Amplification of the therapeutic potential of AMPA receptor potentiators from the nootropic era to today

Abstract

α-Amino-3-hydroxy-5-methyl-4-isoxazolepropionic receptors (AMPA receptors or AMPARs) are involved in fast excitatory neurotransmission and as such control multiple important physiological processes. AMPARs also are involved in the dynamics of synaptic plasticity in the nervous system where they impact neuroplastic responses such as long-term facilitation and long-term potentiation that regulate biological functions ranging from breathing to cognition. AMPARs also regulate neurotrophic factors that are strategically involved in neural plastic changes in the nervous system. As with other major ionotropic receptors, modulation of AMPARs can have prominent effects on biological systems that can include marked tolerability issues. AMPAR potentiators (AMPAkines) are positive allosteric modulators of AMPARs which have therapeutic potential. Medicinal chemistry combined with new pharmacological findings have defined AMPAkines with favorable oral bioavailability and pharmacological safety parameters that enable clinical advancement of their therapeutic utility. AMPAkines are being investigated in patients with diverse neurological and psychiatric disorders including spinal cord injury (breathing and bladder function), cognition, attention-deficit-hyperactivity disorder, and major depressive disorder. The present discussion of this class of compounds focuses on their general value as therapeutics through their impact on synaptic plasticity.

1. Introduction

The excitatory amino acid receptors regulate excitatory neurotransmission in the central nervous system (CNS) and have long been hypothesized to be key protein targets for the medicinal control of most neurological and psychiatric disorders. The ionotropic glutamate receptors, N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA), have been the subject of decades of work to identify small molecule modulators with medicinal value (Meldrum, 1992; Rogawski, 2013; Sang et al., 1998; Yamada, 2000). The primary obstacles in the way of generating medicines from this mechanism has been safety and tolerability issues arising from the direct modulation of these powerful regulators of excitatory synaptic transmission. Memantine, a low affinity NMDA receptor antagonist for dementia has been the only major medicine of this kind (since 2003) until the approval of perampanel, an AMPA receptor antagonist, for epilepsy in 2012. The approval of (S)-ketamine as an adjunct for depression in patients in 2019 opened the door to new glutamatergic therapeutics (Witkin et al., 2023). Updated information on prescribing, safety, and labeling of (S)-ketamine (Spravato) is available (FDA, 2025).

We discuss here a class of compounds called AMPA receptor potentiators (PAMs) or AMPAkines. These compounds were originally studied in the 1970s before their mechanism of action was identified. We make two major conclusions in our discussion of AMPAkines: First, they induce changes in the CNS that produce enduring modifications in synaptic function, a general biological process with global therapeutic implications. Secondly, although early issues of tolerability blocked progress in making AMPAkines into medicines, there is renewed interest in bringing the therapeutic potential of this class of compounds to fruition in both neurology and psychiatry, and compounds with human safety are at the ready (Qneibi et al., 2024).

2. AMPA receptors

AMPARs are one of a class of ionotropic glutamate receptors which mediate the majority of fast excitatory neurotransmission in the central nervous system (CNS) (Dingledine et al., 1999). The work that led the discovery of AMPARs dates to 1954 (see historical overview in McLennan, 1983). It was reported from experiments conducted by Takashi Hayashi that L-glutamate applied to exposed motor cortex produced convulsions. The first reports of effects of glutamate and other excitatory amino acids on the activity of neurons came from the labs of Curtis and Watkins (Curtis et al., 1959). Analogs of glutamate and antagonists were discovered and explored around this time frame, and aided the elucidation of glutamate receptor subtypes (see McLennan, 1983).

Additional functions of AMPARs include the recruitment of voltage-gated ion channels and the facilitation of N-methyl-D-aspartate receptor (NMDAR) activation. Coordinated activity of AMPA and NMDARs are the key regulators of synaptic plasticity (Dingledine et al., 1999; Hansen et al., 2021). AMPARs also play a role in managing the availability of neurotrophic factors (Alt et al., 2006). AMPARs are tetrameric ion channels composed of GLUR1–4 subunit and are additionally functionality modified by auxiliary and scaffolding proteins (Hansen et al., 2021; Kato and Witkin, 2018).

Small molecule modulators of AMPARs have been identified and include agonists, antagonists, negative allosteric modulators, and positive allosteric modulators (PAMs) or potentiators (Golubeva et al., 2022). Since AMPARs are directly or indirectly participate in the neural regulation of all neurological and psychiatric processes, the value of AMPAkines as medicines has long been considered. However, it has been difficult to deliver tolerable drugs to patients. In fact, despite multiple efforts, only one medicine is approved for patients that acts on AMPARs - perampanel is an AMPAR antagonist that was approved for use in 2012 for the treatment of epilepsy (Fan et al., 2023; Lavu et al., 2022). Perampanel is also a potential therapy for pain (Fox et al., 2021) and neurodegeneration (Perversi et al., 2023). The antinociceptive effects of AMPAR antagonists is based upon the control of pain pathways by glutamate release (e.g., Yoshimura and Jessell, 1990; see Knopp et al., 2019 for discussion). Comparable mechanistic rationale for the use of perampanel for neurodegenerative disorders based upon glutamate release has been presented (Perversi et al., 2023).

AMPAR potentiators (also known as AMPAkines) have potential for the treatment of a broad array of neurological and psychiatric disorders including cognition, attention-deficit-hyperactivity disorder, and depression. AMPAkines are also being considered for the treatment of pain (Zhu et al., 2024), symptoms of spinal cord injury such as bladder and breathing dysfunction (Witkin et al., 2024), other breathing disorders (Rana et al., 2024) and opioid-induced respiratory dysfunction (OIRD) (Rana et al., 2024; van der Schier et al., 2014).

AMPAkines (Fig. 1) were discovered through the pharmacological profiling of compounds termed nootropics, non-amphetamine compounds with cognitive enhancing effects (Giurgea, 1972, 1973; Giurgea et al., 1971; Sara and Lefevre, 1972). The nootropic, piracetam was investigated in clinical studies as early as 1972 (Stegink, 1972) but it was not until the 1990s that a putative mechanism of action on AMPARs was reported (Ito et al., 1990; Copani et al., 1992; Nicoletti et al., 1992). The work of Copani et al. in cerebellar granular cell cultures, showed that piracetam and analogs enhanced AMPA-stimulated Ca²⁺ influx and that this potentiation was specific to AMPA receptors (not kainate or NMDA). The elucidation of the molecular mechanism of action of these compounds officially began the field of AMPAkine pharmacology. Since that time, the AMPAkines have been shown to impact key aspects of excitatory neurotransmission that are intimately linked to physiological and pathophysiological functions. As such, these data have significant therapeutic ramifications.

Fig. 1.

Structures of some AMPAR potentiators. Compounds in the top row are low impact AMPAkines. However piracetam has not been fully characterized as such due to its low potency. All other compounds are designated as high-impact.

AMPAkines are typically grouped into two general classes based upon their pharmacology: high impact and low impact. High impact ampakines bind the cyclothiazide modulator site and destabilize the desensitized confirmation of the AMPAR (Arai et al., 2002; Nagarajan et al., 2001). This process of prolonging the opening of the ion channel increases the duration of steady-state currents induced by glutamate. Another action of the high impact ampakines is to increase the affinity of glutamate for AMPARs when in the presence of sub-saturating concentrations of glutamate thereby augmenting peak glutamate currents. When glutamate concentrations are at saturation levels, minimal increases in peak current are produced but the compound (e.g., CX614) markedly offsets receptor desensitization (Arai et al., 2000).

Low-impact AMPAkines differ in their pharmacological properties from high-impact compounds. CX516 accelerates channel opening but exerts little effect on receptor desensitization (Arai et al., 2002). This compound does not fully occupy the cyclothiazide binding site of high-impact AMPAkines (Krintel et al., 2013) and does not produce unwanted side effects observed with predecessor AMPAkines. As such, CX516 was characterized as a sub-class of AMPAkines (low-impact).

The tolerability and safety pharmacology of AMPAkines has been investigated in preclinical and clinical studies. High-impact AMPAkines have been associated with convulsions and thus have been less studied than the low-impact potentiators. Some AMPAkines that strongly offset desensitization (Kunugi et al., 2019; Shaffer et al., 2013) are high efficacy drug candidates but produce convulsions at doses slightly above those needed to produce therapeutic benefits. In an effort to create high efficacy AMPAR PAMs that possess a wider therapeutic window, Takeda Pharmaceuticals developed a series of AMPAR PAMs that, unlike some previously explored compounds (Kunugi et al., 2019), lack intrinsic agonistic activity (Kunugi et al., 2019; Suzuki et al., 2019; Tanaka et al., 2019), a pharmacological property that has been associated with convulsions. In contrast to compounds like LY451646, TAK-137 does not have intrinsic agonistic activity and has a wider therapeutic window (Kunugi et al., 2019).

When compounds like LY451395 (mibampator) were developed, it was believed that they were PAMs devoid of agonistic activity. However, Kunugi et al. (2019) demonstrated in a very elegant study that mibampator possesses agonistic activity in addition to its potentiation properties. It also appears that LY451646 has intrinsic agonistic activities, whereas the comparator, TAK-137 does not (Kunugi et al., 2019). Side-by side comparisons under the same conditions and across a broad concentration range would be useful in clarifying the distinct mode of actions of these compounds.

Generalizations regarding the safety of AMPAkines currently rely on the data from individual compounds. Seizures have not been observed with the low-impact class. The prominent low-impact AMPAkines that have been most explored are CX516 (Arai et al., 2002; Kanju et al., 2008; Radin et al., 2018a, 2018b, 2024a, 2024b); CX717 (Gordillo-Salas et al., 2020; Hampson et al., 2009; Lorier et al., 2010; Porrino et al., 2005; Radin et al., 2024a; Ren et al., 2009, 2013; Thakre et al., 2021, 2022; Turner et al., 2016; Zheng et al., 2011); and CX691 (Org24448) (Mozafari et al., 2018; Radin et al., 2018a; Woolley et al., 2009). Clinical studies have yet to establish their maximally tolerated doses. In preclinical studies, CX1739 exhibited therapeutic effects between 0.03 and 18 mg/kg in various rodent studies and doses up to 2000 mg/kg did not induce convulsions (Radin et al., 2024b). CX717 was similarly safe in seizure studies up to 2000 mg/kg while CX1763, the de-esterified form of CX1942, was safe up to 1500 mg/kg (Haw et al., 2016; Purcell et al., 2018).

There is very little published data on the side effects of certain high impact AMPAkines. CX717 (Boyle et al., 2012), CX1739 (Radin et al., 2024b) and TAK-653 (Dijkstra et al., 2022) have produced headache in the clinical setting. LY451395 is another high impact AMPAkine, which from the available clinical data did not produce a notable side effect profile in patients with Alzheimer’s Disease (Trzepacz et al., 2013), though this study was dose limited and also did not meet its primary endpoint.

Preclinically, high impact AMPAkines typically produce convulsions and excitotoxicity at doses just above those that produce robust AMPA-driven electrophysiological effects. For example, high impact AMPAkines CX1837 and CX1846 augment long-term potentiation at 0.3–1.0 mg/kg (Radin et al., 2016) but also produced seizures at 10 mg/kg (unpublished observations). Low impact ampakines such as CX717 and CX1739 enhance LTP at 1–3 mg/kg (Radin et al., 2024b, 2024c), albeit to a lesser extent than observed with CX1837 and CX1846. However, CX717 produced seizures at 5000 mg/kg and CX1739 did not produce seizures at 2000 mg/kg but was lethal at 3000 mg/kg in rats. Thus, there seems to be greater separation of efficacy and toxicity in preclinical studies with low impact AMPAkines than high impact compounds. Additionally, cellular toxicity might be greater; LY451395 decreased neuronal viability by 50% at 3 μM while 500 μM CX717 did not affect neuronal viability in cell culture (Radin et al., 2024c) suggesting an absence of overt neurotoxicity with low impact ampakines like CX717.

3. A universal mechanism of action

AMPARs are directly involved in the processes of long-term potentiation (LTP) and long-term facilitation (LTF), processes that dictate the path for enduring changes in neural function, synaptic plasticity, and prolongation of biological responses (Beck et al., 2000; Nicoll, 2017). First reported in 1966 (Lømo, 1966) and formalized in 1973 (Bliss and Lomo, 1973), LTP continues to be the dominant mechanistic concept underlying synaptic plasticity (Caya-Bissonnette and Béïque, 2024; Bliss et al., 2013; Lømo, 1966, 2018). Since LTP and LTF are mechanisms producing long-term changes in physiological function, they have major implications for the many therapeutic modalities. The idea of improvement in synaptic plasticity for biological therapeutics is not a new concept (Bliss and Collingridge, 1993; Duman et al., 1997; Haracz, 1984; Lenn, 1987; Siegelbaum and Kandel, 1991) but has a become a primary focus in the thinking about mechanisms of disease therapeutics in the past several years (Brown and Gould, 2024; Dejanovic et al., 2024; Matar et al., 2024; Rana et al., 2024; Witkin et al., 2024). AMPAkines very specifically drive these processes and thus have long been hypothesized to have broad therapeutic value. We will illustrate how these general biological processes impact therapeutics using neurological disorders as examples.

Neurorehabilitation strategies rely on treatments of short duration to produce enduring benefits. One neurological process that operates in such a temporal manner is LTP. LTP may be defined as an enduring increase in the strength of synaptic communication lasting hours or days after high-frequency or tetanic stimulation of afferent pathways (Nicoll, 2017). The existence of LTP in human brain tissue was demonstrated on epileptic tissue resected from epilepsy patients who failed front-line medical management. In these studies, tetanizing stimulation produced LTP in hippocampal slices providing concrete evidence that the phenomenon that was observed in rodent brain tissue was also observable in human tissue (Beck et al., 2000).

Since the discovery of LTP, a substantial effort has been undertaken to understand the molecular etiology of this process in a broader effort to understand how synapses communicate. The ultimate hope is that we might ultimately be able to tune these processes to ameliorate deficits in synaptic communication occurring in neurological disorders (Chang et al., 2012; Golubeva et al., 2022; Qneibi et al., 2024; Yamada, 1998). After tetanizing stimulation, the composition of neuronal synapses is drastically altered to induce persistent expression of LTP. For synapses to maintain their augmented strength, AMPAR are inserted into synapses (Araki et al., 2015). AMPAR membrane insertion has the potential to convert ‘silent’ synapses into active synapses for the purpose of LTP consolidation (Morita et al., 2014). Additionally, the individual AMPARs that are inserted into synapses have higher unitary conductance (Park et al., 2021). These mechanistic findings closely correlate with the observation that mice lacking the Glur1 AMPAR subunit exhibit severely diminished LTP in the CA1 region of the hippocampus (Zamanillo et al., 1999). Other data indicate that AMPAR endocytosis underlies LTP dysfunction (Dong et al., 2015) and that AMPAR reductions are observed in models of Alzheimer’s disease (Chang et al., 2006).

The importance of AMPAR trafficking and conductance in LTP maintenance raises the possibility that positive AMPAR modulation with AMPAkines could rescue deficits in LTP associated with age and in disease states. We have previously shown that aged rats exhibit a decrease in LTP and that a single treatment with the high impact ampakine CX1846 reversed the LTP deficits (Radin et al., 2016). The low impact AMPAkine CX1739 dose-dependently enhanced LTP in young, healthy rats (Radin et al., 2024b) (Fig. 2).

Fig. 2.

Effect of CX1739 on LTP in rats in vivo. (A) Time course of the increase in amplitude of field EPSPs in dentate gyrus following stimulation of the perforant path in rat hippocampus. (B) Mean increases in EPSP amplitude measured over a 10 min period, starting 10 min after post-tetanic stimulation. Data points represent the mean ± SEM (N = 7, vehicle; N = 3, 1 mg/kg; N = 4, 3 mg/kg).

*p < 0.05, ***p < 0.0001.

Figure from Radin et al. (2024b) with permission.

Huntington’s Disease (HD) is an autosomal dominant neurodegenerative disorder involving trinucleotide expansions of CAG repeats in the Huntington gene (Ho et al., 2003). Prior to the onset of motor impairments, memory dysfunction and cognitive impairments can be observed (Ho et al., 2003). In a mouse model of HD, LTP was diminished and treatment with the high impact ampakine CX929 for 4 days was able to reverse the deficits in LTP and corrected deficits in the novel object recognition test, an assessment of long-term memory (Simmons et al., 2009). Improvement has also been reported in a model of Angelman Syndrome (AS). AS is a neurodevelopmental disorder arising from anomalous expression of the maternally inherited UBE3A gene. These patients exhibit profound cognitive and speech impairments and seizures (Oliver et al., 2007). In AS mice, LTP was severely reduced; however, treatment with CX929 for 5 days completely reversed LTP deficits and long-term memory deficits (Baudry et al., 2012). The results of these studies indicate that neurodegenerative/neurodevelopmental conditions underpinned by deficits in LTP may be amenable to treatment with AMPAkines.

From amyotrophic lateral sclerosis (ALS) (Larson et al., 2023) to Parkinson’s disease (Kaczynska et al., 2022) to spinal cord injury (Rana et al., 2024; Witkin et al., 2024), patients suffering from an array of neurological issues can ultimately undergo respiratory compromise. In the case of ALS, ~25 % of deaths are attributed to respiratory failure (Larson et al., 2023). There is an urgent clinical need to address respiratory dysfunction in these patients. A respiratory neurorehabilitation strategy should comprise interventions that can have long-lasting benefits. For many patients with neuromuscular dysfunction, targeted treatments to increase diaphragmatic function could assist with lung mechanics to alleviate hypoxia and hypercarbia. As the diaphragm is innervated by the phrenic nerve, a multitude of studies have been conducted to determine how phrenic nerve output can be boosted to assist patients in maintaining physiological respiratory rhythms. One of the methods under clinical investigation is to use acute intermittent hypoxia (AIH) to evoke phrenic long-term facilitation (LTF) (Sajjadi et al., 2022; Sutor et al., 2021). LTF is conceptually similar to LTP, as it is defined as a persistent increase in neural output after an acute stimulus. In this case, the stimulus is the exposure to a few brief episodes of low oxygen, or AIH (Devinney et al., 2013). Rodents subjected to acute bouts of AIH exhibit sustained increases in phrenic nerve output with associated increased diaphragm activity (Baker and Mitchell, 2000; Gonzalez-Rothi et al., 1985).

Modulating LTF can likely also be applied to other singular and combinatorial respiratory rehabilitative strategies. As AMPAR signaling strongly controls respiratory neurocircuitry (Rana et al., 2024; Ren et al., 2006) a study conducted in 2016 evaluated the potential interaction between positive AMPAR modulation and AIH. In this work, Turner et al. (2016) showed that the LTF of the respiratory output to the upper airway muscles was greatly enhanced when the AMPAkine CX717 was delivered prior to AIH. A subsequent study showed that AMPAkine pretreatment could enable even a single brief exposure to hypoxia to produce LTF of respiratory neural drive to the diaphragm (Wollman et al., 2020). This response occurred when AMPAR modulation occurred immediately prior the AIH (Thakre et al., 2021). The implication from this body of work is that AMPAkine pretreatment could have value in the context of hypoxia-based neurorehabilitation strategies. There has been a considerable amount of work in recent years focusing on AIH in the context of spinal cord injury rehabilitation. Collectively the data from multiple clinical trials has shown that exposure to mild AIH is safe, well-tolerated, and can enhance efficacy of conventional neurologic rehabilitation paradigms (Gonzalez-Rothi et al., 1985). However, not all patients respond to AIH therapy with improvements in motor recovery. Rodent studies have established that the impact of AIH is enhanced by AMPAkine pretreatment (Rana et al., 2024), and the translation of this finding to clinical use could prove of value.

As outlined above, the prominent role of AMPARs in CNS function and their dominant role in synaptic plasticity mechanisms of LTP and LTF, place AMPAkines in a special category of potential global CNS therapeutic agents. Their therapeutic value in many neurological and psychiatric diseases has already been suggested. Here we will use three areas of CNS function, cognition, ADHD, and mood, to illustrate the beneficial use of AMPAkines.

4. Cognition

AMPARs have a central role in excitatory synaptic transmission in the CNS underlying the processes of plasticity and memory with their dysregulation implicated in the mechanisms causing cognitive impairment and psychiatric disorders. As such AMPARs present an attractive target for novel therapies. While direct acting AMPAR agonists suffer from their narrow safety margin, substantial drug development efforts have been devoted to the development of AMPAkines (Kadriu et al., 2021).

The initial work in cognition with these compounds was devoted to the racetam series of nonselective AMPAkines (Copani et al., 1992), a class of drugs that share a pyrrolidone nucleus. Some of the racetams have nootropic properties (piracetam, aniracetam, oxiracetam, pramiracetam and phenylpiracetam) (Gouliaev and Senning, 1994) with piracetam, aniracetam and pramiracetam gaining regulatory approval in some European countries as cognitive enhancers. While the efficacy of nootropic racetams has been questioned (Herrmann and Stephan, 1992; Flicker and Grimley Evans, 2001), extensive meta-analysis reported global efficacy of piracetam in a diverse group of older subjects with cognitive impairment (Waegemans et al., 2002). In the United States, racetams are not approved by the Food and Drug Administration (FDA) but appear as ingredients of cognitive enhancement dietary supplements (Cohen et al., 2020). The FDA has issued a warning that such cognitive enhancement supplements “may be ineffective, unsafe, and could prevent a person from seeking an appropriate diagnosis and treatment” (Commissioner, 2020).

The development of racetams was followed by the development of AMPAkines, selective AMAPAR-PAMs, typically belonging to the benzamide structural class (Arai and Kessler, 2007). Several such ampakines exhibited promising preclinical profiles and entered clinical development (Brogi et al., 2019; Kadriu et al., 2021). The AMPAkines with nootropic properties which entered clinical development are summarized in Table 1.

Table 1.

AMPAkines with cognitive enhancing effects that have entered clinical development.

Compound	Clinical endpoint assessed
CX516 (Ampalex)	MCI, AD, schizophrenia, Fragile X, autism, ADHD
CX691 (Farampator, ORG2448)	Cognitive enhancer, MDD
CX717	Cognitive enhancer, AD, ADHD
CX1739	Cognitive enhancer, ADHD
ORG26576	ADHD, MDD
S47445 (CX1632, Tulrampator)	MCI, AD, MDD
S18986	MCI
LY451395 (Mibampator)	Cognitive enhancer, AD
BIIB-104 (Pesampator, PF-04958242)	Cognitive enhancer, schizophrenia
TAK-137	Cognitive enhancer, ADHD
GSK729327	Cognitive enhancer
NBI-1065845 (Osavampator, TAK-653)	MDD

AD: Alzheimer’s disease; ADHD: attention deficit hyperactivity disorder; MCI: mild cognitive impairment; MDD – major depressive disorder.

CX516 was effective in improving measures of visual, olfactory, and visuospatial memory in healthy volunteers (Ingvar et al., 1997) and delayed recall of nonsense syllables in young and elderly subjects (Lynch et al., 1996, 1997). CX516 was not effective for cognition or for symptoms of schizophrenia when added to clozapine, olanzapine, or risperidone (Goff et al., 2008). CX516 was also ineffective in clinical trials for mild cognitive impairment, autism, AD, ADHD, and Fragile-X, likely due to its low potency and short half-life (Danysz, 2002; Berry-Kravis et al., 2006).

CX691 (ORG-24448, farampator) produced improvement in the short-term memory of healthy elderly volunteers, but appeared to impair episodic memory (Wezenberg et al., 2007). A clinical trial for the treatment of cognitive impairments in schizophrenia was withdrawn (NCT00425815) and a clinical trial in patients with major depression (NCT00113022) failed to progress.

In preclinical models, CX717 enhanced cognitive measures (Hampson et al., 1998) and high doses of CX717 counteracted the effects of sleep deprivation on attention-based tasks in healthy human volunteers (Boyle et al., 2012). However, due to its interference with recovery sleep, the compound was not progressed. CX717 was also investigated for use in Alzheimer’s disease (AD), but the trial was halted, and the data were not disclosed. More recently, CX717 entered clinical development for the treatment of ADHD (NCT03375021) but no results were publicly disclosed (but see Radin et al., 2025).

CX1739 exhibited therapeutic effects in multiple preclinical assays that include cognitive enhancement, attention-deficit hyperactivity disorder (ADHD), and reversal of opiate-induced respiratory depression. In a Phase-1 clinical study, CX1739 displayed good tolerability, safety and pharmacokinetics and is progressing into Phase 2 clinical studies (Radin et al., 2024b).

ORG-26576 produced improvements in executive functioning and speed of cognitive processing in patients with major depressive disorder (MDD) (Nations et al., 2012) and produced some encouraging results in a clinical trial for the treatment of ADHD (Adler et al., 2012). Further development of ORG-26576 was discontinued most likely due to weak efficacy and the occurrence of adverse effects.

S47445 was acquired from Cortex Pharmaceuticals by Servier and displayed procognitive effects in preclinical studies (Morley-Fletcher et al., 2018) but failed to improve cognition in patients with mild to moderate AD with depressive symptoms (Bernard et al., 2019). The large Phase-2 clinical trial also found that the treatment failed to improve overall function (Disability Assessment for Dementia) or to reduce depressive symptoms. S47445 was also evaluated as a as adjunctive treatment of MDD in patients with an inadequate response to antidepressant therapy (NCT02805439) but failed to achieve superiority over placebo (Servier-47445_Synopsis, 2018). S47445 was also evaluated in a Phase-1 clinical trial for mild cognitive impairment (RespireRx/Servier, 2020) but no results were disclosed.

S18986 entered clinical development in patients with mild cognitive impairment (NCT00202540). The objective of this Phase-2 study was to demonstrate long-term efficacy of S18986 versus placebo on episodic memory. This study was prematurely stopped (38 patients included, 450 planned) due to the emergence of severe adverse events (Servier-18986_Synopsis, 2007).

LY451395 (mibampator) is a biarylpropylsulfonamide AMPAkine that has been clinically evaluated in patients with AD. In a Phase-2 clinical trial in patients with mild-to-moderate AD dementia, LY451395 produced some marginal efficacy on the neuropsychiatric inventory secondary measure (NPI) but failed to differentiate from placebo on the primary outcome measure (ADAS-Cog) (Chappell et al., 2007). LY451395 was further evaluated in AD disease patients for treatment of agitation and aggression derived from the neuropsychiatric inventory (NPI-4-A/A), but achieved no separation from placebo on the primary endpoint despite higher doses than used in the initial study (Trzepacz et al., 2013). Other measures of the NPI-4-A/A, including apathy, were also not affected (Trzepacz et al., 2013; Ruthirakuhan et al., 2018) and no further clinical development of this compound has been reported.

BIIB-104 (PF-04958242, pesampator) entered a Phase-1 trial to explore its on brain circuitry associated with working memory in healthy participants using BOLD Functional MRI and Arterial Spin Labeling (NCT04068532). BIIB-104 was also progressed in development to target schizophrenia-associated cognitive impairment (NCT03745820). A Phase-2 study enrolled 195 subjects where primary endpoints (working memory and cognitive battery) and secondary endpoints (cognition, functioning, and psychiatric symptomology) were investigated. The results of the study, however, failed to reach the primary and secondary endpoints and Biogen discontinued the program for this indication (BIIB-104, 2022).

TAK-137 is an AMPAkine with efficacy in animal models of cognitive impairment and schizophrenia (Kunugi et al., 2019; Tanaka et al., 2019). TAK-137 entered Phase-1 clinical studies to characterize its safety and tolerability when administered as multiple oral doses in healthy volunteers and in adults with ADHD (NCT02334982, NCT02163915). No clinical trial data were released, and it is not clear if the compound was progressed in development.

GSK729327 was investigated in a Phase 1 study in healthy volunteers (NCT00448890). A single-blind, randomized, placebo-controlled two-part study was designed to evaluate safety, tolerability, pharmacokinetics (PK) and pharmacodynamics (PD) of single oral escalating doses and repeat doses of GSK729327. PD analysis involved assessment of multiple cognitive endpoints, but no analysis was performed as the study was stopped due to pathological preclinical findings (GSK729327 - GSK Study Register, 2019).

Several novel AMPAkines, including TAK-653 and MDI-222, are showing promising data in preclinical models of cognition (Ward et al., 2020; Suzuki et al., 2021), and TAK-653 (NBI-1065845) has progressed into clinical development for MDD where it met primary and secondary efficacy endpoints and was in general well tolerated (Neurocrine, 2024).

Recent work primarily in the area of antidepressants, has generated data on the neuroplasticity and neural remodeling properties of other, non-AMPAkine compounds. This includes work on NMDA receptor antagonists and psychedelic drugs (Brown and Gould, 2024; Weiss et al., 2025; Zhang et al., 2024).

5. Attention deficit/hyperactivity disorder (ADHD)

ADHD is a commonly diagnosed psychiatric disorder highlighted by hyperactivity, poor impulse control, and inattentiveness (Rappley, 2005). ADHD is typically diagnosed in childhood, yet patients have been shown to exhibit symptoms well into adulthood (Biederman and Faraone, 2005) with the prevalence of adult ADHD being 4–5 % in the general population (Franke et al., 2012). Patients with ADHD exhibit poorer functioning in social, educational and occupational domains (Kessler et al., 2006) and have higher rates of divorce, lower school performance, and higher unemployment (Biederman et al., 2000). ADHD patients also present with ancillary psychological/psychiatric issues such as substance abuse and depression/anxiety (Adler and Chua, 2002).

Medical treatment typically involves the use of stimulants such as amphetamines and methylphenidate and/or non-stimulants like atomoxetine, a norepinephrine reuptake inhibitor. More than 60 % of patients who take prescribed stimulants show improvements in ADHD symptoms (Faraone et al., 2004). However, stimulants produce cardiovascular side effects and patients can become tolerant, dependent and may misuse medications (Hennissen et al., 2017; Kooij et al., 2013; Wojnowski, 2006). Atomoxetine has demonstrated efficacy, though the latency to a durable effect is longer and the effect size is lower than that of stimulants (Michelson et al., 2003) As a result, there exists a significant interest in translating quick acting non-stimulant compounds into ADHD medicines.

In addition to pharmacological interventions to increase the synaptic availability of catecholamines, there is a growing body of evidence to suggest a hypo-glutamatergic etiology of ADHD. In rodent models of ADHD, it has been shown that increasing AMPAR function in the prefrontal cortex normalized ADHD-like behaviors in rats (Cheng et al., 2017). Additionally, infusion of an AMPAR antagonist ablates the therapeutic effects of methylphenidate (Zhang et al., 2023). These results are consistent with findings that rats lacking the transmembrane AMPAR regulatory protein 8 (Tarp-γ8), an auxiliary protein required for proper AMPAR functioning, exhibit hyperactivity, impulsivity and memory deficits (Bai et al., 2022). Associations between certain Tarp-γ8 single nucleotide polymorphisms and ADHD susceptibility has also been observed (Bai et al., 2022). TARP proteins have been shown to modulate the pharmacology of AMPAkines (Radin et al., 2018b). In a magnetic resonance spectroscopy study, glutamate and glutamine levels were decreased in the basal ganglia of patients with ADHD. In those with untreated ADHD, decreased levels of glutamate and glutamine in the basal ganglia were significantly associated with more severe symptoms of inattention (Maltezos et al., 2014).

Given evidence that AMPAR enhancement may ameliorate symptoms of ADHD, several groups have examined whether AMPAR up modulation by AMPAkines might moderate symptoms associated with ADHD. Preclinical studies to screen for compounds that might treat ADHD typically utilize the 5-choice serial reaction time task (5CSRTT) which detects compounds that reduce impulsivity (Callahan and Terry Jr., 2015; Levin et al., 2011; Sanchez-Roige et al., 2012). Bilateral vestibular deafferentation produced cognitive deficits in rats, which translated to deteriorated performance in the 5CSRTT. This reduced performance was partially ameliorated by 20 mg/kg CX717. CX717 also reduced the premature responses in rats, which is a marker of reduced impulsivity (Zheng et al., 2011). Our group found that CX1739 also enhanced performance in the 5CSRTT. Using healthy rats, 3 and 10 mg/kg CX1739 increased the percentage of correct responses and reduced the number of premature responses as with the approved ADHD drug atomoxetine (Fig. 3). CX1739 also was able to decrease hyperactivity in rats (Fig. 4). These preclinical studies support the potential use of low impact AMPAkines in the treatment of ADHD. Additional preclinical studies using the AMPAR modulator TAK-137 found an increase in the number of correct responses in this paradigm, though the number of premature responses was not significantly altered. However, only dose was used in these studies (Tanaka et al., 2019).

Fig. 3.

Effects of CX1739 (left panels) and atomoxetine (right panels) on performance measures (percentage of correct responses, latency to response, percent of premature responses, and percent of response omissions) in the 5-choice serial reaction time test in rats. Each point represents the mean ± SEM of 16 rats. *p < 0.05, **p < 0.01 compared to vehicle values.

Figure from Radin et al. (2024b) with permission.

Fig. 4.

CX1739 dose-dependently reduced the increases in locomotor activity induced by amphetamine in mice. Each data point represents the mean ± SEM (N = 12, saline; N = 16, amphetamine; N = 8, 5.6 mg/kg; N = 16, 10, 18, and 30 mg/kg). *p < 0.05, **p < 0.01, ****p < 0.0001, compared to amphetamine alone. Saline—12, amphetamine—16; 8 animals—5.6 mg/kg, 16 animals—10, 18, and 30 mg/kg each.

Figure from Radin et al. (2024b) with permission.

To our knowledge, there have been only two clinical studies conducted evaluating the effects of AMPAkines on ADHD symptoms in adults. A Phase 2 study using the low impact AMPAkine Org26576 (CX1351) used 100 mg BID and 100–300 mg BID (Adler et al., 2012). In this study, 100 mg BID but not 100–300 mg BID was superior to placebo in reducing symptoms of ADHD in adults. Change from baseline using the ADHD investigator symptom rating scale (AISRS) was used to quantify effects of treatment. For patients treated with Org26576 at 100 mg BID, there was a significant decrease in hyperactivity/impulsivity but not inattentiveness (Adler et al., 2012). The number of patients with at least a 50 % reduction in AISRS score was 5/28 in the 100 mg BID group but 0/27 in the placebo group. For patients taking Org26576 100–300 mg BID, inattentiveness increased compared to those taking placebo and hyperactivity was not decreased. 56 % of patients taking placebo reported at least one adverse event while 71 % of patients taking 100 mg BID and 85 % of patients taking 100–300 mg BID reported at least one adverse event. Those taking Org26576 most commonly reported dizziness, headache and nausea. While Org26576 is no longer being clinically developed, these data highlight that enhancing AMPAR function might reduce some symptoms associated with ADHD.

Another clinical study examined the effects of CX717 in treating symptoms of ADHD. In this double-blinded, randomized crossover study, patients were given placebo or 200 or 800 mg CX717 BID (Adler et al., 2006). Each treatment period lasted three weeks with a two-week washout between treatments. While 200 mg BID was not superior to placebo, 800 mg BID was superior to placebo on the total ADHD rating scale. Furthermore, 800 mg BID reduced inattentiveness and hyperactivity in patients. CX717 was well tolerated at both doses and did not affect cardiovascular parameters. Sleep disturbance and headache were the most frequently reported adverse events. Given the safety and dose-response nature of CX717’s clinical effects, these findings suggest that CX717 and other low impact ampakines should be further investigated to treat ADHD. It is also possible that low impact AMPAkines could be paired with approved medications to boost therapeutic effects and decreasing unwanted side effects.

6. Major depressive disorder

The AMPA receptor hypothesis posits that AMPA receptor potentiation is a primary gateway to the therapeutic effects of antidepressants (Alt et al., 2006; Skolnick, 1999; Skolnick et al., 2001). The AMPA hypothesis is based upon the findings that classical antidepressants signal through Ca²⁺/cAMP response element binding protein (CREB) involved in the induction of brain-derived neurotrophic factor (BDNF). BDNF through its neuroprotective, neuronal sprouting, and neurogenesis effects, is able to relieve damaged stress/depression circuits. This biochemical pathway triggered by conventional antidepressants can also indirectly potentiate AMPARs through the phosphorylation of dopamine- and cAMP-regulated phosphoprotein of Mr. 32,000 (DARP-32). AMPAR potentiation can then induce BDNF through Ca²⁺/calmodulin-dependent protein kinase (CaMKinase) or through mitogen-activated protein kinase (MAPK). The direct route through AMPAR signaling enables antidepressant effects to be generated by non-classical antidepressant compounds such as ketamine and other NMDA receptor antagonists and mGlu_2/3 receptor antagonists (Chaki, 2017; Witkin, 2020).

The AMPAR hypothesis was also built, in part, on the findings that AMPAkines produce antidepressant effects (Alt et al., 2006; Nisenbaum and Witkin, 2010; Skolnick et al., 2001). Importantly, the compounds that most directly potentiate AMPARs are also unique in their pharmacological profiles. In contrast to monoamine-based antidepressants, these compounds produce rapid-acting antidepressant effects and are effective in treatment resistant patients (Witkin et al., 2019). The benefit of the procognitive effects of AMPAkines could be an additional boost to patients with MDD where cognitive dysfunction is a common comorbid disease symptom (Matar et al., 2024). Importantly, as discussed earlier, AMPARs are direct links to the mechanisms supporting synaptic plasticity, a phenomenon of likely paramount importance to antidepressant efficacy (Alt et al., 2006; Brown and Gould, 2024; Duman et al., 1997; Matar et al., 2024).

To date, two novel antidepressants that indirectly potentiate AMPARs have been approved for patients, (S)-ketamine and dextromethorphan/bupropion (Witkin et al., 2023) – these compounds potentiate AMPARs through their relief of GABA inhibition by NMDA receptor antagonism. The hope continues that mGlu_2/3 receptor antagonists will eventually find a way into the hands of patients – these compounds activate AMPARs through their release of glutamate. Although there are currently no AMPAkines approved for depressed patients, there are continued efforts to address this gap. An earlier study showed promise where both mood and executive function were improved in patients with the AMPAkine Org 26,576 (Nations et al., 2012). There were also a number of failures with other AMPAkines to achieve efficacy (see Cognition section above). The types of AMPAkines and the clinical protocols for using them to treat MDD has yet to be determined. As with most novel targets for diseases, multiple, concentrated and iterative clinical studies are generally needed to fully evaluate their potential value and their liabilities (Bespalov et al., 2016).

Additional work with AMPAkines in depression is ongoing (Witkin and Lippa, 2024). In healthy volunteers, the high-impact AMPAkine, TAK-653 (NBI-1065845), was tolerable without serious adverse events and importantly produced cortical activation biomarkers (O’Donnell et al., 2021) associated with antidepressant activity (Dijkstra et al., 2022). The availability of AMPAkines that have demonstrated safety in humans, such as TAK-653, CX717 and CX1739, add additional tools to interrogate the clinical potential of this class of compounds in patients with MDD. Recently reported Phase 2 data with TAK-653 showed efficacy in patients (not adequately treated with standard of care antidepressants) with once daily dosing in a double-blind placebo-controlled study (Neurocrine, 2024). These effects were not dose-dependent making the path to optimization a bit more difficult. Statistically significant effects were observed with one dose at both day 28 and 56 and a trend was observed with the other dose using the Montgomery Åsberg Depression Rating Scale (MADRS). Improvement compared to placebo was −4.3 (p = 0.016) and −7.5 (p = 0.002) at Day 28 and Day 56, respectively. TAK-653 was well tolerated, with no major side effects and minor side effects were comparable to placebo patients (Neurocrine, 2024).

Additional innovative therapeutics arise from data on the structure of AMPAR complexes. The discovery of auxiliary proteins, such as the transmembrane AMPA receptor regulatory proteins (TARPS) might be one such advance in this direction. TARPS are regionally localized in the brain and their targeting by selective AMPAkines could initiate changes in excitatory neurotransmission linked to depression (Alt et al., 2006; Kato and Witkin, 2018). This was suggested in 2006 with the idea of creating a TARP γ−8-selective AMPAkine that could activate hippocampal AMPARs known to be involved in depression (Alt et al., 2006). Impacting necessary neuronal circuits without affecting ancillary brain regions could also provide more tolerable medicines. Tangential proof of principal for this idea comes from data on a TARP γ−8-selective AMPAR antagonist. LY3130481 (CERC-611 or ES-481) blocks forebrain-associated AMPARs without significant antagonism of AMPARs in the cerebellum associated with motor dysfunction and dizziness (Kato et al., 2016). This compound shows remarkably less motoric side effects than non-TARP-selective AMPAR antagonists and, at the same time, produces comparable anticonvulsant efficacy (Kato et al., 2016). A recent clinical trial in drug-resistant epileptic patients provided clinical confirmation of this efficacy/safety proposition (O’Brien et al., 2024).

7. Conclusions

AMPARs play extensive and dominating roles in disorders of the CNS. Their involvement in overall CNS excitation, LTP and LTF, and the regulation of neurotrophins are among the basic biological processes by which positive modulation of AMPARs have an important place in therapeutics. Current work is progressing to establish AMPAkines as new medicines for multiple neurological and psychiatric disorders. The current availability of clinically tolerable compounds and the strong biological rationale for their investigation in disease states suggests the emergence of a new class of medicines.

Data availability

No data was used for the research described in the article.