Subanesthetic Ketamine with an AMPAkine Attenuates Motor Impulsivity in Rats

Abstract

The concept of “impulse control” has its roots in early psychiatry and today has progressed into a well-described, although poorly understood, multidimensional endophenotype underlying many neuropsychiatric disorders (e.g., attention deficit hyperactivity disorder, schizophrenia, substance use disorders). There is mounting evidence suggesting that the cognitive and/or behavioral dimensions underlying impulsivity are driven by dysfunctional glutamate (Glu) neurotransmission via targeted ionotropic Glu receptor (GluR) [e.g., N-methyl-D-aspartate receptor (NMDAR), α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR)] mechanisms and associated synaptic alterations within key brain nodes. Ketamine, a noncompetitive NMDAR antagonist and FDA-approved for treatment-resistant depression, induces a “glutamate burst” that drives resculpting of the synaptic milieu which lasts for several days to a week. Thus, we hypothesized that single and repeated treatment with a subanesthetic ketamine dose would normalize motor impulsivity. Next, we hypothesized that AMPAR positive allosteric modulation, alone or in combination with ketamine, would attenuate impulsivity and provide insight into the mechanisms underlying GluR dysfunction relevant to motor impulsivity. To measure motor impulsivity, outbred male Sprague Dawley rats were trained on the one-choice serial reaction time task. Rats pretreated with single or repeated (3 days) administration of ketamine (10 mg/kg; i.p.; 24-hr pretreatment) or with the AMPAkine HJC0122 (1 or 10 mg/kg; i.p.; 30-min pretreatment) exhibited lower levels of motor impulsivity versus control. Combination of single or repeated ketamine plus HJC0122 also attenuated motor impulsivity versus control. We conclude that ligands designed to promote GluR signaling represent an effective pharmacological approach to normalize impulsivity and subsequently, neuropsychiatric disorders marked by aberrant impulse control.

Introduction

Glutamate (Glu) excitatory neurotransmission and its associated receptor (GluR) systems are profusely expressed throughout the entire brain and heavily involved in executive function. Many neuropsychiatric disorders are marked by Glu system dysfunction. For example, deficient Glu neurotransmission is linked to schizophrenia (Howes et al., 2015; Uno and Coyle, 2019), bipolar disorder and depression (Wise et al., 2018), substance use disorders (Koob and Volkow, 2016), and impulsivity (Anastasio et al., 2019). Impulsivity is an intricate, multidimensional behavioral endophenotype broadly defined as behavior without sufficient foresight that is associated with several neuropsychiatric disorders (Moeller et al., 2001). Importantly, neurodevelopmental psychiatric disorders classified as synaptopathies (i.e., diseases associated with synaptic dysfunction) (Brose et al., 2010) and concurrently associated with impulsivity, such as schizophrenia and autism spectrum disorder (ASD), are characterized by persistent cortical GluR deficits that affect higher order executive functioning (e.g., decision making, cognition, etc.) (Sceniak et al., 2019). This GluR-associated hypofrontality phenomenon extends directly to impulsivity (Davis-Reyes et al., 2019; Murphy et al., 2012), suggesting that the dysfunctional synaptic milieu underlying impulsivity bears a striking resemblance to that of neuropsychiatric disorders characterized by increased impulsivity.

Evidence exists that impulsivity may be governed by GluR hypofunction, particularly related to the ionotropic glutamate N-methyl-D-aspartate receptor (NMDAR) (Cottone et al., 2013; Davis-Reyes et al., 2019; Higgins et al., 2003; Murphy et al., 2012) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) (Nakamura et al., 2000). The NMDAR forms a heterotetrametric structure composed of the obligatory GluN1 subunit and GluN2A-D subunits. Systemically administered NMDAR antagonists elevate impulsivity, and selective antagonism of the NMDAR GluN2B subunit enhances impulsivity (Burton and Fletcher, 2012; Higgins et al., 2003). Furthermore, trait impulsivity predicts responsiveness to D-cycloserine, an NMDAR partial agonist, which corresponds to lower cortical NMDAR subunit expression (Davis-Reyes et al., 2019). The AMPAR consists of a family of tetrameric subunits (GluA1-4, a.k.a. GluR1-4) to regulate fast excitatory synaptic transmission and synaptic plasticity at glutamatergic synapses (Derkach et al., 2007). AMPAR antagonism dose-dependently increases impulsivity, which can be reversed following the co-administration of aniracetam, an AMPAR positive allosteric modulator (PAM, i.e. AMPAkine) (Nakamura et al., 2000). Taken together, this suggests that GluR hypofunction, governed in part by the NMDAR and AMPAR may underlie high impulsivity and pharmacological tools that correct/normalize GluR function represent a viable approach in treating heightened impulsivity and further, its involvement in neuropsychiatric and neurodevelopmental disorders.

Ketamine is a non-competitive NMDAR antagonist originally developed as an anesthetic and subanesthetic doses of S-ketamine (esketamine) are FDA-approved and marketed as antidepressants for treatment-resistant populations (Tibensky et al., 2016). Esketamine has a short half-life (~2-3 hours) (Singh et al., 2018) with maximal antidepressant effects occurring within 24 hours after administration (Canuso et al., 2018). Therefore, the therapeutic efficacy of esketamine is not directly associated with the presence of drug in the system, but rather a result of prominent synaptic resculpting within the brain long after clearance (Duman, 2018; Duman and Aghajanian, 2012). There is strong evidence that ketamine preferentially targets tonically active gamma-aminobutyric acid (GABA) interneurons within the corticolimbic system (Gerhard et al., 2020), resulting in disinhibition of cortical pyramidal glutamatergic neurons and a “burst” of downstream Glu (Duman, 2018; Duman and Aghajanian, 2012). Of note, at anesthetic doses (>50 mg/kg), a Glu “burst” is not observed in animals (Moghaddam et al., 1997). Exposure to a single dose of ketamine is known to increase the levels of proteins related to dendritic and synaptic remodeling, such as AMPAR (Li et al., 2011), in several brain regions, all of which is described as a synaptic plasticity event (Abdallah et al., 2017). Further, a single subanesthetic dose of ketamine increases cortical dendritic spine density for up to two weeks in mice (Phoumthipphavong et al., 2016), highlighting the relative longevity of these neuroadaptations. We proposed that, unlike other non-competitive NMDAR antagonists, which increase impulsivity due to active, ubiquitous blockade of the NMDAR, the long-term neuroplastic effects of ketamine, in the absence of active NMDAR blockade, will normalize impulsivity.

AMPAkines (e.g., aniracetam, piracetam) are clinically relevant compounds that activate the AMPAR on the allosteric site to decrease the likelihood of GluR desensitization and deactivation (Arai et al., 2000; Lynch, 2006) without inducing excitotoxicity (Arai and Kessler, 2007; Jardemark et al., 2012; Lynch, 2006) and enhance the induction of long-term potentiation both in vitro (Arai and Kessler, 2007) and in vivo (Staubli et al., 1994). AMPAkines robustly ameliorate ketamine-induced impairment of working memory (Roberts et al., 2010) and induce pro-cognitive effects in schizophrenia (Jardemark et al., 2012; Tuominen et al., 2005). We identified the AMPAkine HJC0122 using a combined bioinformatics and chemoinformatics approach (Chen et al., 2013). HJC0122 is a new chemical scaffold which promotes neuroprotective effects and prevents neuroapoptosis, both in vivo and in vitro (Chen et al., 2013). Moreover, HJC0122 represents a pharmacological agent for disorders and behavioral conditions marked by GluR hypofunction.

In the present study, we utilized ketamine and HJC0122 as pharmacological tools to normalize motor impulsivity. Motor impulsivity/impulsive action, a subtype of impulsivity, herein is defined as difficulty withholding a prepotent motor response (Hamilton et al., 2015). We tested the effects of ketamine and HJC0122 on motor impulsivity in a preclinical rat model using the 1-choice serial reaction time (1-CSRT) task. While single (acute) ketamine administration consistently elicits antidepressant effects, the results of studies employing repeated administration of ketamine on motor impulsivity are less conclusive (Benn and Robinson, 2014; Nikiforuk et al., 2015; Short et al., 2018). Therefore, we tested single and repeated subanesthetic dosing of ketamine alone, and in combination with HJC0122. We predicted that ketamine and HJC0122, both alone and in combination, would normalize inherent motor impulsivity.

Materials and Methods

Animals

Male, outbred Sprague–Dawley rats (n=37; Envigo, Indianapolis, IN) weighing 250–275 g upon arrival were housed two/cage under a 12-h light–dark cycle with controlled temperature (21–23°C) and humidity (40–50%). Animals were acclimated for seven days to the colony room prior to the start of handling and experimental procedures. During the 1-CSRT task acquisition and maintenance, rats were food restricted to ~90% free-feeding weight; water was available ad libitum except during daily operant sessions. Rats were weighed daily to ensure that their body weights were maintained at ~90% of free-feeding levels. All experiments were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals (2011) and with the University of Texas Medical Branch Institutional Animal Care and Use Committee approval.

Drugs

Ketamine [dl 2-(o-chlorophenyl)-2-(methylamino) cyclohexanone hydrochloride; 100 mg/mL; Columbus, OH; catalog #056344] was purchased from Henry Schein Animal Health and diluted in saline (0.9% NaCl; Baxter Healthcare, Deerfield, IL; catalog #2F7124). HJC0122 [(2R)-Propane-2-sulfonic acid (2-{4-[2-(pyrrolidine-1-carbonyl)pyridin-4-yl]-phenyl}propyl) amide] was synthesized as previously described (Chen et al., 2013) and dissolved in 4% Tween® 80 (polyoxyethylene-20-sorbitan monooleate; Fisher Scientific, Hampton, NH; catalog # CAS 9005-65-6) and 1% DMSO (dimethyl sulfoxide; Sigma, St. Louis, MO; catalog # D8418-500ML) in saline.

1-Choice Serial Reaction Time Task Training

Procedures occurred in standard five-hole nose-poke operant chambers equipped with a house light, food tray, and an external pellet dispenser capable of delivering 45 mg pellets (Bio-Serv, Frenchtown, NJ) housed within ventilated and sound-attenuated chambers (MedAssociates, St Albans, VT). The 1-CSRT task methodology has been described in detail previously (Anastasio et al., 2014; Anastasio et al., 2019; Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019). Briefly, rats were habituated to the test chamber; a nose-poke into the singly illuminated center hole resulted in the delivery of one food pellet into the magazine on the opposite wall of the chamber and simultaneous illumination of the magazine light. During this stage, all responses made in the correctly lit (target) hole resulted in the illumination of the magazine light and presentation of a single food pellet. The latency to retrieve the food pellet is recorded. The training stages thereafter were each comprised of daily sessions of 100 trials to be completed in a maximum of 30 min; each training stage involved incrementally lowering the stimulus duration to 0.5 sec with a 5-sec limited hold and an intertrial interval (ITI) of 5 sec (ITI5). A maximum of 100 correct responses in a session resulted in a maximum of 100 reinforcers earned; incorrect, premature responses or omissions resulted in a 5-s time-out period and a reduction in potential reinforcers obtained. Advancement to the next training stage required rats to meet acquisition criteria: ≥50 correct responses, >80% accuracy [correct responses/(correct + incorrect) × 100] and <20% omissions (omitted responses/trials completed × 100). Time to finish the training session is recorded for each rat.

Premature responses [(total premature responses = target premature responses + non-target premature responses; “non-target” indicates premature response detected outside of the center nose poke hole)] were employed as the primary indication of motor impulsivity (Anastasio et al., 2014; Anastasio et al., 2019; Dalley et al., 2002; Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019; Winstanley, 2011). The number of reinforcers earned provides a measure of task competency and a secondary assessment of motor impulsivity, while percent accuracy was a general indication of attentional capacity (Anastasio et al., 2014; Anastasio et al., 2019; Dalley et al., 2002; Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019; Winstanley, 2011). Percent omissions indicated failures of detection of the visual stimuli in the target hole as well as motivation to perform the task (Anastasio et al., 2014; Anastasio et al., 2019; Dalley et al., 2002; Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019; Winstanley, 2011).

Pharmacological manipulation of motor impulsivity via systemic administration of ketamine alone, HJC0122 alone, or the combination of ketamine plus HJC0122

Two cohorts (Cohort 1 and Cohort 2) of rats were trained on the 1-CSRT task for pharmacological testing. The cohorts could not be collapsed and were analyzed separately because not all treatment groups from Cohort 1 (n=24) were performed in Cohort 2 (n=13). After meeting stability criteria for the final training stage of the 1-CSRT task over three consecutive ITI5 sessions (with <20% variability), all rats were administered a vehicle test (1 mL/kg; i.p.). Following A Simple Practice Guide for Dose Conversion Between Animals and Human (Nair and Jacob, 2016), our selected 10 mg/kg dose of ketamine in rats roughly mirrors subanesthetic doses in humans (ranging from 0.5-2 mg/kg) (Newport et al., 2015). Additionally, the ketamine pretreatment time was selected to mirror human studies which have shown rapid, antidepressant effects 24 hours following a subanesthetic dose of intravenous ketamine hydrochloride (Murrough et al., 2013; Zarate et al., 2006). Dosing and pretreatment time for HJC0122 were selected based on previous studies (Chen et al., 2013; Hu et al., 2018). Rats were treated with vehicle the day before each drug treatment and received a minimum of one-week washout period between all treatments during which 1-CSRT task training persisted. Rats were required to demonstrate stability for three days in task criteria and individual rat performance (determined by behavioral performances within 20% above or below individual mean performance) prior to a subsequent vehicle-drug test session. Pending continued stability on the vehicle and in a balanced, pseudo-randomized order, within subject design, testing was as follows:

1. Cohort 1 rats (n=24) received a single (10 mg/kg; i.p) or repeated (10 mg/kg x 3 days; i.p.) dose of ketamine (“ketamine alone”) immediately after completion of the 1-CSRT task, i.e., 24 hrs prior to behavioral assessment on the 1-CSRT task (see experimental design, Fig. 1A); HJC0122 (1 or 10 mg/kg; i.p.) (“HJC0122 alone”) administration occurred 30-min prior to performance of the 1-CSRT task (see experimental design, Fig. 2A). Additionally, rats were administered a single (10 mg/kg; i.p.) dose of ketamine immediately after completion of the 1-CSRT task and the following day were administered HJC0122 (1 mg/kg; i.p.) 30 min prior to initiation of the task (“ketamine plus HJC0122”) (see experimental design, Fig. 3A). Six to seven rats failed to meet/maintain 1-CSRT task stability criteria or were statistical outliers (i.e., two standard deviations above or below the group mean) and were excluded from subsequent analyses (ketamine efficacy n=17/24 analyzed, HJC0122 efficacy n=18/24 analyzed, ketamine plus HJC0122 n=17/24 analyzed).

2. Cohort 2 rats (n=13) received a repeated (10 mg/kg x 3 days; i.p.) dose of ketamine alone as described above (ketamine alone) (see experimental design, Fig. 4A); HJC0122 (1 mg/kg; i.p.) (HJC0122 alone) administration occurred 30-min prior to performance of the 1-CSRT task (see experimental design, Fig. 4A). Additionally, rats were administered a repeated (10 mg/kg x 3 days; i.p.) dose of ketamine immediately after completion of the 1-CSRT task and the following day were administered HJC0122 (1 mg/kg; i.p.) 30 min prior to initiation of the task (ketamine plus HJC0122) (see experimental design, Fig. 4A). Two rats failed to meet/maintain 1-CSRT task stability criteria or were statistical outliers (i.e., two standard deviations above or below the group mean) and were excluded from subsequent analyses (ketamine efficacy n=11/13 analyzed, HJC0122 efficacy n=11/13 analyzed, ketamine plus HJC0122 n=11/13 analyzed).

Figure 1. Effects of single and repeated ketamine administration on 1-CSRT task performance.

1-CSRT task stability criteria are denoted by dotted line, when applicable. Each open circle on bar graph represents an individual animal. (A) Single (10 mg/kg, i.p.; 24 hr pretreatment; 1 day) and repeated (10 mg/kg, i.p.; 24 hr pretreatment; 3 days) ketamine was administered during maintenance and stability sessions (ITI5). Graphs depict mean ± SEM (B) total premature responses, (C) total reinforcers earned, (D) percent accuracy, and (E) percent omissions. * denotes p<0.05 vs Vehicle

Figure 2. Effects of HJC0122 administration on 1-CSRT task performance.

1-CSRT task stability criteria are denoted by dotted line, when applicable. Each open circle on bar graph represents an individual animal. (A) HJC0122 (1 mg/kg and 10 mg/kg, i.p.; 30 min pretreatment) treatment was administered during maintenance and stability sessions (ITI5). Graphs depict mean ± SEM (B) total premature responses, (C) total reinforcers earned, (D) percent accuracy, and (E) percent omissions. * denotes p<0.05 vs Vehicle

Figure 3. Effects of combination single ketamine and HJC0122 administration on 1-CSRT task performance.

1-CSRT task stability criteria are denoted by dotted line, when applicable. Each open circle on bar graph represents an individual animal. (A) Single ketamine (10 mg/kg, i.p.; 24 hr pretreatment; 1 day) and HJC0122 (1 mg/kg, i.p.; 30 min pretreatment) administered during maintenance and stability sessions (ITI5). Graphs depict mean ± SEM (B) total premature responses, (C) total reinforcers earned, (D) percent accuracy, and (E) percent omissions. Here q-value is a p-value adjusted for false discovery rate (FDR). * denotes q<0.05 vs Vehicle; ^ denotes q<0.05 vs ketamine (10 mg/kg); # denotes q<0.05 vs HJC0122

Figure 4. Effects of combination repeated ketamine and HJC0122 administration on 1-CSRT task performance.

1-CSRT task stability criteria are denoted by dotted line, when applicable. Each open circle on bar graph represents an individual animal. (A) Repeated ketamine (10 mg/kg, i.p.; 24 hr pretreatment; 3 days) and HJC0122 (1 mg/kg, i.p.; 30 min pretreatment) administered during maintenance and stability sessions (ITI5). Graphs depict mean ± SEM (B) total premature responses, (C) total reinforcers earned, (D) percent accuracy, and (E) percent omissions. Here q-value is a p-value adjusted for false discovery rate (FDR). * denotes q<0.05 vs Vehicle; ^ denotes q<0.05 vs HJC0122

Statistical Analyses

Investigators who performed behavioral studies were blinded to all treatment assignments and endpoint statistical analyses. Using a within subject design, each rat within each cohort received all doses of vehicle, ketamine alone, HJC0122 alone, or ketamine plus HJC0122 as described above. To evaluate the relationship between outcome measures of the 1-CSRT task following vehicle, ketamine alone or HJC0122 alone, a one-way repeated measures analysis of variance was performed. A priori Dunnett’s multiple comparisons test was performed to assess differences between vehicle and single and repeated ketamine alone or HJC0122 alone when a main effect was observed. In the combination treatment experiments (single or repeated ketamine plus HJC0122), differences between all treatment groups were assessed using the two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli (Benjamini and Hochberg, 1995) to correct for multiple hypotheses and to control type II errors in large data sets when many significant outcomes are expected; the q-value reported is a p-value adjusted for false discovery rate (FDR). For the repeated ketamine analyses, the data were collected from a single test session 24 hrs following the third administration of ketamine. All statistical analyses were performed in GraphPad Prism 7 with an experiment-wise error rate set at α=0.05.

Results

Single and Repeated Ketamine Administration Alone Reduces Motor Impulsivity

Figure 1A details the experimental design for single and repeated administration of ketamine. A main effect of treatment on premature responses was observed (Fig. 1B F_2,32=5.78; p<0.05); single and repeated ketamine administration reduced premature responses versus vehicle (Fig. 1B; p<0.05). No main effect on reinforcers earned was detected (Fig. 1C; F_2,32=0.40; n.s.). A main effect of treatment on accuracy was detected (Fig. 1D; F_2,32=17.38, p<0.05); repeated, but not single, ketamine administration increased accuracy versus vehicle (Fig. 1D; p<0.05). A main effect of treatment on omissions was observed (Fig. 1E; F_2,32=8.23, p<0.05); repeated, but not single, ketamine administration increased omissions versus vehicle (Fig. 1E; p<0.05). The overall mean increase in omissions did not exceed task stability criteria (<20%).

We assessed the separate vehicle sessions to further demonstrate stability throughout the testing period (Supp. Figs. 1 and 2). For Cohort 1, no main effect of treatment (i.e., saline prior to ketamine or 4% Tween 80 + 1% DMSO prior to HJC0122) on premature responses (Supp. Fig. 1A; F_4,60=1.10; n.s.), reinforcers earned (Supp. Fig. 1B; F_4,60=2.16; n.s) or omissions (Supp. Fig. 1D; F_4,60=0.87; n.s) was observed. A small but significant effect of treatment on accuracy (Supp. Fig. 1C; F_4,60=5.85; p<0.05) was observed; vehicle performance for the repeated ketamine, HJC0122 (10 mg/kg), and ketamine plus HJC0122 was different from the vehicle for single ketamine (Supp. Fig. 1C; p<0.05). This difference in accuracy amounted to <2% which was >96% for all groups. For Cohort 2, no main effect of vehicle (i.e., saline prior to ketamine or 4% Tween 80 + 1% DMSO prior to HJC0122) on premature responses (Supp. Fig. 2A; F_2,18=1.02; n.s.), reinforcers earned (Supp. Fig. 2B; F_2,18=1.27; n.s), accuracy (Supp. Fig. 2C; F_2,18=0.22; n.s) or omissions (Supp. Fig. 2D; F_2,18=3.47; n.s) was observed. Taken together, these results indicate no potential order effects of the prior drug treatment on subsequent drug treatments on 1-CSRT task performance.

AMPAkine HJC0122 Alone Administration Reduces Motor Impulsivity

Figure 2A details the experimental design for HJC0122 alone. A main effect of treatment on premature responses was observed (Fig 2B; F_2,34=13.53; p<0.05); both doses (1 or 10 mg/kg) of HJC0122 reduced premature responses versus vehicle (Fig 2B; p<0.05). There was no main effect of treatment observed on reinforcers earned (Fig 2C; F_2,34=0.27; n.s.) or accuracy (Fig 2D; F_2,34=2.81, n.s.). A main effect of treatment on omissions was observed (Fig 2E; F_2,34=6.96, p<0.05); HJC0122 (10 mg/kg) increased omissions versus vehicle. However, the overall mean increase in omissions did not exceed stability criteria established in this assay.

We also determined if HJC0122 induced behavioral effects 24 hrs after administration. No main effect of treatment on premature responses (Supp. Fig. 3A; F_2,32=1.95; n.s.), reinforcers earned (Supp. Fig. 3B; F_2,32=0.04; n.s.), or accuracy (Supp. Fig. 3C; F_2,32=0.82; n.s.) 24 hrs after administration was observed. There was a main effect of treatment on omissions 24 hrs after administration (Supp. Fig. 3D; F_2,32=3.87); 1 mg/kg HJC0122 modestly increased omissions (p<0.05), which did not exceed task criteria.

Ketamine Plus Acute HJC0122 Reduces Motor Impulsivity

Figure 3A details the experimental design for single administration of ketamine plus acute HJC0122. A main effect of treatment on premature responses was observed (Fig 3B; F_3,48=10.25, q<0.05). The combination of single ketamine administration plus HJC0122 reduced premature responses versus ketamine alone (q<0.05) and HJC0122 alone (q<0.05) (Fig 3B). A main effect of treatment was observed on reinforcers earned (Fig 3C; F_3,48=4.82; q<0.05); combination treatment, but not ketamine alone or HJC0122 alone, increased reinforcers earned versus vehicle (Fig 3C; q<0.05). Additionally, combination treatment increased reinforcers earned versus ketamine alone (q<0.05) and HJC0122 alone (q<0.05) (Fig 3C). No main effect of treatment on accuracy (Fig 3D; F_3,48=2.26, n.s.) or omissions (Fig 3E; F_3,48=0.85, n.s.) was detected.

Figure 4A details the experimental design for repeated administration of ketamine plus acute HJC0122. There was a main effect of treatment on premature responses (Fig 4B; F_3,30=6.32; q<0.05); premature responses were reduced following repeated ketamine administration alone versus vehicle (Fig 4B; q<0.05) as reported in Fig 1. However, premature responses following HJC0122 alone did not differ from vehicle in this cohort (Fig 4B; n.s.). Repeated ketamine administration plus HJC0122 reduced premature responses versus vehicle only (Fig 4B; q<0.05). There was no main effect of treatment observed on reinforcers earned (Fig 4C; F_3,30=2.06; n.s.). A main effect of treatment on accuracy was observed (Fig 4D; F_3,30=3.27; q<0.05); repeated ketamine alone was reduced versus HJC0122 alone (Fig 4D; q<0.05). Of note, the overall mean decrease in accuracy following repeated ketamine alone (97.85%) versus HJC0122 alone (99.54%) was not below the 1-CSRT task stability criteria (>80%). There was a main effect of treatment on omissions (Fig 4E; F_3,30=3.77, q<0.05); repeated ketamine combination treatment group was increased versus vehicle (Fig 4E; q<0.05). The overall mean increase in omissions did not exceed task stability criteria. As in Cohort 1, there was no effect of treatment on premature responses (t₂₀=0.64; n.s.), reinforcers earned (t₂₀=0.32; n.s.), accuracy (t₂₀=0.13; n.s.), or omissions (t₂₀=0.07; n.s.) 24 hrs following HJC0122 (1 mg/kg) alone versus vehicle in Cohort 2 (data not shown).

Discussion

Motor impulsivity was blunted in a population of outbred rats by systemic administration of a single and repeated dose of ketamine. Repeated ketamine treatment more robustly lowered motor impulsivity versus single ketamine treatment (~29.1% vs 14.6% decrease, respectively). Although there was a modest increase in omissions after treatment with repeated ketamine only in Cohort 1, all rats still completed 100 trials during each session (data not shown) and the mean omissions was below the task stability criteria. We also observed a decrease in motor impulsivity following administration of an AMPAkine, HJC0122, alone and in combination with ketamine. Taken together, these data support the neurobiological underpinnings of impulsivity are linked to GluR signaling.

Repeated ketamine administration decreased motor impulsivity which was replicated across two separate cohorts of animals. It is possible that repeated ketamine treatment induced a more vigorous rescultping of the synaptic milieu (e.g., AMPAR trafficking, spine formation, dendritic branching) in key brain regions underlying motor impulsivity, but further studies are warranted to validate this claim. Repeated ketamine administration also increased accuracy versus vehicle, but only in Cohort 1. The increase in accuracy is negligible (vehicle=98.25%; repeated ketamine=99.50%) and occurred in the absence of alterations in the animals’ ability to perform the task as indicated by the high level of accuracy. This reflects a limitation of the 1-CSRT task to output any meaningful changes in accuracy, as animals reliably perform near 100% once the task is acquired (Anastasio et al., 2014; Anastasio et al., 2019; Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019). The 5-CSRT task, in which the attentional demands of the animals are substantially increased (Bari et al., 2008), may be better suited to measure the effects of ketamine on accuracy, if any (Benn and Robinson, 2014; Burton and Fletcher, 2012; Higgins et al., 2003; Murphy et al., 2012; Nikiforuk et al., 2015; Smith et al., 2011).

The small number of increased omissions in the 1-CSRT task following repeated ketamine may be interpreted as an indication that nonspecific motivational impairments to perform the task are waning (but see, Nikiforuk et al., 2015), a characteristic that is very common in animals which exhibit lower motor impulsivity (Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019). In line with this interpretation, subanesthetic ketamine administration decreases sign-tracking in a Pavlovian conditioned approach behavior indicative of decreased incentive-motivational value of reward-related cues (Fitzpatrick and Morrow, 2017). Taken together, it is wholly possible that neurochemical alterations induced by ketamine administration, such as enhanced release of serotonin (5-HT), dopamine and glutamate (Ago et al., 2019; Chatterjee et al., 2012; Lindefors et al., 1997), impacts the neural circuitry that encodes incentive salience. Future studies can be designed to interrogate these interactive mechanisms using both behavioral paradigms of incentive motivation in impulsivity and neurochemical analyses.

All ketamine-mediated behavioral outputs were measured after 24 hours, despite its relatively short half-life suggesting that the behavioral effects of ketamine on motor impulsivity are not directly correlated to active NMDAR blockade, which elicits an opposite effect on impulsivity (Benn and Robinson, 2014; Burton and Fletcher, 2012; Higgins et al., 2003; Murphy et al., 2012; Nikiforuk et al., 2015; Smith et al., 2011). Variance in task and demand parameters between the 5-CSRT and 1-CSRT tasks may underlie distinctions between our present results and those reported (Benn and Robinson, 2014; Burton and Fletcher, 2012; Higgins et al., 2003; Murphy et al., 2012; Nikiforuk et al., 2015; Smith et al., 2011), Briefly, acute (3 or 10 mg/kg) and chronic (30 mg/kg x 10 days) ketamine administration decreases correct responses, increases omissions, and increases latency, suggesting that ketamine impairs task performance and attentional set shifting, but not premature responding, in the 5-CSRT task (Nikiforuk et al., 2015). Systemic ketamine at a dose of 6 mg/kg decreases correct responses, increases omissions, and increases latency in a variable ITI 5-CSRT task (Benn and Robinson, 2014). However, these findings were not replicated when a short, fixed stimulus duration was employed (Benn and Robinson, 2014). As our 1-CSRT task paradigm is optimized to determine levels of motor impulsivity (Hamilton et al., 2015), our findings support the hypothesis that the beneficial effects of ketamine on impulsivity are not directly correlated to acute effects but may manifest due to the psychoplastogenic actions of ketamine at the synapse (Li et al., 2010; Li et al., 2011; Ly et al., 2018).

Blockade of the AMPAR increases impulsivity, and is reversed following the co-administration of aniracetam, an AMPAkine (Nakamura et al., 2000). Our goal herein was to capture the window during which the immediate effects of HJC0122 would be expected (Chen et al., 2013; Hu et al., 2018). We report that acute administration of HJC0122 alone decreased motor impulsivity, wherein the highest dose of HJC0122 tested more robustly suppressed motor impulsivity indicating that positive allosteric modulation of AMPAR transmission is sufficient to dose dependently reduce impulsivity. Unlike our findings with ketamine, we did not observe long-lasting effects of our AMPAkine on impulsivity, suggesting a temporal difference in the effects of these two ligands; additional biochemical and behavioral studies are ongoing to address these properties. Interestingly, the lowest dose of HJC0122 (1 mg/kg) did not significantly decrease motor impulsivity in Cohort 2. This may indicate that 1 mg/kg of HJC0122 is a sub-effective dose to reliably decrease motor impulsivity in a population of outbred rats, in which individual differences from cohort to cohort may influence treatment responsiveness (Davis-Reyes et al., 2019; Fink et al., 2015; Sholler et al., 2019). Lastly, 10 mg/kg of HJC0122 increased omissions, but not above the stability criteria. We propose that AMPAR positive allosteric activation does not pharmacologically immobilize the animals, but rather may lessen drive to perform the 1-CSRT task. We also observed that a combination treatment of ketamine plus HJC0122 increased reinforcers earned suggesting an increase “focus” to perform the 1-CSRT task for a reward. Though this is a nuanced distinction, it is important because the state of action readiness and impulse control are interlocked, both within the confines of this task and in the broader sense of impulsive behavior (Frijda, 2010).

Ketamine-mediated alterations to the synaptic milieu is marked by increased cortical and hippocampal GluA1 and GluA2 subunits (Koike and Chaki, 2014; Li et al., 2010; Li et al., 2011). AMPAR antagonism blocks cellular and behavioral responses to ketamine, and AMPAR stimulation is required for the sustained antidepressant properties of ketamine (Beurel et al., 2016; Koike and Chaki, 2014; Li et al., 2010), highlighting the importance of AMPAR to facilitate the effects of ketamine. Excitingly, the combination of both single and repeated ketamine treatments with HJC0122 elicited a substantial decrease in motor impulsivity compared to vehicle. Further, single, but not repeated, ketamine plus HJC0122 lowered motor impulsivity to a greater degree than ketamine and HJC0122 treatments alone. These data suggest that ketamine plus HJC0122 can substantially normalize motor impulsivity at lower therapeutic doses, which is an important consideration, as repeated/higher dosing regimens are often accompanied by increased risk for negative side effects (Molero et al., 2018).

Taken together, our findings support that the neurobiological underpinnings of motor impulsivity are linked to the ionotropic GluR system. While many studies have demonstrated ketamine-induced neurochemical changes in AMPAR expression/signaling and other markers of synaptic plasticity, future studies are needed to assess these changes directly in the context of motor impulsivity, an active area of research in our laboratory. These apparent resculpting events may have translational value in repairing glutamatergic signaling that is disrupted in psychiatric disorders associated with aberrant impulsivity. As high motor impulsivity is thought to arise from hypoactivity in the prefrontal cortex and loss of top-down control of subcortical regions (Anastasio et al., 2019; Dalley et al., 2011), it is reasonable to speculate that subanesthetic doses of ketamine in combination with an AMPAkine normalize motor impulsivity due to enhanced glutamatergic activity between and within key nodes which govern impulsive responding. Animal models that can discern GluR hypofunction/dysfunction within a population of animals, such as the 1-CSRT task, may even serve as tentative behavioral screening tools to predict the effectiveness of drugs that modulate the GluR system as a novel and effective approach for impulse-control disorders.