Dopamine: The Essential Bridge Mediating Obstructive Sleep Apnea and Hippocampus-Dependent Learning and Memory Impairments.

Save

ABSTRACT

Obstructive sleep apnea is a clinical syndrome that triggers a series of pathophysiologic changes, including disturbed sleep architecture, chronic intermittent hypoxia and hypercapnia, and ultimately severe cognitive dysfunction. The hippocampus plays a key role in various cognitive processes such as learning and memory. Previous studies have focused on inflammatory factors, oxidative stress, and apoptotic changes in the hippocampus.

However, studies on neurotransmitter changes are scarce. Dopamine is an important neurotransmitter that has a direct effect on learning and memory formation in the hippocampus. This article summarizes the relationship between hippocampus‐dependent learning and memory deficits induced by obstructive sleep apnea and dopamine, and explores how dopamine mediates hippocampus‐dependent learning and memory deficits associated with obstructive sleep apnea.

Detailed searches were performed by utilizing open‐ended keywords on PubMed and manually searching the reference lists of the identified articles; all references up to December 31, 2024, were used. Search terms included “obstructive sleep apnea,” “OSA,” “chronic intermittent hypoxia,” “sleep fragmentation,” “sleep deprivation,” “sleep disorder,” “apoptosis,” “inflammation,” “oxidative stress,” “hippocampus,” and “dopamine.” The final article includes all articles that explicitly address dopamine maintenance of memory and how obstructive sleep apnea affects hippocampal dopamine. First, obstructive sleep apnea inhibits hippocampal dopamine synthesis.

Second, downregulated dopamine induces hippocampus‐associated inflammatory responses, oxidative stress, and apoptosis. Finally, intermittent hypoxia may also further exacerbate dopamine deficiency by inducing a positive cycle of dopamine downregulation. The net effect of these processes is a hippocampus‐dependent decline in learning and memory capacity, which is characteristic of obstructive sleep apnea.

Dopamine plays a crucial role in obstructive sleep apnea‐associated hippocampus‐dependent learning and memory deficits, not only due to the neural signaling role of dopamine itself, but also due to obstructive sleep apnea‐associated inflammation, apoptosis, and oxidative stress, which are also closely related to dopamine.

Introduction

Obstructive sleep apnea (OSA) is a clinical syndrome characterized by recurrent upper airway obstruction during sleep, leading to hypoventilation and/or respiratory interruption. This, in turn, leads to a series of pathophysiological changes, including disturbances in sleep architecture, chronic intermittent hypoxia (CIH), hypercapnia, and micro‐arousals [1]. Recent studies have indicated that the global prevalence of OSA is as high as 24%, with an upward trend [2].

OSA is a systemic disease [3], a close relationship has been demonstrated between OSA and cognitive dysfunction, with numerous studies examining the cognitive functioning of patients with OSA [4, 5, 6, 7, 8]. For example, compared to those without OSA, patients with OSA exhibit decreased attention, executive function, spatial memory capacity, working memory, and situational memory capacity. However, speech function is mostly unaffected [4, 5, 6].

According to previous studies, several mechanisms have been associated with impaired neurocognitive function in OSA, including intermittent hypoxemia (IH), sleep fragmentation (SF), and sleep deprivation (SD) [9, 10].

The hippocampus, a component of the limbic system, plays a crucial role in various cognitive processes [11]. Previous studies on the hippocampus have focused on hippocampus‐associated inflammatory factors, oxidative and antioxidant factors, and apoptotic factor changes, while there is a paucity of research on neurotransmitter changes [11]. Dopamine (DA) is a class of neurotransmitters that has been found to directly affect memory formation in the hippocampus [12, 13].

Hippocampal‐dependent synaptic plasticity [14, 15] and hippocampal‐dependent learning [16] represent significant conduits between the aforementioned processes, and research into both of these phenomena has identified DA as a crucial mediator.

The factors contributing to OSA include hypoxemia and hypercapnia, as well as microarousals during sleep, resulting in SF, SD, and excessive daytime sleepiness [17]. Current research suggests a close relationship between OSA and learning and memory function, with SF [18] and SD [18, 19, 20, 21, 22] potentially contributing to impaired learning and memory. Specifically, research has confirmed that SD significantly reduces hippocampal DA levels [23, 24, 25].

Consequently, the downregulation of DA itself is sufficient to induce hippocampus‐dependent learning and memory deficits by interfering with the formation of synaptic plasticity in the hippocampus. However, supplementation with dopamine receptors (DR) agonists reverses the SD‐induced decrease in DA concentration and DR receptor downregulation, alleviating SD‐induced learning and memory deficits [24]. Therefore, the DR receptor alterations are difficult to explain by synaptic plasticity.

The present study aims to provide a new entry point to this ambiguity by investigating whether OSA‐associated hippocampus‐dependent learning and memory deficits are exclusively attributable to DA interference with hippocampal synaptic plasticity or whether DA downregulation also affects hippocampus‐dependent learning and memory functions through alternative pathways.

In this article, we explore the role of DA in mediating OSA‐associated learning and memory deficits. First, DA plays a crucial role in SF‐ or SD‐induced learning and memory deficits, and SF or SD inhibits DA synthesis, which in turn directly inhibits DA‐associated hippocampus‐dependent synaptic plasticity. Second, downregulated DA induces inflammatory responses, oxidative stress, and apoptosis in the hippocampus, factors that ultimately contribute to hippocampus‐associated learning and memory deficits.

In the case of IH, in addition to the oxidative stress and inflammatory response induced by IH itself, IH may further exacerbate DA deficiency by inducing a positive cycle of DA downregulation, and ultimately, these outcomes induce OSA‐associated hippocampus‐dependent learning and memory decline.

Result

The Relationship Between Dopamine and Memory

DA is a catecholamine neurotransmitter that is derived from tyrosine. This is converted to DA by Tyrosine Hydroxylase (TH), primarily in dopaminergic neurons [26]. In addition, DA is also synthesized and metabolized by certain immune cells, including astrocytes and microglia [27].

DR are divided into two subgroups: D1‐like DR (D1DR and D5DR) and D2‐like DR (D2DR, D3DR, and D4DR) [26]. These receptors are widely expressed in neurons and several immune cell populations, for example, NK cells, T cells, macrophages, and dendritic cells [28]. In general, in addition to direct action on target neurons, the elevation of DA levels in the extracellular fluid resulting from DA release spillover is also a mode of dopaminergic action [29, 30].

Synaptic plasticity, defined as the ability of synapses to flexibly change the strength of their connections in response to activity, underlies learning and memory. Long‐term potentiation (LTP) and Long‐Term Depression (LTD) are its primary forms. LTP is a long‐lasting enhancement phenomenon that occurs in the signaling of two neurons, leading to a sustained enhancement of synaptic transmission.

Similar to LTP, LTD can be induced by a variety of protocols, which in turn leads to sustained attenuation of different forms of synaptic transmission [31]. The requirement of dopaminergic inputs for the induction of hippocampal LTP is well documented, and various forms of hippocampal‐dependent learning and memory are contingent on the maintenance of normal dopaminergic activity [32, 33, 34, 35, 36, 37, 38]. In addition to LTP, LTD can also encode fine spatial detail in the rat environment in vivo [39, 40].

Similarly to LTP, dopaminergic input is necessary for LTD to occur in the hippocampus [41, 42, 43].

Specifically, at the functional level, dopamine can have various effects on cognition, including influencing spatial memory, working memory, situational memory, attention, and executive function [12]. However, in the hippocampus, dopamine primarily affects spatial memory, working memory, and situational memory [13]. The specific effects of dopamine on hippocampal memory function have been detailed in a review [13].

Overall, this review primarily suggests that DA signaling can ultimately impact hippocampal memory function by lowering the threshold for synaptic plasticity or modulating the integration of different inputs from hippocampus Cornu Ammonis 1 pyramidal cells [13].

Effects of OSA on Dopamine and Memory

SF and SD

Inflammation Response

It is noteworthy that activation of DR alone does not affect learning and memory when SD is absent; however, activation of DR in the presence of SD alleviates SD‐induced learning and memory deficits [44]. This finding suggests that DA may play a protective role in the presence of additional stimuli that are, at least in part, directly related to SD [44]. As previously mentioned, the volumetric transmission resulting from elevated DA levels is also a mode of DA action [29, 30], but dopamine is also widely expressed in immune cells.

Consequently, DA‐mediated immune cell activation may be a significant cause of OSA‐induced hippocampus‐dependent learning and memory dysfunction [28]. Further studies have found that SD‐ and SF‐related hippocampus‐dependent learning and memory deficits are usually accompanied by an inflammatory response, mainly an increase in pro‐inflammatory cytokines in the brain and activation of immune cells [45, 46, 47], which ultimately manifests as hippocampus‐related learning and memory impairments [45, 46, 47].

Specifically, cytokines are primarily associated with the immune system; however, they can also lead to critical changes in learning and memory. Specifically for the hippocampus, SF increases the expression of IL‐1β, TNF‐α, and Transforming Growth Factor‐β (TGF‐β) [46, 47]. The induction of IL‐1β and TNF‐α in the hippocampus ultimately results in memory impairment [22, 25, 48].

IL‐1β induces concentration‐dependent memory impairment when injected directly into the hippocampus [49]. Furthermore, TNF‐α also inhibits LTP in the CA1 region of the hippocampus, thereby triggering memory impairment [48]. Therefore, SD and SF can induce the expression of pro‐inflammatory cytokines, which in turn induce learning and memory impairments.

The supplementation of DA can inhibit neuroinflammation in the hippocampus of mice, decrease the levels of pro‐inflammatory cytokines (e.g., TNF‐α and IL‐1β) in the hippocampus, and prevent hippocampus‐dependent learning and memory impairments [44]. However, the fundamental question remains: by what means does DA modulate inflammatory factor levels, ultimately affecting hippocampus‐dependent learning and memory?

It is found that lipopolysaccharide (LPS)‐induced microglia overexpress D1DR and D4DR in vitro [50]. In addition, LPS also induces the release of TNF‐α and nitric oxide (NO) from macrophages, and exogenous DR agonist supplementation alleviates this effect [51]. Further studies indicate that SD increases hippocampal IL‐1β and Interleukin‐6 (IL‐6) levels, accompanied by astrocyte activation in CA1 [52].

These results suggest that it is at least theoretically plausible that DA can bind to DR on immune cells, thereby inducing altered cytokine expression and ultimately leading to learning and memory deficits. Microglia are resident immune cells in the central nervous system that express DR, and there is a direct correlation between SD‐induced impaired learning and memory function and microglia activation [45]. Therefore, microglia are one of the important entry points for us to study “DA‐inflammation‐memory impairment.” It is found that DA alone has no effect on IL‐6, TNF‐α, or NO production, but DA inhibits LPS‐induced microglia activation and NO overexpression [53].

DA induces microglia to express D1DR and D5DR in large amounts and inhibits NO production, whereas antagonists reverse the DA‐induced NO inhibition [53]. Further studies reveal that DA modulates pro‐inflammatory responses in LPS‐stimulated microglia by regulating Extracellular Regulated protein Kinases (ERK) and Nuclear Factor kappa‐B (NF‐κB) signaling, which plays a role in DA‐mediated inhibition of NO production by altering Nitric Oxide synthase (iNOS) transcription [53]. Thus, DA may ultimately alleviate OSA‐induced learning and memory dysfunction by inhibiting the inflammatory response of immune cells and reducing cytokine levels (see Table S1 for species, sex, and age in different studies).

Oxidative Stress

Furthermore, SF has been demonstrated to induce hippocampal oxidative stress, as evidenced by elevated levels of lipid peroxidation and malondialdehyde (MDA) [18] and decreased hippocampal Superoxide Dismutase (SOD), glutathione peroxidase (GPx) activities, and decreased MDA levels [22]. It is noteworthy that the neurological damage caused by the pro‐oxidant effects of DA remains a subject of study. However, DA itself has antioxidant activity, and it may have neuroprotective effects.

Indeed, the neuroprotective effect of DA is not to be disregarded, and this effect can be principally categorized as neuroprotection induced by receptor‐ and non‐receptor‐mediated mechanisms [54, 55]. The non‐receptor‐mediated mechanism is attributable to the antioxidant activity of DA itself, while the receptor‐mediated mechanism necessitates the binding of DA to DR. In vitro studies have demonstrated that D4DR agonists offer protection to hippocampal neurons against oxidative stress [54, 56], a finding that cannot be fully explained by their intrinsic antioxidant properties alone [54].

In addition, in the kidney, DA binding to DR receptors inhibits NADPH oxidase and stimulates antioxidant enzymes [57]. In the striatum, DA alleviates oxidative stress by enhancing the levels of antioxidants, GSH, catalase, and SOD via D2DR [55]. However, it would appear that the neuroprotective effects of DA are more complex than this.

Apomorphine, a nonselective DR agonist, has been shown to induce oxidative stress and inflammatory responses in the hippocampus in the absence of stressors, which ultimately leads to neuronal damage [58]. Haloperidol, a D2DR antagonist, has been observed to reduce the antioxidant capacity of the hippocampus in the absence of stressors and to induce oxidative stress damage to neurons [59]. Based on these results, we hypothesize that the maintenance of homeostatic DA concentrations within the hippocampus is imperative and that fluctuations in DA concentration, whether elevated or decreased, induce oxidative stress.

For instance, in the absence of stressors, the administration of either DR agonists or antagonists leads to alterations in the relative concentration of DA, either elevated or decreased, which in turn induces oxidative stress in hippocampal neurons [58, 59]. Conversely, in the presence of stressors, DA concentration declines, thereby inducing oxidative stress, and it is in such circumstances that DA supplementation becomes pivotal [54, 56]. Since fluctuations in DA concentration can induce oxidative stress, balancing the effects of DA treatment on hippocampal oxidative stress may be an issue requiring further clarification.

We hypothesize that sustained dopaminergic stimulation, minimizing abrupt fluctuations in DA concentration, is a critical point. Furthermore, maintaining the minimal DA supplemental dose while controlling oxidative stress is essential. Additionally, adjusting the DA supplemental dose promptly according to the intensity of stressors is also necessary.

However, these perspectives require further research to validate. Consequently, DA has the potential to mitigate OSA‐induced learning and memory dysfunction by attenuating SF‐ and SD‐induced oxidative stress.

Apoptosis

Apoptosis is defined as a process of cell death that is often secondary to oxidative stress and inflammatory responses. Studies have shown that SD upregulates levels of the proapoptotic Bcl‐2‐associated X protein (Bax) and cleaved caspase‐3 in the hippocampus, and decreases the antiapoptotic protein B‐cell lymphoma‐2 (Bcl‐2), thereby promoting apoptosis in neurons [60, 61, 62, 63, 64]. However, SF and SD can induce inflammatory responses and oxidative stress, which themselves can lead to hippocampal neuronal apoptosis, which does not seem to be directly related to DA.

Therefore, we would like to clarify whether DA plays a role in SF‐ and SD‐induced apoptosis without oxidative stress and inflammatory response.

The cAMP response element binding (CREB) protein is located in the nucleus, where it acts as a transcription factor. This protein is thought to play a key role in learning and memory processes [65, 66]. CREB can regulate gene expression through a variety of signaling pathways, including cellular processes such as proliferation, differentiation, and survival.

CREB is regulated by protein kinase A (PKA), mitogen‐activated protein kinase (MAPK), and Ca/calmodulin‐dependent protein kinases (CaMK), which serve as the primary upstream pathways. CREB activation ultimately leads to phosphorylation, activating downstream pathways and antagonizing apoptosis [67, 68]. SD has been shown to downregulate CREB expression and inhibit CREB phosphorylation, ultimately promoting neuronal apoptosis in the hippocampus and leading to learning and memory deficits [25, 65].

Conversely, DR activation significantly reverses SD‐induced reductions in CREB gene and protein phosphorylation and ameliorates SD‐induced learning and memory deficits in the hippocampus [37].

Studies on the upstream signaling pathway have shown that SD reduces the phosphorylation of ERK [25], which is one of the MAPKs, leading to impairment of the MAPK signaling pathway. In addition, SD upregulates levels of phosphodiesterase (PDE), which degrades cAMP, leading to an impaired PKA pathway [65]. However, D1DR agonists increase the expression of genes and proteins in the PKA pathway and alleviate SD‐induced learning and memory deficits [37].

For CaMK, activation of phosphatidylinositol‐linked dopamine receptors (PLDRs) promotes calcium release, which further activates CaMK II and CaMK IV, ultimately promoting CREB phosphorylation [68]. SD decreases CaMK IV levels, but D1DR receptor agonists dramatically reverse the downregulation of CaMK IV gene and protein expression, alleviating learning and memory deficits [37]. In addition to these classic signaling pathways, SD inhibits activation of the PI3K/AKT/GSK‐3β pathway [63, 64], ultimately promoting hippocampal neuronal damage via the Bcl‐2/Bax/Caspase‐3 pathway [63, 64].

However, DR agonists upregulate the PI3K/AKT/GSK‐3β signaling pathway, inhibiting apoptosis of hippocampal neurons [69, 70]. Notably, activation of the PI3K/AKT/GSK‐3β pathway under normal conditions shows antiapoptotic properties but does not ameliorate cognitive deficits, possibly because there is no significant hippocampal neuronal damage [63]. Therefore, DA may alleviate OSA‐induced hippocampus‐dependent learning and memory deficits through the classical PKA, MAPK, and CaMK signaling pathways.

In addition, activation of the PI3K/AKT/GSK‐3β pathway is also important for DA to improve OSA‐related learning and memory deficits (Figure 1).

CIH

Numerous studies have shown that CIH can induce oxidative stress and inflammatory responses that lead to hippocampus‐related learning and memory deficits. However, little is known about how DA is involved in CIH‐induced learning and memory deficits. No study has directly investigated how hippocampal DA levels are altered by CIH, so in this section, we will discuss whether CIH can induce changes in DA and its metabolism in other brain regions to provide a reference for hippocampal‐related DA changes.

A study using microdialysis to measure changes in DA concentration found that IH decreases extracellular DA levels in the striatum [71]. Another study using cell homogenization shows that IH induces an increase in DA concentration in the frontal cortex [72]. In addition, IH also causes a large increase in DA levels in medullary cell homogenates [73].

In vitro experiments demonstrate that IH exposure increases the DA content of PC‐12 [74]. Thus, from these results, it seems unclear to us how IH induces changes in DA concentration. Based on the fact that only extracellular DA is the real functioning DA, we assumed that DA can be divided into total DA, intracellular DA, and extracellular DA, for example, the DA concentration measured by cell homogenate as mentioned above is the total DA concentration [72, 73], the DA content measured by microdialysis is the extracellular DA content [71], and the cellular DA content measured by in vitro experiments is the intracellular DA [74].

So, how IH mediates the changes in DA concentration is our next question to be answered. It has been shown that IH induces upregulation of striatal vesicular monoamine transporter‐2 (VMAT‐2) protein expression, which is responsible for transporting synthesized DA into vesicles [75]. In addition, IH induces an upregulation of TH phosphorylation and TH activity in the medulla oblongata.

TH phosphorylation induces TH activation [73]. In vitro experiments have shown that IH also induces an upregulation of TH activity in PC‐12 [74]. However, it does not alter the expression level of TH protein, which is comparable in normoxic and IH cells [74].

Given that TH acts as a key enzyme in DA synthesis and that phosphorylation of TH activates TH, we wonder whether IH could alter DA synthesis by regulating TH phosphorylation. Interestingly, protein phosphatase 2A (PP2A) dephosphorylates substrate proteins, but IH reduces the activity and expression of PP2A, ultimately leading to an increase in TH activity [73]. In addition, in vitro studies have shown that PKA and CaMK are also involved in TH phosphorylation and subsequent activation in PC12 cells [74].

In detail, CaMK and PKA inhibitors attenuate IH‐induced TH activation [74]. Taken together, IH downregulates the activity and expression of PP2A, or through the PKA and CaMK pathways, both of which promote TH phosphorylation, and that phosphorylated TH promotes the synthesis of DA, which in turn undergoes eventual translocation into vesicles via the upregulated VMAT‐2. Thus, IH ultimately induces an increase in intracellular DA and total DA.

However, how the extracellular DA concentration decreases remains controversial. As we all know, hippocampal dopaminergic synapses are mainly composed of axons from VTA and LC projections and postsynaptic membranes from hippocampal neurons that form synaptic structures [13, 76]. Based on previous statements [13, 27, 37, 76], we simply divide the cells within the hippocampus into three categories.

Class 1 is dopaminergic neurons from VTE and LC, which provide the presynaptic membrane; class 2 is hippocampal neurons, which provide the postsynaptic membrane; and Class 3 is immune cells in the hippocampus. Activation of the D2DR has been shown to upregulate the activity of PP2A [44], and the D2DR acts as a presynaptic receptor to inhibit the synthesis and release of DR [77]. In addition, activation of PLDR promotes calcium release and further activates CaMK [68].

It is therefore likely that DA mediates IH‐induced changes in DA through these pathways (Figure 2). For example, IH induces a decrease in extracellular DA levels [71], and the decreased DA attenuates binding to presynaptic D2DR in dopaminergic neurons, downregulates PP2A levels, and facilitates DA synthesis, consistent with upregulated VMAT‐2. However, IH also induces changes in synaptic structure [78], resulting in an inability to release oversynthesized DA [78], again consistent with a decrease in extracellular DA levels.

The decrease in extracellular DA attenuates binding to the D1DR on immune cells, downregulates PKA or CAMK, and inhibits phosphorylation of TH, which ultimately attenuates the synthesis and release of DA from immune cells. These combined results lead to a decrease in extracellular DA concentration. The relationship between DA reduction and oxidative stress, inflammatory response, and apoptosis, and how this ultimately leads to hippocampus‐dependent learning and memory deficits, has been described above and will not be repeated here.

Conclusion

Focusing on the hippocampus and DA, our article summarizes the role of DA in OSA‐associated hippocampus‐dependent learning and memory deficits. Notably, although impaired spatial, working, and situational memory in patients with OSA is strongly associated with the hippocampus [4, 5, 6], impaired attention and executive function in OSA do not appear to be associated with the hippocampus but are primarily mediated by the frontal cortex [79]. Although our article also does not mention whether DA is associated with OSA‐related impaired attention and executive function, OSA can induce changes in DA levels in the frontal cortex [72].

Therefore, whether DA still mediates OSA‐associated reductions in attention and executive function through the various pathways we mentioned is a point that could be further investigated in the future. In conclusion, our article suggests an important role for DA in OSA‐induced cognitive deficits and provides a theoretical basis for further basic and clinical studies in the future.

Author Contributions

Rui Fan: conceptualization, investigation, writing – original draft. Tao Li: conceptualization, funding acquisition, writing – review and editing. Yan Yan: conceptualization, project administration, writing – review and editing.

Ethics Statement

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Related articles

PubMed

Machine learning prediction of the change in sleepiness following continuous positive airway pressure in obstructive sleep apnea: a multi-cohort analysis.

Abstract Study Objectives To predict the improvement in sleepiness following continuous positive airway pressure treatment for obstructive sleep apnea using baseline characteristics. Materials and Methods Data from five polysomnography cohorts (HomePAP, BestAIR, ABC, STAGES, and MESA; total n = 2332) were analyzed. Associations between polysomnographic metrics and Epworth Sleepiness Scale were assessed using multivariable linear regression. Cross-validated Least Absolute Shrinkage and Select...

PubMed

Ticagrelor-Associated Central Sleep Apnea: Mechanisms, Clinical Evidence, and Implications for Sleep Medicine.

Ticagrelor is a reversible P2Y12 receptor antagonist used for antiplatelet therapy and is well recognized to cause dyspnea. Emerging concern has focused on a possible association between ticagrelor and central sleep apnea, Cheyne-Stokes respiration, and other manifestations of ventilatory instability. This narrative review examines the clinical evidence for ticagrelor-associated central sleep apnea and the respiratory-control mechanisms proposed to underlie this association. A structured literat...