Eat, seek, rest? An orexin/hypocretin perspective

Seeking and ingesting nutrients is an essential cycle of life in all species. In classical neuropsychology these two behaviours are viewed as fundamentally distinct from each other, and known as appetitive and consummatory, respectively. Appetitive behaviour is highly flexible and diverse, but typically involves increased locomotion and spatial exploration. Consummatory behaviour, in contrast, typically requires reduced locomotion. Another long‐standing concept is “rest and digest”, a hypolocomotive response to calorie intake, thought to facilitate digestion and storage of energy after eating. Here, we note that the classical seek➔ingest➔rest behavioural sequence is not evolutionarily advantageous for all ingested nutrients. Our limited stomach capacity should be invested wisely, rather than spent on the first available nutrient. This is because nutrients are not simply calories: some nutrients are more essential for survival than others. Thus, a key choice that needs to be made soon after ingestion: to eat more and rest, or to terminate eating and search for better food. We offer a perspective on recent work suggesting how nutrient‐specific neural responses shape this choice. Specifically, the hypothalamic hypocretin/orexin neurons (HONs) – cells that promote hyperlocomotive explorative behaviours – are rapidly and differentially modulated by different ingested macronutrients. Dietary non‐essential (but not essential) amino acids activate HONs, while glucose depresses HONs. This nutrient‐specific HON modulation engages distinct reflex arcs, seek➔ingest➔seek and seek➔ingest➔rest, respectively. We propose that these nutri‐neural reflexes evolved to facilitate optimal nutrition despite the limitations of our body.

efficient.6][7][8][9][10][11][12] To be evolutionarily optimal, the bidirectional switch between appetitive and consummatory behaviour needs to be intelligently tuned to context.In food intake, this context can be internal (e.g., current level of nutrients in the body) or external (e.g., level of danger in the environment).This article is concerned with the internal context, specifically with how ingested nutrients govern the switch between eating and exploring.It focusses on the recent evidence that this switch involves hypothalamic hypocretin/orexin neurons (HONs).3][24][25][26][27][28][29][30][31] Here, we review these properties of HONs in the context of recent evidence indicating that they are important components of a nutrient-controlled switch between appetitive and consummatory behaviours.We focus on acute (<1 h) effects and acute responses to food, rather than the longer-term, circadian role of HONs in shaping feeding rhythms that was previously described and discussed. 32,33

| HYPOCRETIN/OREXIN NEURONS AS DRIVERS OF MOBILE BEHAVIOURS
5][36][37][38] Of note, the hypoactivity phenotype of HON deficiency is not due to increased sleep: the total daily sleep is not altered. 38This implies that HONs are important for appropriate timing of mobile behaviours.
Are natural activity patterns of HONs consistent with their role as drivers of mobile behaviour?0][41][42][43][44][45][46][47][48] However, HON activity in vivo is rapidly modulated, consistent with their brainwide neural inputs. 49[51][52][53][54][55][56] While functional subpopulations of HONs most likely exist, 45,57,58 multiphoton imaging of 1000s of HONs in vivo indicates that rapid dynamics of HONs in vivo is typical of most, if not all, HONs, though not all HONs (in)activate together and some do so in antiphase with each other. 55,59,60Furthermore, optogenetic evidence implies that this HON activity contributes to the initiation of locomotion.HON optostimulation produces frequency-dependent increase in locomotion probability, while their optosilencing makes both self-paced and sensory-evoked locomotion less likely. 55e ability of HONs to drive mobile behaviours may also involve an emotional aspect.In real-time place preference tests, HON optosilencing makes mice spend more time in the silencing-associated experimental chamber. 61In other experiments, HON optosilencing makes mice more likely to enter areas of experimental arena that they otherwise find innately aversive, that is, brightly and/or open areas. 61[63] How may HONs drive mobile behaviours?Locally in the lateral hypothalamus, HONs activate a distinct neural population, the GAD65-expressing neurons. 64,65Selective, bidirectional chemogenetic modulation of these GAD65 neurons reveals that their activity is sufficient to induce locomotion, and necessary for keeping self-paced locomotion at a high level. 65The GAD65 cell activity precedes the initiation of spontaneous running bouts by ≈1-2 s, and is engaged by an acute stress (forced immobilisation) in an orexin-dependent manner (i.e., response blocked by orexin receptor antagonist). 65[68] Can HON-dependent movement patterns be viewed as "appetitive behaviours"?To answer this question, these movement patterns need to have a known goal.When complex, non-locomotor movements (skilled pulling on a robotic handle using one forelimb) motivated by a food goal were recently examined, it was found that they were also associated with HON activation. 69,70However, this HON activation did not seem to control the amount of food consumed, but instead facilitated movement patterns that ensured steady food supply over time. 69Overall, since these studies suggest that HON activity promotes mobile behaviours that either increase or maintain the likelihood of food discovery or acquisition, we propose that HON activity can be viewed as a driver for appetitive behaviours.This is a distinct function from food intake, as discussed further below.

| HONS AS INTRINSIC SENSORS OF GLUCOSE AND AMINO-ACIDS: IN VITRO DATA
Electrophysiological (patch-clamp) recordings of single HONs in mouse acute hypothalamic brain slices indicate that they are hyperpolarized and electrically silenced by D-glucose, mannose, and 2-deoxyglucose (but not galactose, L-glucose, alpha-methyl-D-glucoside, or fructose). 44,71,72These responses involve glucose-induced opening of plasma membrane "background" potassium channels. 73A similar mechanism was found in glucose-inhibited neurons of invertebrates. 74The molecular identify of glucose-activated channels in HONs remains elusive.6][77] Interestingly, HON glucose responses in brain slices are dose-dependently suppressed by lactate, puryvate, and ATP, 78 as well as diet-relevant mixes of amino acids. 79 this is of relevance in vivo, this may imply HONs' glucose-sensing may be somewhat contingent on body energy levels.
Apart from glucose, key dietary nutrients are amino acids.They may directly signal to the brain since ingested proteins elevate amino acid levels in the blood and brain [80][81][82][83][84][85] on a time scale of tens of minutes.In brain slices, HONs are electrically depolarized and excited by dietary relevant amino acids mixtures. 79This happens via a dual biophysical mechanism, involving activation of depolarizing electrogenic membrane amino acid transporters inhibitable by MeAIB ("system-A" transporters), and concurrent shutting down of membrane K-ATP channels resulting mitochondrial generation of ATP upon usage of amino acids as fuels after they enter the cell. 79Concordant with the preferred substrates of system-A transporters, 86,87 nonessential amino acids excited HONs while essential ones had little effect. 79,88These responses have therefore been proposed to couple dietary non-essential amino acids to HON-regulated physiology and behaviour. 89Next, we consider recent in vivo evidence for this.

| DIFFERENTIAL RESPONSES OF HONS TO DIFFERENT INGESTED NUTRIENTS IN BEHAVING MICE
Hypothalamic levels of amino acids go up after amino acids ingestion, and hypothalamic levels of glucose follow blood levels of glucose 82,84,[90][91][92] .Thus, HONs have direct access to these nutrients, which in theory can be used for direct sensing.The ability of HONs to distinguish between ingested nutrients has recently been examined by monitoring HON activity using HON-targeted cytosolic calcium indicators and fiberoptic fluorescence imaging in behaving mice, in combination with direct intragastric infusions of different dietary macronutrients. 59,88The intragastric infusions were performed via an implanted catheter, for precise control over volume and composition of ingested nutrients, and to avoid any caveats from oral taste receptors that may activate HONs.In the fast frequency domain (seconds), HON activity of behaving mice was dominated by rapid behaviours such as running, as expected from previous data and from the brainwide direct neural inputs to HONs. 10,49,55,59,88However, in the slower frequency domain (minutes), HONs displayed differential responses to some of the intragastrically administered nutrients.
Intragastric infusions of dietary-relevant mixtures of nonessential amino acids evoked a slowly building excitation of HON population activity, which peaked 10s of min after the injection. 88is time-course resembles that of amino acid concentration in the lateral hypothalamus after an oral gavage of amino acids. 82In further concordance with HON in vitro responses to amino acids described above, the in vivo HON response to intragastric amino acids in behaving mice (1) was not affected by destruction of vagal afferents; (2) was specific for non-essential amino acids, and (3) was not explained by stomach distention or caloric value of infused nutrients.This demonstrates that HONs can detect and interpret the intragastric appearance of non-essential amino acids inside the organism of behaving mice.The simplest current explanation is that this happens by diffusion of the intragastric amino acids into the lateral hypothalamus (as described in 82 ), where they directly excite HONs as shown in, 79 but direct experiments verifying this simple interpretation remain to be performed.
Intragastric infusions of glucose produced a very different HON response. 59HONs were rapidly inhibited following the intragastric glucose infusion. 59,88Unlike amino acids, glucose level deviations are rapidly counter-regulated inside the body.Therefore, to interpret the in vivo HON response to glucose in relation to the glucose "stimulus", it was necessary to measure the blood glucose concentration concurrently with the HON response.When this was done by combining blood glucose telemetry with HON activity monitoring, the surprising observation was made that peak HON inhibition occurred minutes before the highest level of glucose was reached. 59Intriguingly, this anticipatory-like HON response closely followed the temporal gradient (first derivative) of blood glucose, that is, HONs appeared to act as detectors of glucose trends rather than of absolute levels. 59This "derivative detection" by HONs makes evolutionary sense because it enables efficient, maximal response to be mounted before maximal glucose deviations.In addition to these cue-based anticipatory strategies well studied in neurophysiology since the times of Pavlov, 93 this represents a fundamentally different type of anticipatory strategy that is cue-independent, and involves generating responses based on derivative (trend) of the input. 94This derivative detection is well established as a way of generating anticipatory-like signals in engineering. 94,95

| EVIDENCE FOR THE ROLE OF HONS IN BEHAVIOURAL RESPONSES TO DIETARY NUTRIENTS
Do HON responses to glucose and non-essential amino acids guide any aspects of behaviour?If this were the case, then one would predict that some behavioural responses to deviations in nutrient levels inside the body should be abolished when HONs are not there and reproduced by targeted HON modulation in the absence of the nutrient deviations.These predictions were recently directly tested in experiments using HON-ablated mice and targeted optogenetic HON stimulation.
When non-essential amino acids were infused into the stomachs of mice, the animals responded with rapid reprioritization of their ongoing behaviours. 88Specifically, they reduced their food intake and increased their mobile, exploratory-like behaviours.This refocusing of innate behaviours -away from eating and towards spatial exploration -was mimicked by targeted optogenetic stimulation of HONs in the absence of amino acid infusions, and abolished when the amino acid infusion experiments were repeated in HON-ablated mice. 88These data are consistent with the interpretation that ingested amino acids stimulate HONs, and this HON stimulation is essential for the normal behavioural response to ingested nonessential amino acids, which is to reprioritize exploration and seeking of more essential foods.
In contrast, when glucose was infused in behaving mice, the animals responded with rapid depression of their spontaneous locomotor activity and general body mobility. 59This behavioural response was mimicked by targeted optogenetic inhibition of HONs in the absence of the glucose infusions, 55 occluded (via a "ceiling effect") by tonic optogenetic HON stimulation, and abolished when glucose infusions were repeated in HON-ablated mice. 59

| SUMMARY AND EVOLUTIONARY PERSPECTIVE
A fundamental behavioural cycle of seeking and ingesting nutrients is central to survival of mammals and other organisms.Highly diverse and flexible appetitive (seeking) behaviours have therefore evolved, but they all have in common a drive to explore, which typically causes animals to move and cover large distances.Consuming nutrients, on the other hand, typically cannot happen optimally on the go, and requires a downregulation of movement.A further downregulation of movement ("rest and digest") is traditionally envisioned to be required after nutrient ingestion, to facilitate digestion of nutrients.These classical concepts typically overlook nutrient diversity and the physical limitations of our body, specifically the finite volume of our stomachs and gastrointestinal (GI) tracts, as well the time limitation for finding, consuming, and digesting the food.
If the stomach is allowed to be filled to the limit with "non-essential" nutrients, when essential nutrients are just around the corner, the latter may be eaten by someone else by the time the stomach is free again to receive food.Evolutionarily speaking, the classical seekàingestàrest behavioural sequence is thus not advantageous for all ingested nutrients.We speculate that the brain evolved mechanisms that rapidly sample nutrient content of ingested food, and based on this information make the decision of whether to stay with this food or explore further to find (potentially) better food.Specifically, HONs could be central in such a mechanism since they drive spatial exploration and are activated by non-essential amino acids and inhibited by glucose.This nutrientspecific HON modulation would route behaviour along one of the two reflex arcs, seekàingestàseek and seekàingestàrest, respectively (Figure 1).We propose that these nutri-neural reflexes evolved to facilitate optimal nutrition despite the limitations of our body.

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
Data sharing is not applicable to this article as no new data were created or analyzed in this study.F I G U R E 1 Hypocretin/orexin neurons as a nutrient-gated switch between appetitive and consummatory behaviour.Some nutrients (e.g., non-essential amino acids, nAAs) activate HONs thus inducing a suppression of eating and reingnition of exploration (appetitive behaviour).Other nutrients (glucose) suppress HONs, thus facilitating consummatory behaviours as well as "rest and digest" physiology.
These data are consistent with the interpretation that ingested glucose inhibits HONs, and HON inhibition is essential for the normal behavioural response to ingested glucose.Overall, the recent behavioural experiments thus indicate that HONs are essential for normal locomotor stimulation and suppression induced by ingested non-essential amino acids and glucose, respectively; as well as for the broader nutrient-specific reconfiguration of behavioural repertoire by ingested nutrients.It may be that alteration of HON activity by nutrients is a simple consequence of the need to tune exploration based on the predictions of incoming/outgoing energy and nutrient balance.It remains to be seen whether it is possible to directly test this hypothesis in situations where food seeking, consumption, and ingestion are uncoupled from exploration and locomotion.