Visual Overview
Abstract
The consumption of Δ9-tetrahydrocannabinol (THC)- or cannabis-containing edibles has increased in recent years; however, the behavioral and neural circuit effects of such consumption remain unknown, especially in the context of ingestion of higher doses resulting in cannabis intoxication. We examined the neural and behavioral effects of acute high-dose edible cannabis consumption (AHDECC). Sprague-Dawley rats (six males, seven females) were implanted with electrodes in the prefrontal cortex (PFC), dorsal hippocampus (dHipp), cingulate cortex (Cg), and nucleus accumbens (NAc). Rats were provided access to a mixture of Nutella (6 g/kg) and THC-containing cannabis oil (20 mg/kg) for 10 minutes, during which they voluntarily consumed all of the provided Nutella and THC mixture. Cannabis tetrad and neural oscillations were examined 2, 4, 8, and 24 hours after exposure. In another cohort (16 males, 15 females), we examined the effects of AHDECC on learning and prepulse inhibition and serum and brain THC and 11-hydroxy-THC concentrations. AHDECC resulted in higher brain and serum THC and 11-hydroxy-THC levels in female rats over 24 hours. AHDECC also produced: 1) Cg, dHipp, and NAc gamma power suppression, with the suppression being greater in female rats, in a time-dependent manner; 2) hypolocomotion, hypothermia, and antinociception in a time-dependent manner; and 3) learning and prepulse inhibition impairments. Additionally, most neural activity and behavior changes appear 2 hours after ingestion, suggesting that interventions around this time might be effective in reversing/reducing the effects of AHDECC.
SIGNIFICANCE STATEMENT The effects of high-dose edible cannabis on behavior and neural circuitry are poorly understood. We found that the effects of acute high-dose edible cannabis consumption (AHDECC), which include decreased gamma power, hypothermia, hypolocomotion, analgesia, and learning and information processing impairments, are time and sex dependent. Moreover, these effects begin 2 hours after AHDECC and last for at least 24 hours, suggesting that treatments should target this time window in order to be effective.:
Introduction
Cannabis is consumed mostly through smoking and inhalation (Borodovsky et al., 2016); however, the use of edible cannabis, consumed by oral ingestion, has increased in recent times (Weiss, 2015). Possible reasons for this increase include the fact that there is no exposure to smoke, the convenience and discreteness of edibles, and their longer-lasting effect compared with other forms of intake (Giombi et al., 2018). However, this increase in popularity has raised concerns about accidental overconsumption of edible products. This arises from the delayed onset of subjective effects, which depends on factors such as body weight, metabolism, and eating habits, resulting in difficulty with dose estimation and titration. These factors often lead people to unintentionally consume more edible products than desired, resulting in accidental cannabis intoxication (Giombi et al., 2018). Moreover, the incidence of accidental cannabis intoxication after high-dose edible cannabis has increased among children and pets (Amissah et al., 2022; Tweet et al., 2023), who mistake edible cannabis for other cannabis-free foods (Wang et al., 2016; Lewis et al., 2020). This emphasizes the need to investigate how edible cannabis affects the body after consumption. Most clinical signs of acute high-dose cannabis consumption are neurologic (Janczyk et al., 2004; Claudet et al., 2017), occurring due to the interaction between Δ9-tetrahydrocannabinol (THC) and type 1 cannabinoid receptors (CB1Rs) expressed in brain regions, including the cingulate cortex (Cg), prefrontal cortex (PFC), hippocampus, and nucleus accumbens (NAc). The THC in edible cannabis is metabolized into 11-hydroxy-Δ9-tetrahydrocannabinol (11-OH-THC), which might be more potent than THC (Barrus et al., 2016) and cause stronger and longer lasting effects (Favrat et al., 2005). Oral THC administration also results in high brain THC levels in rats (Hložek et al., 2017), which may contribute to the high incidence of edible cannabis–induced intoxication (Hudak et al., 2015); however, brain THC levels after edible exposure have not been assessed in humans.
Intraperitoneal THC administration decreased gamma power in the rat hippocampus, a brain region critical for learning (Robbe et al., 2006; Jenkins and Khokhar, 2021), possibly through the inhibition of presynaptic neurotransmitter release (Cortes-Briones et al., 2015). Gamma power refers to the magnitude of neural activity within the gamma frequency range (30–150 Hz). Brain oscillations within this range are called gamma oscillations and are necessary for associative learning, and their alteration may therefore lead to learning deficits (Uhlhaas et al., 2008). In our previous work, THC vapor administration decreased gamma power in the rat dorsal striatum, orbitofrontal cortex, and PFC, regions implicated in the cognitive and psychotic effects of THC (Nelong et al., 2019). Although THC inhalation and injection lead to decreased brain oscillations, few studies have investigated similar effects after edible cannabis consumption. Prepulse inhibition (PPI), a behavioral measure of sensory-motor gating (Jones et al., 2014), is impaired in cannabis users (Kedzior and Martin-Iverson, 2007) and in rodents administered with THC (Wegener et al., 2008); deficits in sensory-motor gating have been reported after disruptions in gamma oscillations, suggesting an association between exposure to THC and gamma power alteration (Jones et al., 2014). PPI can be evaluated using the acoustic startle reflex test (ASR) (Swerdlow and Geyer, 1998). Learning deficits can also be evaluated using the active avoidance task (AAT), and in animal models acute intraperitoneal THC administration impairs AAT performance (Mishima et al., 2001). Acute high-dose edible cannabis consumption (AHDECC) may also produce such deficits. Although THC exposure via vapor, injection, and gavage induces specific behavioral effects referred to as the cannabis tetrad, which includes catalepsy, hypolocomotion, hypothermia, and antinociception (Moore and Weerts, 2022), few studies have evaluated similar behavioral effects using edible cannabis (Moore et al., 2021; English et al., 2022). Like humans (Fogel et al., 2017; Sholler et al., 2021; Graves et al., 2023), sex differences in sensitivity to the effects of THC and its metabolism have also been identified in rodents (Craft et al., 2013; Wiley and Burston, 2014); therefore, it may be worth investigating these differences with respect to AHDECC. Based on the above discussion, we hypothesized that AHDECC in rats would lead to sex-dependent disruptions in gamma power, learning, and sensorimotor gating as well as marked cannabinoid tetrad effects.
This is an important topic given that most patients who report to the emergency unit due to cannabis intoxication report consuming edible cannabis (Noble et al., 2019). Understanding the underlying neural mechanisms and the pharmacokinetics of THC via edible consumption could help prevent and/or identify effective treatments for cannabis intoxication. Therefore, the objective of this study was to investigate the effects of AHDECC on neural activity, behavior, and serum and brain THC and 11-OH-THC levels.
Materials and Methods
Animals.
Forty-four 56-day-old Sprague-Dawley rats (body weight: 150–250 g; males: n = 22; females: n = 22) were purchased from Charles River for the study. These rats were divided into two cohorts: cohort 1 (males: n = 6; females: n = 7) and cohort 2 (males: n = 16; females: n = 15). The rats were singly housed and allowed to habituate for 1 week. Male and female rats were housed in the same room but on separate racks, which were adjacent to each other. To motivate rats to consume a mix of either Nutella (a brand of sweetened hazelnut cocoa spread; Ferrero USA Inc., Somerset, NJ) and medium-chain triglyceride (MCT) oil (N-MCT; vehicle for dissolving and diluting THC-containing cannabis oil; Alpha Supreme MCT Oil; Assured Natural Distribution Inc., British Columbia, Canada) or Nutella and THC-containing cannabis oil (N-THC; edible), rats were maintained on 85%–90% free-feeding body weight on standard rodent chow for the entire study period except the first week after arrival and the week after stereotaxic surgery (for rats that underwent surgery). The MCT oil was derived from coconut oil, whereas the THC oil was diluted in olive oil. Rats were maintained on a 12-hour light/dark cycle with lights on at 07:00 hours. Experiments were performed during the light phase. All procedures were approved by the University of Guelph Institutional Animal Care and Use Committee (approval number: 4789). Additionally, experiments were performed in accordance with guidelines set by the University of Guelph Animal Care Committee and guidelines in the Guide to the Care and Use of Experimental Animals.
Preclinical Model Development.
On the seventh day after arrival, each rat was given 3 ml Nutella (density: 1.2 g/ml) for 3 hours (Fig. 1A) in a small petri dish placed on the floor in the home cage to familiarize the animals to the Nutella in their home environment. For all subsequent N-MCT/N-THC accesses, each rat was placed alone in an empty cage with a lid and only corn cobs as bedding material before N-MCT/N-THC was provided in a petri dish. With the exception of the duration of Nutella access, which was 3 hours for the seventh day, all subsequent accesses to N-MCT or N-THC was for only 10 minutes. During this 10-minute period, all the rats consumed all of the N-MCT or N-THC provided. Food restriction began afterward and lasted until at least the 13th day. On the ninth and 11th days, rats received N-MCT (Nutella: 6 g/kg; MCT oil: 20 mg/kg) for 10 minutes. High-dose edible cannabis (N-THC) was prepared similar to N-MCT; however, THC-containing cannabis oil [Five Founders THC Oil (30 mg THC/g oil); Ontario Cannabis Store] was added at 20 mg/kg (adjusted to animal’s body weight; this dose was calculated using the formula provided in the “Rat Equivalent Dose Calculation” section of the Supplemental Material file). On the 13th day, rats either were given 10-minute access to N-MCT or underwent stereotaxic surgery.
Experiment Paradigm.
The first rat cohort was used for electrophysiology experiments. These rats underwent cannabis exposure as described above for the development of the preclinical model for AHDECC except that on day 13 they underwent stereotaxic surgery for multielectrode array implantation (described below). One week after surgery, rats were rehabituated to the N-MCT and tetrad behavior test, and the results were used as baseline (time point 0) for comparison with the tetrad test results from the AHDECC experiment. Afterward, rats were habituated to the recording chamber, where they were tethered to the headstage for 15 minutes. The data recorded during headstage habituation were used as baseline local field potential (LFP) activity. During the THC test, performed 2 days later, rats were given N-THC for 10 minutes. Subsequently, the cannabis tetrad test followed by LFP activity recording 15 minutes later was performed 2, 4, 8, and 24 hours after AHDECC (Fig. 1B). Rats were subsequently euthanized and their brains retrieved for electrode location verification.
The second rat cohort was used for behavioral [active avoidance test (AAT) and acoustic startle response (ASR)] and pharmacokinetics experiments (Fig. 1A). Upon arrival, these rats received Nutella for 3 hours on day 7, received N-MCT for 10 minutes on day 9, and received N-MCT for 10 minutes as well as underwent the baseline cannabis tetrad on day 11. On days 13 and 15, rats received N-MCT and underwent habituation and baseline measurements for PPI in the ASR operant box. Subsequently, rats in each sex group were divided into the control [(ctrl) male: n = 8; female: n = 8] and the test [(test) male: n = 8; female: n = 7] groups. On day 17, ctrl rats received N-MCT, whereas the test rats received N-THC. Two and a half hours later, both groups (ctrl and test) underwent the AAT (Fig. 1B). For the test groups, saphenous blood was also collected 4, 8, and 24 hours after AHDECC. The brains of the test rats were also retrieved after the 24-hour blood collection (Fig. 1B). After 5 days (day 22), the remaining ctrl rats (males: n = 8; females: n = 8) received N-THC and 3.5 hours later underwent the ASR. Similarly, saphenous blood was collected 4, 8, and 24 hours after AHDECC (Fig. 1B). Test rat brains were retrieved after the 24-hour blood collection. In this study, rats were euthanized through exposure to isoflurane first, leading to unconsciousness, followed by exposure to carbon dioxide. For experiments that required brain retrieval, rats were additionally decapitated.
Tetrad Behavior Evaluation.
The tetrad comprised tests for hypothermia, analgesia, catalepsy, and hypolocomotion, similar to that described in a previous study (Moore and Weerts, 2022). To test for hypothermia, the rectal temperature of rats was measured using a microprobe thermometer (model BAT-4; Physitemp Instruments Inc.), which was inserted into the rectum for 5 seconds. To test for analgesia (thermal pain sensitivity), the tails of rats were placed in the groove on the tail flick analgesia meter (Columbus Instruments, Columbus, OH) containing a radiant heat source and latency to tail flexion was recorded. The tail flick analgesia meter was calibrated to an intensity setting of 10. The tails of rats were removed from the groove after 15 seconds (if the rat did not do so by itself) to prevent tissue damage. Catalepsy was evaluated using an open-source automated catalepsy bar apparatus (Luciani et al., 2020). During the test, the two front paws of the rats were placed on the catalepsy bar with the hindlimbs on the floor of the apparatus. The bar (diameter: 1.27 cm) was set to a height of approximately 12 cm from the floor of the apparatus. In the catalepsy test, a cutoff of 15 seconds was used. Due to the absence of catalepsy, the test lasted for approximately 2 seconds. Hypolocomotion was measured over a 10-minute period in an open field box 45 × 45 cm in size. Rats were placed in the middle of the box at the start of the test, and their total distance moved was recorded using EthoVision XT 16.0 video tracking software (Noldus Information Technology).
Two-Way Active Avoidance Test.
The AAT was conducted using a standard two-way shuttle box (model H10-11R-SC; Coulbourn Instruments, Allentown, PA) placed in a ventilated isolation chamber (height: 51 cm; width: 53 cm; length: 80 cm; model H10-24; Coulbourn Instruments) with a grid floor made of stainless-steel bars. A metallic wall partition with a 9 × 9–cm door separated the shuttle box, which contained signal lights, a house light, and an infrared sensor to detect transitions between chambers, into two identical chambers. Scrambled electrical foot shocks were delivered via the grid floor by a precision animal shocker (model H13-15; Coulbourn Instruments). The Graphic State software version 5.9 (Coulbourn Instruments) was used to program the experimental protocol.
Rats were placed individually into the shuttle box and allowed to habituate for 30 seconds. They underwent 70 signaled avoidance trials, with intertrial intervals ranging from 15 to 60 seconds. The trials were subsequently divided into seven blocks of 10 trials each. Each trial consisted of a conditioned stimulus (CS; asynchronously flashing signal and house lights at a frequency of 2 Hz) and an unconditioned stimulus (US; 0.5 mA foot shock). The CS was presented for 10 seconds, and the US was applied during the last 2 seconds of the CS presentation. Moving to the opposite chamber during the CS prevented the delivery of foot shock (avoidance), whereas moving to the opposite chamber during the US (shock) delivery attenuated the foot shock (escape). The number of avoidance and escape behaviors within each block was detected and expressed as percentages per block.
Prepulse Inhibition of the Acoustic Startle Reflex.
The ASR was performed using Med Associates Acoustic Startle Chambers (MED-ASR-PRO1). Each chamber was soundproof and equipped with a ventilation fan, house light, load cell platform, and two speakers for acoustic stimuli (white noise) delivery. The platform was calibrated by adjusting the load cell amplifier gain, which ranged from −2047 to 2047 arbitrary units to 200 arbitrary units with a standard weight of 300 g. In the chamber, rats are restrained using a plexiglass cylinder mounted on the platform. A 70-dB white noise was used as background noise during the experiment.
During the baseline session, rats were placed in the plexiglass cylinders in the startle reflex chambers for 15 minutes while white noise (background noise) was on. Five acoustic startle sounds were played between 5 and 10 minutes. In the test session, rats were allowed 5 minutes to acclimate to the chamber while the 70-dB background noise was on. Subsequently, five consecutive startle stimuli (120 dB) were presented, followed by 50 trials separated into 10 blocks, and then five more startle stimuli. Each block comprised five trials in a randomized order: 1) startle stimuli alone; 2) a 73-dB prepulse and startle stimuli; 3) a 76-dB prepulse and startle stimuli; 4) an 82-dB prepulse and startle stimuli; and 5) no stimulus (only background noise). These prepulse intensities were selected because they did not elicit significant startle reflex when applied alone. The startle stimulus lasted for 40 milliseconds, whereas the prepulse stimulus lasted for 20 milliseconds. The prepulse stimulus was applied 120 milliseconds prior to the onset of the startle stimulus. The intertrial interval ranged from 15 to 30 seconds. In this study, PPI was defined as a decrease in the magnitude of the startle reflex elicited by a startling stimulus when it is preceded by a nonstartling stimulus (prepulse). The PPI for each prepulse intensity was calculated as follows:
Serum and Brain THC and 11-OH-THC Concentration Quantification.
Quantification of serum and brain levels of THC and 11-OH-THC were performed as previously described (Baglot et al., 2021).
Materials
Reference standards of 11-OH-THC and THC and their deuterated internal standards 11-OH-THC-D3 and THC-D3 were purchased from Sigma-Aldrich Canada (Oakville, ON, Canada). Liquid chromatography mass spectrometry (LC-MS)-grade acetonitrile, methanol, and formic acid were purchased from Fisher Scientific (Fisher Chemical Optima grade). Ammonium formate was purchased from Sigma (St. Louis, MO), and water was obtained from the Milli-Q system (Millipore, Bedford, MA).
Serum Sample Preparation.
To extract 11-OH-THC and THC from rat serum, a Captiva enhanced matrix removal lipid (EMR-Lipid) 96-well plate (Agilent, Santa Clara, CA) was used. Briefly, 250 μl acetonitrile (acidified with 1% formic acid) was added to each well, and then 50 μl rat serum and 20 μl (10 μg/ml) internal standard solution were added. After the sample passed through under positive pressure at 3 psi, the extraction plate was washed with 150 μl of a mixture of water/acetonitrile (1:4; v:v) and passed through under gradually increasing pressure up to 15 psi. The effluent was evaporated under nitrogen at 40°C, and the residual was reconstituted with 100 μl mobile phase for subsequent liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis. Calibration standards (2–1000 ng/ml) and quality controls (3 ng/ml and 800 ng/ml) were prepared on the day of analysis by spiking standard working solutions into blank rat serum.
Brain Sample Preparation.
The brain tissue samples were removed from a −80°C freezer and immediately placed into a cooled rat brain matrix. A coronal slice between 6 and 7 mm posterior to the frontal tip of brain was cut, weighed, and placed into the ice-cooled glass tubes for manual homogenization with a glass tissue homogenizer. Acetonitrile (2 ml) and 50 ml internal standard solution (10 μg/ml) were added, and the tissue was homogenized completely. Samples were sonicated in an ice bath for 30 minutes and stored overnight at −20°C. The next day, samples were centrifuged at 1800 rpm at 4°C for 4 minutes, and the supernatant were transferred to a new glass tube and evaporated under nitrogen. The tube sidewalls were washed with 250 μl acetonitrile and evaporated again. The samples were reconstituted in 100 μl of the mobile phase and centrifuged for 20 minutes (15,000 rpm, 4°C), and the supernatants were transferred to the vials. Calibration standards (1–4000 ng/ml) and quality controls (10 ng/ml and 2000 ng/ml) were prepared on the day of analysis by spiking standard working solutions into the mobile phase.
Liquid Chromatography Tandem Mass Spectrometry Method.
Serum and brain concentration of 11-OH-THC and THC were determined using an LC-MS/MS method. The LC separation was achieved on a Thermo Scientific Vanquish Flex UHPLC system. Five microliters of sample extracts was injected and separated on an ACQUITY UPLC BEH C18 Column (1.7 μm, 2.1 mm × 50 mm; Waters, Ireland) connected with a VanGuard UPLC BEH C18 Pre-column (Waters, Ireland). The auto sampler was kept at 4°C, and column temperature was at 35°C. The mobile phase containing A: 10 mM ammonium formate with 0.1% formic acid aqueous solution and B: acetonitrile with 0.1% formic acid was at a flow rate of 400 μl/min under a gradient mode. The gradient conditions were: from 0.1 to 4 minutes ramp from 40% to 95% of mobile phase B, maintain 95% B for 2 minutes, and then ramp back to 40% B. 11-OH-THC and THC were eluted at 3.5, 4.0, and 4.6 minutes, respectively, with a total run time of 7 minutes. MS analysis was conducted with a Thermo Q Exactive Focus Orbitrap mass spectrometer equipped with an Ion Max source in positive electrospray ionization (ESI) mode. The source conditions were optimized as spray voltage 3.5 kV, capillary temperature 300°C, and auxiliary gas heater temperature 425°C.
Data were acquired and processed in parallel‐reaction monitoring (PRM) mode using Thermo Scientific TraceFinder software. In this PRM mode, protonated 11-OH-THC ion [mass/charge (m/z) 331.23] and THC ions (m/z 315.23) were selected as precursors and then fragmented in the higher-energy collisional dissociation cell at collision energy of 20 eV for 11-OH-THC and 25 eV for THC. The resulting MS/MS product ions were detected in the Orbitrap at a resolution of 17,500 (full width at half maximum at m/z of 200) with AGC target set at 1e5. The most abundant fragments from the MS/MS spectra (m/z 313.22 for 11-OH-THC and m/z 193.12 for THC) were selected as the quantifying ions. Other specific fragments, m/z 193.12 for 11-OH-THC and m/z 259.17 for THC, were selected as the confirming ions. The resulting chromatograms were extracted and reconstructed with a mass accuracy of 5 ppm for quantification and confirmation. The optimized MS/MS compound parameters are summarized in Table 1.
Stereotaxic Surgery.
After habituation to the N-MCT, rats underwent stereotaxic surgery for electrode implantation, which begins with anesthesia using isoflurane, as previously described (Thériault et al., 2021; Quansah Amissah et al., 2023). A custom-built, 18-channel microelectrode array was constructed using Delrin templates and stainless-steel wires insulated with polyimide tubing. Four regions were targeted bilaterally with eight electrodes: the medial PFC [anterior-posterior (AP): +3.24 mm; mediolateral (ML): ±0.6 mm; (dorsoventral (DV): −3.8 mm], Cg (AP: +1.9 mm; ML: ±0.5 mm; DV: −2.8 mm), NAc (AP: +1.9 mm; ML: ±1.2 mm; DV: −6.6 mm), and dorsal hippocampus [(dHipp) AP: −3.5 mm; ML: ±2.5 mm; DV: −2.6 mm], as per the coordinates found in the Paxinos rat atlas (6th edition) (Paxinos and Watson, 2009). After the surgery, rats were allowed 1 week to fully recover before experimentation began. All electrodes were confirmed to be accurately placed within the brain regions of interest in the rats at the end of the experiment.
Electrophysiological Data Acquisition and Analysis.
LFP data were acquired at a sampling rate of 1 kHz using a TDT RZ10X multichannel acquisition system (Tucker Davis Technologies, Florida). The data were filtered between frequencies of 0.1 to 300 Hz. A notch filter was used to remove any 60-Hz frequency noise present in the signal during recording. The recording lasted for 15 minutes in awake, freely behaving rats in an open field box. LFP analysis was performed using MATLAB (R2020a; The MathWorks) and routines from the open source Chronux package (Mitra and Bokil, 2008). The data were preprocessed using detrending and denoising routines in Chronux. Subsequently, the data were low-pass filtered to eliminate frequencies higher than 100 Hz. Afterward, band pass filters were used to separate the LFP data into its constituent brain rhythms: delta (0.1–4 Hz), theta (>4–12 Hz), beta (>12–30 Hz), low gamma (>30–59 Hz), and high gamma (>60–100 Hz). Continuous multispectral power was calculated for each of the brain rhythms in each brain region, and coherence between regions was determined.
Statistical Analysis.
The results are presented as means ± standard error of means. The Shapiro-Wilk test was performed to evaluate normality before subsequent statistical tests were performed. The power spectral density for each brain region and the coherence between pairs of brain regions after AHDECC were compared in male and female rats using the two-way repeated measures (RM) analysis of variance (ANOVA). When sex differences were significant, subsequent one-way RM ANOVA was used to analyze data within each sex. The two-way RM ANOVA was also performed to compare the means of results obtained during the tetrad behavioral tests, ASR, and serum THC and 11-OH-THC concentrations between male and female rats. Comparison of the brain THC and 11-OH-THC concentrations in male and female rats was performed using the two-tailed unpaired Student’s t test. The three-way RM ANOVA was performed to compare the means of the percentage avoidance in rats during the AAT, with sex, trial block, and treatment as independent variables and percentage avoidance as the dependent variable. This was followed by a two-way RM ANOVA with treatment and trial block as independent variables based on the lack of sex differences. All ANOVA tests were followed by a post hoc Bonferroni test to correct for multiple comparisons. Statistical analyses were performed using GraphPad version 6.01 (GraphPad Software Inc., La Jolla CA), and statistical significance was set at P < 0.05.
Results
Serum and Brain THC and 11-OH-THC Levels after AHDECC.
There was a significant main effect of sex on serum THC concentration (F(1,29)=21.05, P < 0.0001) after AHDECC. In males serum THC concentrations did not differ at any time point, whereas in females AHDECC increased serum THC concentration at the 4-hour time point compared with the 24-hour time point (P = 0.0463). Moreover, AHDECC increased the serum THC concentration at the 4-hour time point in females compared with males (P = 0.0019) (Fig. 2A). The main effects of sex, time, and their interaction on serum 11-OH-THC levels were significant (F(1,29)=86.00, F(2,58)=23.86, and F(2,58)=21.25, respectively; all P < 0.0001). In males there were no time-related differences in serum 11-OH-THC concentration after AHDECC, whereas in females the 11-OH-THC concentration at the 4-hour time point was higher than at the 8-hour (P = 0.0220) and 24-hour (P < 0.0001) time points (Fig. 2B) after AHDECC. Moreover, females had higher serum 11-OH-THC concentration than males at the 4-hour (P < 0.0001) and 8-hour (P < 0.0001) time points after AHDECC. Comparison using t tests showed that AHDECC increased brain THC and 11-OH-THC levels in females (P = 0.0159 and P = 0.0085, respectively) compared with males at the 24-hour time point (Fig. 2, C and D).
Effects of AHDECC on Gamma Oscillations.
Although AHDECC had some effect on delta, theta, and beta oscillations, the low and high gamma oscillations were the most consistently affected, consistent with previous findings with THC and cannabis (Nelong et al., 2019; Jenkins et al., 2022). Therefore, only the results for the effects of AHDECC on PFC, dHipp, Cg, and NAc gamma oscillations at the different time points (Fig. 3) will be described. The same data have been plotted as percentage change from baseline and presented in Supplemental Fig. 2 to allow for comparison despite the baseline differences.
There was a significant main effect of time on PFC (low gamma) LG power (F(4,96)=17.59, P = 0.0001). AHDECC decreased PFC LG power at all time points compared with baseline in males (P < 0.05) but only at the 2-hour, 4-hour, and 8-hour time points in females (P < 0.0001). At the 24-hour time point, the effect of AHDECC had worn off in female rats, with the PFC LG power returning to baseline levels (Fig. 3, A–D). The main effects of time and its interaction with sex on PFC high gamma (HG) power were significant (F(4,96)=23.42 and F(4,96)=3.169, respectively; all P < 0.05). AHDECC similarly decreased PFC HG power at all time points compared with baseline in males (all P < 0.001) but only at the 2-hour, 4-hour, and 8-hour time points in females (P < 0.0001). Once again, at the 24-hour time point, the effect of AHDECC had worn off in females, causing the PFC HG power to return to baseline levels.
Sex, time, and their interaction had significant main effects on the dHipp LG power (F(1,24)=24.40, F(4, 96)=5.540, and F(4,96)=4.897, respectively; all P < 0.05). dHipp LG power was higher in females at baseline and at the 24-hour time point and similar between sexes at other time points (Fig. 3E). In males, according to the two-way RM ANOVA, AHDECC did not affect dHipp LG power at any time point compared with baseline (Fig. 3E); however, the one-way RM ANOVA revealed that AHDECC decreased dHipp LG power at the 4-hour and 8-hour time points (all P < 0.05; Supplemental Fig. 1). In females, AHDECC decreased dHipp LG power at the 2-hour, 4-hour, and 8-hour time points (P < 0.0001) compared with baseline: however, at the 24-hour time point, the effect of AHDECC had decreased, causing the dHipp LG power to return to baseline levels. The dHipp LG power in females at baseline and the 24-hour time point was higher than that in males. Similarly, the main effects of sex, time, and their interaction on dHipp HG power were significant (F(1,24)=14.38, F(4,96)=5.5, F(4,96)=4.294, respectively; all P < 0.05). In males, according to the two-way RM ANOVA, AHDECC did not affect dHipp HG power at all time points compared with baseline; however, the one-way RM ANOVA revealed that AHDECC decreased dHipp HG power at all time points (P < 0.01; Supplemental Fig. 1). In females, AHDECC decreased dHipp HG power at the 2-hour, 4-hour, and 8-hour time points (P < 0.001) compared with baseline. dHipp HG power was higher in females at baseline and the 24-hour time point (Fig. 3F) compared with males.
Only the main effects of sex and time on the NAc LG power were significant (F(1,24)=8.994 and F(4,96)=4.122, respectively; all P < 0.01). Females had higher NAc LG power than males at baseline (P = 0.0109) and the 24-hour (P = 0.0071) time point (Fig. 3G) after AHDECC. In males, according to the two-way RM ANOVA, AHDECC did not affect the NAc LG power (Fig. 3G); however, the one-way RM ANOVA revealed that AHDECC decreased NAc LG power at the 2-hour, 4-hour, and 8-hour (all P = 0.05) time points compared with baseline (Supplemental Fig. 1C). In females, AHDECC decreased NAc LG power at the 2-hour (P = 0.0130) and 8-hour (P = 0.0009) time points. Sex and time had significant main effects on the NAc HG power (F(1,24)=5.544 and F(4,96)=7.510, respectively; all P < 0.05). Females had higher HG power than males at baseline (P = 0.0357) and the 24-hour (P = 0.0384) time points (Fig. 3H). Similarly, according to the two-way RM ANOVA, AHDECC did not affect HG power at any time point in males (Fig. 3H); however, the one-way RM ANOVA revealed that AHDECC decreased NAc LG power at the 4-hour, 8-hour, and 24-hour (P < 0.01) time points (Supplemental Fig. 1D). In females, AHDECC decreased the HG power at the 2-hour, 4-hour, and 8-hour (all P < 0.001) time points but not the 24-hour time point.
The main effects of sex, time, and their interaction on Cg LG power were significant (F(1,24)=14.69, F(4,96)=8.707, and F(4,96)=4.506, respectively; all P < 0.005). In males, according to both the one-way and two-way RM ANOVA, AHDECC did not affect Cg LG power at any time point (Fig. 3I and Supplemental Fig. 1E, respectively). AHDECC decreased Cg LG power in females at the 2-hour, 4-hour, and 8-hour (all P < 0.0001) time points but not the 24-hour time point. Females had higher Cg LG power than males at baseline (P < 0.0001) and the 24-hour time point (P = 0.0002). No differences were found at the other time points (Fig. 3I). There were significant main effects of sex, time, and their interaction on Cg HG power (F(1,24)=14.92, F(4,96)=7.109, F(4,96)=3.107, respectively; all P < 0.0190). In males, according to the two-way RM ANOVA, AHDECC did not affect Cg HG power at any time point (Fig. 3J); however, the one-way RM ANOVA revealed that AHDECC decreased HG power at the 4-hour, 8-hour, and 24-hour (all P < 0.05) time points (Supplemental Fig. 1F). AHDECC decreased Cg HG power in females at the 2-hour, 4-hour, and 8-hour (all P < 0.05) time points but not the 24-hour time point. Males had lower Cg HG power than females at baseline and the 4-hour and 24-hour (all P = 0.01) time points (Fig. 3J).
The coherence between pairs of brain regions within the gamma frequency ranges was also evaluated. However, unlike the power spectral density analysis, the results (Supplemental Fig. 3) were inconsistent and will not be described.
Effects of AHDECC on Tetrad Behavior.
Although we evaluated the four cannabis tetrad behaviors (Moore and Weerts, 2022), there were no observable cataleptic effects of AHDECC in either sex; therefore, results will only be presented for hypolocomotion, hypothermia, and antinociception.
The main effects of time and its interaction with sex on rectal temperature were significant (F(4,44)=26.91 and F(4,44)=3.295, respectively; all P < 0.05). In both males and females, AHDECC decreased rectal temperatures at the 2-hour, 4-hour, and 8-hour time points compared with baseline (all P < 0.05). However, at the 24-hour time point, although AHDECC decreased rectal temperature in males (P < 0.05), it did not affect the temperature in females (Fig. 4A). Therefore, females had higher 24-hour temperatures than males (P = 0.014).
Time and its interaction with sex had significant main effects on tail flick latency (F(4,44)=11.92 and F(4,44)=5.517, respectively; all P < 0.01). AHDECC increased tail flick latency in males at all time points (all P < 0.01) compared with baseline. In females, AHDECC increased the tail flick latency at the 2-hour (P = 0.0209) and 8-hour (P < 0.0001) time points but not at the 4-hour and 24-hour time points (Fig. 4B). After AHDECC, the tail flick latency was longer in females than males at the 8-hour time point (P = 0.0091) but was not different at the 24-hour time point (P = 0.0743).
The main effects of time and its interaction with sex on distance were significant (F(4,44)=32.90 and F(4,44)=2.453, respectively; all P < 0.05). AHDECC decreased the distance moved compared with baseline at all time points in either sex (all P < 0.01) (Fig. 4C).
The data for body temperature, antinociception, and distance moved have been plotted as percentage change from baseline and presented in Supplemental Fig. 4 to allow for comparison despite the baseline differences between the sexes.
Effects of AHDECC on Active Avoidance Learning and Prepulse Inhibition.
Three-way ANOVA in the AAT revealed significant effects of trial block (F(4.322,109.5)=30.14, P < 0.0001), treatment (F(1,26)=15.94, P = 0.0005), trial block × treatment interaction (F(6,152)=10.29, P < 0.0001), and trial block × sex interaction (F(6,152)=4.318, P = 0.0005) on percentage avoidance. No significant effects were found for sex, sex × treatment interaction, or trial block × treatment × sex interaction on percent avoidance. The two-way ANOVA revealed significant main effects of trial block, treatment, and their interaction on percentage avoidance in male rats (F(6,78)=7.037, F(1,13)=4.701, and F(6,78)=3.396, respectively; all P < 0.050). Male THC rats had lower percentage avoidance than control rats during trial blocks 5 (P = 0.0474) and 7 (P = 0.0098) (Fig. 5A). Similarly, male THC rats had higher percentage escape than male control rats during trial blocks 3 (P < 0.0275), 5 (P < 0.0477), 6 (P < 0.0470), and 7 (P < 0.0077) (Supplemental Fig. 5A). There were significant main effects of trial block, treatment, and their interaction on percentage avoidance in female rats (F(6,74)=27.47, F(1,13)=12.79, and F(6,74)=7.552; all P < 0.05). Female THC rats had lower percentage avoidance than control rats during trial blocks 5 (P < 0.0001), 6 (P = 0.0004), and 7 (P = 0.0001) (Fig. 5B). Similarly, female THC rats had higher percentage escape than female control rats during trial blocks 5 (P < 0.0297), 6 (P < 0.0190), and 7 (P < 0.0390) (Supplemental Fig. 5B).
No sex-specific differences in the effect of sound intensity on percentage PPI were observed (results not shown). THC rats had lower averaged percentage PPI compared with control rats in either sex (male: P = 0.0209, female: P = 0.0037) (Fig. 5C; significant main effects – treatment: F(1,14)=13.92, P = 0.0022).
Discussion
We established and characterized a rat model of AHDECC showing marked sex differences in the observed behavioral, pharmacokinetic and neural outcomes. Female rats exhibited higher serum and brain THC and 11-OH-THC levels compared with male rats after AHDECC despite consuming the same N-THC dose. The study also revealed sex- and time-dependent decreases in dHipp, Cg, and NAc gamma power over a 24-hour period starting 2 hours after AHDECC. We also observed time-dependent changes in cannabinoid tetrad behaviors and disruptions in active avoidance and PPI.
The observed differences in serum THC levels between sexes may be partly due to the rapid redistribution of THC to fatty tissues, which tends to be more pronounced in male rats (Rubino and Parolaro, 2011), due to their higher body fat content (Cortright et al., 1997). Consistent with previous studies in rats (Ruiz et al., 2021), we found sex differences in serum 11-OH-THC levels, suggesting differences in THC metabolism in rodents. This difference can be attributed to the sexually dimorphic expression of cytochrome P450 (P450) enzymes (Gerges and El-Kadi, 2023), leading to the preferential metabolism of THC into 11-nor-9-carboxy-Δ9-THC (11-COOH-THC) in males and into 11-OH-THC in females (Torrens et al., 2020, 2022), contributing to the higher serum 11-OH-THC levels in females. Although the exact mechanisms may differ in humans, women also exhibit higher plasma levels of both THC and its metabolite compared with men (Nadulski et al., 2005; Matheson et al., 2020; Sholler et al., 2021). Furthermore, we also observed higher brain THC and 11-OH-THC levels in females, potentially linked to their higher serum THC levels. Relatedly, male rat brains may be protected against THC through increased expression of blood-brain barrier proteins (including claudin-5) (Torrens et al., 2020, 2022). Since the rat brain also expresses P450 enzymes capable of metabolizing THC (Watanabe et al., 1988), which are known to play an important role in mediating the effects of centrally acting drugs (Khokhar and Tyndale, 2014), it is plausible that brain-expressed P450s in female rat brains metabolize more THC into 11-OH-THC, resulting in higher brain 11-OH-THC levels (Wiley and Burston, 2014). Both serum and brain 11-OH-THC levels decreased gradually over 24 hours, possibly due to the excretion of 11-OH-THC (Wall et al., 1983) or rapid oxidation into 11-COOH-THC (Andrenyak et al., 2017). The higher brain and serum THC and 11-OH-THC levels in female rats (Craft et al., 2017) may also explain the observed sex differences in cannabis effects observed in rats (Craft et al., 2013; Wiley and Burston, 2014).
LFP power reflects neural synchrony, with higher power indicating greater synchrony and vice versa (Luo and Guan, 2018). Gamma power and synchrony are modulated by parvalbumin-expressing (PV) interneurons (Volman et al., 2011), which in turn are modulated by cholecystokinin-expressing (CCK) interneurons (Karson et al., 2009). These CCK interneurons are the only cortical and hippocampal CB1R-expressing interneurons (Tsou et al., 1999; Bodor et al., 2005). THC binding to CB1Rs on CCK interneurons impacts PV interneuron activity, leading to disruptions in gamma oscillations and a subsequent decrease in gamma power. This phenomenon may explain the decreased gamma power observed after AHDECC in our rats and in previous studies assessing the effects of vaporized THC and cannabis (Nelong et al., 2019; Jenkins and Khokhar, 2021; Jenkins et al., 2022). The observation of decreased gamma power in the dHipp, Cg, NAc, and PFC is not surprising given the high densities of CB1Rs in these areas. In our study, significant effects of AHDECC on gamma power manifested after 2 hours, consistent with the peak subjective behavioral effects of edible cannabis observed between 1.5 and 3 hours after ingestion (Vandrey et al., 2017), as well as our pharmacokinetic findings. Additionally, we noted sex differences in gamma power in the NAc, Cg, and dHipp regions. These differences may partly stem from the higher estrogen levels in female rats (Bao et al., 2017). Elevated estrogen levels enhance estrogen binding to estrogen β receptors expressed on PV interneurons, leading to increased firing, greater inhibition, and increased gamma activity (Sohal et al., 2009; Clemens et al., 2019). This mechanism may clarify the elevated gamma power in female rats, potentially contributing to their quicker recovery from AHDECC-induced suppression, as evidenced by the absence of AHDECC effects at the 24-hour time points alongside potential pharmacokinetic differences or tolerance. In future studies, we will use machine learning–based approaches to uncover LFP features that contribute meaningfully to the behavioral outcomes observed here (Dwiel et al., 2019).
The similarity in AHDECC effects on antinociception (tail flick latency) between male and female rats observed in our study aligns with findings in rats after THC vapor administration (Moore et al., 2021). Pain relief is one of the most commonly cited reasons for medical cannabis use (Reinarman et al., 2011). The periaqueductal gray, a brain region that contains CB1R-expressing somatodendritic structures and presynaptic terminals, is implicated in pain modulation (An et al., 1998; Wilson-Poe et al., 2012) and may play a role in the decreased pain sensitivity that we observed after AHDECC. Additionally, AHDECC induced hypothermia in both male and female rats, likely due to the interaction between THC and the preoptic area of the hypothalamus, known for its role in temperature regulation and high density of CB1Rs (Ripamonte et al., 2020). Interestingly, in female rats, antinociception and body temperature had returned to baseline levels by the 24-hour time point, likely due to tolerance development to the high serum and brain THC and 11-OH-THC levels or the differential time course of effects between THC and its metabolite. Previous studies have shown similar sex differences, with female rats exhibiting greater tolerance to the antinociceptive effects of THC compared with male rats (Wakley et al., 2014). This might also explain the earlier return of the gamma power in the females to baseline levels, despite the higher brain and serum THC and metabolite levels. Although no sex differences in hypolocomotion were observed in our study, hypolocomotion, previously reported (Rodríguez de Fonseca et al., 1998), may stem from the interaction between THC, 11-OH-THC, and CB1Rs on cerebellar basket cells (Patel and Hillard, 2001). Interestingly, AHDECC-induced tetrad effects began around the 2-hour time point, similar to the observed neural effects of AHDECC in this study, coinciding with the time point at which cognitive and psychomotor deficits were observed in humans after edible cannabis ingestion (Schlienz et al., 2020). This suggests that interventions to reverse the effects of AHDECC should target the 2-hour time window. Notably, the present study revealed a dramatic sex difference in serum and brain THC and 11-OH-THC levels, sometimes not paralleled by sex differences in behavior. One possible explanation is a potential ceiling effect with the behavioral measures, likely due to our use of a relatively high dose of THC. This aligns with findings in humans, where high-potency cannabis concentrates produce significantly higher levels of THC (nearly 2×) but do not elicit higher subjective effects or impairments compared with flower (Bidwell et al., 2020).
The inability of rats that underwent AHDECC to avoid the foot shock, leading to high escape rates during the AAT, may be attributed to the decreased hippocampal gamma power, which is necessary for learning associations between stimuli (Uhlhaas et al., 2008). This finding aligns with previous studies that found learning deficits in cannabis users (Skosnik et al., 2012) and THC-treated rats (Mishima et al., 2001). Although low-dose THC administration (0.3–3 mg/kg) had no effect on PPI (Malone and Taylor, 2006), a higher dose (10 mg/kg) disrupted PPI (Nagai et al., 2006), similar to our observations in rats that underwent AHDECC (20 mg/kg). This suggests a potential dose-dependent effect of THC on PPI. A previous study reported decreased PPI in rats after direct CB1R agonist infusion into the hippocampus and PFC (Wegener et al., 2008). The study concluded that CB1R activation in the hippocampus and PFC modulates neural GABA and glutamate release, influencing the activities of PPI-mediating structures like the NAc and ventral tegmental area (Wegener et al., 2008). This conclusion aligns with our finding of decreased PFC and dHipp gamma power after AHDECC, potentially impacting NAc activity. Moreover, a recent clinical study (Karunakaran et al., 2024) found that intoxicating oral THC doses in adults reduced spectral power in the PFC. This aligns with findings from our animal model after AHDECC, supporting the use of electrophysiological signals as a potential translational biomarker for aiding the diagnosis and treatment of cannabis intoxication.
Although we present several novel findings, it is important to acknowledge several limitations. We could only administer the N-THC once, as rats voluntarily consume the Nutella but abstain from it after N-THC exposure. Additionally, blood sampling to determine serum THC and 11-OH-THC levels began at the 4-hour time point, immediately after the AAT and ASR. This was done to ensure that the behavioral effects of AHDECC remained unaffected by the blood sampling; however, our assessment may have missed the peak THC and 11-OH-THC levels. Moreover, although male rats metabolize THC preferentially into 11-COOH-THC, we did not evaluate serum and brain 11-COOH-THC levels but will do so in future studies. Although we chose to use full-spectrum THC oil to closely mimic human edible cannabis products, we could not assess the impact of additional cannabinoids and terpenes present in the cannabis oil on the measured outcomes. Future studies exploring various cannabis oils with varying cannabinoid and terpenoid levels may shed light on these effects. Furthermore, we did not assess the estrous cycle phase in female rats, which may have impacted the behavioral and electrophysiological effects of AHDECC. Future studies will be adequately powered to assess the impact of estrous cycle on AHDECC. We also did not evaluate multiple doses, preventing us from establishing a dose-response relationship for this study, which could have significantly increased the impact of this study. A dose-response assessment would have also helped clarify the potential ceiling effects observed in our studies, where sex differences in brain and plasma levels of THC and its metabolite are not mirrored in behavioral studies. Additionally, the use of a single dose precludes our ability to discern whether the observed sex differences are related to differences in THC potency or efficacy.
In the present study, the absence of a control group for some experiments makes it challenging to conclusively attribute the observed effects (e.g., hypolocomotion) solely to AHDECC, as other factors such as habituation could potentially contribute. However, a comparison with baseline effects alongside previous studies reporting similar effects (Rodríguez de Fonseca et al., 1998) suggests that the effects may indeed be attributable to AHDECC. Future studies will incorporate adequate controls for all experiments to ensure accurate interpretation of our results. Furthermore, although the two contexts (active avoidance and the acoustic startle reflex) and apparatuses are distinct, prior experience in the active avoidance task could potentially influence performance in the acoustic startle reflex test (Götz and Janik, 2011). This limitation in our study design should be addressed in future studies by employing study designs that minimize or control for (via randomized crossover designs) interference between behavioral tests. Additionally, the catalepsy findings might be influenced by the timing of the test, which might have been too early to detect measurable levels of catalepsy, considering that previous studies show a more pronounced catalepsy response 330 minutes after oral THC dosing (Moore and Weerts, 2022). We opted for MCT oil as the vehicle because of its tasteless and odorless nature; however, since coconut oil is slightly more calorically dense compared with olive oil, this might have introduced a confounding factor not accounted for in our design. Although the rats consumed all provided N-MCT and N-THC throughout the study, suggesting that the caloric content of the vehicle did not affect their consumption, potential unknown interactions cannot be ruled out. Lastly, although the proposed model will be valuable, these findings should be extended to younger rats to better model AHDECC in children.
In conclusion, AHDECC decreased gamma power, produced hypolocomotion, hypothermia, antinociception, and learning deficits, and impaired PPI. Although neural and behavioral effects from cannabis consumption are expected, this study is the first to demonstrate these neural effects after voluntary consumption of edible cannabis, the most commonly cited cannabis product associated with cannabis intoxication, especially in children and pets. Interestingly, the neural and behavioral effects of AHDECC began after 2 hours, coinciding with the period when peak effects of edible cannabis on cognition were reported. This suggests that administering interventions during this time window may be key to reversing AHDECC effects. Moreover, the observed sex differences in gamma power and serum and brain THC and 11-OH-THC levels after AHDECC suggest sex differences in the effects of THC and its metabolism, which should be considered in the development of effective treatments.
Data Availability
The neural power spectral density and coherence data as well as the raw behavioral data that support the findings of this study are openly available in the Open Science Framework at DOI: 10.17605/OSF.IO/3KFZX. All other data, including data for behavior and brain and serum THC and 11-OH-THC concentrations, are contained within the manuscript and Supplemental Material.
Authorship Contributions
Participated in research design: Quansah Amissah, Kayir, Urban, Khokhar.
Conducted experiments: Quansah Amissah, Talhat, Hassan.
Contributed new reagents or analytic tools: Gu, Johnson.
Performed data analysis: Quansah Amissah, Kayir, Gu, Johnson.
Wrote or contributed to the writing of the manuscript: Quansah Amissah, Kayir, Urban, Khokhar.
Footnotes
- Received October 26, 2023.
- Accepted May 13, 2024.
This work was supported by a Natural Sciences and Engineering Research Council Alliance grant [Grant ALLRP 549529] (to J.Y.K.) and a MITACS Accelerate Fellowship [Grant IT27597] (to R.Q.A. and J.Y.K.) in partnership with Avicanna Inc. The funders, including Avicanna Inc., had no role in the design, data collection and analysis, decision to publish, or preparation of the manuscript.
The authors have declared a conflict of interest. Dr. Urban is an employee of Avicanna Inc., during which time she has received stock options. None of the other authors have any disclosures.
↵This article has supplemental material available at jpet.aspetjournals.org.
Abbreviations
- AAT
- active avoidance task
- AHDECC
- acute high-dose edible cannabis consumption
- AP
- anterior-posterior
- ASR
- acoustic startle reflex test
- CB1R
- type 1 cannabinoid receptor
- CCK
- cholecystokinin-expressing
- Cg
- cingulate cortex
- 11-COOH-THC
- 11-nor-9-carboxy-Δ9-THC
- CS
- conditioned stimulus
- Ctrl
- control rats
- dHipp
- dorsal hippocampus
- DV
- dorsoventral
- HG
- high gamma
- LC-MS/MS
- liquid chromatography tandem mass spectrometry
- LFP
- local field potential
- LG
- low gamma
- MCT
- medium-chain triglyceride
- ML
- mediolateral
- m/z
- mass/charge
- NAc
- nucleus accumbens
- N-MCT
- Nutella and MCT oil
- N-THC
- Nutella and THC-containing cannabis oil
- 11-OH-THC
- 11-hydroxy-THC
- P450
- cytochrome P450
- PFC
- prefrontal cortex
- PPI
- prepulse inhibition
- PRM
- parallel‐reaction monitoring
- PV
- parvalbumin-expressing
- RM
- repeated measures
- Test
- test rats
- THC
- Δ9-tetrahydrocannabinol
- US
- unconditioned stimulus
- Copyright © 2024 by The Author(s)
This is an open access article distributed under the CC BY-NC Attribution 4.0 International license.