Hypothalamic Supramammillary Control of Cognition and Motivation

SuM and theta rhythm

Theta oscillations are a prominent pattern of neural activity observed in the mammalian brain, in particular when an animal is actively exploring an environment or in the state of rapid eye movement (REM) sleep (Buzsaki, 2002). Several studies reported a positive correlation between the theta power and the subsequent formation of associative memory (Landfield et al., 1972; Herweg et al., 2020); and its plausible circuit mechanism, synaptic plasticity in the hippocampus has been shown to be modulated depending on the time of synaptic inputs to the theta phase (Huerta and Lisman, 1995). Theta oscillations are also suggested to play a key role in information coding in the brain. Place cells in the hippocampus, for example, not only increase their firing rates as an animal crosses a particular location, but also change their spike times relative to the theta phase, enabling temporal coding of spatial positions (O'Keefe and Recce, 1993; Huxter et al., 2003). To better understand the roles of the theta rhythm, it is important to clarify the mechanism of its generation and modulation. A central focus in this line of research has been the SuM and the MS.

Differential contributions of the MS and the SuM to the hippocampal theta rhythm have been described in different brain states. In anesthetized animals, silencing of the SuM resulted in a considerable reduction of the theta frequency in the hippocampus (Kirk and McNaughton, 1993). However, lesioning of the SuM failed to abolish the hippocampal theta rhythm in awake-behaving rats (Thinschmidt et al., 1995). By contrast, similar lesioning of the MS largely eliminated hippocampal theta oscillations in behaving rats (Mitchell et al., 1982). Subsequent studies in behaving animals revealed the major impact of the MS on theta-mediated information processing in the hippocampal-entorhinal regions (Brandon et al., 2011; Koenig et al., 2011; Petersen and Buzsaki, 2020; Etter et al., 2023).

However, the fact that SuM lesions do not abolish hippocampal theta oscillations does not necessarily mean that the SuM has no contribution to the theta rhythm. Considering the ability of SuM neurons to generate theta-rhythmic spiking independently of the MS (Kirk and McNaughton, 1991), the brain's theta rhythm may not be a unitary phenomenon driven by a single oscillator, but rather a result of interferences between multiple oscillators. If so, what could be the functional need for the brain to possess multiple theta oscillators? Because of the differences in neural projections between the MS and the SuM (Gaykema et al., 1990; Vertes, 1992), each oscillator may have differential impacts on individual brain regions, potentially allowing for desirable theta-rhythm coordination or synchrony between a particular region pair.

Oscillatory synchrony has been thought to support the integration of distributed information across the brain (Singer, 1993; Fries, 2015), for example, by dynamically changing the synaptic efficacy between neurons depending on the time of synaptic inputs relative to phases of local oscillations (Laurent, 2002). During spatial navigation, information encoded in distant brain regions, including the hippocampus and the PFC, must be integrated for coherent planning and execution of behaviors, and oscillatory synchrony likely plays a pivotal role here. For example, neurons in the mPFC fire preferentially at a particular phase of the hippocampal theta rhythm during navigation (Siapas et al., 2005), and this spike-phase modulation changes dynamically depending on the demand of trajectory decisions (Jones and Wilson, 2005; Benchenane et al., 2010), which is thought to facilitate information transfer between the regions.

These observations led to a question of how the brain can increase interregional theta-rhythm coordination specifically at the time of trajectory decisions. Ito et al. (2018) suggested a plausible role of the SuM in this aspect (Ito et al., 2018). The study focused on the flow of information of an animal's movement direction on a T-maze alternation task. A previous study demonstrated that this information is transferred from the mPFC to the hippocampus by mediating the thalamic nucleus reuniens as a relay (Ito et al., 2015); and in support of this idea, neurons in both mPFC and reuniens exhibit enhanced spike-time coordination to the hippocampal theta rhythm during trajectory decisions. Here, the optogenetic silencing of SuM neurons largely eliminated this enhancement of theta-rhythm coordination during trajectory decisions, while theta oscillations themselves were maintained in the hippocampus, pointing to the role of the SuM in behavior-dependent modulation of interregional coupling via the theta rhythm.

While this study serves as evidence for the SuM's contribution to the theta rhythm in behaving animals, it is still largely unclear how the theta rhythm in the SuM is precisely coordinated with that in the MS in accordance with behavioral demands. Furthermore, several studies reported the SuM's roles in exploratory behavior beyond theta generation (Pan and McNaughton, 2002; Aranda et al., 2006; Chen et al., 2020; Farrell et al., 2021), as described in this review as well. How different types of information can be encoded in the SuM on top of the theta rhythm requires further investigation, which will provide us with a better understanding of the brain as a dynamical system driven by multiple oscillators.

SuM and reward-seeking

Some of the earliest reports of a role for the SuM in reward date back to the seminal work by Olds and colleagues in the 1950s and 1960s while surveying brain regions supporting electrical simulation-mediated reward or aversion (Olds, 1956, 1962; Olds and Olds, 1963). However, it was not known what actual physical elements were involved in producing such reward, and little follow-up on this issue ensued, possibly because of stronger interest in regions adjacent to the SuM (e.g., the dopaminergic neurons of the VTA or lateral hypothalamic [LH] neurons controlling feeding motivation). Consequently, the SuM remained understudied in this respect until the early 2000s when Ikemoto et al. (2004) serendipitously found that rats would self-administer excitatory drugs into the SuM, including the glutamate receptor agonist AMPA, the GABA_A receptor antagonist picrotoxin (Ikemoto, 2005), and nicotine (Ikemoto et al., 2006). They also found that intra-SuM AMPA increases extracellular concentrations of DA in the NAc (Ikemoto et al., 2004). Interestingly, SuM neurons are activated, as indicated by c-Fos induction, by other rewarding manipulations, including intra-VTA infusions of carbachol (Ikemoto et al., 2003) and LH electrical stimulation (Arvanitogiannis et al., 1997). These studies strongly implicated the SuM in reward/reinforcement processes, yet how excitation of the SuM influenced canonical reward systems, such as the VTA-to-NAc DA pathway, or what role SuM neurons play in natural reward-seeking behaviors remained elusive.

Following up on earlier work by the Ikemoto group, Kesner et al. (2021) found that mice would press a lever for optogenetic excitation of SuM neurons, and that the reinforcing effects of this stimulation were most likely mediated by SuM glutamate neurons projecting to the MS. These SuM projections excite MS glutamate neurons (which also support optogenetic self-stimulation), which in turn project to the VTA and activate VTA-to-NAc DA neurons. They also performed small animal fMRI imaging experiments and found similar patterns of brain activity during optogenetic stimulation of either SuM-to-MS or VTA-to-NAc pathways, which, when taken together with experiments showing that systemic injection of DA receptor antagonists similarly attenuated optogenetic self-stimulation of these two pathways, suggests a common circuitry linking SuM-mediated reward with canonical DA mesolimbic reward circuitry. In their ongoing investigation, Arima and Ikemoto (2023) found that SuM-to-MS neurons have extensive collateral projections to the lateral preoptic area (LPO) among others, and the stimulation of the SuM-to-LPO pathway also reinforces behavior and activates VTA-to-NAc DA neurons. These results suggest that multiple projections of SuM neurons are involved in reinforcement. The finding of SuM efferent collateralization is consistent with other emerging (Holloway et al., 2022) and published observations (Vertes and McKenna, 2000). We also note that separate neural populations may be involved in negative emotional effects from the SuM neurons involved in positive reinforcement (Arima and Ikemoto, 2023; see also the next section). Further research is needed to understand functional implications of these collaterals and projection-defined neural populations.

To home in on a natural role for the SuM in reward-seeking, Kesner et al. (2021) performed single-unit recordings of SuM neurons during an operant task and found that different populations of SuM neurons responded to various appetitive behaviors (e.g., lever pressing) and reward predictive-stimuli (e.g., tones predicting availability of sucrose reward), but essentially all SuM neurons greatly reduce activity once mice began consuming rewards, and reactivated once consummatory behaviors ceased (Kesner et al., 2021). These observations led to the hypothesis that SuM activity is critical for environmental interaction toward obtaining goals (i.e., seeking-behaviors), and SuM activity is less involved once goals are obtained. To test this, Kesner et al. (2021) trained mice to perform a sucrose-seeking behavior involving lever pressing and reward predictive cues and found inhibition of SuM activity via intra-SuM infusions of the inhibitory GABA receptor agonists baclofen and muscimol greatly attenuated operant responding for rewards and attentiveness to reward predictive cues. But when sucrose reward was made freely available, the mice consumed similar amounts whether or not the SuM was inhibited (Kesner et al., 2021). Taken together, these SuM inhibition studies support a role for SuM in appetitive, but not consummatory, processes of reward-seeking behaviors. This role is also consistent with past studies using neurotoxic agents to perform microlesions of the SuM, where such lesions mainly disrupt reward-seeking behaviors related to behavioral inhibition, where animals must refrain from an appetitive response to earn rewards (Pan and McNaughton, 2002).

SuM and uncertainty/stress/aversion

In addition to rewarding properties of SuM stimulation described above, stimulation of specific SuM pathways can also be aversive. Profound real-time place aversion was found during optogenetic stimulation of SuM projections to the paraventricular nucleus of the thalamus (PVT) (Kesner et al., 2021). The PVT is a small, mediodorsal thalamic subnucleus that has received much interest in the past decade for its role in many behavioral states, including those involved with motivation (Choi and McNally, 2017; McGinty and Otis, 2020; Penzo and Gao, 2021) and aversive states during drug withdrawal (Zhu et al., 2016). Understanding the extent of SuM's functional influence of PVT-related circuitry is an area ripe for further research. Indeed, much less is known about how inputs to the PVT are involved during these behavioral processes, and few researchers may be aware of SuM's aversion-related inputs to this region.

Moreover, research involving c-Fos as a marker for neural activation suggests that SuM neurons are activated by stimuli that are not necessarily rewarding, per se. For example, c-Fos is strongly induced in SuM neurons by the stimuli, including novel environments (Wirtshafter et al., 1998); noxious stimuli (Bullitt, 1990); anxiety-provoking environments (Silveira et al., 1993); taste cues associated with sickness (Yasoshima et al., 2005); contexts or discrete stimuli paired with aversive stimuli (Sandner et al., 1992; Beck and Fibiger, 1995; Silveira et al., 1995; Day et al., 2004; Yasoshima et al., 2005); swim and restraint stress (Cullinan et al., 1995); contexts that allow hungry rats to anticipate food (Le May et al., 2019); and appetitive tasks that require spatial working memory (Vann et al., 2000). Moreover, experiments with fiber-photometry Ca²⁺ signals indicated that SuM glutamate neurons projecting to the LPO (whose activation is reinforcing as discussed above) were activated by novel stimuli, footshock, and footshock-paired cues, but not water reward or water-paired cues to which thirsty mice were exposed over a few days (Arima and Ikemoto, 2023). These results are consistent with other recently presented data (Holloway et al., 2022) and the idea that SuM glutamate neurons are activated by stimuli that demand attention during adaptive environmental interactions.

Therefore, the SuM participates in both approach and avoidance aspects of motivated behavior, characteristics which it shares with regions, such as the VTA and LH, with well-established roles in such processes. In a recent perspective article, Kesner et al. (2022) synthesize knowledge related to SuM-to-MS-to-VTA and SuM-to-hippocampal circuitries' roles in motivated behaviors and propose that these circuitries, rooted in the SuM, contribute to the motivation to seek information. They propose a framework for understanding information seeking behavior based on the concept of an environmental prediction error (Kesner et al., 2022), where salient stimuli, whether positively or negatively valanced, influence SuM activity that can then coordinate multiple brain systems toward processes involving adaptive behavior and learning/memory. Further research is needed to parse out SuM's roles in both reward and aversion and will likely be fruitful in advancing our understanding of SuM's overall role in cognitive processes related to environmental interaction.

SuM and neurotransmitter corelease in the DG

Although it has long been known that SuM neurons project to the DG and CA2 region (Segal and Landis, 1974; Amaral and Cowan, 1980; Haglund et al., 1984; Vertes, 1992), the precise synaptic connections between SuM and hippocampus have only recently begun to be understood. Previous anatomic studies demonstrated that axon terminals from the SuM in the DG form asymmetrical synapses onto dentate granule cells (DGGCs) (Dent et al., 1983; Magloczky et al., 1994; Halasy et al., 2004), suggesting that SuM inputs are excitatory. Recent anatomic studies have shown that SuM terminals in the DG form symmetrical synapses as well as asymmetrical synapses onto DGGCs and, more interestingly, its terminals contain both VGluT2 and VGAT (Boulland et al., 2009; Soussi et al., 2010; Root et al., 2018; Billwiller et al., 2020), implying corelease of glutamate and GABA. In agreement with this morphologic evidence, recent studies using electrophysiology combined with optogenetics elucidated that glutamate and GABA are coreleased at SuM-to-DGGC synapses (Pedersen et al., 2017; Hashimotodani et al., 2018; Billwiller et al., 2020; Chen et al., 2020; Li et al., 2020; Ajibola et al., 2021). Furthermore, DG interneurons, but not mossy cells, are also targeted by SuM axons, which corelease glutamate and GABA (Hashimotodani et al., 2018; Ajibola et al., 2021). Interestingly, in the CA2 region, glutamate is exclusively released at SuM-to-CA2 pyramidal neuron synapses (Chen et al., 2020; Robert et al., 2021). Unique synaptic transmission of corelease of glutamate and GABA in the DG exerts complex effects of SuM inputs on the DG network activity, such as direct excitation or inhibition of DGGCs, feedforward inhibition to DGGCs, and disinhibition of DGGCs (Segal, 1979; Mizumori et al., 1989; Carre and Harley, 1991; Nakanishi et al., 2001; Hashimotodani et al., 2018; Ajibola et al., 2021). These diverse modulations could be mediated by differential target cell-specific cotransmission balance of glutamate and GABA. Therefore, regulation of cotransmission balance of glutamate and GABA at SuM-to-DG synapses could have a strong impact in the DG network and learning and memory.

SuM and DG synaptic potentiation

It is now established that glutamatergic and GABAergic cotransmission balance at SuM-to-DGGC synapses is dynamically modulated in an activity-dependent manner (Hirai et al., 2022; Tabuchi et al., 2022). By depolarization of DGGCs, glutamatergic cotransmission at SuM-to-DGGC synapses exhibits LTP (Tabuchi et al., 2022). This depolarization-induced LTP of excitatory transmission (depol-eLTP) is NMDA receptor-independent and postsynaptically expressed. Importantly, GABAergic cotransmission is not modulated by DGGC depolarization. Therefore, depol-eLTP shifts excitation and inhibition balance of SuM inputs to excitation, thereby enhancing DGGC output. Mechanistically, burst firing of DGGCs triggered by perforant-path inputs induces depol-eLTP, indicating that depol-eLTP is heterosynaptically induced regardless of SuM activity. In contrast, NMDA receptor-dependent Hebbian LTP is also observed at SuM-to-DGGC synapses by a pairing of EPSPs of SuM-to-DGGC synapses with GC spikes (Hirai et al., 2022). A pairing of brief timing window of pre-post order induces spike timing-dependent LTP (t-LTP) of SuM-to-DGGC excitatory synapses. Similar to depol-eLTP, t-LTP is also selectively induced of glutamatergic, but not GABAergic, cotransmission, increasing net excitatory drive of SuM inputs. Given that neural activity between SuM and hippocampus is often synchronized (Ito et al., 2018; Li et al., 2020; Vicente et al., 2020; Farrell et al., 2021), coincident activity of SuM inputs and DGGCs could elicit t-LTP at SuM-to-DGGC synapses. By shifting glutamatergic and GABAergic cotransmission balance, LTP at SuM-to-DGGC synapses modulates DG network activity and might contribute to SuM-to-DG circuit-related learning and memory (Shahidi et al., 2004; Aranda et al., 2008; Gutierrez-Guzman et al., 2012; Chen et al., 2020; Li et al., 2020; Qin et al., 2022).

SuM and adult hippocampal neurogenesis

Recent studies have established the role of SuM in regulating spatial memory retrieval through the SuM-to-DG pathway (Li et al., 2020). The DG is a unique brain structure that not only participates in learning and memory, but also produces new neurons from neural stem cells (NSCs) in adulthood through a process known as adult hippocampal neurogenesis (AHN). The role of SuM in regulating AHN has not been investigated until recently. Despite mature DGGCs receiving SuM glutamate/GABA coreleasing inputs (Hashimotodani et al., 2018; Chen et al., 2020; Li et al., 2020; Ajibola et al., 2021), using optogenetics-guided slice recording, the Song group showed that developing adult-born cells receive differential inputs from SuM; either SuM-glutamate inputs (neural stem cells [NSCs]) or SuM-GABA inputs (adult born neurons [ABNs]). The dual SuM inputs start to appear when ABNs reach 4 weeks of cell age, suggesting that SuM glutamate or GABA may differentially regulate distinct neurogenesis stages of the multistage AHN process. They went on to find that patterned stimulation of SuM neurons promotes self-renewal and neurogenic proliferation of NSCs through SuM glutamate inputs yet promotes dendritic/spine development and maturation of ABNs through SuM GABA inputs. Therefore, patterned SuM stimulation across multiple neurogenesis stages from NSCs to ABNs collectively contributes to increased production of ABNs with improved properties (Li et al., 2022). Importantly, chemogenetic manipulation of the activity of these SuM-enhanced ABNs (vs control) modulates memory performance and anxiety-like behavior. These results highlight activity-dependent contribution of SuM-enhanced ABNs in hippocampal function. Stimulating the activity of SuM neurons by optogenetics could be artificial, raising the question of how SuM neurons respond to natural stimuli. Interestingly, SuM neurons exhibit increased firing frequency, calcium dynamics, and c-Fos expression when animals are exposed to a novel environment (Li et al., 2022). Importantly, SuM neurons are required for environmental novelty-induced neurogenic effects, as ablation of SuM neurons abolishes these effects (Li et al., 2022). These results raise an exciting possibility that stimulating SuM can mimic environmental novelty-induced enhancement of AHN and, therefore, may facilitate clinical benefits associated with environmental stimulation.

The Song group has gone on to ask whether this novel AHN-promoting strategy can be applied to diseased brains to restore functions, with the focus on Alzheimer's disease (AD). Using the 5xFAD mouse model, they found impaired cognitive and affective deficits, along with reduced AHN and SuM activity (Li et al., 2023). By application of patterned SuM stimulation as described above, the number and developmental properties of ABNs in AD mice are both restored. Importantly, acute chemogenetic activation of a small population of SuM-enhanced ABNs in AD mice is sufficient to improve memory performance, reduce anxiety/depressive-like behaviors, and promote microglia phagocytosis of plaques. By contrast, SuM stimulation alone or activation of ABNs without SuM modification fails to restore behavioral deficits, suggesting that both SuM stimulation and activation of SuM-modified ABNs are essential for functional restoration in AD mice. Furthermore, quantitative phosphoproteomics analyses of the whole hippocampal tissues reveal activation of the canonical pathways related to synaptic plasticity and microglia phagocytosis of plaques following acute chemogenetic activation of SuM-enhanced (vs control) ABNs (Li et al., 2023). These results establish activity-dependent contribution of SuM-enhanced ABNs in modulating AD-related deficits and inform signaling mechanisms mediated by activation of SuM-enhanced ABNs. Future studies are underway to further decipher cell type-specific mechanisms underlying these beneficial effects mediated by activation of SuM-enhanced ABNs in AD mouse models.

SuM and social recognition memory via the CA2 region

Within the realm of episodic memory, social recognition memory holds particular significance as it aids individuals in remembering past social encounters and associating them with specific individuals (Ferguson et al., 2002; Bielsky and Young, 2004; Penn and Frommen, 2010). The formation of social recognition memory entails the integration of sensory cues, emotional experiences, and social interactions, engaging specific neural networks, such as the hippocampus, amygdala, and PFC (Ferguson et al., 2001; Gur et al., 2014; Tanimizu et al., 2017; Wang and Zhan, 2022). Through the intricate interplay of these brain regions, social recognition memory contributes to our ability to navigate and thrive in social interactions.

In a recent study investigating the processing and routing of different types of novelty in the brain, Chen et al. (2020) identified the SuM as a crucial hub for novelty processing, including social novelty. They discovered that the SuM not only responds broadly to novel stimuli but also selectively directs specific types of information to distinct regions of the hippocampus, namely, the DG and CA2 regions. This team created a transgenic mouse line that expresses Cre-recombinase driven by the CSF2RB gene, which interestingly has notable, but functionally unclear, regional specificity to the SuM. Thus, the mouse line was named SuM-Cre; and using it, the researchers found that SuM neurons projecting to the DG are activated in response to contextual novelty, while SuM neurons projecting to the CA2 region are preferentially activated by novel social encounters. By manipulating these neural circuits, they demonstrated that different routing of novelty signals in these projections can modify contextual or social memory. Building on these findings, Robert et al. (2021) investigated the impact of SuM inputs on the local circuitry in the hippocampus. They found that stimulating SuM axons in the CA2 region increased excitatory responses in basket cells, primarily parvalbumin-expressing inhibitory neurons. These basket cells were responsible for the feedforward inhibitory drive from SuM to CA2. The researchers also observed that modified CA2 output, resulting from SuM stimulation, caused polysynaptic inhibition in the CA1 region, reducing firing rates. Together, these studies provided new insights into the hypothalamus–hippocampus connection, emphasizing the role of SuM in processing novelty and social recognition memory.

Intrigued by these findings, Thirtamara Rajamani et al. (2022) investigated how oxytocin in the SuM influences social recognition memory in rats. Their research was motivated by a substantial body of prior studies highlighting the crucial involvement of oxytocin in social recognition memory (Popik et al., 1992; Ferguson et al., 2000; Takayanagi et al., 2005), along with earlier evidence indicating the presence of oxytocin fibers and receptors in the SuM (Yoshimura et al., 1993; Kremarik et al., 1995; Gould and Zingg, 2003; Cumbers et al., 2007). The researchers commenced their study by confirming the presence of oxytocin fibers in the SuM and tracing their specific origin back to oxytocin neurons in the paraventricular nucleus of the hypothalamus (PVH). Moreover, they verified the existence of oxytocin receptors in the SuM, noting their expression on both glutamatergic and GABAergic neurons. They found that blocking oxytocin signaling in the SuM resulted in impaired social memory, thereby providing substantial evidence for the involvement of oxytocin within the SuM in the processes underlying social memory. Building on these findings, Thirtamara Rajamani et al. (2022) proposed a working model that outlines the functioning of the PVH^oxytocin-to-SuM pathway. According to their model, this pathway enhances the salience of social stimuli, while the activation of the SuM-to-CA2 pathway facilitates the processing of social information to support the formation and retention of social memory (Thirtamara Rajamani et al., 2022).

Further research is needed to understand the specific roles of the SuM to CA2 pathway in the different stages of social recognition memory and to uncover the molecular and cellular mechanisms by which oxytocin influences SuM and CA2 neurons. This understanding could lead to the identification of new targets for treating social behavior deficits seen in psychiatric disorders, such as schizophrenia and autism spectrum disorder.

SuM, sleep, and feeding

There are several other current avenues of research which have found exciting roles for the SuM. One such role is in sleep. Owing to its robust connections to basal forebrain and cortical areas involved in sleep-wake vigilance states (Saper, 1985; Saper and Fuller, 2017), the SuM has recently garnered attention from sleep researchers and has been shown to play an important role in REM sleep (Renouard et al., 2015; Luppi et al., 2017; Pedersen et al., 2017). REM sleep is a critical component of learning and memory, and it was recently shown that the aforementioned roles for SuM in social memory may be because of SuM-to-CA2 activity during REM sleep (Qin et al., 2022). Another topic where SuM has begun garnering an emerging role is in glucagon-like-peptide-1 (GLP-1) mediated feeding behaviors. GLP-1 is released centrally in nucleus tractus solitarius and the intermediate reticular nucleus of the medulla oblongata (Merchenthaler et al., 1999), and is highly implicated in feeding behavior and glucose metabolism (Smith et al., 2019; Chen et al., 2021). SuM neurons express GLP-1 receptor (Lopez-Ferreras et al., 2018), and activation of these neurons with GLP-1 microinjections (Vogel et al., 2016) or chemogenetic approaches (Lopez-Ferreras et al., 2018) reduces food-seeking behaviors.

SuM as a target for clinical intervention

Much of our knowledge about SuM has come from preclinical studies using rodents. In humans, the SuM is quite small and situated deep and medial, likely making it a difficult target for invasive interventions, such as deep brain stimulation (DBS). That being said, regions, such as the subthalamic nucleus and globus pallidus, are also small, deep-seated regions and are often targeted for DBS in several neurologic disorders (Perlmutter and Mink, 2006). Conversely, as we have reviewed, the SuM exerts important influence on regions with more feasible therapeutic access, such as hippocampal, basal forebrain, and cortical areas. For example, the MS has already been a target for DBS treatments related to oscillopathies, and stimulation there is well tolerated in patients (Takeuchi et al., 2021). So, understanding the functional influence of SuM on these downstream regions could help design DBS or transcranial magnetic stimulation protocols with therapeutic potential. From a pharmaceutical perspective, drugs targeting the GLP-1 system have gained extreme interest given the blockbuster results in treating a variety of conditions ranging from diabetes/obesity to substance use disorders using GLP-1 agonist drugs (Wilding et al., 2021; Klausen et al., 2022), and as such, the emerging role of SuM in GLP-1 mediated behavioral neurobiology is quite exciting. Similarly, as mentioned in the section on SuM and reward-seeking, rats actively self-administer nicotine into the SuM (Ikemoto et al., 2006), making this region one of very few where nicotine directly reinforces behavior. Understanding the extent that SuM plays a role in reinforcing properties of systemic nicotine administration (e.g., with tobacco or e-cigarette substance use disorders) may prove fruitful in helping individuals to stop using these drugs. In general, we have decades of research on canonical reward circuitry yet have very little translational success targeting these systems to treat psychiatric disorders arising from maladaptation in reward-seeking (e.g., depression, anxiety, substance use disorders). So, the emerging role for the SuM and its rather noncanonical influence on these behaviors may provide new avenues for treating these persistent distressing human conditions. Finally, as described in the section on regulation of hippocampal plasticity, learning, and memory, understanding SuM's profound influence on hippocampal processes related to learning and memory could provide useful strategies for treating disorders of these processes as seen in many psychiatric disorders, such as AD and dementias, and treat other types of learning/memory impairments.

In conclusion, the SuM is a small but mighty brain region. With its extensive network of connections, the SuM exerts its influence over various brain regions that play well-established roles in cognition. This review delves into several important aspects of SuM's role in behavior and cognition, such as spatial navigation and exploration, reward and aversion, learning and memory, and social interaction. It is important to note that the roles highlighted in this review are not exhaustive, and ongoing research suggests that the SuM may have additional, yet undiscovered roles. With continued investigation, future research has the potential to expand our knowledge on the SuM's functionality and augmenting the repertoire of brain regions that hold therapeutic promise.