Preclinical characterization of a water-soluble low-impact ampakine prodrug, CX1942 and its active moiety, CX1763

Abstract

Aim: AMPA-glutamate receptor (AMPAR) dysfunction mediates multiple neurological/neuropsychiatric disorders. Ampakines bind AMPARs and allosterically enhance glutamate-elicited currents. This report describes the activity of the water-soluble ampakine CX1942 prodrug and the active moiety CX1763.

Results: CX1763 and CX1942 enhance synaptic transmission in hippocampi of rats. CX1763 increases attention in the 5CSRTT in rats and reduces amphetamine-induced hyperactivity in mice. CX1942 potently reverses opioid-induced respiratory depression in rats. CX1942/CX1763 was effective at 2.5–10 mg/kg. CX1763 lacked epileptogenicity up to 1500 mg/kg in rats.

Conclusion: These data document that CX1942 and CX1763 are active and without prominent side effects in multiple pre-clinical assays. CX1942 could serve as a prodrug for CX1763 with the advantage of high water solubility as in an intravenous formulation.

Plain language summary

Article highlights.

• The water-soluble glycine ester pro-drug of CX1763, CX1942 is potent and active against alfentanil-induced respiratory depression in vivo

• CX1942 and CX1763 produce a durable increase in hippocampal EPSP in vivo

• CX1763 increases metrics of attention in rodents while reducing hyperactivity associated with amphetamine, indicating multiple potential benefits for ADHD treatment

• CX1763/CX1942 is therapeutically active at doses of 2.5–10 mg/kg and CX1763 lacks epileptic and fatal toxicities up to 1500 mg/kg, demonstrating a notable safety margin, justifying further preclinical and clinical explorations

1. Introduction

AMPA-glutamate receptors (AMPAR) govern a large majority of excitatory synaptic transmission in the mammalian central nervous system( CNS) [1]. The ability to safely augment AMPAR conductance would therefore be useful in several psychiatric and neurological disease states in which excitatory signaling is compromised [2,3]. AMPAR positive allosteric modulators, AMPAkines, are a class of small molecule compounds that bind to the AMPAR and allosterically enhance the agonistic effects of the endogenous ligand glutamate [4–7]. In preclinical studies, AMPAkines have shown efficacy in several neurodegenerative and neurodevelopmental disorders including Parkinson’s, Huntington’s, Angelman and Rett Syndrome [8–12]. Clinical explorations have yielded positive data showing that AMPAkines reduces symptoms of ADHD [13] and diminish opiate-induced respiratory depression (OIRD) without compromising the analgesic effects of opiates [14].

The promising pre-clinical results of AMPAkines were offset by the propensity of certain AMPAkines to produce seizures at therapeutic doses [15], drastically halting their clinical translation. Promising results were published by Arai et al., however, regarding two ampakines (CX546 and CX516, Figure 1) that exhibited significantly different electrophysiological properties even though they had very similar chemical structures [6]. CX546 interfered with desensitization producing steady-state AMPAR currents in neurons by 200%, while CX516 produced a maximal increase in steady-state currents of only 25%. Both CX516 and CX546 enhanced deactivation response to brief 1-ms glutamate pulses. However, CX516 enhanced current amplitudes to a greater extent than did CX546 [6]. This discovery led to AMPAkines being grouped into high- and low-impact classifications, based upon their pharmacological properties. High-impact AMPAkines like cyclothiazide were shown to be seizurogenic, probably due to their block of desensitization and thus were not clinically developed [16–18]. When tested in humans, early low impact AMPAkines, such as CX516, produced positive clinical findings without exhibiting the unwanted side effects of the high impact AMPAkines. Unfortunately, the low potency and poor pharmacokinetic properties of CX516 [19] caused the termination of its clinical development.

Figure 1.

Ampakine chemical structures. Structure of the previously characterized low and high impact ampakines CX516 and CX546, respectively. Chemical structures of N-(4-trans-Hydroxycyclohexyl)-N-methyl-[2,1,3]-benzoxadiazole-5-carboxamide (CX1763) and trans-4-[(2,1,3-benzoxadiazol-5-ylcarbonyl)(methyl)amino]cyclohexyl glycinate hydrochloride (CX1942).

Over the past several years RespireRx Pharmaceuticals has developed later generation, low-impact AMPAkines to treat OIRD [14,20–22]. Low-impact AMPAkines augment AMPAergic tone in the pre-Botzinger complex, a pool of neurons in the brainstem that controls inspiration in mammals [14,21,23,24]. Multiple low-impact ampakines such as CX717 and CX1739 have been shown to reverse OIRD in pre-clinical models [20–25] and in the clinical setting [14,26] without interfering with the analgesic properties of the opiates. Not only do these studies demonstrate target site engagement at AMPA receptors with the doses used, but also encourage expansion into a number of different clinical indications. As further work is done to translate low-impact AMPAkines into the clinical setting, it would be of considerable value to bring to develop a low-impact AMPAkine with improved water solubility. For example, an oral liquid formulation could be developed for pediatric indications where patients may have problems swallowing pills (ADHD). A soluble, but not necessarily liquid formulation, might also be developed to enhance absorption and design controlled release formulations.

In the present work, we describe the preclinical pharmacology of a novel low-impact AMPAkine, CX1942 [27]. CX1942 is a water-soluble glycine ester pro-drug of CX1763, an active low-impact AMPAkine. In this paper we report the ability of CX1942 and CX1763 to enhance excitatory postsynaptic potentials (EPSP) in the hippocampi of rats. As AMPAkines have been shown to improve attention in preclinical models [22,28] and in humans [13], we have examined the ability of this compound to improve measures in the 5-Choice Serial Reaction Time Task (5CSRTT), an assay used to screen for compounds for the potential ability to treat ADHD. We also have studied the interaction of CX1763 with amphetamine, a drug used currently in the clinic to treat ADHD. The ability of CX1942 to reverse OIRD was investigated as well. Finally, we assessed the epileptogenic capacity of high-doses of CX1763, in an effort to gauge the compound's therapeutic ratio.

2. Materials & methods

Animal studies were carried out in compliance with the NIH Guide for the Care and Use of Laboratory Animals and using protocols approved by the Institutional Animal Care and Use Committee of University California, Irvine (Irvine, CA, USA) (study approval number #61–05–05B). The 5CSRTT studies were done in accordance with all local and national ethical requirements in compliance with the Animals (Experimental Procedures) Act, 1986 under license from the UK Home Office. We made efforts to reduce animal suffering and the number of animals used was reduced whenever possible. We selected animal strains and species based upon previous characterizations in the studies described below.

2.1. Ampakine preparation & solubilization

CX1763 (N-(4-trans-Hydroxycyclohexyl)-N-methyl-[1-3]-benzoxadiazole-5-carboxamide) and CX1942 (trans-4-[(2,1,3-benzoxadiazol-5-ylcarbonyl)(methyl)amino]cyclohexyl glycinate Hydrochloride) (Figure 1) were synthesized by Norchim (France). NMR of CX1763 and CX1942 are supplied in Supplementary Figure S1. CX1763 was 99.56% pure while CX1942 was 99.3% pure by High Performance Liquid Chromatography (HPLC). CX1942 was prepared for in vivo studies by solubilization in 0.9% saline while CX1763 was prepared for in vivo studies by solubilization in 0.9% saline with the addition of 10% hydroxypropyl-P-cyclodextrin. Both drug solutions were further prepared with sonication.

2.2. Stability measurements of CX1763 glycine ester pro-drug CX1942 in rat, dog & human whole blood

CX1942, the glycine ester of CX1763, was evaluated for its susceptibility to hydrolysis in whole blood from various species. CX1942 was added to whole blood and incubated at 37°C, room temperature and on ice. Aliquots of incubation samples were taken over a 2–60 min time course to determine the rate of loss of CX1942 and the production of CX1763. Conditions for stabilization of CX1942 in whole blood were established such that PK and TK studies could commence.

CX1942 powder was dissolved at 1 mM in injectable saline and was added to whole blood for incubation at 37°C, room temperature and on ice. Aliquots of incubation samples were taken over a 15–60 min time course to evaluate the loss of CX1942 and production of CX1763 by LC-MS and spiked into whole blood to produce a final concentration of 25 μM in 0.5 ml aliquots of whole blood. Bioreclamation Inc. (Westbury, NY, USA) was the source of human and dog whole blood anti-coagulated with heparin or K3EDTA and shipped at 4°C. Human blood was 48 h old upon receipt and beagle dog blood was 24 h old upon receipt. Rat blood was procured in house and used fresh (1–3 h old). Aliquots of blood (50 μl) were removed at various time points, converted to plasma and extracted to monitor the progress of CX1942 hydrolysis.

Short column LC–MS based methods were developed to separate the glycine ester pro-drug CX1942 from its ester hydrolysis product CX1763. The methodology developed employs two different HPLC C18 columns and two gradient systems with the same mobile phase. Detection was by both UV and MS. Conditions of analysis utilized acid washed glassware and or polypropylene tubes and HPLC vials to minimize ester hydrolysis during the analysis. Minimal contact hydrolysis of CX1942 occurs when using storage tubes and HPLC vials made of polypropylene. Both the Phenomenex Gemini C18 column and the Thermo BetaBasic C18 columns give acceptable separation of CX1763 from CX1942 on the order of ≥2 min within a total 10 min run time. The conditions of analysis are described below.

Samples of CX1942 were made up in nanopure water. CX1942 standard was made up at 2.5 mM and stored in 2 ml acid washed HPLC vials at 4°C. CX1763 standard was made up at 2.5 mM in 50% acetonitrile water and stored in glass. From the 2.5 mM working solution 10 uM CX1942 or CX1763 samples were made by dilution with water and stored in acid washed HPLC vials at 4°C or placed in an autosampler held at 10°C for LC–MS analysis.

A MicroMass Quattro II mass spectrometer operated in the electrospray mode was used as the most sensitive detector for both CX1942 and CX1763. UV detection was made with an in line Waters Photodiode Array detector. To evaluate the separation of CX1942 and its hydrolysis product CX1763 short HPLC columns consisting of a Phenomenex Gemini C18 column 3 × 50 mm (3u) and a Thermo BetaBasic C18 column 3 × 50 mm (3u) was used. All analysis were conducted using Methanol with 0.1% formic acid as the mobile phase. The aqueous phase consisted of nanopure water with 0.1% formic acid. The mobile Phase flow was 0.4–0.5 ml/min. Gradient LC conditions are listed in Supplementary Tables S1–S4. As shown in Supplementary Figure S2 using the Phenomenex Gemini column CX1942 is baseline separated from CX1763 under gradient conditions described in Supplementary Table S1. Similarly with the Thermo Beta Basic column CX1942 is baseline separated from CX1763 under gradient conditions described in Supplementary Table S2. Both HPLC run times are on the order of 10 min. The analytical HPLC system utilized here did not attribute to CX1942 ester hydrolysis.

2.3. Electrically evoked field excitatory postsynaptic potentials (fEPSPs)

Evaluation of ampakine effects on hippocampal EPSP were described previously [29]. Male Long Evan rats (weight of 250–350 g) were used in this study. They were given an intraperitoneal (IP) injection of pentobarbital (60 mg/kg). Two catheters made from PE10 polyethylene tubing were inserted into the femoral vein and artery for ampakine administration and blood sampling. Anesthesia in these animals was maintained by pentobarbital infusion with a rate of 2–4 mg/kg/h. Animals were then placed into a stereotactic frame. Burr holes drilled into the left hemisphere of the skull allowed for positioning of a stimulating electrode (-7.8 to -8.1 from bregma, 4.2 to 4.4 lateral to midline) and a recording electrode (-3.0 to -3.3 from bregma, 1.6 to 2.2 lateral to midline). The monopolar stainless steel stimulating electrode (formvar insulated, 175 μm) was lowered to the perforant path together with a platinum/iridium recording electrode (75 μm) into the hilus of the dentate gyrus in the hippocampus. The current used to produce the evoked potential was adjusted to produce a response size between 50–60% of the spike-free maximal amplitude. Evoked hilar excitatory postsynaptic field potentials (EPSPs) were recorded in response to single pulse stimulation which was delivered with a frequency of one pulse per 20 s to the perforant path. Approximately 20–30 min after a stable baseline was established, CX1942 or CX1763 was injected intravenously (IV) and field potentials were recorded continuously every 20 seconds for approximately another 2 h. Data were acquired and subsequently analyzed using available software (NAC and NACSHOW). The amplitude, half-width and area of the EPSPs were measured for each stimulation pulse, and the effects of CX1942 and CX1763 on EPSPs were compared with baseline EPSPs using a two-tailed, t-test.

2.4. The 5-choice serial reaction time task (5CSRTT)

Male hooded Wistar rats (250–310 g) were utilized in this study (Charles River Labs, Kent, UK). The rats were trained as described previously by Hahn et al. [30]. One second duration light flashes were presented randomly in one of the nose holes after a 15 second inter trial interval (ITI). If a rat poked its nose into an illuminated hole or within five seconds after the light was extinguished, a 45 mg sucrose food pellet was delivered into a food tray. These responses were subsequently counted as correct responses. Nose pokes into a hole that was not illuminated produced a timeout period of 2 s during which time the experimental apparatus was dark and responses had no scheduled consequences. When a rat did not respond until the end of the limited hold, this was then recorded as an omission error. When there was a response during the ITI, this was recorded as a premature response. Premature responses were quantified as a percentage premature responses previously calculated by Robinson et al. [31]. All rats in these studies had to yield a stable criterion whereby their performances had to include less than 20% omissions and greater than 70% correct responses on the aforementioned task parameters before testing with CX1763 took place. Sixteen rats were used in the study and repeated tests were conducted with graded doses utilizing a randomized sequence to ensure all rats received vehicle and both doses of CX1763 [22].

The experiments were conducted in an experimental apparatus inside of noise-cancelling enclosures (Med-Associates, Fairfax, VT, USA). In the apparatus, five square 2.5 cm openings 5 cm above the grid floor and 5 cm deep were positioned on the curved back wall of the chambers. In the holes, there was contained an infra-red photocell beam detector which monitored their entrance and a green light at the back. On the opposite wall and equidistant from the rear wall openings was the tray onto which food pellets could be delivered. The apparatus was lit from the ceiling. CX1763 was dosed IP and then 5CSRTT testing sessions took place and lasted for 30 min.

For the described measures of attentional and cognitive performance, dose-response changes were assessed via one-way ANOVA followed by Dunnett multiple comparison tests.

2.5. CX1763 effects on amphetamine-induced locomotor activity (LMA)

Low-impact ampakines have the ability to attenuate the hyperactivity associated with amphetamine [22,32]. In order to assess the interaction between CX1763 and amphetamine, we utilized an assay we recently described [22]. Briefly, CD1 mice (Charles River Labs) were placed into polycarbonate animal cages with two photobeam arrays to detect LMA (26 cm × 48 cm × 20 cm; W × L × H). Mice were allowed to habituate in these cages for 20 min prior to treatment with vehicle or CX1763 (1–18 mg/kg). Five min later they were injected IP with amphetamine 2 mg/kg. Mice were returned to the cages 10 min later. Their activity was measured for 15 min. In these cages, LMA was monitored continuously and subsequently recorded in a computer for all experimental cages (Photobeam Activity System, San Diego Instruments, San Diego, CA, USA).

2.6. Antagonism of opioid-induced respiratory depression in vivo studies

Alfentanil-induced respiratory depression studies were conducted as recently described [22]. Twenty-four hour prior to experimentation, male Charles River Rats underwent surgery to have two cannulas (PE30) inserted into the right jugular vein. This minimizes rodent handling during the experiments. On the day of experimentation, rats were placed into a whole-body plethysmograph (Buxco, St. Paul, MN, USA) chamber for one to 2 h to acclimate to the environment. A stable breathing rhythm baseline was established for 20 min, after which time alfentanil was introduced through one of the cannulas by an infusion pump at 250 μg/kg/20 min for 60 min, which was provided to reduce breathing by approximately 50%. Twenty minutes after starting alfentanil, CX1942 was injected as a bolus via the second cannula at a dose of 2.5, 5 or 10 mg/kg. When alfentanil infusion was stopped, respiration was monitored for another 20 min. Data were collected within the Buxco plethysmograph system. Minute volume, measured by ml/min of rat respiration, was determined by multiplying respiration rate (breaths per minute) by the tidal volume (ml). Minute volume was normalized to that prior to alfentanil infusion. The reduction and rescue by CX1942 was expressed as a percentage of the individual animal’s baseline. Data analysis was determined by ANOVA followed with Dunnett’s multiple comparison test

2.7. Single dose toxicity in adult rats

In order to assess acute epileptogenic effects of CX1763, small groups (n = 2–5 rats) of young adult male Sprague–Dawley rats (250–350 g) (Harlan Sprague–Dawley) were treated with CX1763 by oral gavage and observed for seizure and death for 2 h after dosing. On each subsequent day for 2 weeks, animals were observed for 10 min. Dosages were 1000, 1500 and 2000 mg/kg. Seizures were rated by the scale 0–5 discussed previously [22]. No mice at any dosages experienced a seizure, resulting in all mice getting a seizure score of 0 during the treatment periods.

2.8. Data analysis

Data with dose-responses were typically analyzed with ANOVA followed by Dunnett’s multiple comparison test using Prism. Comparing two different groups was done with student's t-test. Alpha value was set at 0.05.

3. Results

CX1942 is the glycine ester of the active low-impact ampakine CX1763 (Figure 1). The solubility of CX1942 was 99.9% water soluble at 70 mg/ml in physiological saline, In contrast, the solubility of CX1763 was only 0.1% under the same conditions.

Evaluation of CX1942 stability in whole blood from rat, dog and human at 37°C indicated a rapid loss of CX1942 by hydrolysis (Figure 2A). The calculated half-life for CX1942 was 2 min in rat (n = 3), 4 minutes in dog (n = 1) and 10–15 min in human (n = 3) at 37°C. The hydrolysis of CX1942 correlated with the rapid formation of an equimolar concentration of CX1763 (Figure 2A). As shown in Figure 2B & C, Addition of citric acid at a concentration of 50 mM (a maximal concentration not causing red blood cell lysis) was found to inhibit CX1942 hydrolysis and limit CX1763 formation in rat and human plasma: percentage of CX1763 formation in rat plasma was reduced from 99 to 12.8% after 15 min and in human plasma from 85 to 5% after 30 min. While it is still unclear as to why citric acid significantly inhibited CX1942 hydrolysis, it may be due to the acid reducing blood pH to a point where serum esterases lose enzymatic activity or simply by a chelation effect of the tricarboxylic moieties in the acid.

Figure 2.

Examination of CX1942 hydrolysis and CX1763 formation in human, dog and rat whole blood in the presence and absence of 50 mM citric acid. (A) CX1942 stability and CX1763 formation in rat, dog and human whole blood incubated at 37°C. (B) CX1942 stability and CX1763 formation in human blood with or without 50 mM citric acid. (C) CX1942 stability and CX1763 formation in rat blood with or without 50 mM citric acid.

CX1763 and CX1942 were tested for their ability to facilitate synaptic responses in vivo in anesthetized animals following peripheral administration. Following IV administration of 10 mg/kg CX1763 and CX1942, the amplitude of EPSP increased and peaked approximately 15 min after drug administration. The increases in amplitudes were long-lasting, and declined to baseline approximately 100–120 min after drug administration (Figure 3A). Two-way ANOVA was used to compare the effects between CX1763 and CX1942 on EPSP facilitation over time. There was a significant effect of time on EPSP amplitudes (F(405, 2430) = 14.17, p < 0.0001), but there was no significant difference between treatment overall (F(1, 6) = 1.712, p = 0.239). Regarding the interaction between time and treatment, the amplitude increase of CX1763 tended to persist longer than that of CX1942, and this interaction closely approximated statistical significance (F(405, 2430) = 1.127, p = 0.0527). Significant potentiation of the amplitude of the EPSP was achieved by both ampakines 15 min after administration is shown in Figure 3B.

Figure 3.

Effect of IV CX1763 and CX1942 on excitatory postsynaptic potentials in rat hippocampus in vivo. (A) The time course of the measured increase in the amplitude of the EPSP in dentate gyrus following 10 mg/kg IV administration of CX1763 or CX1942. (B) Potentiation of EPSP amplitudes 15 min after administration of each compound. Bars represent the mean ± SEM from 4–5 animals. *p < 0.05; **p < 0.01 t-test compared with baseline.

EPSP: Excitatory postsynaptic potential; SEM: Standard Error of Means.

We and others have shown that low-impact ampakines improve metrics of attention in preclinical models and in the clinical setting [13,22,28]. In the 5CSRTT, CX1763 did not significantly improve the percentage of correct responses. The indices that were mainly affected by CX1763 were the latency to a correct response, the percentage of premature responses and the omission errors (Figure 4). 1 mg/kg CX1763 did not significantly alter these parameters, though 5 mg/kg did significantly decrease correct response latency and the percent of premature responses. 5 mg/kg CX1763 did significantly increase the omission errors as well (Figure 4). As reported previously, the number of premature responses when the ITI was 5 s was low, but when the ITI was changed from 5 s to 15 s, there was a significant increase in the premature responses [22]. Robinson et al. reported an increase in premature responses upon the ITI interval change as well [31]. These findings illustrate that CX1763 may possess the ability to increase attentional attributes, in line with other low-impact ampakines [22].

Figure 4.

Effect of CX1763 in the rat 5-CSRTT. CX1763 was administered IP prior to testing sessions. Data represent mean ± SEM from 16 rats per treatment group. Statistical differences were calculated using a one-way ANOVA followed by Dunnett’s multiple comparison test. *p < 0.05 compared with vehicle.

SEM: Standard Error of Means.

For translational relevance, it is necessary to probe the interaction between CX1763 and amphetamine, a standard of care therapy for ADHD. Mice were treated with increasing doses of CX1763 and then treated with amphetamine to stimulate hyperactivity. Amphetamine approximately quadrupled the number of beam breaks compared with saline treatment (Figure 5). CX1763 dose-dependently diminished the amphetamine-induced hyperactivity with an AD₅₀ of 2 mg/kg (Figure 5). CX1763 at doses of 3–18 mg/kg significantly perturbed hyperactivity. Eighteen mg/kg CX1763 completely reduced amphetamine’s effects back to baseline values.

Figure 5.

IP CX1763 significantly reduces amphetamine-stimulated locomotion bars represent the mean ± SEM from 12–16 mice. For mice, statistical differences were determined using a one-way ANOVA (p < 0.01 for each panel) followed by Dunnett’s multiple comparison test versus control; **p < 0.01; ****p < 0.0001, compared with amphetamine alone.

SEM: Standard Error of Means.

We also investigated the ability of CX1942 to reverse OIRD in rats. The efficacy and timing of action of CX1942 was examined in vivo using adult rats and plethysmographic recording. Figure 6 shows data for alfentanil-induced suppression of minute volume and the subsequent antagonism of OIRD, in a dose-dependent manner, by CX1942. Additional population data for changes in respiratory frequency and minute volume induced by increasing doses of CX1942 are shown in Figure 6B. 5 mg/kg CX1942 completely reversed OIRD, albeit approximately 10–12 min after IV administration. 10 mg/kg CX1942 produced a more rapid complete resolution of OIRD approximately 6–7 min after administration.

Figure 6.

IV CX1942 dose-dependently reverses opioid-induced respiratory depression in rat. (A) Changes in minute volume of adult rats in response to alfentanil infusion and vehicle or CX1942 administration. (B) Population data showing changes in minute volume in response to alfentanil infusion with administration of vehicle (control) or 2.5, 5, 10 mg/kg CX1942. Data points represent the mean ± SEM from at least 5 rats per group. *p < 0.05; **p < 0.01 following analysis using a one-way ANOVA and Dunnett’s multiple comparison test versus control.

SEM: Standard Error of Means.

We concluded our studies by treating rats orally with high doses of CX1763 (1000, 1500 and 2000 mg/kg) and assessing for the onset of epileptogenic activity. After administration of 1000 and 1500 mg/kg we observed that rats had reduced activity, but all five rats in each treatment group survived to day 14 post-treatment. However, after treatment with 2000 mg/kg, one of the two rats was found dead the next day. The other rat was found 24 h after treatment severely hypoactive, so the decision was made to euthanize the rat (Table 1). Therefore, the minimum lethal dose in rats was between 1500 and 2000 mg/kg. Regardless, these findings illustrate that the active ampakine CX1763 is pharmacologically active and devoid of epileptogenic effects well within therapeutic doses and has a markedly encouraging therapeutic ratio.

Table 1.

Single dose toxicity of high-dose CX1763.

Dose (mg/kg)	# of rats seizing	# of Rats Surviving to 14 Days
1000	0/5	5/5
1500	0/5	5/5
2000	0/2	0/2

4. Discussion

Positive allosteric modulation of AMPAR by AMPAkines has the potential to yield substantial benefits in neurological/psychiatric diseases in which glutamatergic tone in the CNS is compromised. However, the seizurogenic activity of experimental, high impact AMPAR modulators like cyclothiazide [16–18] has significantly stalled the translational efforts of these compounds. Fortunately, low-impact AMPAkines such as CX717 and CX1739 exhibit efficacy in a myriad of preclinical [20–22,24,25,33–41] and clinical settings [14,26]. Here, we describe the activity of the low impact AMPAkine CX1763 and the glycine ester pro-drug CX1942. IV administration of both compounds produced a significant, long-lasting increase in the amplitudes of hippocampal EPSPs at 10 mg/kg. Future studies can fully assess blood brain barrier permeation kinetics, but since both compounds increase EPSPs within 10 min of IV administration, this would suggest modestly rapid penetration into the CNS. Both AMPAkines were active in other rodent assays at doses between 2.5 and 5 mg/kg. The lack of epileptic activity at high multiples of therapeutic doses demonstrates that the therapeutic effects of low-impact AMPAkines can indeed be differentiated from unwanted side effects and that this subclass of AMPAkines may have a profound therapeutic ratio.

Our findings also build upon prior results illustrating the beneficial effects low-impact AMPAkines may have on attention and cognition. The 5CSRTT procedure tests rodent attention and cognition and is used as an animal model to study the efficacy of compounds that are used to treat ADHD in the clinic. CX1763 dose-dependently improves latency to initiate a correct response and reduces anticipatory responses (Figure 4). It is interesting to note that unlike CX1739 [22], CX1763 did not increase the percent correct responses in the 5CSRTT (Figure 4). The reason(s) for this difference in the overall therapeutic effects of low-impact AMPAkines is uncertain and requires further study. Nevertheless, CX1763 and CX1942 may be useful in improving attention and ameliorating cognitive dysfunction in humans that occur as a result of neurological disease or damage. This result is consistent with the idea that patients with ADHD have less glutamate in certain areas of their brains [13,42,43]. Thus, enhancing glutamatergic, specifically AMPAergic neurotransmission, may be critical in reversing some of the symptoms associated with ADHD.

We expanded upon the utility of CX1763 to treat ADHD by examining its interaction with amphetamine, a stimulant frequently prescribed to patients with ADHD. CX1763 reduces amphetamine-induced LMA with an AD₅₀ of 2 mg/kg (Figure 5). These data show that at concentrations required to improve performance in the 5CSRTT, CX1763 reduced the common side effect of stimulant medication. In this assay, CX1763 was substantially more potent than CX1739 [22]. 10 mg/kg CX1739 minimally reduced amphetamine-induced LMA whereas 10 mg/kg CX1763 reduced LMA by approximately 90% (Figure 5). Taken together, it may be of interest to investigate the effects of administering CX1942 and a stimulant together to bolster therapeutic effects in treating ADHD while minimizing side effects of stimulant use. Nevertheless, data from these animal studies are in line with previously reported clinical studies in which earlier low impact AMPAkines reduced ADHD scores in adult ADHD patients [13].

Finally, we asked whether CX1942, the water-soluble pro-drug of CX1763, could reverse alfentanil-induced respiratory depression in vivo. In the CNS of mammals, breathing rhythms are governed by glutamatergic tone in the pre-botzinger complex. Like other low impact AMPAkines that reverse OIRD in the preclinical [20–22,24,25,27,44,45] and clinical settings [14,26], CX1942 potently and fully reversed OIRD at 5 mg/kg (Figure 6), making it four-times more potent than CX1739 [22]. The increase in potency over CX1739 in OIRD is paralleled by its increased potency in attenuating amphetamine-induced LMA (Figure 5). CX1942 is soluble in saline at >50 mg/ml and is rapidly de-esterified to CX1763 in the blood. Thus, the current data suggests that it may be possible for CX1942 to be used by emergency medical technicians to assist in treating patients that have overdosed on opiates. Since low impact AMPAkines counteract OIRD without affecting analgesia, future research will look into whether the antagonism of OIRD produced by CX1942 will occur without the induction of withdrawal.

The lack of profound effects on glutamate-induced AMPAR activation makes low-impact AMPAkines a more attractive chemical class for clinical translation but also hinders mechanistic studies that could further define their mechanism of action. Fortunately, work has been done to elucidate how CX1763 affects AMPAergic signaling in comparison to high-impact AMPAkines. Previously, Dai et al. [45] published mechanistic data of CX1763 (termed LCX001) from studies using rats. They found that CX1763 could reverse respiratory depression due to opiates and barbiturates. They also found CX1763 had its own analgesic effects and that it did not compromise the analgesic effects of morphine [45]. Notably, CX1763 increased the binding of [³H]AMPA to rat brain membranes at clinically meaningful concentrations. When comparing the effects of CX1763 to those of the high-impact AMPAkine CX614 on glutamate-induced currents, they found that CX614 was more effective at inhibiting desensitization and augmenting steady-state currents but that both AMPAkines increased peak current amplitude. They also found that CX1763 lacks intrinsic agonistic activity at concentrations up to 100 μM [45]. These data provide further support that low-impact AMPAkines like CX1763 preferentially accelerate channel opening but do not offset desensitization to the extent observed with high-impact AMPAkines. Additionally, they also reported that CX1763, not CX614, increased membrane levels of Glur2(R) but not Glur1 [45]. Glur2(R)-containing AMPAR are impermeable to calcium [46,47]. The increase in Glur2(R)-containing AMPAR on the surface of neurons by CX1763 may explain the safe augmentation in glutamatergic currents without producing calcium-dependent excitotoxity in neurons. Further studies are warranted to see how other low-impact AMPAkines affect cell surface expression of specific AMPAR subunits and how other low-impact AMPAkines affect specific aspects of glutamate-induced currents.

5. Conclusion

Taken together, our findings indicate that the low-impact AMPAkine CX1763, and the glycine ester waster-soluble pro-drug CX1942 are safe and active in models of ADHD, reducing amphetamine-induced LMA, in attenuating OIRD, and in facilitating synaptic transmission in the hippocampus. The data presented here also add to accumulating evidence that the drug class of benzoxadiazole low-impact AMPAkines have the potential to be effective therapies in a host of neuropsychiatric and cognitive disorders in which AMPAergic synaptic function may be compromised with limited worry for epileptogenic effects. A water-soluble, low-impact AMPAkine, with excellent safety margins, has the capacity to serve a vitally needed role in combatting the opiate epidemic and greatly assist emergency medical technicians in the field. It may also be used in hospitals to preserve breathing in patients undergoing anesthesia who cannot take ampakines orally.

Supplemental material

Supplemental data for this article can be accessed at https://doi.org/10.1080/17568919.2024.2401312

Financial disclosure

Daniel Radin, Rok Cerne, Jeffrey Witkin, and Arnold Lippa are associated with RespireRx, where A.L. is acting CEO, and D.P.R., R.C. and J.M.W. are non-paid researchers who occasionally conduct studies on these compounds.

Competing interests disclosure

The authors are employees of RespireRx Pharmaceuticals, developers of low impact ampakines for the treatment of OIRD, ADHD and spinal cord injuries. The authors have no other competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript apart from those disclosed.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.