USING POSTMEAL MEASURES AND MANIPULATIONS TO INVESTIGATE HIPPOCAMPAL MNEMONIC CONTROL OF EATING BEHAVIOR

Abstract

Episodic meal-related memories provide the brain with a powerful mechanism for tracking and controlling eating behavior because they contain a detailed record of recent energy intake that likely outlasts the physiological signals generated by feeding bouts. This review briefly summarizes evidence from human participants showing that episodic meal-related memory limits later eating behavior and then describes our research aimed at investigating whether hippocampal neurons mediate the inhibitory effects of meal-related memory on subsequent feeding. Our approach has been inspired by pioneering work conducted by Ivan Izquierdo and others who used posttraining manipulations to investigate memory consolidation. This review describes the rationale and value of posttraining manipulations, how Izquierdo used them to demonstrate that dorsal hippocampal (dHC) neurons are critical for memory consolidation, and how we have adapted this strategy to investigate whether dHC neurons are necessary for mnemonic control of energy intake. I describe our evidence showing that ingestion activates the molecular processes necessary for synaptic plasticity and memory during the early postprandial period, when the memory of the meal would be undergoing consolidation, and then summarize our findings showing that neural activity in dHC neurons is critical during the early postprandial period for limiting future intake. Collectively, our evidence supports the hypothesis that dHC neurons mediate the inhibitory effects of ingestion-related memory on future intake and demonstrates that post-experience memory modulation is not confined to artificial laboratory memory tasks.

In many species, the hippocampus (HC) is vital for memory (Moscovitch et al 2016, Murray et al 2018), particularly for episodic memories of personal experiences (Baudry 2020, Eichenbaum 2013, Squire & Dede 2015). More recent evidence suggests that the HC also controls feeding behavior. HC neurons receive neural signals regarding ingestion, such as taste and stomach distention (Kanoski & Grill 2015). They also express receptors for virtually all pre- and postprandial food-related signals, such as leptin (Mercer et al 1996, Shioda et al 1998), insulin (Hami et al 2014, Unger et al 1989), ghrelin (Ferrini et al 2009), and melanocortins (Kishi et al 2003, Shen et al 2013), often at high levels that parallel or exceed the hypothalamus (Gantz et al 1993, Unger et al 1989). Moreover, HC neurons project to most brain areas critical for energy regulation (Tannenholz et al 2014), and a substantial body of evidence implicates the HC in feeding behavior in humans and other species (Kanoski & Grill 2015, Stevenson & Francis 2017).

Why does the HC control two vital functions (i.e., memory and feeding,) that may appear to be unrelated? One possibility is that evolution has co-opted episodic memory to control eating behavior, for instance, by allowing memory of recent ingestion to influence decisions about future eating behavior. Interestingly, when scientists describe episodic memory, they often use recall of a recently eaten meal as an example (Tromp et al 2015), and meal-related memories are likely a prime example of incidental episodic memories that are automatically encoded rather than intentionally (Wang & Morris 2010). Episodic meal-related memories provide the brain with a powerful mechanism for tracking and controlling eating behavior because they contain a detailed record of where, when and what has been eaten (Banquet et al 2021, Sugar & Moser 2019) that likely outlasts the physiological signals generated by feeding bouts. The brain has evolved many mechanisms to promote food-seeking and consummatory behaviors; however, neural controls are also needed to inhibit these behaviors, for instance, when sufficient energy has been obtained, to avoid predation, and to promote reproduction. This review briefly summarizes evidence from human participants showing that episodic meal-related memory limits later eating behavior and then describes our research aimed at investigating whether HC neurons mediate the inhibitory effects of meal-related memory on subsequent feeding. Our approach has been inspired in large part by pioneering work conducted by Ivan Izquierdo and others who used posttraining manipulations to investigate the neurochemical and molecular bases of memory consolidation. This review describes the rationale and value of posttraining manipulations, how Izquierdo used them to demonstrate that HC neurons are critical for memory consolidation, and how we have adapted this strategy to investigate whether HC neurons are necessary for mnemonic control of energy intake.

Episodic memory inhibits later intake in humans

Research from human participants suggests that episodic memories of recently eaten meals limit subsequent ingestive behavior (Higgs & Spetter 2018). For instance, disrupting the encoding of the memory of a meal by having participants watch television or play computer games increases the amount that is consumed during the next eating episode; conversely, manipulations that enhance memory of a recently consumed meal decrease the amount that is ingested subsequently. Influencing memory encoding during eating has a smaller effect on the amount that is consumed during that bout than on the amount that is consumed during the next one, suggesting that episodic memory of a meal plays primarily influences future eating behavior (Robinson et al 2013). This ability of episodic memory to control future eating consumption is further corroborated by the finding that episodic memory of how much participants think they ate rather than the actual amount ingested predicts hunger hours later (Brunstrom et al 2012). Episodic memory deficits are associated with uncontrolled eating in healthy adults (Martin et al 2018), and patients with amnesia for personal events, including Patient HM, do not remember eating, eat larger quantities of food and will eat an additional meal when presented with food, even when they have just eaten to satiety (Hebben et al 1985, Higgs et al 2008, Rozin et al 1998).

The findings from healthy human adults suggest that episodic meal-related memory limits later eating behavior, although the contribution of other processes such as attention cannot be ruled out. Nonetheless, these findings suggest that the HC limits energy intake given that episodic memory requires the HC. The findings from patients with amnesia provides supporting evidence for the possibility that the HC and episodic meal-related memory limits later eating behavior given that the HC is usually included in the brain pathology that produces amnesia, although multiple brain areas are typically impacted. As a result, mechanistic studies in nonhuman models are needed to determine whether HC neurons are involved in mnemonic control of feeding.

Using postmeal measures and manipulations to study mnemonic control of eating behavior

In rodents, the HC can be functionally divided into dorsal (dHC) and ventral (vHC) regions. dHC and vHC have different anatomical connections, cellular and circuit properties and patterns of gene expression (Barkus et al 2010, Bienkowski et al 2018, Dong et al 2009, Fanselow & Dong 2010, Moser & Moser 1998, Strange et al 2014, Thompson et al 2008). dHC appears to be primarily responsible for the cognitive functions associated with the HC, namely spatial and episodic-like memories (Barbosa et al 2012, Drieskens et al 2017, Hunsaker et al 2008, Panoz-Brown et al 2018, Zhou et al 2012).

We hypothesize that dHC neurons mediate the inhibitory effects of ingestion-related memory on future intake based on their well-documented role in episodic-like memory. To begin to test our hypothesis, we drew our inspiration from research using posttraining manipulations to investigate memory consolidation. Posttraining treatments are based on the assumption that learning induces reverberating activity in the brain that persists after a learning experience, that this activity is essential for establishing neural connections underlying the memory of that learning (Hebb 1949) and that a memory trace is initially labile and gains stability and permanence through consolidation (Muller & Pilzecker 1900). Izquierdo became an early adopter of the use of posttraining manipulations to study memory consolidation after his close friend and collaborator James McGaugh demonstrated that administering drugs immediately after learning influences later retention performance (McGaugh 1966, McGaugh & Petrinovich 1965). They both provided a wealth of evidence showing that peripheral or central treatments given immediately after learning can enhance or impair the consolidation of the memory of that event (Izquierdo 1989, Izquierdo & McGaugh 2000, McGaugh 2000). Not only are posttraining manipulations a simple yet powerful approach for targeting this labile memory consolidation period, but they also avoid non-mnemonic effects on performance by ensuring that an animal is not under the influence of that manipulation during training or on the memory test, with the latter being performed hours after the effects of the drug have worn off (McGaugh 1989).

Amongst his many significant scientific contributions, Izquierdo played a pivotal role in demonstrating that the dHC is essential for memory consolidation. He did so chiefly by administering a variety of compounds into the dHC of rats immediately after training, most commonly in a 1-trial inhibitory avoidance (IA) task. Compared to other behavioral tests of dHC-dependent memory that require many trials for learning to occur (e.g., spatial water maze, win-shift in the radial arm maze, place learning in the Barnes or plus maze), the power of 1-trial IA is that it allows for clear delineation of when the memory consolidation period begins, which is right after the one learning trial is given. Izquierdo and others demonstrated that IA training activates the neurochemical and molecular events necessary for synaptic plasticity in the dHC, and importantly, that posttraining administration of drugs into the dHC that interfere with or promote these neurochemical and molecular processes impacts subsequent retention performance (Bekinschtein et al 2010, Izquierdo et al 2008, Izquierdo & Medina 1993).

Stimulated by these findings, we set out to determine whether dHC neurons mediate the inhibitory effects of ingestion-related memory on future intake. Our general approach was to determine whether ingestion influences dHC neural activity during the early postprandial period and whether postmeal inhibition of dHC neurons increases subsequent intake. In these experiments, rats were allowed to eat a meal composed of either standard chow, a sucrose solution, or the noncaloric sweetener saccharin. The rats were exposed to the sweetened solutions more than once to avoid any effects of novelty, and time of day and context were kept constant. We then measured biomarkers of synaptic plasticity or manipulated dHC neural activity during the early postprandial period, when the memory of the meal is presumably undergoing consolidation.

We have shown that ingestion engages the molecular events implicated in synaptic plasticity and memory in principal dHC glutamatergic neurons during the early postprandial period. The most common form of HC synaptic plasticity depends on glutamate NMDA receptor (NMDAR)-dependent increases in intracellular calcium that activate many cellular events that act collectively to increase glutamatergic AMPA receptor (AMPAR) function in the postsynaptic cell, thereby increasing glutamate signaling and synaptic strength (Bengtson & Bading 2012, Clark et al 2015, Czerniawski et al 2012, Kent et al 2007, Kutlu & Gould 2016, Maggio et al 2015, Portero-Tresserra et al 2014, Xia & Storm 2012, Xu et al 2005, Zhang et al 2001, Zhu et al 2014). Phosphorylation of glutamate AMPAR GluA1 subunits on serine 831 (pSer⁸³¹) and serine 845 (pSer⁸⁴⁵) residues are two posttranslational modifications that occur early in the induction of synaptic plasticity. We found that consuming a sucrose meal increases AMPAR pSer⁸³¹ but not AMPAR pSer⁸⁴⁵ in dHC neurons (Ross et al 2019). AMPAR pSer⁸³¹ improves synaptic efficacy by increasing AMPAR conductance and by promoting insertion of AMPARs in postsynaptic spines (Derkach et al 1999, Hu et al 2007, Kristensen et al 2011). Interestingly, the effects of sucrose ingestion on AMPAR pSer⁸³¹ and pSer⁸⁴⁵ are strikingly similar to those produced by 1-trial IA learning (Whitlock et al 2006). Like sucrose ingestion, 1-trial IA training increases AMPAR pSer⁸³¹ in dHC neurons, and the magnitude of sucrose- and IA-induced increases in AMPAR pSer⁸³¹ are comparable. Also like sucrose ingestion, IA training increases dHC AMPAR pSer⁸³¹ but not AMPAR pSer⁸⁴⁵ (Whitlock et al 2006). Thus, these findings suggest that ingestion-induced increases in dHC AMPAR pSer⁸³¹ are part of the molecular events underlying the memory of a meal.

Learning events also stimulate the expression of activity-regulated cytoskeleton-associated protein (Arc) in dHC glutamatergic neurons (Descalzi et al 2019, Guzowski et al 2001a, Guzowski et al 2001b, Hudgins & Otto 2019, Tripathi et al 2020). As in the case of learning, we found that sucrose or saccharin ingestion increases Arc expression in dHC neurons (Henderson et al 2017, Henderson et al 2016). Like AMPAR pSer⁸³¹, Arc expression is typically dependent on NMDAR activation (Czerniawski et al 2011, Link et al 1995, Lyford et al 1995, Steward et al 1998), and Arc is considered a master regulator of synaptic plasticity that is necessary for memory consolidation (Bramham et al 2008, Korb & Finkbeiner 2011, Shepherd & Bear 2011).

Our results also suggest that the degree of familiarity with the ingested food impacts the ability of that eating episode to activate the molecular events required for synaptic plasticity. Specifically, increasing the amount of previous experience with a sucrose solution prevents the ability of ingesting that solution to induce AMPAR pSer⁸³¹ in dHC (Ross et al 2019), and dHC Arc expression also appears to diminish with sucrose familiarity (Henderson et al 2016). These findings are consistent with evidence showing that extensive behavioral training in learning and memory tasks or repeated exposure to the same environment decreases dHC expression of several molecules critical for memory, including Arc (Gardner et al 2016, Guzowski et al 2006, Guzowski et al 2001b, Kelly & Deadwyler 2002, Kelly & Deadwyler 2003), phosphorylated cAMP response element-binding protein (pCREB)(Moncada & Viola 2006) and protein kinase M-ζ (Moncada & Viola 2008). Habitual consumption of one energy source in the same context may reduce the amount of dHC activity required to remember that meal and/or diminish the role of the dHC in remembering that meal, which is consistent with evidence showing that habit memory is mediated by the dorsolateral striatum (Goodman 2020). The possibility that repeated consumption of one energy source diminishes dHC activation has implications for dHC involvement in reward-based memory tasks. For instance, if place-reward training involves the same food reward for each trial, then dHC should only be necessary during the early phases of training. This possibility is consistent with findings showing that repeated training in appetitively-motivated spatial tasks diminishes the necessity of dHC in these tasks (Cammarota et al 2005, Layfield et al 2020a, Packard & McGaugh 1996), and raises the intriguing prospect that changing the food reward during training in a spatial memory task will mitigate the effects of overtraining and promote continued dHC involvement.

We have also shown that pharmacological and optogenetic inhibition of dHC neurons given during the early postprandial period stimulates later eating, both by promoting meal initiation and increasing meal size. Specifically, intra-dHC infusions of the GABA-A agonist muscimol administered after the end of a sucrose meal accelerates the initiation of the next meal, doubles the amount rats consume during that next meal, increases the number of meals ingested during the next hour after the injection, and disrupts the postprandial correlation, which is the positive relationship the size of a meal and the amount of time that passes before the next one is eaten (i.e., the postprandial intermeal interval) (Henderson et al 2013). The findings from Izquierdo’s group that posttraining intra-dHC muscimol infusions produce retrograde amnesia (Izquierdo et al 1992b, Oliveira et al 2010) supports the hypothesis that postmeal dHC muscimol infusions increase later intake by disrupting the consolidation of the preceding meal. However, unlike memory tasks where one can wait days for the effects of a drug to wear off before testing memory, one cannot wait this long to assess the effects of postmeal infusions on subsequent intake. Muscimol inhibits neural activity for 12–24 hr (Majchrzak & Di Scala 2000, Martin 1991, Martin & Ghez 1999) and the average intermeal interval for chow is ~3 hours and is shorter for sucrose and sacharin (Hannapel et al 2019b, Snowdon 1969). As a result, postmeal muscimol-induced hyperpolarization of dHC persists throughout the postprandial period and during intake of all meals measured during the 1-hr period after the injection. Consequently, it is unclear whether these postmeal manipulations increase consumption via a non-mnemonic effect on intake during subsequent ingestion rather than by disrupting memory-based processes during the postprandial period. Also, muscimol infusions will affect all cells bearing GABA-A receptors in the vicinity of the injection, such as interneurons (Glykys et al 2007, Mann & Mody 2010, Semyanov et al 2003), and thus the phenotype of the cells that contribute to the muscimol-induced increase in intake is unknown.

We hypothesize that the ability of dHC muscimol infusions to increase subsequent intake is mediated by principal dHC glutamatergic neurons. Muscimol hyperpolarizes these neurons (Majchrzak & Di Scala 2000, Rombo et al 2016), which play a central role in synaptic plasticity (Ashby et al 2021, Bartsch & Wulff 2015) and dHC-dependent memory (Baker & Kim 2002, Izquierdo et al 2008b, Morris 2013, Zamorano et al 2018). Our findings showing that ingestion increases both AMPAR pSer⁸³¹ and Arc mRNA also implicate glutamatergic neurons. To test the necessity of dHC glutamatergic neurons, we used an activity-guided optogenetic approach to inhibit these cells before, during, or after the consumption of a meal (Hannapel et al 2019b). We used a construct containing a calmodulin-dependent protein kinase II (CaMKII)α promoter that limits opsin expression to glutamatergic neurons while sparing interneurons, fibers of passage, and glia (Butler et al 2016, Han & Boyden 2007, Kohl et al 2011, Stark et al 2013, Stark et al 2014, Weitz et al 2015, Yizhar et al 2011). This allowed us to determine when neural activity in dHC glutamatergic neurons is critical for limiting future intake. We confirmed with electrophysiological measures that neural activity in these neurons returned to baseline as soon as photostimulation was terminated (Hannapel et al 2019b). This allowed us to inhibit dHC glutamatergic neurons for a finite period (i.e., 10 min) and assess subsequent intake when the neurons were no longer hyperpolarized, which is critical for testing the role of memory. We found that inactivation given after the end of a chow, sucrose, or saccharin meal significantly hastened the initiation of the next meal and increased the amount consumed during that next meal when the neurons were no longer inhibited. Importantly, inactivation given before or during intake of the first meal did not affect the amount consumed during that meal or during the next bout, and photoillumination of a control virus in dHC for the same duration did not affect feeding behavior when given before, during, or after a meal (Hannapel et al 2019b). These data demonstrate that neural activity in dHC glutamatergic neurons is critical during the early postprandial period for limiting future intake, which is consistent with the hypothesis that these neurons inhibit future intake by consolidating the memory of the preceding meal.

The effects of postmeal manipulations are time-dependent

Extensive evidence shows that the effects of posttraining manipulations are time-dependent, such that the efficacy of posttraining treatments diminishes as the training-treatment interval increases (McGaugh 1966a, McGaugh & Izquierdo 2000). These findings support the idea that new memories are initially labile and gain stability and permanence through consolidation. Izquierdo and other researchers have shown that the effects of posttraining dHC manipulations on memory are time-dependent (Izquierdo et al 1992a, Jerusalinsky et al 1992, Lorenzini et al 1996, Oliveira et al 2010, Riedel et al 1999). For example, temporarily inactivating dHC with tetrodotoxin (TTX) immediately after IA training impairs memory tested 48 hr after acquisition; whereas, dHC infusions of TTX given 6 hr after training have no effect (Lorenzini et al 1996). Inhibiting dHC glutamate receptors after IA learning also produces retrograde amnesia in a time-dependent manner (Bonini et al 2003, Jerusalinsky et al 1992).

If dHC glutamatergic neurons limit future intake through a process that requires memory consolidation, then the effects of postmeal optogenetic inhibition of dHC glutamatergic neurons should be time-dependent. Inhibition should increase subsequent intake when given soon after the end of a meal, but delayed inhibition should not have an effect. In support, we found that postmeal inhibition of dHC glutamatergic neurons given after the end of a meal increased the likelihood that rats would consume a second meal and significantly increased the amount consumed during that next meal when the neurons were no longer inhibited (Briggs et al 2021b). Critically, delayed inhibition given 80 min after the end of the meal did not affect the probability of consuming a second meal or the amount eaten during that second meal. These findings are consistent with several previously published findings, many from Izquierdo’s group, showing that dHC neurons are involved in memory consolidation in a time-dependent manner (Izquierdo et al 1992a, Jerusalinsky et al 1992, Lorenzini et al 1996, Oliveira et al 2010, Riedel et al 1999). The time-dependent nature of posttraining manipulations depends on the type of learning experience and the neurotransmitter or molecular process that is being targeted (Jerusalinsky et al., 1992). Our finding that inhibition given 80 min after saccharin ingestion did not increase subsequent intake is analogous to the time-course for the effects of posttraining inactivation of dHC NMDAR receptors. Specifically, dHC infusions of the NMDAR antagonist AP5 impair memory when given immediately after IA training, but not when given 90 min later (Jerusalinsky et al., 1992). These findings suggest that NMDAR mechanisms may be critical during the early postprandial period for limiting future intake, which will need to be investigated. Combined with our prior evidence showing that inhibition given before or during the intake of a first meal does not affect the amount consumed during that meal or the next (Hannapel et al 2019a), these findings indicate that neural activity in dHC glutamatergic neurons is necessary during the early postprandial period for limiting future intake. The fact that inhibition increases intake only when given at certain times relative to a meal argues against the possibility that inhibiting dHC increases eating via some non-specific effect, such as by increasing arousal or activity levels. Rather, it provides further support for the hypothesis that dHC glutamatergic neurons inhibit future intake through a process that involves memory consolidation.

Impaired memory vs. interoceptive deficits

A striking finding in patients who have amnesia is that they appear to be unable to identify whether they are hungry or satiated, even after they have just consumed a large meal (Hebben et al 1985, Higgs et al 2008, Rozin et al 1998). A relationship between interoception and HC functioning is also observed in healthy individals (Dudley & Stevenson 2016, Edwards-Duric et al 2020, Stevenson et al 2018), and HC lesions impair the ability to discriminate hunger signals in rats (Davidson et al 2009, Davidson & Jarrard 1993). Thus, it is possible that neural activity in dHC glutamatergic neurons is necessary during the early postprandial period for limiting future intake because that is when critical interoceptive signals generated by ingestion are most prominent (Page et al 2012, Waise et al 2018), and postmeal inactivation may increase subsequent intake by disrupting the processing of internal visceral cues rather than by impairing the consolidation of the memory of that meal. Our finding that postmeal inhibition of dHC neurons increases later intake when the neurons are no longer inhibited does not support this interpretation because interoceptive cues are presumably intact when the neurons are no longer inhibited. The possibility remains, however, that postmeal inhibition causes a weaker meal-related memory to be formed by disrupting interoception.

We investigated the necessity of interoceptive signaling in dHC control of energy intake by determining whether postmeal dHC inactivation increases saccharin consumption. In these experiments, animals were given limited exposure to saccharin to reduce the likelihood of any conditioned responses to the taste. Under such conditions, saccharin has minimal postingestive gastric consequences (Foletto et al 2016, Mook et al 1980, Renwick 1985, Renwick 1986, Sclafani & Nissenbaum 1985), and unlike sucrose and chow, saccharin meal timing and size are determined primarily by oropharyngeal rather than gastrointestinal processes (Kushner & Mook 1984, Renwick 1985, Renwick 1986, Sclafani & Nissenbaum 1985). As in the case of chow and sucrose intake, we found that postmeal dHC inactivation accelerates the onset of the next saccharin meal (Hannapel et al 2019b) and doubles the amount of saccharin consumed during the next meal (Briggs et al 2021b, Hannapel et al 2019b). These findings suggest that the ability of dHC neurons to control future intake does not require postprandial interoceptive visceral signals and are consistent with the interpretation that postmeal manipulations increase future intake by disrupting the consolidation of the memory of the meal. This conclusion is further supported by our findings showing that saccharin ingestion increases dHC expression of Arc mRNA (Henderson et al 2016) and by our results showing that inhibition increases subsequent intake when given soon after the end of a meal but not when the postmeal inhibition is delayed (Briggs et al 2021b).

Conclusions

Collectively, our results implicate dHC glutamatergic neurons in the ability of meal-related memory to inhibit future intake. Of note, our findings show that neural activity in dHC glutamatergic neurons during the early postprandial period is critical for inhibiting future intake. Specifically, photoinhibition of dHC glutamatergic neurons before or during intake of a meal or inhibition that is given many minutes after a meal do not influence the amount eaten during the next meal (Figure 1A). Critically, only inactivation given during the early postprandial period, when the memory of the meal is expected to be undergoing consolidation, accelerates the onset of the next meal and increases the amount eaten during that next meal (Figure 1B). We have also shown that ingestion activates molecular processes required for synaptic plasticity during this early postprandial period (Figure 1 C). The comparable effects of learning and ingestion on dHC expression of synaptic plasticity biomarkers and memory suggest that increased dHC pSer⁸³¹ and Arc expression are molecular events involved in forming the memory of a meal. Admittedly, there are no actual tests of memory per se in our experiments; rather, we measure and manipulate neural activity during the early postprandial period when the memory of the meal is undergoing consolidation and use biomarkers of synaptic plasticity and meal timing and size as proxies of that memory. We already know based on decades of research using food-motivated tasks that manipulations that interfere with dHC function impair more direct measures of food-related memory. For instance, dHC neurons are necessary for memories of the location of food (Layfield et al 2020, Lee & Kesner 2003a, Lee & Kesner 2003b, McDonald & White 2013, Packard & McGaugh 1996, Potvin et al 2006), the conditions under which food will be available (Crystal et al 2013) and when food is expected to arrive (Tam et al 2015). Our findings suggest that dHC neurons may use these kinds of memories to control meal initiation and termination, although more evidence is needed to support that possibility. Relatedly, an important question that still needs to be addressed is, “What is being remembered?”. Our findings with saccharin suggest calories and introception are not necessary for the formation of a meal-related memory. There is still a remote chance that dHC inhibition disrupts saccharin intake by interfering with the processing of any mechanical stimulation produced by the saccharin solution in the gut (Waise et al 2018), but the fact that saccharin meal timing and size are determined primarily by oropharyngeal rather than gastrointestinal processes does not support that interpretation (Kushner & Mook 1984, Renwick 1985, Renwick 1986, Sclafani & Nissenbaum 1985). The finding that sensory-specific satiety is not disturbed in patients with episodic amnesia suggests that memory of taste and/or pleasure are also not necessary. Perhaps the actual act of ingesting (seeing the food, bringing to the mouth, orosensation, chewing, swallowing) are the essential components of a meal-related memory.

Fig. 1.

Neural activity in dHC glutamatergic neurons during the early postprandial memory consolidation period is critical for limiting future intake and ingestion activates the molecular processes required for synaptic plasticity and memory formation during the early postprandial period. (A) Photoinactivation of dHC glutamatergic neurons before or during intake of a meal of inactivation that is given in the late postprandial period do not affect the amount eaten during the next meal. (B) Only inactivation given during the early postprandial memory consolidation period increases the amount eaten during the next meal and also promotes the initiation of the next meal. (C) Ingestion promotes two molecular events associated with synaptic plasticity and memory formation (phosphorylation of serine 831 residues on AMPARs and transcription of Arc mRNA) in glutamatergic neurons during the early postprandial period.

Whether the duration of the consolidation period for meal-related memory and the persistence and resilience of meal-related memories vary as a function of the food stimulus is also unknown. For IA learning, training with a higher footshock intensity produces a shorter consolidation period, activates different molecular processes, and produces a more resilient and persistent memory than does training with a lower footshock intensity (Diaz-Trujillo et al 2009, Gonzalez-Franco et al 2017, Lovitz & Thompson 2015, Medina et al 2019, Morena et al 2014, Parent & McGaugh 1994, Prado-Alcala et al 2020). Collectively, these findings suggest that emotional memories are consolidated more quickly, are more durable, and last longer than less emotional memories, raising the possibility the palatability of a food influences the consolidation, duration, and resilience of the memory of that food. For example, ingestion of palatable sucrose concentrations (Smith 2000) may activate different molecular mechanisms, require a shorter consolidation period to form a memory of that ingestive bout and produce a more persistent memory than ingestion of less preferred sucrose concentrations. Along these lines, foods that cause malaise may have mnemonic properties analogous to highly palatable foods, given extensive evidence showing that conditioned taste aversions can be established in one trial and are extremely long-lasting (Bernstein 1999, Molero-Chamizo & Rivera-Urbina 2020, Welzl et al 2001).

The circuitry through which dHC neurons control intake also remains to be determined. Evidence suggests that dHC glutamatergic neurons limit intake via effects on neurons in the septal area (Azevedo et al 2019), and whether these neurons are the same ones that are critical during the postprandial period should be determined. vHC neurons could also contribute to the representation of a meal given their involvement in appetitive and affective processes (Fanselow & Dong 2010, Strange et al 2014). We have shown that ingestion increases Arc in vHC during the early postprandial period and that postmeal vHC inactivation promotes meal initiation and increases meal size (Hannapel et al 2017, Hannapel et al 2019). Like dHC inactivation, vHC inactivation given before or during ingestion does not impact later intake (Hannapel et al 2017a). We also demonstrated recently that blocking vHC NMDARs and Arc increase eating (Briggs et al 2021a), which is consistent with the idea that the molecular mechanisms required for HC synaptic plasticity influence eating behavior. Additional experiments are needed to determine whether these events are also necessary for dHC control of future intake. Our findings also show that ingestion does not affect AMPAR pSer⁸³¹ nor AMPAR pSer⁸⁴⁵ in vHC neurons, suggesting that some of the molecular mechanisms that regulate ingestion may be different in dHC vs. vHC (Ross et al 2019).

To the best of our knowledge, our research group is the first to perform postmeal manipulations, allowing us to show that postmeal inhibition of dHC neurons increases future energy intake that is both homeostatic, i.e., standard chow, and hedonic, i.e., sucrose and saccharin. Our results are strikingly similar to those observed with IA learning and show that post-experience modulation of memory is not confined to artificial memory tasks in the laboratory, but rather applies to behaviors and experiences that are relatively more natural and vital for survival.

Episodic meal-related memory limits later eating behavior

Our research suggests dorsal hippocampal (dHC) glutamate neurons mediate the inhibitory effects of memory on future eating

dHC glutamate neurons are critical during the early postprandial memory consolidation period for limiting future intake

Ingestion activates molecules critical for memory in dHC glutamate neurons during the early postprandial period

Ivan Izquierdo’s pioneering research using posttraining manipulations to study memory consolidation inspired our approach