Imagine that you have just finished Thanksgiving dinner. You are completely full, having eaten large portions of turkey, green beans and mashed potatoes. Yet, despite feeling full, you still find yourself tempted by a slice of pie for dessert, maybe even with ice cream on top. Why is it that in such a state of fullness, you desire a slice of pie but not, say, another helping of green beans?

The answer may lie in the way the brain responds to food when we do not need any more calories. At such times, your brain drives you to continue eating only those foods that are tasty and calorie-dense. This preference for fatty and sweet foods may have been helpful in the past when we could not be certain where our next meal would come from. But in modern times, the widespread availability of food makes this preference potentially harmful. For example, the drive to consume fatty and sweet foods even when not hungry may now be contributing to soaring levels of obesity and type 2 diabetes.

What exactly is happening inside the brain to produce this behavior? Previous work has implicated a structure called the nucleus accumbens. When scientists activated proteins called mu opioid receptors within the nucleus accumbens, animals ate more of the foods that they find tasty. However, they were not as interested in eating more of the foods that they are more ambivalent towards.

Caref and Nicola now show that preventing opioid binding makes rats unwilling to respond to a cue to obtain cream, an appetizing, high-fat reward. It also abolishes the brain activity that drives the rats to respond the cue. Crucially, however, this effect only occurs in rats that are not hungry.

It therefore appears that opioid binding in the nucleus accumbens drives animals to approach and eat high-fat foods, but only when they do not need the calories. That is, it increases fat consumption in animals that are not actually hungry. A drug that selectively blocks mu opioid receptors in the nucleus accumbens may reduce this behavior. Such a drug could potentially help to prevent obesity and the health problems associated with it.

People often seek and consume calorie-dense food in the absence of hunger, and this behavior has profound implications for human health. Although preference for sweet and fatty foods may once have been adaptive, it now very likely contributes to epidemic rates of obesity and diabetes. Thus, understanding the neural mechanisms that guide seeking of highly palatable foods is an important step in the search for novel therapies that could combat these diseases by reducing caloric intake. One candidate neural substrate is the brain’s opioid system, particularly in the ventral striatum. A role for this circuitry is supported by observations that the ventral striatum, and in particular the nucleus accumbens (NAc), is richly endowed with both opioid peptides and their respective receptors (Mansour et al., 1988), and that activation of NAc µ-opioid receptors (MORs) selectively augments consumption of palatable food (Bakshi and Kelley, 1993; Mucha and Iversen, 1986; Zhang et al., 1998; Zhang and Kelley, 1997) and of preferred flavors (Woolley et al., 2006). Moreover, activation of NAc MORs increases hedonic taste reactions to palatable food (Peciña and Berridge, 2000). Thus, opioids are thought to contribute primarily to the encoding of hedonic responses to food, which in turn reinforces the assignment of incentive salience to cues associated with palatable reward (Berridge, 2009; Castro and Berridge, 2014).

However, several observations indicate that this view is incomplete. First, blockade of NAc MORs does not consistently reduce calorie-dense food consumption (Bodnar et al., 1995; Kelley et al., 1996; Lardeux et al., 2015; MacDonald et al., 2003; Ward et al., 2006). In addition, activation of NAc MORs increases certain measures of reward-seeking behavior, including breaking point on a progressive-ratio task (Zhang et al., 2003) and lever pressing in the presence of food-predictive cues in a Pavlovian-to-instrumental transfer task (Peciña and Berridge, 2013). These studies suggest that NAc MOR activation could promote food-seeking behavior directly, instead of (or in addition to) doing so by enhancing the hedonic or reinforcing effects of the food. Consequently, we hypothesize that when people or animals are sated, their preferences shift toward palatable food because endogenous ligands of NAc MORs selectively promote seeking of calorie-dense foods. This idea has not yet been tested because few studies have examined the contribution of NAc MORs activated by endogenous ligands to appetitive (food-seeking) as opposed to consummatory behaviors.

Here, we address this gap in our knowledge by using a conditioned-stimulus (CS) task that disambiguates appetitive from consummatory behavior. In this task, rats perform an approach response to a reward-predictive cue to obtain a highly palatable, calorie-dense liquid food (cream). NAc neurons encode both cued approach and reward consumption phases of such behaviors (Ambroggi et al., 2011; du Hoffmann and Nicola, 2014; McGinty et al., 2013; Morrison et al., 2017; Nicola, 2010; Nicola et al., 2004a; 2004b; Taha and Fields, 2005), and cue-evoked excitations are necessary for the approach response (Ambroggi et al., 2008; du Hoffmann and Nicola, 2014; Yun et al., 2004). By simultaneously recording from NAc neurons and injecting a MOR antagonist into the NAc, we show that NAc MOR activation is required for both behavioral responding to reward-predictive cues and the neural encoding of those cues by NAc neurons. Importantly, these effects were observed in ad libitum chow-fed rats but not in those that had been food restricted. This striking dichotomy indicates that activation of NAc MORs promotes the approach to palatable food only in the absence of a homeostatic need for calories – i.e., hunger – suggesting that these receptors contribute to a neural mechanism that drives intake of calorie-dense food specifically in the state of satiety.

In states of relatively high satiety, humans and animals greatly favor calorie-dense foods over less palatable options – a preference that likely contributes to overconsumption and obesity. Our results reveal a potential neural mechanism underlying this preference. We find that blockade of MORs in the NAc core attenuates both cue-evoked approach to high-fat food and the encoding of those cues by NAc neurons, and that these effects are observed only in rats fed ad libitum chow, and not in food-restricted (relatively hungry) rats. These effects could not be attributed to changes in consumption-related behavior or firing. Notably, NAc cue-evoked excitations are causal to cued approach (du Hoffmann and Nicola, 2014), suggesting that a novel and fundamentally important role of the NAc opioid system is to promote approach to highly palatable food specifically when there is no immediate homeostatic need for calorie intake. Together, these findings suggest NAc MORs as a target for development of treatments that limit overeating, consistent with the present use of drugs that block MORs as viable therapeutic options for the treatment of obesity (Apovian, 2016; Ziauddeen et al., 2013).

Although the NAc opioid system has long been implicated in the regulation of food intake (Castro and Berridge, 2014; Kelley et al., 2005; Nicola, 2016; Peciña et al., 2006; Selleck and Baldo, 2017), the MOR effects identified here are characterized by several features that were not necessarily predicted by previous studies. We find that activation of NAc MORs by endogenous ligands promotes appetitive behavior by increasing neural activity that drives approach to food, whereas NAc MORs appear to contribute little (if at all) to neural activity related to consumption. The latter conclusion appears to contrast with prior evidence that activation of these receptors by exogenous ligands increases hedonic taste reactions (Peciña and Berridge, 2000), which should be controlled by NAc neuronal activity occurring during consumption. However, because we targeted our electrodes to the NAc core, whereas hedonic taste reactions are promoted by MOR agonist injection in a very specific zone of the NAc shell (Peciña and Berridge, 2000), our results do not preclude the possibility that endogenous opioids promote taste reactivity (and perhaps hedonia) by influencing the consumption-related firing of NAc shell neurons.

On the other hand, injection of MOR agonists into either the NAc core or shell increases consumption of palatable food (Bakshi and Kelley, 1993; Katsuura and Taha, 2014; Mucha and Iversen, 1986; Ward et al., 2006; Woolley et al., 2006; Zhang et al., 1998; Zhang and Kelley, 1997), which must be due to promotion of some form of NAc neuronal activity that drives consumption. Our results suggest that, at least in the core, this form of neural activity is the pre-movement firing of NAc neurons, which can be activated by cues (McGinty et al., 2013; Morrison et al., 2017) and which drives initiation of approach to calorie-dense food (du Hoffmann and Nicola, 2014). Further supporting this idea, activation of MORs in the NAc core by exogenous agonists increases consumption of a high-fat liquid in part by increasing the number of licking bouts (perhaps as a result of increasing the number of approaches to the lickometer) (Katsuura et al., 2011; Lardeux et al., 2015). Moreover, exogenous activation of NAc MORs promotes operant behavior for food reward (Zhang et al., 2003), and in fact is sufficient to increase operant responses to cues predictive of high-calorie food in a Pavlovian-instrumental transfer (PIT) test (Peciña and Berridge, 2013). The latter observation in particular supports the hypothesis that MOR activation directly promotes approach behavior without an intermediate effect on neural activity during consumption because the PIT test is performed in extinction. Further arguing against an intermediate effect on consumption, we observed previously that injection of CTAP into the NAc does not reduce consumption of a high-fat liquid in free fed rats (Lardeux et al., 2015). Finally, we find that unilateral infusion of CTAP into the NAc greatly reduces cue-evoked firing (Figure 6), an effect that could not have been due to reduced consumption or other performance deficit because cued approach performance was unaffected by unilateral infusions (Figure 4). Thus, we conclude that NAc core MORs act primarily to promote food seeking rather than consumption itself.

Our results do not, however, rule out the possibility that MORs contribute to some aspect of the consumption- or reinforcement-related firing of NAc neurons. Indeed, excitations in a small population of NAc neurons encode the value of liquid rewards during consumption (Taha and Fields, 2005); because we did not vary reward value, we may have missed an effect of CTAP on this form of encoding. However, more likely contributors to consumption are the large population of NAc neurons that are inhibited in proportion to the rate of licking during consumption (Taha and Fields, 2005). Together with findings that experimental silencing of NAc neurons drives consumption (Reynolds and Berridge, 2001; Stratford and Kelley, 1997) whereas consumption is interrupted by brief excitation of the NAc (Krause et al., 2010), these observations suggest that naturally occurring reductions in the firing of a population of NAc neurons (possibly containing D1 dopamine receptors and projecting to the lateral hypothalamus; O'Connor et al., 2015) drive consummatory behavior. Our observation that consumption-related inhibitions are not reduced by CTAP indicates that these inhibitions do not depend on MORs, consistent with previous findings that generalized inhibition of the NAc results in nonspecific increases in food intake whereas MOR activation results in preferential increases in consumption of calorie-dense, and especially high-fat, food (Katsuura and Taha, 2014; Ward et al., 2006; Woolley et al., 2006; Zhang et al., 1998; Zhang and Kelley, 1997). This specificity in the case of MOR agonists may be due in part to increased approach to calorie-dense food, perhaps via enhanced firing of NAc core neurons that drive such approach behaviors.

Strikingly, CTAP reduced cued approach when the reward was cream, but not liquid sucrose. The latter observation is consistent with previous studies indicating that activation of NAc MORs induces a preference for fat over carbohydrates (Taha, 2010; Zhang et al., 1998); however, similar manipulations also induce greater preference for the already-preferred flavor of two foods with equivalent nutritional content (Woolley et al., 2006). Although we cannot rule out the possibility that relatively greater preference for (or palatability of) cream vs sucrose was instrumental to the much greater effects of CTAP when the cue predicted cream as opposed to sucrose, the sucrose concentration (3%) was chosen such that the animals’ behavior was similar to that observed with cream reward (Figure 1C,D), suggesting that it was the difference in nutrient content, not preference, that dictated the difference in dependence on NAc MORs. Although cream contains, in addition to fat, small quantities of certain nutrients that are absent from sucrose solution (e.g., protein, lactose), these minor components of cream are unlikely, on their own, to support appetitive approach and consumption in the free-fed state. Thus, our results suggest that NAc MORs specifically promote approach to high-fat foods. Even if flavor-based preference or palatability played a role, we note that preferred palatable foods tend to be calorie dense, supporting a role for NAc MORs in overconsumption that leads to obesity.

Such a role is further supported by the remarkable observation that CTAP affected neither cued approach behavior nor cue-evoked neural activity in food-restricted animals, despite markedly reducing both in rats given ad libitum access to chow. To our knowledge, this is the first report of a satiety state-dependent contribution of endogenous NAc opioids to food-seeking behavior. Indeed, few studies have examined whether blockade of NAc MORs (as opposed to activation with exogenous agonists) impacts food-seeking. One exception is the observation that β-funaltrexamine (β-FNA), a long-lasting MOR antagonist, reduces rats’ speed during runway approach to calorie-dense food (Shin et al., 2010). However, NAc injection of β-FNA has also been shown to reduce spontaneous locomotion (Kelley et al., 1996), suggesting that it may have had non-specific effects. Such effects could potentially also explain the reduction in calorie-dense food consumption after β-FNA injection in the NAc (Bodnar et al., 1995; Kelley et al., 1996; Lenard et al., 2010; Shin et al., 2010). Intriguingly, the MOR antagonist we used, CTAP, does not impair spontaneous locomotion when injected in the NAc (Figure 4—figure supplement 1D), and is also apparently less effective than β-FNA in reducing consumption (Katsuura et al., 2011; Lardeux et al., 2015) although a study in rabbits reports greater reductions in consumption (Ward et al., 2006). One possibility consistent with our results is that CTAP impairs approach to food as opposed to consumption itself; differences in CTAP effects on amount of freely available food consumed could be due to differences in experimental conditions such as size of the test chamber (and thus degree of approach required), species, nutrient content, and satiety state.

The stark difference in effects of CTAP in free fed and restricted animals raises three important topics for further research. The first is to determine the degree of restriction that is sufficient to eliminate the dependence of cued approach on NAc MORs. Although MOR antagonists can reduce food consumption after mild (<24 hr) restriction (Bodnar et al., 1995; Kelley et al., 1996), this may not be the case for cued approach behavior. If endogenous opioids in the NAc promote cued approach to fatty foods when restriction is much less severe than the chronic restriction used here, it would imply that this neural system contributes to caloric intake when meal patterns are more natural than the extremes employed here (freely available chow and severe restriction).

The second question is the mechanism whereby endogenous MOR ligands promote cued approach. As observed previously (Ambroggi et al., 2011; du Hoffmann and Nicola, 2014; McGinty et al., 2013; Morrison et al., 2017; Nicola et al., 2004a), we found that prominent populations of NAc neurons are excited and inhibited by cues that evoke approach behavior (Figure 2A,B). Previously, we established that these changes in firing begin prior to initiation of approach movement, and that the magnitude of the firing changes predicts the latency and speed of approach (McGinty et al., 2013; Morrison et al., 2017). Cue-evoked excitations, but not inhibitions, are dopamine-dependent; because injection of dopamine receptor antagonists reduces both cued approach and cue-evoked excitations (du Hoffmann and Nicola, 2014), the excitations are likely causal to the subsequent approach behavior. These observations suggest that activation of MORs by endogenous opioids could increase cue-evoked excitation via a direct action on NAc neurons, on the glutamatergic terminals that likely drive the excitation, or on inhibitory interneurons or inputs that limit the magnitude of cue-evoked excitation. Because MOR effects tend to be inhibitory, the latter hypothesis is most likely.

Alternatively, MOR activation by endogenous opioids could promote the release of dopamine, which could, in theory, promote greater cue-evoked firing and hence increase the probability of an approach response. This idea is consistent with previous findings that exogenous activation of MORs can increase dopamine levels (Borg and Taylor, 1997; Hipólito et al., 2008; Hirose et al., 2005; Okutsu et al., 2006; Yoshida et al., 1999), and with observations that dopamine release can be modulated at the terminal (Cachope and Cheer, 2014; Wenzel and Cheer, 2018). Moreover, dopamine neurons are strongly regulated by the state of caloric need, with greater activation in higher need states (Meye and Adan, 2014; Nicola, 2016), and in fact dopamine release evoked by food-predictive cues is greater in food-restricted rats than free-fed rats (Aitken et al., 2016; Cone et al., 2014) – an observation that could explain the present finding that NAc cue-evoked excitations are greater in restricted than free-fed animals (Figure 2). According to this hypothesis, dopamine levels in the free-fed state are insufficient to raise cue-evoked firing above the threshold for reliably obtaining a cue-evoked approach response. However, when the subject is in an environment in which calorie dense and/or high-fat food is available, neurons that release the endogenous agonist of NAc MORs in the NAc are activated to release the opioid, and the resulting activation of MORs raises the dopamine level such that the magnitude of cue-evoked firing is sufficient to evoke a behavioral response. In contrast, in food-restricted subjects, the dopamine level is so high that either further increases are not possible, or the cue-evoked firing of NAc neurons is maximal such that further increases in dopamine are without effect.

The third question is the source and nature of the opioid peptides that activate MORs to promote food-seeking. Presumably, the endogenous ligand for NAc MORs is enkephalin released by the large population of D2 receptor-expressing spiny neurons (Gerfen et al., 1990; Mansour et al., 1995). The peptide could be released by the extensive axon collaterals of these neurons within the NAc; alternatively, while it has not been demonstrated in the NAc, opioids can be released somatodendritically and act as a retrograde messengers (Iremonger and Bains, 2009; Wagner et al., 1993; Wamsteeker Cusulin et al., 2013). The conditions under which enkephalin release in the NAc is increased are unknown; however, in the dorsomedial striatum, enkephalin levels are elevated during meal onset (DiFeliceantonio et al., 2012), suggesting that information about the availability of food drives release of the peptide. One possibility is that enkephalin release is tonically promoted when the subject is in an environment in which fatty foods are available; another is that release occurs precisely at cue onset in response to discrete cues that predict fat, but not carbohydrates. Further investigation of the hypothesis that NAc enkephalin release is regulated by fat availability, and the mechanisms by which this could occur, is clearly warranted.

The mechanism we propose here – that enkephalin levels are elevated by fat availability, and these high enkephalin levels promote dopamine release that in turn increases cue-evoked excitations that drive cued approach to high-fat food – is partially speculative, but it is fully consistent with the present and previous results and provides a starting point for further exploration. Importantly, our results indicate that the neural mechanisms underlying appetitive behavior must be considered when studying the contribution of opioids (and other neuromodulators) in the forebrain to food intake regulation. Cued approach is only one form of appetitive behavior, but our demonstration that endogenous opioids bias this form of behavior towards fat seeking suggests that opioids may have similar effects on the neural mechanisms that control other, more complex and/or more cognitive appetitive behaviors, such as deciding among simultaneously-available food options. Although MOR antagonists are currently used to treat obesity (Apovian, 2016; Ziauddeen et al., 2013), a more refined understanding of the impact of endogenous opioids on appetitive behaviors is required to understand how these drugs work, and to identify future targets for more specific and effective treatments that reduce preference for calorie-dense or high-fat foods.

1. 1. Kevin Caref

   (2018) Data from: Endogenous opioids in the nucleus accumbens promote approach to high-fat food in the absence of caloric need

   Available at Dryad Digital Repository under a CC0 Public Domain Dedication.

   https://doi.org/10.5061/dryad.t8v2gg7

1. 1. Aitken TJ
   2. Greenfield VY
   3. Wassum KM

   (2016) Nucleus accumbens core dopamine signaling tracks the need-based motivational value of food-paired cues

   Journal of Neurochemistry 136:1026–1036.

   https://doi.org/10.1111/jnc.13494

   • PubMed
   • Google Scholar
2. 1. Ambroggi F
   2. Ghazizadeh A
   3. Nicola SM
   4. Fields HL

   (2011) Roles of nucleus accumbens core and shell in incentive-cue responding and behavioral inhibition

   Journal of Neuroscience 31:6820–6830.

   https://doi.org/10.1523/JNEUROSCI.6491-10.2011

   • PubMed
   • Google Scholar
3. 1. Ambroggi F
   2. Ishikawa A
   3. Fields HL
   4. Nicola SM

   (2008) Basolateral amygdala neurons facilitate reward-seeking behavior by exciting nucleus accumbens neurons

   Neuron 59:648–661.

   https://doi.org/10.1016/j.neuron.2008.07.004

   • PubMed
   • Google Scholar
4. 1. Apovian CM

   (2016) Naltrexone/bupropion for the treatment of obesity and obesity with Type 2 diabetes

   Future Cardiology 12:129–138.

   https://doi.org/10.2217/fca.15.79

   • PubMed
   • Google Scholar
5. 1. Bakshi VP
   2. Kelley AE

   (1993)

   Feeding induced by opioid stimulation of the ventral striatum: role of opiate receptor subtypes

   The Journal of Pharmacology and Experimental Therapeutics 265:1253–1260.

   • PubMed
   • Google Scholar
6. 1. Berridge KC

   (2009) 'Liking' and 'wanting' food rewards: brain substrates and roles in eating disorders

   Physiology & Behavior 97:537–550.

   https://doi.org/10.1016/j.physbeh.2009.02.044

   • PubMed
   • Google Scholar
7. 1. Bodnar RJ
   2. Glass MJ
   3. Ragnauth A
   4. Cooper ML

   (1995) General, mu and kappa opioid antagonists in the nucleus accumbens alter food intake under deprivation, glucoprivic and palatable conditions

   Brain Research 700:205–212.

   https://doi.org/10.1016/0006-8993(95)00957-R

   • PubMed
   • Google Scholar
8. 1. Borg PJ
   2. Taylor DA

   (1997) Involvement of mu- and delta-opioid receptors in the effects of systemic and locally perfused morphine on extracellular levels of dopamine, DOPAC and HVA in the nucleus accumbens of the halothane-anaesthetized rat

   Naunyn-Schmiedeberg's Archives of Pharmacology 355:582–588.

   https://doi.org/10.1007/PL00004987

   • PubMed
   • Google Scholar
9. 1. Cachope R
   2. Cheer JF

   (2014) Local control of striatal dopamine release

   Frontiers in Behavioral Neuroscience 8:188.

   https://doi.org/10.3389/fnbeh.2014.00188

   • PubMed
   • Google Scholar
10. Software

    1. Caref K

    (2018) neural-analysis, version ccd672e 

    GitHub.

    https://github.com/kcaref/neural-analysis
11. 1. Castro DC
    2. Berridge KC

    (2014) Advances in the neurobiological bases for food 'liking' versus 'wanting'

    Physiology & Behavior 136:22–30.

    https://doi.org/10.1016/j.physbeh.2014.05.022

    • PubMed
    • Google Scholar
12. 1. Cone JJ
    2. McCutcheon JE
    3. Roitman MF

    (2014) Ghrelin acts as an interface between physiological state and phasic dopamine signaling

    Journal of Neuroscience 34:4905–4913.

    https://doi.org/10.1523/JNEUROSCI.4404-13.2014

    • PubMed
    • Google Scholar
13. 1. DiFeliceantonio AG
    2. Mabrouk OS
    3. Kennedy RT
    4. Berridge KC

    (2012) Enkephalin surges in dorsal neostriatum as a signal to eat

    Current Biology 22:1918–1924.

    https://doi.org/10.1016/j.cub.2012.08.014

    • PubMed
    • Google Scholar
14. 1. du Hoffmann J
    2. Kim JJ
    3. Nicola SM

    (2011) An inexpensive drivable cannulated microelectrode array for simultaneous unit recording and drug infusion in the same brain nucleus of behaving rats

    Journal of Neurophysiology 106:1054–1064.

    https://doi.org/10.1152/jn.00349.2011

    • PubMed
    • Google Scholar
15. 1. du Hoffmann J
    2. Nicola SM

    (2014) Dopamine invigorates reward seeking by promoting cue-evoked excitation in the nucleus accumbens

    Journal of Neuroscience 34:14349–14364.

    https://doi.org/10.1523/JNEUROSCI.3492-14.2014

    • PubMed
    • Google Scholar
16. 1. du Hoffmann J
    2. Nicola SM

    (2016) Activation of dopamine receptors in the nucleus accumbens promotes sucrose-reinforced cued approach behavior

    Frontiers in Behavioral Neuroscience 10:144.

    https://doi.org/10.3389/fnbeh.2016.00144

    • PubMed
    • Google Scholar
17. 1. Gerfen CR
    2. Engber TM
    3. Mahan LC
    4. Susel Z
    5. Chase TN
    6. Monsma FJ
    7. Sibley DR

    (1990) D1 and D2 dopamine receptor-regulated gene expression of striatonigral and striatopallidal neurons

    Science 250:1429–1432.

    https://doi.org/10.1126/science.2147780

    • PubMed
    • Google Scholar
18. 1. Hipólito L
    2. Sánchez-Catalán MJ
    3. Zanolini I
    4. Polache A
    5. Granero L

    (2008) Shell/core differences in mu- and delta-opioid receptor modulation of dopamine efflux in nucleus accumbens

    Neuropharmacology 55:183–189.

    https://doi.org/10.1016/j.neuropharm.2008.05.012

    • PubMed
    • Google Scholar
19. 1. Hirose N
    2. Murakawa K
    3. Takada K
    4. Oi Y
    5. Suzuki T
    6. Nagase H
    7. Cools AR
    8. Koshikawa N

    (2005) Interactions among mu- and delta-opioid receptors, especially putative delta1- and delta2-opioid receptors, promote dopamine release in the nucleus accumbens

    Neuroscience 135:213–225.

    https://doi.org/10.1016/j.neuroscience.2005.03.065

    • PubMed
    • Google Scholar
20. 1. Iremonger KJ
    2. Bains JS

    (2009) Retrograde opioid signaling regulates glutamatergic transmission in the hypothalamus

    Journal of Neuroscience 29:7349–7358.

    https://doi.org/10.1523/JNEUROSCI.0381-09.2009

    • PubMed
    • Google Scholar
21. 1. Katsuura Y
    2. Heckmann JA
    3. Taha SA

    (2011) mu-Opioid receptor stimulation in the nucleus accumbens elevates fatty tastant intake by increasing palatability and suppressing satiety signals

    American Journal of Physiology-Regulatory, Integrative and Comparative Physiology 301:R244–R254.

    https://doi.org/10.1152/ajpregu.00406.2010

    • PubMed
    • Google Scholar
22. 1. Katsuura Y
    2. Taha SA

    (2014) Mu opioid receptor antagonism in the nucleus accumbens shell blocks consumption of a preferred sucrose solution in an anticipatory contrast paradigm

    Neuroscience 261:144–152.

    https://doi.org/10.1016/j.neuroscience.2013.12.004

    • PubMed
    • Google Scholar
23. 1. Kelley AE
    2. Baldo BA
    3. Pratt WE
    4. Will MJ

    (2005) Corticostriatal-hypothalamic circuitry and food motivation: integration of energy, action and reward

    Physiology & Behavior 86:773–795.

    https://doi.org/10.1016/j.physbeh.2005.08.066

    • PubMed
    • Google Scholar
24. 1. Kelley AE
    2. Bless EP
    3. Swanson CJ

    (1996)

    Investigation of the effects of opiate antagonists infused into the nucleus accumbens on feeding and sucrose drinking in rats

    The Journal of Pharmacology and Experimental Therapeutics 278:1499–1507.

    • PubMed
    • Google Scholar
25. 1. Krause M
    2. German PW
    3. Taha SA
    4. Fields HL

    (2010) A pause in nucleus accumbens neuron firing is required to initiate and maintain feeding

    Journal of Neuroscience 30:4746–4756.

    https://doi.org/10.1523/JNEUROSCI.0197-10.2010

    • PubMed
    • Google Scholar
26. 1. Lardeux S
    2. Kim JJ
    3. Nicola SM

    (2015) Intermittent-access binge consumption of sweet high-fat liquid does not require opioid or dopamine receptors in the nucleus accumbens

    Behavioural Brain Research 292:194–208.

    https://doi.org/10.1016/j.bbr.2015.06.015

    • PubMed
    • Google Scholar
27. 1. Lenard NR
    2. Zheng H
    3. Berthoud HR

    (2010) Chronic suppression of μ-opioid receptor signaling in the nucleus accumbens attenuates development of diet-induced obesity in rats

    International Journal of Obesity 34:1001–1010.

    https://doi.org/10.1038/ijo.2009.297

    • PubMed
    • Google Scholar
28. 1. MacDonald AF
    2. Billington CJ
    3. Levine AS

    (2003) Effects of the opioid antagonist naltrexone on feeding induced by DAMGO in the ventral tegmental area and in the nucleus accumbens shell region in the rat

    American Journal of Physiology-Regulatory, Integrative and Comparative Physiology 285:R999–R1004.

    https://doi.org/10.1152/ajpregu.00271.2003

    • PubMed
    • Google Scholar
29. 1. Mansour A
    2. Fox CA
    3. Akil H
    4. Watson SJ

    (1995) Opioid-receptor mRNA expression in the rat CNS: anatomical and functional implications

    Trends in Neurosciences 18:22–29.

    https://doi.org/10.1016/0166-2236(95)93946-U

    • PubMed
    • Google Scholar
30. 1. Mansour A
    2. Khachaturian H
    3. Lewis ME
    4. Akil H
    5. Watson SJ

    (1988) Anatomy of CNS opioid receptors

    Trends in Neurosciences 11:308–314.

    https://doi.org/10.1016/0166-2236(88)90093-8

    • PubMed
    • Google Scholar
31. 1. McGinty VB
    2. Lardeux S
    3. Taha SA
    4. Kim JJ
    5. Nicola SM

    (2013) Invigoration of reward seeking by cue and proximity encoding in the nucleus accumbens

    Neuron 78:910–922.

    https://doi.org/10.1016/j.neuron.2013.04.010

    • PubMed
    • Google Scholar
32. 1. Meye FJ
    2. Adan RA

    (2014) Feelings about food: the ventral tegmental area in food reward and emotional eating

    Trends in Pharmacological Sciences 35:31–40.

    https://doi.org/10.1016/j.tips.2013.11.003

    • PubMed
    • Google Scholar
33. 1. Morrison SE
    2. McGinty VB
    3. du Hoffmann J
    4. Nicola SM

    (2017) Limbic-motor integration by neural excitations and inhibitions in the nucleus accumbens

    Journal of Neurophysiology 118:2549–2567.

    https://doi.org/10.1152/jn.00465.2017

    • PubMed
    • Google Scholar
34. 1. Mucha RF
    2. Iversen SD

    (1986) Increased food intake after opioid microinjections into nucleus accumbens and ventral tegmental area of rat

    Brain Research 397:214–224.

    https://doi.org/10.1016/0006-8993(86)90622-0

    • PubMed
    • Google Scholar
35. 1. Nicola SM
    2. Yun IA
    3. Wakabayashi KT
    4. Fields HL

    (2004a) Cue-evoked firing of nucleus accumbens neurons encodes motivational significance during a discriminative stimulus task

    Journal of Neurophysiology 91:1840–1865.

    https://doi.org/10.1152/jn.00657.2003

    • PubMed
    • Google Scholar
36. 1. Nicola SM
    2. Yun IA
    3. Wakabayashi KT
    4. Fields HL

    (2004b) Firing of nucleus accumbens neurons during the consummatory phase of a discriminative stimulus task depends on previous reward predictive cues

    Journal of Neurophysiology 91:1866–1882.

    https://doi.org/10.1152/jn.00658.2003

    • PubMed
    • Google Scholar
37. 1. Nicola SM

    (2010) The flexible approach hypothesis: unification of effort and cue-responding hypotheses for the role of nucleus accumbens dopamine in the activation of reward-seeking behavior

    Journal of Neuroscience 30:16585–16600.

    https://doi.org/10.1523/JNEUROSCI.3958-10.2010

    • PubMed
    • Google Scholar
38. 1. Nicola SM

    (2016) Reassessing wanting and liking in the study of mesolimbic influence on food intake

    American Journal of Physiology-Regulatory, Integrative and Comparative Physiology 311:R811–R840.

    https://doi.org/10.1152/ajpregu.00234.2016

    • PubMed
    • Google Scholar
39. 1. O'Connor EC
    2. Kremer Y
    3. Lefort S
    4. Harada M
    5. Pascoli V
    6. Rohner C
    7. Lüscher C

    (2015) Accumbal D1R Neurons Projecting to Lateral Hypothalamus Authorize Feeding

    Neuron 88:553–564.

    https://doi.org/10.1016/j.neuron.2015.09.038

    • PubMed
    • Google Scholar
40. 1. Okutsu H
    2. Watanabe S
    3. Takahashi I
    4. Aono Y
    5. Saigusa T
    6. Koshikawa N
    7. Cools AR

    (2006) Endomorphin-2 and endomorphin-1 promote the extracellular amount of accumbal dopamine via nonopioid and mu-opioid receptors, respectively

    Neuropsychopharmacology 31:375–383.

    https://doi.org/10.1038/sj.npp.1300804

    • PubMed
    • Google Scholar
41. Book

    1. Paxinos G
    2. Watson C

    (2007)

    The Rat Brain in Stereotaxic Coordinates (Sixth ed)

    Elsevier, Inc.

    • Google Scholar
42. 1. Peciña S
    2. Berridge KC

    (2000) Opioid site in nucleus accumbens shell mediates eating and hedonic 'liking' for food: map based on microinjection Fos plumes

    Brain Research 863:71–86.

    https://doi.org/10.1016/S0006-8993(00)02102-8

    • PubMed
    • Google Scholar
43. 1. Peciña S
    2. Berridge KC

    (2013) Dopamine or opioid stimulation of nucleus accumbens similarly amplify cue-triggered 'wanting' for reward: entire core and medial shell mapped as substrates for PIT enhancement

    European Journal of Neuroscience 37:1529–1540.

    https://doi.org/10.1111/ejn.12174

    • PubMed
    • Google Scholar
44. 1. Peciña S
    2. Smith KS
    3. Berridge KC

    (2006) Hedonic hot spots in the brain

    The Neuroscientist 12:500–511.

    https://doi.org/10.1177/1073858406293154

    • PubMed
    • Google Scholar
45. 1. Reynolds SM
    2. Berridge KC

    (2001) Fear and feeding in the nucleus accumbens shell: rostrocaudal segregation of GABA-elicited defensive behavior versus eating behavior

    The Journal of Neuroscience : The Official Journal of the Society for Neuroscience 21:3261–3270.

    https://doi.org/10.1523/JNEUROSCI.21-09-03261.2001

    • PubMed
    • Google Scholar
46. 1. Selleck RA
    2. Baldo BA

    (2017) Feeding-modulatory effects of mu-opioids in the medial prefrontal cortex: a review of recent findings and comparison to opioid actions in the nucleus accumbens

    Psychopharmacology 234:1439–1449.

    https://doi.org/10.1007/s00213-016-4522-4

    • PubMed
    • Google Scholar
47. 1. Shin AC
    2. Pistell PJ
    3. Phifer CB
    4. Berthoud HR

    (2010) Reversible suppression of food reward behavior by chronic mu-opioid receptor antagonism in the nucleus accumbens

    Neuroscience 170:580–588.

    https://doi.org/10.1016/j.neuroscience.2010.07.017

    • PubMed
    • Google Scholar
48. 1. Stratford TR
    2. Kelley AE

    (1997) GABA in the nucleus accumbens shell participates in the central regulation of feeding behavior

    Journal of Neuroscience 17:4434–4440.

    https://doi.org/10.1523/JNEUROSCI.17-11-04434.1997

    • PubMed
    • Google Scholar
49. 1. Taha SA
    2. Fields HL

    (2005) Encoding of palatability and appetitive behaviors by distinct neuronal populations in the nucleus accumbens

    Journal of Neuroscience 25:1193–1202.

    https://doi.org/10.1523/JNEUROSCI.3975-04.2005

    • PubMed
    • Google Scholar
50. 1. Taha SA

    (2010) Preference or fat? Revisiting opioid effects on food intake

    Physiology & Behavior 100:429–437.

    https://doi.org/10.1016/j.physbeh.2010.02.027

    • PubMed
    • Google Scholar
51. 1. Wagner JJ
    2. Terman GW
    3. Chavkin C

    (1993) Endogenous dynorphins inhibit excitatory neurotransmission and block LTP induction in the hippocampus

    Nature 363:451–454.

    https://doi.org/10.1038/363451a0

    • PubMed
    • Google Scholar
52. 1. Wamsteeker Cusulin JI
    2. Füzesi T
    3. Inoue W
    4. Bains JS

    (2013) Glucocorticoid feedback uncovers retrograde opioid signaling at hypothalamic synapses

    Nature Neuroscience 16:596–604.

    https://doi.org/10.1038/nn.3374

    • PubMed
    • Google Scholar
53. 1. Ward HG
    2. Nicklous DM
    3. Aloyo VJ
    4. Simansky KJ

    (2006) Mu-opioid receptor cellular function in the nucleus accumbens is essential for hedonically driven eating

    European Journal of Neuroscience 23:1605–1613.

    https://doi.org/10.1111/j.1460-9568.2006.04674.x

    • PubMed
    • Google Scholar
54. 1. Wenzel JM
    2. Cheer JF

    (2018) Endocannabinoid Regulation of Reward and Reinforcement through Interaction with Dopamine and Endogenous Opioid Signaling

    Neuropsychopharmacology 43:103–115.

    https://doi.org/10.1038/npp.2017.126

    • PubMed
    • Google Scholar
55. 1. Woolley JD
    2. Lee BS
    3. Fields HL

    (2006) Nucleus accumbens opioids regulate flavor-based preferences in food consumption

    Neuroscience 143:309–317.

    https://doi.org/10.1016/j.neuroscience.2006.06.067

    • PubMed
    • Google Scholar
56. 1. Yoshida Y
    2. Koide S
    3. Hirose N
    4. Takada K
    5. Tomiyama K
    6. Koshikawa N
    7. Cools AR

    (1999) Fentanyl increases dopamine release in rat nucleus accumbens: involvement of mesolimbic mu- and delta-2-opioid receptors

    Neuroscience 92:1357–1365.

    https://doi.org/10.1016/S0306-4522(99)00046-9

    • PubMed
    • Google Scholar
57. 1. Yun IA
    2. Wakabayashi KT
    3. Fields HL
    4. Nicola SM

    (2004) The ventral tegmental area is required for the behavioral and nucleus accumbens neuronal firing responses to incentive cues

    Journal of Neuroscience 24:2923–2933.

    https://doi.org/10.1523/JNEUROSCI.5282-03.2004

    • PubMed
    • Google Scholar
58. Book

    1. Zar JH

    (1999)

    Biostatistcal Analysis (Fourth Edition)

    Prentice Hall.

    • Google Scholar
59. 1. Zhang M
    2. Balmadrid C
    3. Kelley AE

    (2003) Nucleus accumbens opioid, GABaergic, and dopaminergic modulation of palatable food motivation: contrasting effects revealed by a progressive ratio study in the rat

    Behavioral Neuroscience 117:202–211.

    https://doi.org/10.1037/0735-7044.117.2.202

    • PubMed
    • Google Scholar
60. 1. Zhang M
    2. Gosnell BA
    3. Kelley AE

    (1998)

    Intake of high-fat food is selectively enhanced by mu opioid receptor stimulation within the nucleus accumbens

    The Journal of Pharmacology and Experimental Therapeutics 285:908–914.

    • PubMed
    • Google Scholar
61. 1. Zhang M
    2. Kelley AE

    (1997) Opiate agonists microinjected into the nucleus accumbens enhance sucrose drinking in rats

    Psychopharmacology 132:350–360.

    https://doi.org/10.1007/s002130050355

    • PubMed
    • Google Scholar
62. 1. Ziauddeen H
    2. Chamberlain SR
    3. Nathan PJ
    4. Koch A
    5. Maltby K
    6. Bush M
    7. Tao WX
    8. Napolitano A
    9. Skeggs AL
    10. Brooke AC
    11. Cheke L
    12. Clayton NS
    13. Sadaf Farooqi I
    14. O'Rahilly S
    15. Waterworth D
    16. Song K
    17. Hosking L
    18. Richards DB
    19. Fletcher PC
    20. Bullmore ET

    (2013) Effects of the mu-opioid receptor antagonist GSK1521498 on hedonic and consummatory eating behaviour: a proof of mechanism study in binge-eating obese subjects

    Molecular Psychiatry 18:1287–1293.

    https://doi.org/10.1038/mp.2012.154

    • PubMed
    • Google Scholar

Thank you for submitting your article "Endogenous opioids in the nucleus accumbens promote approach to high-fat food in the absence of caloric need" for consideration by eLife. Your article has been favorably evaluated by a Senior Editor and three reviewers, one of whom, Richard D Palmiter (Reviewer #1), is a member of our Board of Reviewing Editors.. The following individual involved in review of your submission has also agreed to reveal their identity: Roger Adan (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

This is a tightly constructed paper showing that opioid action in the nucleus accumbens core (NAc) promote approach to calorie-dense foods during periods of satiety. By combining behavioral measures, infusion of an opioid antagonist and electrophysiological recording, the authors were able to demonstrate that opioids primarily affect cued approach behavior rather than consumption. Importantly, the opioid system had little effect when rats were hungry. Thus, the authors conclude that opioid action in the NAC contributes to non-homeostatic feeding, which likely contributes to obesity.

Essential revisions:

(Points 1, 3, 4, 6, 7 are essential; points 2 and 5 would be nice additions if data are available.)

1) Justify use of different doses of CTAP for different experiments.

2) Consider adding an experiment in which authors investigate the effect of CTAP when mice go from restricted feeding back to ad-lib feeding.

3) Discuss possible sources and timing of the opioid receptor agonist release and the implications.

4) Provide more details regarding whether the same neurons that respond during cue presentation also respond during reward delivery/consumption.

5) Adding data for one night of free feeding of the restricted rats would be nice.

6) Comment on why the yield of recorded neurons is different for ad lib and restricted groups of rats.

7) Is there a ceiling effect for feeding by the restricted rats such that opioids no longer have an effect?

See entire reviews below for more details on essential revisions.

Reviewer #1:

This is a tightly constructed paper showing that opioid action in the nucleus accumbens core (NAc) promote approach to calorie-dense foods during periods of satiety. By combining behavioral measures, infusion of an opioid antagonist and electrophysiological recording, the authors were able to demonstrate that opioids primarily affect cued approach behavior rather than consumption. Importantly, the opioid system had little effect when rats were hungry. Thus, the authors conclude that opioid action in the NAC contributes to non-homeostatic feeding, which likely contributes to obesity.

This study is well designed, the experiments are described clearly and the results discussed in a cogent manner. I have no major concerns.

Reviewer #2:

Caref and Nicola seek to determine the role of endogenous NAc opioids in reward-directed behavior and reward encoding. This work is built on previous findings that Mu opioid receptor activation in the NAc promotes food intake and intake of high fat foods in particular. Much of this work has focused on the delivery of Mu agonists. Here the authors utilize a Mu receptor antagonist to probe contributions from endogenous Mu ligand activity. They utilize a cue-induced approach task and find that, only in ad libitum fed rats, Mu receptor activation is required for cue-evoked behavior and NAc excitations. The latter finding is consistent with previous work from Nicola supporting that cue-evoked activations of NAc neurons are critical for conditioned approach. The authors model their data to demonstrate that the modulation of cue evoked NAc excitations by hunger cannot be attributed to differences in approach latency or motor behavior. Moreover, the dependency of NAc cue-evoked excitations on Mu receptors is not a secondary effect due to alterations in palatability. The work is elegant, the analyses are well-performed and the manuscript makes for an important contribution to our understanding of cue-driven feeding on high fat foods in the absence of caloric need. Thus, the findings enhance our understanding of neural processes that contribute to the obesity epidemic. I have the following questions that hopefully can be addressed to strengthen the manuscript:

1) There is a marked difference in the dose of CTAP used in behavioral experiments (2 or 4ug) versus those experiments that included electrophysiological recordings (8ug). No rationale for dose choice is given nor is there an explanation is given for the much higher dose used during the recordings. Do dose values reflect what was delivered to each side?

2) Given the striking differences between fed and restricted states, do the authors have any data from restricted rats that returned to ad libitum feeding? A demonstration of 'toggling on' the effect of CTAP in rats that were previously food restricted would be very powerful. This could be done at the strictly behavioral level.

3) In the Discussion, the authors speculate that MOR ligand may be sourced from NAc neurons themselves. Are the authors suggesting that the initiation of cue-evoked excitations could lead to enkephalin release that would then, in some feed-forward manner, enhance the ongoing response to the cue. Indeed, some of the data reflect fewer bins as time proceeds from cue onset that were statistically elevated following CTAP administration. Or do the authors speculate that MOR receptor activation is occurring in a less time-restricted manner? As there are no effects on baseline firing rate, the latter idea seems unlikely. Some deeper discussion of the source of MOR ligand and timing of MOR receptor activation is warranted.

4) Concerning the populations of neurons that responded during reward consumption, it is unclear if these are neurons with sustained responses from cue onset or whether some of these neurons were only modulated during reward consumption. In the "Analysis of neural data" section of the Materials and methods, only details regarding cue responses are described. If there are some neurons that respond exclusively during consumption, then modulation of this population by CTAP could potentially contribute to reinforcement. This effect could be masked by the population with ongoing heightened activity from the cue period. Regardless, more detail concerning the identification of consumption-related neurons and their relationship to cue-related neurons is warranted.

5) Food restriction appears to result in several effects on MSN activity: increases the proportion of neurons responding with an excitation, increases the magnitude of excitations, increases the duration of excitations (Figure 2C-F). However, the data from free-fed pre-injection rats demonstrates a relatively higher percentage of neurons responding with excitations and a percentage that is not different from restricted, re-injected rats. Thus, the proportion of neurons exhibiting excitations may not be a reliable predictor of free versus fasted state. Again, it would be interesting to see the effects of one night of free-feeding would be on these measures in restricted rats.

Reviewer #3:

In their article, Caref and Nicola describe an interesting and physiologically important observation that the MOR antagonist CTAP reduced CS responsivity in a food- and state-dependent manner, by means of in vivo pharmacology and single unit recordings. The authors should be applauded for the extensive analysis of their ephys experiment, and the many control experiments that were included. We do, however, have one major concern that we think should be addressed, which applies to some of their ephys findings.

In their Results, in the first paragraph of the subsection “Neural encoding of reward-predictive cues by NAc neurons is different in free-fed and food-restricted rats”, they say that in the ad lib condition, 83 neurons were recorded from 12 rats, whereas in the food restricted condition, 122 neurons were recorded from 5 rats. This is a very big difference in yield between these condition; ~7 neurons/animal ad lib, versus ~24 neurons/animal during food restriction. We suspect that the baseline firing rate of NAcc neurons increases due to the food restriction, which may have resulted in a selection bias of neurons in the ad lib condition. As such, recorded neurons in the food restricted condition will consist of a higher variety of neurons than in the ad lib fed condition, in which potentially only from a subpopulation of highly active neurons was recorded.

The authors should clarify why the yield per animal is so different between the conditions, and otherwise experimentally or analytically show that the population of neurons in the ad lib condition are similar to those recorded from in the food restricted condition. If this is not the case, the comparison that is made between neural activity between the two conditions (e.g. in Figure 2) should be interpreted with caution.

Related to this there may be a ceiling effect of increased responding in food restricted rats such that motivation in this task is so high that opioids do not modulate it.

We thank Dr Palmiter for his supportive comments.

Reviewer #2:

[…] 1) There is a marked difference in the dose of CTAP used in behavioral experiments (2 or 4ug) versus those experiments that included electrophysiological recordings (8ug). No rationale for dose choice is given nor is there an explanation is given for the much higher dose used during the recordings. Do dose values reflect what was delivered to each side?

We thank reviewer #2 for the supportive comments. While we show in Figure 1C that the 4 µg/side dose significantly attenuates responding to the CS+, it does not block it completely, particularly early in the behavioral session (i.e., immediately following the drug injection). Thus, when performing CTAP injections during our electrophysiology experiments, we were concerned that the drug effect might be masked in the ~33-minute post-injection window. To mitigate this possibility, we used a higher dose (8 µg/side), which is in line with doses used previously in the literature (5 – 8 µg) (Lardeux, Kim, and Nicola, 2015; Perry and McNally, 2013; Ward, Nicklous, Aloyo, and Simansky, 2006). Moreover, because unilateral injections have no effect on behavior, the use of a lower dose that less consistently impacts behavior when injected bilaterally would have made it more difficult to be certain that any given injection is biologically effective (i.e., it would have had a behavioral effect had both sides been injected). In contrast, bilateral injections at the 8 µg dose strongly and consistently attenuated behavioral responding, and so we can be certain that injections of this higher dose were biologically effective.

We have amended the Materials and methods section to include this justification, and we have clarified that these doses reflect the amount of drug delivered to each hemisphere.

2) Given the striking differences between fed and restricted states, do the authors have any data from restricted rats that returned to ad libitum feeding? A demonstration of 'toggling on' the effect of CTAP in rats that were previously food restricted would be very powerful. This could be done at the strictly behavioral level.

We agree with reviewer #2 that the demonstration of a toggle in the requirement of NAc MOR activation for cue responding would represent a powerful insight into the underlying mechanism. However, this experiment is tied up with the broader questions of time course and degree of restriction: how long must animals be food restricted before responding becomes CTAP-insensitive? (Hours? Days? An entire week, as used here?) How long must previously restricted rats be ad libitum-fed before responding becomes CTAP-sensitive? How much restriction (90% of free feeding weight? 95%?) is required for responding to become CTAP-insensitive? These are important questions to address in further studies, and the answers are necessary to design a “toggle” experiment. Because performing these experiments would be a major undertaking, we feel that it is best saved for a separate study.

3) In the Discussion, the authors speculate that MOR ligand may be sourced from NAc neurons themselves. Are the authors suggesting that the initiation of cue-evoked excitations could lead to enkephalin release that would then, in some feed-forward manner, enhance the ongoing response to the cue. Indeed, some of the data reflect fewer bins as time proceeds from cue onset that were statistically elevated following CTAP administration. Or do the authors speculate that MOR receptor activation is occurring in a less time-restricted manner? As there are no effects on baseline firing rate, the latter idea seems unlikely. Some deeper discussion of the source of MOR ligand and timing of MOR receptor activation is warranted.

As suggested, we have included a deeper discussion of the source of NAc enkephalin and the timing with which it carries out its effects on cue responding and cue-evoked excitation (next-to-last paragraph of the Discussion). Clearly the nature of the MOR ligand and the regulation of its release are important questions, but as there is very little data on which to base hypotheses, we have listed some possibilities without committing to any of them.

4) Concerning the populations of neurons that responded during reward consumption, it is unclear if these are neurons with sustained responses from cue onset or whether some of these neurons were only modulated during reward consumption. In the "Analysis of neural data" section of the Materials and methods, only details regarding cue responses are described. If there are some neurons that respond exclusively during consumption, then modulation of this population by CTAP could potentially contribute to reinforcement. This effect could be masked by the population with ongoing heightened activity from the cue period. Regardless, more detail concerning the identification of consumption-related neurons and their relationship to cue-related neurons is warranted.

The reviewer raises the plausible possibility that some of the consumption-modulated firing might be due to residual cue-evoked firing, and thus may be differentially affected by CTAP than firing that is exclusively consumption-modulated. To address this, we identified and excluded from analysis neurons that met all of the following criteria:

1) Neuron exhibited significant cue-evoked excitation as per our stated criteria;

2) Neuron was significantly excited in the 400-ms bin prior to rewarded receptacle entry;

3) Neuron was significantly excited in the first 400-ms bin following rewarded entry.

We defined significant excitation in the 400-ms bins as firing that exceeded the 99% confidence interval of a Poisson distribution comprised of 10 s of pre-cue baseline activity; this description has been added to the Materials and methods section under “Analysis of neural data”.

In fact, very few neurons met all 3 of these criteria for either excitations or inhibitions for each of the neural populations examined in Figure 8, suggesting that excitations during consumption are rarely continuations of excitations begun at cue onset. These “continuation” neurons are summarized in the following table:

Excluding the “continuation” neurons from the analysis did not statistically affect our results; therefore, we left them in.

To further investigate the possibility that consumption-related activity was the result of continued cue-evoked excitation, we focused on neurons that exhibited excitation just prior to receptacle entry by sorting the heat map histograms of firing aligned to receptacle entry by the magnitude of firing rate increase in the 1 sec prior to entry (relative to the pre-cue baseline; see heat maps on the left side of Figure 8—figure supplement 2, which we added in this revision of the paper). We found that these excitations tended to terminate prior to the animal entering the reward receptacle, and therefore these pre-entry excitations are not responsible for consumption-related excitations (which are observed, typically, in other neurons). Next, we sorted the entry-aligned histograms by the magnitude of excitation just after cue onset (between 100-300 ms), and found that those neurons that had the strongest post-cue excitations (i.e., those at the top of the heat maps on the right side of Figure 8—figure supplement 2) tended to also show pre-receptacle entry excitations, which terminated just prior to receptacle entry.

Together, these observations confirm that although cue-evoked excitations can often last until receptacle entry, excitations during reward consumption are rarely, if ever, the result of prolonged cue-evoked excitations, and instead are found in different neurons.

5) Food restriction appears to result in several effects on MSN activity: increases the proportion of neurons responding with an excitation, increases the magnitude of excitations, increases the duration of excitations (Figure 2C-F). However, the data from free-fed pre-injection rats demonstrates a relatively higher percentage of neurons responding with excitations and a percentage that is not different from restricted, re-injected rats. Thus, the proportion of neurons exhibiting excitations may not be a reliable predictor of free versus fasted state. Again, it would be interesting to see the effects of one night of free-feeding would be on these measures in restricted rats.

The reviewer is correct that in bilaterally-injected free-fed rats, the proportion of cue-excited neurons is higher than in the larger population of neurons recorded from free-fed rats that were not injected with drug. This is one reason why we did not make statistical comparisons of the proportions of cue-excited neurons in free-fed and food-restricted populations. We grant that with more recordings in each population, a valid statistical trend might emerge, but we are uncomfortable with stating such a result given this discrepancy (which may in part be a consequence of the lower number of recorded neurons for the bilaterally-injected population).

Additionally, we agree with reviewer #2 that a one-night free-feeding of food-restricted rats could provide valuable information on the effects of food restriction on NAc neuronal activity. However, the question of a single night’s free-feeding effects is part of the larger question, raised above in the response to point #2, about the time courses of restriction and ad libitum feeding with respect to CTAP effects, and we feel that it would be best answered in the context of a more thorough examination.

Reviewer #3:

[…] The authors should clarify why the yield per animal is so different between the conditions, and otherwise experimentally or analytically show that the population of neurons in the ad lib condition are similar to those recorded from in the food restricted condition. If this is not the case, the comparison that is made between neural activity between the two conditions (e.g. in Figure 2) should be interpreted with caution.

Related to this there may be a ceiling effect of increased responding in food restricted rats such that motivation in this task is so high that opioids do not modulate it.

We thank Dr Adan for his supportive comments. To address whether a difference in baseline firing rate might underlie the discrepancy between the number of neurons recorded per animal in free-fed vs. food-restricted populations, we performed the additional analysis shown in Figure 2—figure supplement 2. This figure shows the superimposed distributions of baseline firing rates for both free-fed and food-restricted rats. As is evident from the histograms, the free-fed histogram overlaps cleanly with the food-restricted histogram. The colored arrows in the figure represent the median baseline firing rate of each population, and these are nearly identical. Unsurprisingly, a Wilcoxon rank sum test revealed no statistical difference between the two populations. Thus, if a difference in baseline firing rate underlies the observed discrepancy in neurons per animal, then that would imply the existence of a separate population of completely quiescent neurons in free-fed rats. This seems unlikely to us given that the recorded neurons with the lowest baseline firing rates were all derived from food-restricted rats (see the left-most bar in the figure).

We think the best explanation for the discrepancy is the unpredictable nature of in vivo electrophysiological recordings. Neuronal yield simply varies greatly across implants, for reasons that are not clear.

Author details

1. Kevin Caref

   Department of Neuroscience, Albert Einstein College of Medicine, Bronx, United States

   Contribution

   Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-6424-4272

2. Saleem M Nicola

   1. Department of Neuroscience, Albert Einstein College of Medicine, Bronx, United States
   2. Department of Psychiatry, Albert Einstein College of Medicine, Bronx, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Methodology, Project administration, Writing—review and editing

   For correspondence

   saleem.nicola@einstein.yu.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9582-6312

• 2,338

  Page views

• 249

  Downloads

• 15

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.