Cholecystokinin (CCK) comprises a family of intestinal peptide hormones that share the same five C-terminal amino acids as gastrin (Table 1). CCK exists in several forms depending on the number of amino acids it contains and the presence in most of them of a sulfate group attached to a tyrosine located seven residues from the C-terminus (denoted with the letter S) [1]. The secretion of the longest forms [CCK-22S (sulfated 22-aa CCK), CCK-33S, CCK-58S] is linked to the upper gut; meanwhile, sulfated CCK octapeptide (CCK-8S) is expressed in higher quantities than any other neuropeptide in the brain [1–3]. CCK functions through two receptor subtypes: the “alimentary” CCK₁r, largely expressed in the gastrointestinal tract, and the “brain” CCK₂r, which is predominant in the brain [4, 5] (Table 2).

CCK is a complex and multifaceted messenger that has undergone over 600 million years of evolutionary history [1]. In the central nervous system, CCK-mediated neurotransmission regulates feeding behavior [6], modulation of opioid-mediated analgesia [7–9], memory, and cognition [10–12]. Interestingly, alterations of the brain CCK system have been linked to the physiopathology of schizophrenia [13–15], major depression [16], suicide [17], addiction [14, 18–20], and particularly anxiety [13, 21, 22]. Despite the high hopes of the preclinical evidence, most of the CCK antagonists have been shown to be unsuccessful in psychiatry clinical trials and as pain medicine [23–26] (see comparative Table 3). These disappointing outcomes sparked contentious debates on the possible therapeutic benefits of drugs that target CCK₂r-mediated neurotransmission specifically [27]. Even while interest in CCK’s therapeutic potential subsequently waned, recent discoveries on its importance in the central nervous system have reignited it [3]. In light of this field’s resurgence, and to encourage further clinical research, the goal of this review was to propose a new function of CCK neurotransmission: the regulation of neuronal homeostasis during stress.

Bioactive peptides of varying lengths with different N-terminal extensions [CCK tetrapeptide (CCK-4), CCK-8, CCK-12, CCK-22, CCK-33, and CCK-58] are synthesized by enzymatic processing of the human CCK preproprotein (115 aa residues) [1]. CCK possesses a sulfated tyrosine at the 7th amino acid residue from the C-terminus. Notice that the C-terminal phenylalanine residue is amidated. The C-terminal sequence (Trp-Met-Asp-Phe-NH₂) is highly conserved across different CCK peptides and gastrin, and it’s important for its biological activity.

This review was based on documents retrieved from the Scopus and PubMed databases as of April 1, 2025. All research publications addressing CCK in the nervous systems that were written in English and published in peer-reviewed journals between 1975—the year when CCK was first discovered in the brain—and 2025 met the inclusion criteria. For advanced research, the following keywords were adopted to sight documents: TITLE-ABS-KEY [(“Cholecystokinin”) AND (“CCK-A receptor” OR “CCK-B receptor” OR “central nervous system” OR “neurotransmission” OR “brain” OR “neurons” OR “amygdala” OR “hippocampus” OR “cortex” OR “hypothalamus” OR “anxiety” OR “learning” OR “memory” OR “satiety” OR “pain” OR “peripheral nervous system” OR “gastrointestinal”)] AND PUBYEAR > 1975 AND PUBYEAR < 2025 AND [LIMIT-TO (DOCTYPE, “ar”) OR LIMIT-TO (DOCTYPE, “re”)]. Eligible articles were based on the author’s own experience with the topic.

Several explanations have been proposed to explain the disheartening clinical trials [23–26]. The most plausible one is “dynamic neuromodulation” [10], which means that CCK release is triggered in response to high-frequency neuronal firing [28] to regulate the activity of other neurotransmitters. Another possibility is that CCK could interact with both CCK₁r and with CCK₂r in the brain [29], producing opposite effects in most cases [30, 31]. The mesolimbic system is the most well-known case of this [32]. CCK colocalized with dopamine in neurons of the tegmental ventral area projecting to the medial nucleus accumbens [33]. In the anterior part of the nucleus accumbens, CCK inhibits dopamine release via CCK₂r, whereas CCK via CCK₁r promotes DA in the posterior nucleus accumbens [34]. CCK controls dopamine neurotransmission in the limbic system, affecting motivation [35, 36], reward, and anxiety [28]. In the conditioned place preference (CPP) test in the rat, CCK₂r and CCK₁r antagonists enhance and decrease respectively the rewarding effects of morphine [37]. The close neuroanatomical distribution of CCK with opioids in the limbic system raises the possibility of an opioid-CCK functional link [38, 39]. Remarkably, the activation of CCK₁r and CCK₂r by endogenous CCK may also have opposite effects in the regulation of antidepressant effects induced by endogenous enkephalins [31]. Herein, the reciprocal relationships between the two CCK receptor subtypes further underscore the neuromodulatory relevance of CCK by fine-tuning the impact of opioid- and dopamine-mediated neurotransmission.

CCK influences brain-wide structural-functional networks across the isocortex [40]. CCK’s function in memory relies on the hippocampal neuronal circuitry [41–44]. For instance, in the developing dentate gyrus, cortical activity guides the formation of the CCK⁺ basket cell network, which preserves the inhibitory to excitatory balance in the hippocampus, a crucial aspect of learning and memory [45]. More importantly, the evidence shows that CCK release interacts with CCK₂r to promote high-frequency stimulation-induced long-term potentiation caused by NMDA receptors [46–48]. CCK is heavily present in neurons of the hippocampus and subiculum, sending fibers to the septum and hypothalamus [49]. In the dorsomedial nucleus of the hypothalamus, CCK shifts the plasticity of GABA synapses from long-term depression to long-term potentiation [50]. There is proof that CCK-containing interneurons play a crucial role in the regulation of place-cell temporal coding and the development of contextual memories [51]. Given the evidence that CCK₂r antagonists have carry-over effects in the baseline for anxiety after the drug is cleared [52], it is conceivable that CCK may generate plastic changes in the brain [3, 53–56].

CCKergic pathways appear to be interwoven with key components of the anxiety/fear, memory, and pain networks such as the amygdala, cortex, hippocampus, periaqueductal grey (PAG), and hypothalamus [57] (Figure 1). Intravenous administration of full CCK₂r agonists such as CCK-4 [58] and the sinthetic analogue pentagastrin [59] is panicogenic in healthy volunteers and worsens symptoms in panic attack patients. Functional magnetic resonance imaging in humans points to a cerebral activation in anxiety-related brain regions following CCK-4-induced panic challenge [60, 61]. In rodents, CCK-induced anxiety is linked to CCK₂r activation at the basolateral amygdala (BLA) [28, 62, 63–65] and cerebral cortex [29, 66].

Glutamate-GABA harmony plays a critical role in anxiety [67, 68], pain [69], and memory [70–72]. CCK is interspersed in the excitatory-inhibitory neural circuits of limbic cortices [73, 74]. Enhanced neuronal excitability may be one of the mechanisms by which the selective CCK₂r activation causes anxiety and panic attacks in humans [75]. GABA release is crucial for regulating anxiety and fear processing in BLA [62, 76], hippocampus [77, 78], and cerebral cortex [48]. GABAergic inhibition is modulated by CCK₂r [62, 79, 80] and to a lesser degree by CCK₁r [29, 62]. CCK controls glutamate [81, 82] and GABA [83] release in the hippocampus. When anxiety is expressed, CCK controls electrical activity in the cortex [29]. Similar CCK-mediated mechanisms play an important role in cognition [11, 84]. By facilitating glutamate release and gating GABAergic basket cell activity in the hippocampus, CCK regulates memory rather than encoding it [82].

In inflammatory pain, the CCK/CCK₂ system of the central amygdala switches from an anxiogenic to analgesic role that implicates descending control to the spinal cord [85]. CCK takes part in the descending pain facilitation system, particularly in the rostral ventromedial medulla and spinal cord [86, 87]. CCK contributes to pain hypersensitivity and is implicated in the nocebo effect [8]. This effect is likely to be mediated by the CCK input from the anterior cingulate cortex to the lateral PAG [88]. Interestingly, the CCK₂r is thought to be responsible for anxiety-induced hyperalgesia states in this structure [89], since it mediates the anxiogenic effect of CCK [90–92]. Integrating aversive memories and mediating defensive and emotional states, including fear, anxiety, and pain, may be important functions for CCK in the PAG [93].

The hypothalamic CCK plays a role in mediating stress responses, particularly the hypothalamic-pituitary-adrenal (HPA) axis [94–96] and stress-induced suppression of appetite [97]. Corticotropin-releasing hormone (CRH) and CCK are strongly related in the human CNS [98]. There is also evidence that the hypothalamus acts as the primary coordinator of memory updating [99]. HPA alterations impact on memory [100]. The paraventricular nucleus of the hypothalamus (PVN) is a major site of CCK concentration in the hypothalamus and where CRH neurons express CCK [101]. The CRH type 1 receptor or CRHR1 interacts with CCK to trigger anxiety [102]. Therefore, it is not extrange that PVN, a node for the CCK-regulated stress responses [103], is also one of the significant sites of glucocorticoid negative feedback regulation of the HPA axis [104]. Besides glucocorticoid feedback via bloodstream, PVN receives projections from the hindbrain neurons in the nucleus of the solitary tract (NTS) [105], the port of entry of vagal afferents.

The vagus nerve, which presents both CCK₁r and CCK₂r [106], is the route that uses intraperitoneal CCK to enhance memory retention [107, 108]. CCK activating vagal afferent C fibers enhances memory consolidation and retention involved in long-term visceral negative affective state like in irritable bowel syndrome [109]. Peripheral CCK may work partially through centrally projecting neurons from the nucleus tractus solitarius [110], since it has been connected to the activation of brain stem neurons, amygdala, and hypothalamus [111]. In contrast to the central CCK, which uses CCK₂r found in the hippocampus to produce its mnemonic effects [47, 48, 56, 112, 113], peripheral CCK may aid in memory formation via CCK₁r [114, 115]. Brain-derived neurotrophic factor is likely to be an intermediate of vagus nerve/CCK-1R-mediated memory [116].

Growing preclinical evidence points to an impact of CCK on memory [41, 45, 51, 53, 55, 56, 81, 112, 117–119]. NMDA receptors promote CCK release in the cerebral cortex [120] and the hippocampus, where it switches long-term potentiation [121]. Through neuroplasticity, memory offers tools to rewire anxious brain patterns, lowering hypervigilance, and encouraging more balanced responses [122–124]. Similarly, rather than merely being a moment-to-moment appraisal of a nociceptive input, perception of pain is a dynamic process that is influenced by prior experience and basal anxiety levels [8, 125]. Trace fear memory development is facilitated by neuroplasticity processes through CCKergic projections terminals of the anterior cingulate cortex into the lateral amygdala [126]. Anticipatory stress may be impacted by CCK’s role in fostering associative memory [47, 127, 128]. This implies that, via modulating memory and neural plasticity, CCK may have a great impact on anxiety/fear [126] and nocebo pain effect [8].

Neuronal homeostasis refers to the nervous system’s ability to maintain a stable internal environment and regulate neuronal firing, ensuring proper function and adaptation to changing conditions. CCK would carry out its neuronal homeostatic function in four ways: (1) Homeostatic plasticity mechanisms; (2) interaction with retrograde endocannabinoid signaling; (3) neuroprotection; and (4) dynamic neuromodulation of CCK release occurring after high-frequency neuronal firing (for a summary, see Table 4).

Synaptic scaling and intrinsic excitability changes are crucial components of neuronal homeostatic plasticity, since they stabilize firing and overall network function around a set-point value in the face of activity fluctuations [129]. CCK can influence synaptic transmission, potentially through synaptic mechanisms (affecting glutamate and GABA release) or postsynaptically altering receptor sensitivity. CCK can alter the properties of ion channels, impacting intrinsic excitability, that is, the neurons’ firing threshold and firing rate. In the context of chronic pain, CCK-8 enhances acid-sensing ion channel currents in rat primary sensory neurons [130]. Through its remarkable cell-specific modulation of excitatory and inhibitory signals and synaptic transmission, the CCK system can influence these mechanisms, contributing as an essential molecular switch to regulate the functional output of neural circuits [74, 131]. Synaptic plasticity is an interesting aspect of neuronal homeostasis in which endogenous CCK release could hypothetically operate to ameliorate the impairment induced by stress to synaptic plasticity [132].

The endocannabinoid system plays a significant homeostatic role in brain functions [133]. Cortical CCK⁺-GABA basket cells, which exert perisomatic inhibition of pyramidal cells [73, 74, 134], are downstream of the activation of cannabinoid type-1 (CB1) receptors in the forebrain [135]. The endocannabinoid system is involved in the generation and stability of rhythmic synchronous network activity of the CA1 region of the hippocampus that impacts cognitive processes, which is mediated by the chemical and electrical coupling of CCK interneurons co-expressing CB1 receptors [136]. Additionally, CB1 and CCK₂ receptors work together to modulate cortical GABAergic release in opposite ways in the cortex, making them relevant to anxiety [80]. The similar thing occurs with CCK₁r in the PAG that can both oppose and reinforce opioid and cannabinoid modulation of pain and anxiety within this brain structure [137]. Lastly, the amygdala projection CCK⁺-glutamatergic neurons to the nucleus accumbens, which regulates mood stability, have CB1 receptors [138]. Thus, CB1 receptors widely mediate endocannabinoid effects on glutamatergic and GABAergic transmission to modulate cortical networks and the expression of anxiety and fear [139]. It is likely that fear-related psychiatric diseases may be the result of the dysfunctional CCK-CB1 homeostatic interactions [140–142].

Homeostatic mechanisms protect neurons from harm, particularly during times of stressful and chronic pain situations, as well as prolonged and intense cognitive efforts accompanied by stress (i.e., cognitive overload). Chronic stress and pain have detrimental effects on the limbic system [143]. Oxidative stress may be a major component of anxiety pathology [144], while chronic pain leads to the weakening or loss of these synaptic connections, leading to maladaptive changes in the brain [145]. Neuroprotection by CCK can occur through an anti-oxidative stress mechanism [146] or by the anti-inflammatory inactivation of microglia through CCK₂r [147]. Preclinical evidence suggests that CCK could even help with depression, Parkinson’s and Alzheimer’s diseases through CCK₂r [76, 82, 148–150] and cerebral ataxia via CCK₁r [151].

The fact that CCK release mostly monitors neuronal activity and that no single neurotransmitter directly causes it reinforces CCK’s role in modulating certain aspects of neuronal homeostasis. It is known that the interaction of CCK with serotonin in the cortex under aversive conditions [152], and that serotonin functions as a strong CCK release factor in the cerebral cortex and nucleus accumbens by activating 5-HT₃ receptors on the CCK-releasing terminals [153]. In the ventral tegmental area, CCK is released from the somato-dendrites of dopamine neurons, triggering long-term potentiation of GABAergic synapses onto those same dopamine neurons [154] and dopamine release in the nucleus accumbens and the amygdala via CCK₁r [155]. GABA regulates CCK release in the cortex [156], while dopamine controls CCK release in the neostriatum [157], and opioids mediate CCK release in the spinal cord [158] and frontal cortex [159]. Through its ability to colocalize and interact with these neurotransmitters, CCK is actually in the position to modulate a broad range of behaviors and functions.

The most illustrative example is the homeostasis of the opioid system by CCK [11, 39]. CCK₂r activity accounts for neuropathic pain [160] and the development of opioid tolerance and/or dependence after chronic administration of opioids [158]. CCK₁r also contributes to the anti-opioid action of CCK [161, 162] and visceral pain at the level of dorsal root ganglia [163], but the cooperative stimulation of both CCK₁r and CCK₂r produces modest opioid-like effects [164, 165]. CCK-opioid interactions are in the onset and manifestation of stress-induced hyperalgesia [89], morphine withdrawal-induced stress [166], and addiction [37].

The dynamic neuromodulatory action of CCK through two receptors [10], its neuroplasticity role [53], and the involvement in overlapping brain processes (anxiety, pain, and memory) suggests that the CCKergic system is a network of three main components (CCK, CCK₁r, and CCK₂r) interacting together to restore baseline neuronal function. The brain produces C-terminal sulfated octapeptide fragments of CCK-8 or CCK-8S, one of the major neuropeptides in the brain, followed to a lesser degree by CCK-4 [167], which is released in distinct limbic regions under anxiety [168]. The distribution of CCK₁r and CCK₂r across the peripheral and central nervous system differs, as do their affinities for these fragments. Though sparsely distributed in the brain, CCK₁r is highly selective for the CCK-8S, whereas CCK₂r is more common in the brain but less selective due to its interchangeable binding with CCK-8S and CCK-4 [169]. The components of the CCKergic system would be positioned conveniently over several network subdivisions responsible for different brain processes, resulting in a particular pattern of CCK₁r and CCK₂r activation across subdivisions depending on where and how CCK-8S and CCK-4 are released (Figure 1). Any imbalance in the CCKergic system could change how the brain processes memory, pain, and anxiety.

Several neural pathways were connected in this model to bolster the CCKergic system hypothesis. One of them is the HPA axis, which is regulated by CCK in the human brain [98], and whose disruption can lead to alteration of neuronal homeostasis [170]. The HPA axis plays a crucial role in regulating body stress response, including its impact on anxiety and memory [100]. CCK function on the HPA axis might be accomplished through CCK₂r [94, 95] (Figure 1). The vagus nerve, which contains both CCK receptors [103], and the NTS, which expresses CCK [171], form the brain-gut axis, the main brain-body communication [172]. Since the NTS is an essential autonomic integration center with reciprocal connections with the PAG, the HPA axis, and the amygdala [173], its participation in CCK-mediated anxiety and memory cannot be excluded [107–109, 174] (Figure 1). It should come as no surprise that endogenous peripheral and brain CCK, each activating unique neural circuits, have anxiogenic and anxiolytic effects respectively through the CCK₂r [174, 175], whereas the inhibition of peripheral CCK impairs memory [174].

In the cerebral cortex, NMDA receptors trigger CCK release [47]. Although CCK₂r is regarded as the brain-specific receptor, the electrical activity of local neuronal networks in the fronto-parietal neocortex [176] and hippocampus [177] is under the control of diffuse populations of CCK₁r. The intercalation of CCK-expressing neurons with excitatory and inhibitory neuronal circuits of the limbic system controls dopamine-mediated neurotransmission [14] (Figure 1), which in turn modulates anxiety-like behaviors [178] and memory [179]. Rat anxiety-like behaviors are undeniably mediated by CCK₂r- [180], while CCK₁r antagonists also have anxiolytic effects [181, 182]. Both of these effects seem to depend on cortical CCK receptors [29]. The enhancing effects of CCK in memory [183] also require the participation of both receptors, albeit through different pathways [114]. Thus, CCK₁r agonists and CCK₂r antagonists both enhance memory in an olfactory recognition test in the rat [184]. This could be the reason why the injection of the selective CCK₂r agonist BC264 into the nucleus accumbens impairs memory in the rat [185]. Short-term memory is affected in healthy volunteers by the panicogenic agent CCK-4, a full CCK₂r agonist [186]. While CCK-4 is raised during stress [168], high levels of CCK-8S during the induction of stress can mitigate the detrimental effects of stress on hippocampal synaptic plasticity and memory [147]. This demonstrates how CCK₁r and CCK₂r work in conjunction to support CCK function at the cortical intersection of anxiety and memory [29].

The cortical CCKergic system may contribute to pain modulation [187] through the CCK/CCK₂r system within the amygdala [85] and likely by the connections of the anterior cingulate cortex to lateral PAG [88]. In the rat, CCK microinjection into the ventrolateral and dorsolateral PAG produces anxiolytic-like and anxiogenic-like effects, respectively [93]. Additionally, through the activation of CCK₂r, CCK exerts its pronociceptive and anxiety-induced hyperalgesia effects in the PAG [89, 188]. Spinal CCK₁r also contributes to the anti-opioid action of CCK [161, 162]. Because CCK₁r and CCK₂r produce modest opioid-like effects [164, 165], it is anticipated that coordinated activation of both receptors across the cortex-PAG-spinal cord axis will play a role in the nexus of anxiety and pain (Figure 1).

Neuronal homeostasis over a wide range of temporal and spatial scales requires dynamic plastic changes of neuronal and circuit activity [189]. Because of their active neuromodulatory and neuroplasticity roles, the elements of the CCKergic systems work together to support cognition and affective regulation [10, 48, 53, 64, 76]. During neuronal homeostasis, compensatory plastic changes of the CCKergic system could go awry. Therefore, it should not come as a surprise that certain cortical areas of the schizophrenic brain exhibit abnormal CCK mRNA expression [190, 191], and that the cerebral cortex of suicide victims shows abnormally elevated CCK₂r binding [192]. In the rat, interindividual differences in “novelty-seeking”—a behavioral trait associated with anxiety and addiction—are influenced by varying expression of CCK elements across the limbic system. [193, 194]. Restoring the proper function of the entire CCK system, rather than just a single component, may be necessary for the correction of certain mental conditions.

The link between anxiety and CCK₂r expression is not as straightforward as it was first thought [195]. Even if intravenous administration of full CCK₂r agonists [58] is panicogenic in healthy volunteers and worsens symptoms in panic attack patients, CCK₂r antagonists have not been proven to alleviate panic attacks [24, 25]. They are also ineffective in generalized anxiety disorder [23]. Even worse, patients with panic disorder receiving long-term treatment with the drug had an unforeseen higher baseline incidence of panic attacks than those receiving a placebo [24]. The findings are also controversial when it comes to pain management. The potent CCK agonist ceruletide is characterized as a robust analgesic [196], whereas the CCK antagonist proglumide ameliorates neuropathic pain [197], potentiates opioid analgesia [198, 199], and inhibits the nocebo effect [8]. In contrast to preclinical predictions, morphine-induced analgesia is not increased by a full CCK₂r antagonist [26]. It follows that the intended therapeutic effect might not be achieved by specifically targeting the brain-specific CCK₂r. It might be more appropriate to take into account additional components of the CCKergic system, such as CCK₁r [200, 201]. The scientific evidence gathered in this review regarding the intricate neuromodulatory and neuroplasticity functions of the CCK neuropeptide supports the idea that certain affective, pain, and cognitive disorders, and even neurological conditions like epilepsy [202, 203] may be associated with dysregulations of the homeostatic CCKergic system, rather than an overactive CCK₂r-mediated neurotransmission.

Novel compounds that have a combination of CCK agonist and antagonist activities could be a good addition to existing mental health drugs. These compounds have the ability to simultaneously affect multiple pathways within the CCKergic system. Combining the two activities may be able to get around the drawbacks of single-target strategies [204] and provide more extensive therapeutic advantages. Given that CCK₁r can control imbalances in dopaminergic neurotransmission [14, 155], a compound with CCK₁r agonism and CCK₂r antagonism may potentially lessen anxiety while influencing other aspects of mood regulation [192]. In a similar vein, these compounds could offer a fresh strategy for managing psychotic symptoms [13–15]. Furthermore, different individuals may exhibit varying degrees of CCK-ergic activity [193, 194] and receptor subtype activity. Compounds with combined agonist/antagonist activity could potentially be tailored to address the individual differences.

Finally, a fascinating but little-known pharmacological feature that merits additional investigation is the discovery of positive allosteric modulators that target the physiologic spatial and temporal engagement of CCK₁r by CCK [205]. The synthesis of a single CCK-based ligand with several off-targets is another intriguing research avenue. For example, bifunctional peptides that act as agonists on δ and μ opioid receptors as agonists and with CCK receptors as antagonists provide a good alternative for the treatment of chronic neuropathic pain [206].

In conclusion, the pharmacological development of such putative CCKergic agents is associated with abnormal dopamine and opioid neurotransmitters like schizophrenia [14, 191], depression [16, 192], and addiction [207].