was reduced by NTS OT delivery (G). *P < 0.05 and error bars represent SE. NTS, nucleus tractus solitaries; OT, oxytocin.

Experiment 2b: NTS OT delivery reduced motivation to work for palatable food.

0.25 µg OT.

There was a significant effect of NTS OT treatment on operant responding (dose × lever interaction [F(1,6) = 9.012, P = 0.0239]). Figure 4B shows that NTS delivery of 0.25 µg/100 nL OT reduced active-lever presses and not inactive-lever presses when compared with the vehicle (P = 0.01). Rats treated with 0.25 µg/100 nL OT also earned significantly fewer pellets than those treated with the vehicle [t(6) = 3.437 P = 0.0139] (Fig. 4C). Of the 11 rats that underwent all treatments and experimental conditions, 7 animals that had cannula correctly targeted at the NTS were included in the final analyses.

1 µg OT.

There was a significant effect of OT treatment on operant responding (dose × lever interaction [F(1,6) = 6.517, P = 0.0433]). Figure 4D shows that NTS delivery of 1 µg OT reduced active-lever presses when compared with vehicle treatment (P = 0.021). There was no significant difference in inactive-lever presses between treatments. Rats treated with 1 µg/100 nL OT earned significantly fewer pellets than those treated with the vehicle [t(6) = 2.545, P = 0.0438] (Fig. 4E). Of the 11 rats that underwent all treatments and experimental conditions, 7 animals that had cannula correctly targeted at the NTS were included in the final analyses.

Experiment 2c: NTS OT delivery reduced palatable food-seeking behavior.

The reinstatement paradigm was first validated by comparing active-lever presses under the extinction phase versus vehicle treatment during reinstatement testing. There was an effect of phase on operant responding (phase × lever interaction [F(1,7) = 27.01, P = 0.0013]). As shown in Fig. 4F, rats treated with the vehicle during reinstatement testing significantly increased active-lever presses compared with lever presses under the last 2 days of extinction (P = 0.0004), and there was no significant difference in inactive-lever presses during extinction and when treated with the vehicle during reinstatement testing. Analysis of NTS OT on reinstatement of palatable food seeking revealed a significant effect of treatment on operant responding (lever × dose interaction [F(2,14) = 20.96, P < 0.0001]). Figure 4G shows that NTS parenchymal injection of OT reduced active-lever presses during reinstatement testing at 0.25 µg/100 nL (P = 0.017) and 1 µg/100 nL (P = 0.029) compared with vehicle responding. There was no significant difference in inactive-lever presses between vehicle and either dose of OT during reinstatement testing.

Experiment 3: VTA

Experiment 3a: VTA OT delivery reduced chow intake.

There was a main effect of OT treatment on food intake [F(2,20) = 4.144, P = 0.031]. Pairwise comparisons showed that 1 µg/100 nL OT (P = 0.0185) significantly reduced chow intake when compared with the vehicle (Fig. 5A). There was no effect of 0.25 µg/100 nL OT on chow intake. Of the 13 rats that underwent all treatments and experimental conditions in experiments 3a and 3b, 11 animals that had cannula correctly targeted at the VTA were included in the final analyses.

Fig. 5.

VTA OT delivery reduced food intake, food motivation, and food seeking. VTA OT delivery reduced chow intake (A) and active-lever presses (B) and pellets earned (C) in a progressive ratio schedule of reinforcement. Pellet plus cue priming increased active-lever presses during reinstatement testing (D), which was reduced by VTA OT delivery (E). *P < 0.05 and error bars represent SE. VTA, ventral tegmental area; OT, oxytocin.

Experiment 3b: VTA OT delivery reduced motivation to work for palatable food.

There was a significant effect of VTA treatment on operant responding (dose × lever interaction [F(2,18) = 12.69, P = 0.0004). Figure 5B shows that VTA delivery of 1 µg/100 nL OT but not 0.25 µg/100 nL OT significantly reduced active-lever presses when compared with the vehicle (P < 0.0001). There was also no difference in inactive-lever presses between the vehicle and both 0.25/100 nL and 1 µg/100 nL OT treatment. There was an effect of OT on pellets earned [F(2,18) = 11.45, P = 0.0006], where OT reduced the number of pellets earned in PR at 1 µg/100 nL (P = 0.0004) but not 0.25 µg/100 nL when compared with the vehicle (Fig. 5C).

Experiment 3c: VTA OT delivery reduced palatable food-seeking behavior.

The reinstatement paradigm was validated by first comparing active-lever presses under extinction versus vehicle treatment during reinstatement testing. There was a significant effect of phase on operant responding (phase × lever interaction [F(1,5) = 35.25, P = 0.0019]). As shown in Fig. 5D, animals treated with the vehicle during reinstatement testing significantly increased active-lever presses compared with lever presses during the last 2 days of extinction (P = 0.006). There was no significant difference in inactive-lever presses during extinction and when treated with the vehicle during reinstatement testing. Analysis of VTA OT on reinstatement of palatable food seeking revealed a significant effect of treatment on operant responding (lever × phase interaction [F(1,5) = 12.70, P = 0.0162]). As shown in Fig. 5E, OT reduced active-lever presses at 1 µg/100 nL (P = 0.006) when compared with vehicle, and there was no significant difference in inactive-lever presses between vehicle and OT treatment during reinstatement testing. Of the 11 rats that underwent all treatments and experimental conditions in experiment 3c, 6 animals that had cannula correctly targeted at the VTA were included in the final analyses.

DISCUSSION

This study investigated the hypothesis that the reduction in food intake triggered by central OT signaling is mediated by decreasing food seeking and food motivation. In support of this hypothesis, central OT delivery achieved with LV injection as well as targeted OT delivery to the NTS and VTA reduced operant responding in both progressive ratio and reinstatement paradigms. Data support the hypothesis that reduced food motivation and food seeking contribute to central OT-induced feeding inhibition.

The finding that targeted NTS OT administration reduces food motivation and food seeking is notable for two reasons. First, NTS contributions to feeding have typically centered on its role in the processing and integration of vagally mediated satiation signals (24), rather than on influencing motivated feeding behavior. The current results complement a growing base of evidence that peptidergic signaling within the NTS, including glucagon-like peptide 1 (GLP-1) (2) and leptin (34) receptor signaling, contributes to intake inhibition via reductions in food motivation and food seeking. Together, these studies expand on the conventional view of NTS contributions in intake inhibitory control. Second, the current findings support that OT reduces food intake by acting through more than one mechanism, as previous studies show that OT’s intake-inhibitory effects are mediated by amplifying the neural processing of GI satiation signals in NTS neurons (12, 14, 30, 52, 53, 56). The contribution of both mechanisms to OT-induced intake inhibition may be akin to the multimodal control of intake inhibition resulting from the central actions of GLP-1 (2, 3, 27). The current data, therefore, expand OT’s intake inhibitory action in the NTS to include not only amplifying the neural processing of satiation signals but also reducing food motivation and food seeking.

The reduction in motivated behavior and amplification of satiation signal processing contributions to OT’s intake inhibitory actions may also be interrelated. Given that animals in this study consumed sucrose or chocolate pellets during PR and pellet-primed reinstatement tests (respectively), and that such consumption likely engaged postingestive GI processes, it is possible that the effects of OT on aspects of appetitive control are either secondary to or at least influenced by OT action to amplify GI satiation signal processing. The possibility that pellet consumption coupled with NTS OT delivery engaged such processing is supported by experiments showing that OT delivery to the NTS amplifies the intake-inhibitory effects of consuming a mixed-meal preload (52). Although the link between these two mechanisms is not known, electrophysiological work by Peters et al. (56) identified vagal afferent drive and also PVH OT projections to NTS neurons as the signaling mechanism responsible for amplifying GI signals (32), providing insight into what could be a potentially shared synaptic mechanism. The study by Peters et al. (56) as well as the work described here involved broad targeting across the anterior-posterior extent of the NTS, and it is important to note that different populations of NTS neurons serve some distinguishable functions. More specifically, rostral NTS neurons are innervated by cranial nerve afferents from oral sensors that transmit information about sweet taste (16, 25, 26), whereas caudal NTS neurons are innervated by vagal afferents transmitting information from the GI tract (23). Future work is required to determine whether different populations of NTS neurons are responsible for each mechanism underlying OT-induced intake inhibition or whether they overlap.

Targeted OT delivery to the VTA also reduced food motivation and food seeking, a finding consistent with the well-known contribution of the VTA to reinforcing and motivated behaviors, as well as responding to cues that predict rewards (21, 59, 60, 67). These data are also well supported by findings that many other energy status- and intake-relevant signals like GLP-1 (19), leptin (17), amylin (45), and insulin (39) also interact with the mesolimbic system to reduce aspects of motivated feeding behaviors. The mechanisms through which VTA OT reduces food motivation and food seeking are unknown, but recent work by Xiao et al. suggests a promising mechanism. Xiao et al. (69) found that VTA OT-R activation initiates retrograde endocannabinoid signaling that dampens glutamatergic transmission onto VTA neurons, which is a synaptic mechanism known to underlie VTA insulin receptor-induced reductions in appetitive feeding behavior (39). Although promising, future studies are required to link this synaptic mechanism with OT-induced reductions in food seeking and food motivation in the VTA.

Targeted delivery of OT to the NTS and the VTA reduced food motivation and food seeking, and although this suggests a level of functional redundancy across multiple OT-R-expressing sites, VTA dose requirements were greater than those in NTS. The highest dose tested (1 µg/100 nL) was required to reduce operant responding in the VTA, whereas NTS delivery of 0.25 µg/100 nL was sufficient (although also observed with 1 µg). These findings align with previous work showing that a ventricle-effective dose (1 µg) was required to reduce sucrose intake after targeted OT delivery to the VTA (47), as well as other sites outside of the hindbrain like the nucleus accumbens core (NAcc) (29), the basolateral (BLA), and central (CEA) (37) nucleus of the amygdala. Although the basis for the higher VTA dose requirement is unknown, it appears that OT-R expression might be greater in the NTS than in the VTA (and NAcc BLA and CEA), increasing NTS sensitivity to lower doses, although the available literature on this topic is mixed and at this point inconclusive (22, 33, 66). Although there is a wealth of evidence to suggest that OT’s intake inhibitory actions are a result of actions at the OT-R (6–8, 12, 14, 38, 53, 68, 71, 72), justifying our speculation about relative OT-R expression, it is worth noting here the possibility that the differential dose sensitivity here as well as our OT-induced reductions in food motivation and food-seeking behaviors could result from interactions at vasopressin receptors. The current experiments focus on oxytocin, the endogenous ligand for the OT-R; however, oxytocin also has equimolar affinity to the vasopressin receptor (22, 33), which is also expressed in the NTS and the VTA (66). Future work should not only investigate relative OT-R expression but also the relative receptor density of vasopressin receptors to better characterize the differences in dose sensitivity in these two nuclei as well as the receptors underlying the behavioral effects described.

There is support for the perspective that the VTA effects described here are site specific and feeding relevant and are not solely due to pharmacological levels of OT. Direct VTA injection of two different OT-R antagonists increases sucrose intake (47), suggesting that VTA OT-Rs are physiologically relevant to intake inhibition and the current effects are unlikely to be nonspecific to VTA, OT-Rs, or feeding. It is also unlikely that the higher OT dose inadvertently targeted OT-Rs expressed elsewhere in proximity to the VTA, like the substantia nigra (SN) (70). Targeted SN OT delivery reduces physical activity (5), but the intraVTA injections here did not significantly reduce inactive lever responding, a result inconsistent with drug spread to the SN. It is also unlikely that 1 µg OT induced visceral illness, another factor that reduces activity, given the nonsignificant inactive lever effects and the strong evidence that central OT does not induce visceral malaise (13, 29, 32, 35, 37, 48, 71). Such evidence suggests that the current VTA results are meaningful and highlights that more work is needed to more fully evaluate the functional contributions of VTA OT-Rs to motivated feeding behavior.

NTS and VTA OT administration also reduced chow intake requiring the higher OT dose (1 µg/100 nL) for both sites. In the NTS, the higher dose requirement for chow inhibition is discordant with the lower dose requirement (0.25 µg/100 nL) for operant responding. The differential NTS dose requirements may result from differences in the palatability of the test food (i.e., chow vs. sucrose vs. chocolate pellets) used across the experiments. A growing body of work in rodents suggests that the intake-inhibitory effects of OT are specific to reducing palatable carbohydrate intake (36), as rodent models with deficiencies in OT and OT-R signaling overeat carbohydrate solutions (4, 38, 61, 68) when compared with intact controls. In addition, peripheral OT administration has been shown to reduce sucrose palatability but not the palatability of nonsweet tastants (63).The PR and reinstatement paradigms used here measured operant responding for pellets high in sucrose (sucrose pellets, 64.37%; chocolate pellets, 49.60%) that contrast with the low sucrose content in chow (3.7%). The fact that the same dose of OT delivered to the NTS led to a reduction in responding for food high in sucrose and did not reduce chow intake could further support that central OT-mediated feeding effects are somewhat carbohydrate selective. However, other studies show that central OT reduces high-fat diet intake in rodents (13, 18, 43, 44, 46, 58, 71, 72) and in humans (55), suggesting that OT’s intake effects are not limited to high-sucrose foods or to palatable carbohydrates. Future studies are needed to determine whether OT-induced reductions in food seeking and food motivation are macronutrient dependent.

Perspectives and Significance

This study suggests a new mechanism through which central OT reduces food intake—via reductions in food seeking and food motivation. The data also suggest that OT’s effects on feeding are mediated by more than one mechanism. Results also showed that OT delivery to the NTS and VTA reduces food motivation and food seeking, highlighting that the role of OT in motivated feeding behavior is neuroanatomically distributed, and exhibits some functional redundancy. As OT is receiving attention as a potential target for obesity therapy (10, 40, 41, 51, 65) for its enhanced intake inhibitory efficacy in animals with diet-induced obesity (13, 58) and in humans with obesity (65), further investigations into the mechanisms of central OT action, especially as they pertain to palatable food intake and motivated behavior, are warranted in animals with diet-induced obesity to guide the development of effective OT-based pharmacotherapies.

GRANTS

This work was supported by National Institutes of Health Grants DK-21397 (to H. J. Grill and M. R. Hayes) and F31-120162 (to H. S. Wald).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

H.S.W., Z.Y.O., and H.J.G. conceived and designed research; H.S.W., A.C., and A.K. performed experiments; H.S.W., A.C., and A.K. analyzed data; H.S.W., Z.Y.O., M.R.H., and H.J.G. interpreted results of experiments; H.S.W. prepared figures; H.S.W. drafted manuscript; H.S.W., Z.Y.O., M.R.H., and H.J.G. edited and revised manuscript; H.S.W., A.C., A.K., Z.Y.O., M.R.H., and H.J.G. approved final version of manuscript.