Introduction
Cannabis use is a risk factor for psychosis (Colizzi & Bhattacharyya, Reference Colizzi, Bhattacharyya, Thompson and Broome2020; Schoeler et al., Reference Schoeler, Monk, Sami, Klamerus, Foglia, Brown and Bhattacharyya2016a, Reference Schoeler, Petros, Di Forti, Pingault, Klamerus, Foglia and Bhattacharyyab). Its main psychoactive ingredient delta-9-tetrahydrocannabinol (Δ9-THC) impacts the endocannabinoid system through its partial agonist effect at the cannabinoid receptor type 1 (CB1) in the brain (Pertwee, Reference Pertwee2008). Δ9-THC can stimulate neuronal firing of mesolimbic dopamine neurons and elevate striatal dopamine levels (Sami, Rabiner, & Bhattacharyya, Reference Sami, Rabiner and Bhattacharyya2015), possibly via CB1-mediated dysregulation of glutamate signaling (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a), resulting in a psychosis-like state (Colizzi, Weltens, McGuire, Van Oudenhove, & Bhattacharyya, Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyya2019b).
In addition to their predominant neuronal location, CB1 receptors are also located on glial cells, particularly astrocytes (Stella, Reference Stella2010). Glia have a prominent role in a number of brain processes, including maintaining homeostasis and modulating synaptic activity, immune response, and nervous system restoration after an insult (Garcia-Segura & McCarthy, Reference Garcia-Segura and McCarthy2004; Kurosinski & Götz, Reference Kurosinski and Götz2002). Preclinically, chronic cannabinoid-induced CB1 receptor activation has been shown to disrupt glial function (Rubino et al., Reference Rubino, Realini, Braida, Guidi, Capurro, Viganò and Parolaro2009), in turn affecting glutamate signaling and behavior (Han et al., Reference Han, Kesner, Metna-Laurent, Duan, Xu, Georges and Zhang2012). Myo-inositol is an astroglial marker whose levels, especially in the medial prefrontal/anterior cingulate cortex (ACC), have also been shown to be reduced in psychosis patients (Das et al., Reference Das, Dey, Sabesan, Javadzadeh, Théberge, Radua and Palaniyappan2018) and otherwise healthy cannabis users (Blest-Hopley et al., Reference Blest-Hopley, O'Neill, Wilson, Giampietro, Lythgoe, Egerton and Bhattacharyya2019; Prescot, Locatelli, Renshaw, & Yurgelun-Todd, Reference Prescot, Locatelli, Renshaw and Yurgelun-Todd2011). Glial cells in the medial prefrontal cortex have also been shown to be involved in behavioral responses to stress and in determining differential susceptibility to stress (Bonnefil et al., Reference Bonnefil, Dietz, Amatruda, Wentling, Aubry, Dupree and Liu2019).
Rodent research and human studies also support a role for the endocannabinoid system in regulating hypothalamic-pituitary-adrenal (HPA) axis activity (Appiah-Kusi et al., Reference Appiah-Kusi, Leyden, Parmar, Mondelli, McGuire and Bhattacharyya2016), with downstream effects on the developing brain as well as neuronal plasticity in the adult brain (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010). In preclinical studies, exogenous cannabinoid administration results in the release of corticotrophin-releasing hormone (CRH), initiating a cascade of events along the HPA-axis, culminating in the release of glucocorticoids from the adrenal cortex into the bloodstream (Pagotto, Marsicano, Cota, Lutz, & Pasquali, Reference Pagotto, Marsicano, Cota, Lutz and Pasquali2006). Similarly, acute exposure to Δ-9-THC in healthy individuals induces a dose-dependent increase in cortisol plasma levels (Ranganathan et al., Reference Ranganathan, Braley, Pittman, Cooper, Perry, Krystal and D'Souza2009). Glucocorticoids act through glucocorticoid receptors to modulate glutamate signaling (Hillard, Beatka, & Sarvaideo, Reference Hillard, Beatka and Sarvaideo2016) and behavior (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010). HPA axis dysregulation with higher baseline cortisol levels and blunted cortisol reactivity to acute stress has been reported in psychosis (Borges, Gayer-Anderson, & Mondelli, Reference Borges, Gayer-Anderson and Mondelli2013) and, less consistently, in chronic cannabis users (Cservenka, Lahanas, & Dotson-Bossert, Reference Cservenka, Lahanas and Dotson-Bossert2018). Emerging evidence also suggests a role for blunted cortisol reactivity in exacerbating anxiety responses to social stress in people at risk of psychosis (Appiah-Kusi et al., Reference Appiah-Kusi, Petros, Wilson, Colizzi, Bossong, Valmaggia and Bhattacharyya2020).
Even though perturbations of the glial and glucocorticoid (HPA-axis) systems have been reported in psychosis (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010) and in cannabis use (Blest-Hopley et al., Reference Blest-Hopley, O'Neill, Wilson, Giampietro, Lythgoe, Egerton and Bhattacharyya2019; Cservenka et al., Reference Cservenka, Lahanas and Dotson-Bossert2018; Prescot et al., Reference Prescot, Locatelli, Renshaw and Yurgelun-Todd2011), their biological relevance for the psychosis-inducing effects of Δ-9-THC remains unclear. Despite a role for glial-glucocorticoid interactions in the pathological or neuroprotective responses to stressful insults (Pearson-Leary, Osborne, & McNay, Reference Pearson-Leary, Osborne and McNay2015), with glial cells critically regulating stress responses (Pearson-Leary et al., Reference Pearson-Leary, Osborne and McNay2015) and glucocorticoids, in turn, acting on glial cells (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010), the precise nature of the interaction between these two systems remains unclear, especially in experimental conditions (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010). Characterizing them in the same subjects may help elucidate how their alteration may relate to differential sensitivity to the emergence of a Δ-9-THC-induced transient psychosis-like state.
Not everyone experiences psychotic symptoms under the influence of Δ9-THC (Colizzi et al., Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyya2019b) and differential sensitivity to its acute psychotomimetic effects may be moderated by genetic (Bhattacharyya et al., Reference Bhattacharyya, Atakan, Martin-Santos, Crippa, Kambeitz, Prata and McGuire2012) and neurophysiological factors (Bhattacharyya et al., Reference Bhattacharyya, Sainsbury, Allen, Nosarti, Atakan, Giampietro and McGuire2018). Although preclinical and human evidence suggests that Δ9-THC acutely modulates glial function (Han et al., Reference Han, Kesner, Metna-Laurent, Duan, Xu, Georges and Zhang2012) and levels of cortisol (Ranganathan et al., Reference Ranganathan, Braley, Pittman, Cooper, Perry, Krystal and D'Souza2009), the key neuroendocrine marker of the HPA-axis, whether interindividual differences in sensitivity to the acute psychotomimetic effects of Δ9-THC is associated with differential effects of Δ9-THC on glial function and HPA-axis response has never been tested. Using data from our previous study reporting the acute effects of Δ9-THC on psychotomimetic symptoms (Colizzi et al., Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyya2019b) and brain glutamate levels (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a), we examined whether sensitivity to the acute and transient psychotomimetic effects of Δ9-THC is moderated by the acute effects of a single dose of Δ9-THC on myo-inositol levels as well as time-course of change in day-time cortisol levels. Myo-inositol levels were measured in the ACC, the key brain region where myo-inositol alterations have been associated with cannabis use (Blest-Hopley et al., Reference Blest-Hopley, O'Neill, Wilson, Giampietro, Lythgoe, Egerton and Bhattacharyya2019; Prescot et al., Reference Prescot, Locatelli, Renshaw and Yurgelun-Todd2011) and psychosis (Das et al., Reference Das, Dey, Sabesan, Javadzadeh, Théberge, Radua and Palaniyappan2018). As medial prefrontal/ACC myo-inositol is found to be reduced in psychosis (Das et al., Reference Das, Dey, Sabesan, Javadzadeh, Théberge, Radua and Palaniyappan2018) and cannabis use (Blest-Hopley et al., Reference Blest-Hopley, O'Neill, Wilson, Giampietro, Lythgoe, Egerton and Bhattacharyya2019; Prescot et al., Reference Prescot, Locatelli, Renshaw and Yurgelun-Todd2011), and medial prefrontal glial cells mediate stress response (Bonnefil et al., Reference Bonnefil, Dietz, Amatruda, Wentling, Aubry, Dupree and Liu2019), we predicted that Δ9-THC administration will result in a more pronounced myo-inositol reduction in those sensitive to compared to those not sensitive to the acute psychotomimetic effects of Δ9-THC. As Δ9-THC administration induces a cortisol increase (Ranganathan et al., Reference Ranganathan, Braley, Pittman, Cooper, Perry, Krystal and D'Souza2009) and higher cortisol levels associated with a blunted cortisol reactivity to stress have been reported in psychosis (Borges et al., Reference Borges, Gayer-Anderson and Mondelli2013) and cannabis use (Cservenka et al., Reference Cservenka, Lahanas and Dotson-Bossert2018), we further predicted Δ9-THC administration to result in a more blunted decline in morning serum cortisol levels in those sensitive to compared to those not sensitive to the psychotomimetic effects of Δ9-THC.
Methods
Procedure and participants
The experimental procedure, psychopathological assessment, image acquisition, magnetic resonance spectroscopy (1H-MRS) quantification, and statistics have been previously detailed (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a; Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyyab). Briefly, we employed a double-blind, randomized, placebo-controlled, repeated-measures, within-subject design, with counterbalanced order of drug administration, using an established protocol (Bhattacharyya et al., Reference Bhattacharyya, Atakan, Martin-Santos, Crippa, Kambeitz, Prata and McGuire2012; Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a). Sixteen right-handed, English-speaking, healthy (confirmed by physical examination) participants (seven males), with no personal or family history of psychiatric illness in first-degree relatives and no history of alcohol abuse, nicotine dependence, or illicit drug use, were assessed on two different occasions separated by at least a 2-week interval, with each session preceded by intravenous administration of Δ9-THC (1.19 mg/2 ml) or placebo. A dose of 1.19 mg was used, as previous work has suggested that an intravenous dose range 0.015–0.03 mg/kg body weight is consistently associated with the induction of psychotomimetic symptoms (Radhakrishnan, Wilkinson, & D'Souza, Reference Radhakrishnan, Wilkinson and D'Souza2014). Prior to each study visit, participants were advised to remain fast and get at least 6–8 h sleep overnight, and to refrain from smoking for 4 h, to take caffeine for 12 h, and alcohol for 24 h. Also, they had been abstinent from cannabis for at least 6 months before the first study visit and were advised to abstain from using any substance throughout the duration of the study. Abstinence was confirmed on each study day by a negative urinary drug screen for most commonly used drugs. All female participants had a negative pregnancy test; also, all of them were consistently using a reliable contraceptive method, apart from a single subject who underwent both study visits in the first week of the menstrual cycle. Blood samples for serum cortisol measurements (2 ml) were collected in serum-separating tubes at the beginning of the study visit (~9.30 a.m.) and at 20 min (~10.50 a.m.) and 2.5 h (~13.00 p.m.) after drug administration (~10.30 a.m.). Similarly, psychopathological ratings were recorded by an expert clinical researcher, using the Positive and Negative Syndrome Scale (PANSS), a well-established scale used for measuring symptom severity of individuals with psychosis (Kay, Fiszbein, & Opler, Reference Kay, Fiszbein and Opler1987), at the same time points before and after drug challenge. Starting at 11.00 a.m., 1H-MRS spectra (Point RESolved Spectroscopy – PRESS; TE = 30 ms; TR = 3000 ms; 96 averages) were acquired on a 3 Tesla MR system in the left caudate head, ACC, and hippocampus, employing the standard GE PROBE (proton brain examination) sequence with CHESS (Chemically Selective Suppression) water suppression, and analyzed with LCModel version 6.3-1L. Metabolite levels were corrected for voxel tissue content. Voxel segmentation and spectral quality have been reported before as well as the values scaled to creatine (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a). A detailed description of the image acquisition and 1H-MRS quantification is provided in the Supplementary methods.
Statistical analyses
Sensitivity to the psychotomimetic effects of Δ9-THC was based on the manifestation of clearly detectable primary symptoms of psychosis (⩾2-point increase in PANSS delusions, hallucinations, unusual thought content, suspiciousness, and grandiosity items), as drawn from previous factor analytic work (Fulford et al., Reference Fulford, Pearson, Stuart, Fisher, Mathalon, Vinogradov and Loewy2014) as well as previous work to characterize acute sensitivity to Δ9-THC (Bhattacharyya et al., Reference Bhattacharyya, Sainsbury, Allen, Nosarti, Atakan, Giampietro and McGuire2018). Data were normally distributed. Repeated measures ANOVA was used to estimate the main effect of Δ9-THC on myo-inositol as well as changes in cortisol levels from baseline to the post-drug injection time points. Independent t tests were used to estimate whether the effect of Δ9-THC on myo-inositol levels as well as changes in cortisol levels from baseline to the post-drug injection time points differed between subjects sensitive to and those not sensitive to the psychotomimetic effects of Δ9-THC. A correlation analysis was used to explore the association between ACC myo-inositol and cortisol levels.
Ethics
The study was approved by the Joint South London and Maudsley (SLaM) and Institute of Psychiatry, Psychology & Neuroscience (IoPPN) National Health Service Research Ethics Committee (PNM/13/14-38), and the investigators had a license to use Δ9-THC for research purposes.
Results
Study participants and psychotomimetic effects of Δ9-THC
Study participants had a mean age of 24.44 (s.d.: 4.29) years. All except three (with self-described mixed ethnic origin) of the volunteers were white Europeans. They had 16.94 ± 2.84 years (M ± s.d.) of education.
As expected, administration of Δ9-THC was associated with acute induction of transient psychotic symptoms. Qualitative (Colizzi et al., Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyya2019b) and quantitative (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a) descriptions of psychotic symptoms under the influence of Δ9-THC in the study sample have been extensively reported before. Eleven subjects (69%) were identified as sensitive to the psychotomimetic effects of Δ9-THC as determined on the basis of ⩾2 point increase in the relevant PANSS items (as described in Methods) (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a). They had a 5.91 (±4.18) point increase in the primary symptoms of psychosis compared to a 0.6 point increase (±0.55) for the remaining subjects (drug effect, t (14) = 4.13, p = 0.002).
Myo-inositol and cortisol levels as a function of sensitivity to the psychotomimetic effects of Δ9-THC
As previously reported (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a), there was no main effect of acute drug administration on myo-inositol levels in the left ACC. However, the Δ9-THC-induced change in myo-inositol levels differed significantly between those sensitive to (Δ9-THC minus placebo; M = −0.251, s.d. = 1.242) and those not sensitive (M = 1.615, s.d. = 1.753) to the psychotomimetic effects of the drug (t(14) = 2.459, p = 0.028; Fig. 1). In those sensitive to the psychotomimetic effects of Δ9-THC, there was a decrease in myo-inositol levels following Δ9-THC administration compared to the placebo condition, while there was an opposite effect of Δ9-THC administration on myoinositol levels in those not experiencing the psychotomimetic effects.
Acute effect of Δ9-THC on myo-inositol measures in the left Anterior Cingulate Cortex as a function of sensitivity to psychotomimetic effects. The effect of drug administration on myo-inositol levels was obtained subtracting myo-inositol levels post-placebo injection from myo-inositol post-Δ9-THC injection (Δ9-THC minus placebo); the effect of the drug was then compared between individuals sensitive to the psychotomimetic effects of Δ9-THC and individuals not sensitive to the psychotomimetic effects of Δ9-THC by independent t test; *, significant effect at p < 0.05; error bars show mean and standard deviations; PLB, placebo; Δ9-THC, delta-9-tetrahydrocannabinol.
As expected, compared to placebo (baseline minus 2.5 h post-drug injection; M = 202.33, s.d. = 138.43), Δ9-THC administration (M = 43.47, s.d. = 251.75) resulted in a more blunted cortisol decrease over the study period (F(1, 14) = 5.463, p = 0.035). Further, compared to placebo, the Δ9-THC-induced change in cortisol levels over the study period (baseline minus 2.5 h post-drug injection) differed significantly between those sensitive to (Δ9-THC change minus placebo change; M = −275.4, s.d. = 207.519) and those not sensitive (M = 74.2, s.d. = 209.281) to the psychotomimetic effects of the drug (t(13) = 3.068, p = 0.009; Fig. 2).
Acute effect of Δ9-THC on cortisol levels over time as a function of sensitivity to psychotomimetic effects. Change in cortisol levels over time was obtained subtracting cortisol levels 2.5 h post-drug injection from cortisol levels at baseline (baseline minus 2.5 h post-drug injection); the effect of drug administration on change in cortisol levels over time was obtained subtracting the change in cortisol levels over time in the placebo condition from the change in cortisol levels over time in the Δ9-THC condition (Δ9-THC change minus placebo change); the effect of the drug was then compared between individuals sensitive to the psychotomimetic effects of Δ9-THC and individuals not sensitive to the psychotomimetic effects of Δ9-THC by independent t test; **, significant effect at p < 0.01; error bars show mean and standard errors; PLB, placebo; Δ9-THC, delta-9-tetrahydrocannabinol.
As sensitivity to the Δ9-THC-induced transient psychosis-like symptoms seemed to be associated with a reduction of ACC myo-inositol levels as well as a disruption of the physiological decrease of cortisol over time, we further examined their association. We found that the lower the ACC myo-inositol values under Δ9-THC the more blunted was the cortisol decrease over time following Δ9-THC administration (Δ9-THC change minus placebo change), (rS:0.468; p = 0.039; Fig. 3).
Association between the acute effect of Δ9 THC on myo-inositol measures in the left Anterior Cingulate Cortex and the changes in cortisol levels post-drug injection. Change in cortisol levels post-drug injection was obtained subtracting cortisol levels 2.5 h post-drug injection from cortisol levels 20 min post-drug injection (20 min post-drug injection minus 2.5 h post-drug injection); the effect of drug administration on change in cortisol levels post-drug injection was obtained subtracting the change in cortisol levels post-drug injection in the placebo condition from the change in cortisol levels post-drug injection in the Δ9-THC condition (Δ9-THC change minus placebo change); the effect of the drug was then correlated with myo-inositol values under Δ9-THC in the whole group by Spearman's correlation; p 1-tailed; PLB, placebo Δ9-THC, delta-9-tetrahydrocannabinol.
Discussion
To the best of our knowledge, this is the first human study to investigate whether sensitivity to the acute psychotomimetic effects of Δ9-THC moderates its acute effects on both myo-inositol and cortisol levels under experimental conditions. After a single dose of Δ9-THC, subjects developing transient psychosis-like symptoms had lower myo-inositol levels in the ACC and a disruption of the physiological cortisol decrease with cortisol levels remaining even higher for a prolonged time. Also, such Δ9-THC-induced alterations appeared to be interrelated.
Glutamate signaling has been previously shown to be disrupted following acute Δ9-THC administration (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a; Mason et al., Reference Mason, Theunissen, Hutten, Tse, Toennes, Stiers and Ramaekers2019), extending data of chronic cannabis exposure (Colizzi, McGuire, Pertwee, & Bhattacharyya, Reference Colizzi, McGuire, Pertwee and Bhattacharyya2016) and providing a mechanistic neurochemical explanation underlying the acute psychotomimetic effects of Δ9-THC (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a) that resemble those observed in patients with psychosis (Colizzi et al., Reference Colizzi, Weltens, McGuire, Van Oudenhove and Bhattacharyya2019b). Glial cells have been involved in inflammation, homeostasis, neurotransmission, and signal transduction, including glutamate metabolism (Verkhratsky, Steardo, Parpura, & Montana, Reference Verkhratsky, Steardo, Parpura and Montana2016). Consistent, experimental preclinical evidence indicates that Δ9-THC can impair synaptic function via a CB1-mediated reduction of glutamate uptake by glutamate transporters in astrocytes (Chen et al., Reference Chen, Zhang, Fan, Teng, Wu, Yang and Chen2013), resulting in sustained elevation and accumulation of extracellular glutamate (Han et al., Reference Han, Kesner, Metna-Laurent, Duan, Xu, Georges and Zhang2012). Also, complementary independent human evidence suggests lower levels of myo-inositol in the context of chronic cannabis use directly correlating with glutamate levels, where such relationship is absent in non-using subjects (Blest-Hopley et al., Reference Blest-Hopley, O'Neill, Wilson, Giampietro, Lythgoe, Egerton and Bhattacharyya2019). It is therefore biologically plausible that glia cells might exert a modulating role in the psychotomimetic effects of cannabis. Consistent with this, these results suggest that loss of glial function may underlie sensitivity to the acute psychotomimetic effects of Δ9-THC, whereas such Δ9-THC-mediated disruption of glial function is absent in subjects not developing psychosis-like symptoms under its influence.
The normal diurnal rhythm of cortisol involves a peak in the levels of the glucocorticoid observed after awakening which then declines progressively (Fries, Dettenborn, & Kirschbaum, Reference Fries, Dettenborn and Kirschbaum2009). Evidence also suggests that acute Δ-9-THC administration may reduce the normal diurnal decline, resulting in cortisol levels that are either the same as or higher than the baseline (Ranganathan et al., Reference Ranganathan, Braley, Pittman, Cooper, Perry, Krystal and D'Souza2009). Results from this study extend such findings, indicating that Δ-9-THC administration may interfere with the normal diurnal decline of cortisol only in those subjects experiencing transient psychotomimetic effects under its influence. In fact, while subjects who did not develop Δ-9-THC-induced psychosis-like symptoms showed a pattern of cortisol decline that was indistinguishable between the Δ-9-THC and placebo conditions, subjects experiencing transient psychotomimetic effects under the influence of Δ-9-THC presented with cortisol levels that were higher than baseline up to 2.5 h following Δ-9-THC exposure, while the same individuals displayed the usual declining pattern under placebo condition. Administration of Δ-9-THC may affect HPA-axis activation by multiple mechanisms. One potential explanation is through activation of presynaptic CB1 receptors on glutamatergic inputs onto corticotrophin-releasing hormone (CRH)-expressing neurons in the paraventricular nucleus of the hypothalamus, in turn affecting glutamate release (Di, Malcher-Lopes, Halmos, & Tasker, Reference Di, Malcher-Lopes, Halmos and Tasker2003; Di, Malcher-Lopes, Marcheselli, Bazan, & Tasker, Reference Di, Malcher-Lopes, Marcheselli, Bazan and Tasker2005). Consistent with this, preclinical evidence indicates that high-affinity synthetic CB1 agonists mimic and occlude the inhibitory effect of glucocorticoids on glutamate release, suggesting a modulatory cross-talk between glucocorticoids and endocannabinoids that leads to reduced excitation of the paraventricular nucleus neurons, with relevance to the central control of stress response (Di et al., Reference Di, Malcher-Lopes, Halmos and Tasker2003).
A large body of evidence suggests that glial cells and stress hormones exert synergistic effects and their joint disruption is relevant to several pathological conditions of the central nervous system (Murphy-Royal, Gordon, & Bains, Reference Murphy-Royal, Gordon and Bains2019). On one hand, converging evidence from animal and human postmortem studies strongly supports a role of deficits in astrocyte density and function in the limbic regions of the brain in the pathology of stress and glucocorticoid overproduction (Fuchs & Flügge, Reference Fuchs and Flügge2003). On the other, independent preclinical evidence indicates that sustained exposure to stress results in the depletion of gliogenesis in limbic regions including the hippocampus and prefrontal cortex (Czéh et al., Reference Czéh, Müller-Keuker, Rygula, Abumaria, Hiemke, Domenici and Fuchs2007). Further, evidence indicates that excessive glutamatergic activity may be deleterious for neurons and contribute to the atrophy of apical dendrites seen under stress conditions in animal models (Conrad, Reference Conrad2006). Astrocytes modulate glutamate availability through the activity of the glial glutamate transporter (GLT-1) and enzyme glutamine synthetase (GS) (Vardimon, Ben-Dror, Avisar, Oren, & Shiftan, Reference Vardimon, Ben-Dror, Avisar, Oren and Shiftan1999). Complementing such evidence, additional studies indicate that glial GLT-1 (Zschocke et al., Reference Zschocke, Bayatti, Clement, Witan, Figiel, Engele and Behl2005) and GS (Vardimon et al., Reference Vardimon, Ben-Dror, Avisar, Oren and Shiftan1999) are highly regulated by glucocorticoids and affected by chronic stress (Autry et al., Reference Autry, Grillo, Piroli, Rothstein, McEwen and Reagan2006; Reagan et al., Reference Reagan, Rosell, Wood, Spedding, Muñoz, Rothstein and McEwen2004). Glucocorticoids also inhibit glucose uptake in the brain, whose energy is essential to the costly task of high-affinity glutamate reuptake (Ritchie, De Butte, & Pappas, Reference Ritchie, De Butte and Pappas2004), potentially increasing the vulnerability to glutamatergic dysfunction (Li, Yang, & Lin, Reference Li, Yang and Lin2018). Altogether, evidence suggests a complex relationship between glucocorticoids and glia, with implications for the glutamate signaling (Jauregui-Huerta et al., Reference Jauregui-Huerta, Ruvalcaba-Delgadillo, Gonzalez-Castañeda, Garcia-Estrada, Gonzalez-Perez and Luquin2010) and the bidirectional communications between neurons and glia (Hinwood, Morandini, Day, & Walker, Reference Hinwood, Morandini, Day and Walker2012). Neuron–glia interactions are critical for preserving the homeostatic environment in the central nervous system (Szepesi, Manouchehrian, Bachiller, & Deierborg, Reference Szepesi, Manouchehrian, Bachiller and Deierborg2018) and evidence suggests that they are disrupted in psychiatric disorders such as schizophrenia (Laskaris et al., Reference Laskaris, Di Biase, Everall, Chana, Christopoulos, Skafidas and Pantelis2016). Our results extend such evidence, indicating a relationship between the severity of the disrupting effects of Δ9-THC on glial function and cortisol diurnal variations underlying sensitivity to the psychotomimetic effects of the drug.
We employed a study design that allowed us to avoid the confounding effect of cannabis withdrawal, dependence, or intoxication, as we only recruited participants with a lifetime history of minimal cannabis use, abstinent from cannabis for a minimum period of 6 months prior to visit, and with a negative urine drug screen for the presence of Δ9-THC at the time of the visit. As all participants had also a lifetime history of negligible use of other common substances of abuse including alcohol, tobacco, and stimulants, it is unlikely that the observed effects on glia and cortisol are attributable to them. Moreover, a minimum interval of 14 days between the two study visits enabled us to avoid any potential carryover effects of Δ9-THC, whose elimination half-life ranges between 18 h and 4.3 days (Kelly & Jones, Reference Kelly and Jones1992). However, a number of limitations also needs to be considered. Foremost, caution is warranted in light of the modest sample size studied here. Further, out of a total of 16 participants, only five were deemed as not sensitive to the acute psychotomimetic effects of Δ9-THC. However, we employed a priori previously reported criteria (Bhattacharyya et al., Reference Bhattacharyya, Sainsbury, Allen, Nosarti, Atakan, Giampietro and McGuire2018) to determine sensitivity to the psychotomimetic effects of Δ9-THC. Nevertheless, we were able to identify significant differences in Δ9-THC-induced changes in myo-inositol and cortisol levels in minimal cannabis users sensitive to the acute and transient psychotomimetic effects of Δ9-THC compared to those not sensitive, using a within-subject repeated measures experimental design. The strict inclusion criteria, while being advantageous in terms of a controlled sample, also warrant caution in terms of generalizability of the results of the study to the wider context of recreational cannabis use. Similarly, administering Δ9-THC intravenously to study participants allowed much more consistent inter-individual Δ9-THC blood levels compared to other routes of administration (D'Souza et al., Reference D'Souza, Perry, MacDougall, Ammerman, Cooper, Wu and Krystal2004), but may affect the generalizability of these results to the effects of recreational cannabis use whose main routes of administration are smoking, inhaling, and swallowing (Baggio et al., Reference Baggio, Deline, Studer, Mohler-Kuo, Daeppen and Gmel2014). Further, even though evidence supports relatively high reproducibility and test-retest reliability of measures of neurochemicals such as glutamate and myo-inositol at 3 Tesla in healthy subjects (Gasparovic et al., Reference Gasparovic, Bedrick, Mayer, Yeo, Chen, Damaraju and Jung2011), we did not examine test-retest reliability of the MRS measures for the regions investigated within the context of the present study. However, using a within-subject design allowed us to avoid the confounding effect of between-subject differences in neurochemicals as well as cortisol measures.
In conclusion, it is likely that the transient psychosis-like effects of Δ9-THC are the result of its effects not only on glutamate signaling (Colizzi et al., Reference Colizzi, Weltens, McGuire, Lythgoe, Williams, Van Oudenhove and Bhattacharyya2019a) but also differential effects linked to sensitivity to acute psychotomimetic effects on other neurobiological systems including glial function and stress response (Colizzi & Bhattacharyya, Reference Colizzi and Bhattacharyya2018), in line with the evidence for multiple neurobiological abnormalities accounting for the complex symptom profile of psychosis (Lisman et al., Reference Lisman, Coyle, Green, Javitt, Benes, Heckers and Grace2008). In this regard, Δ9-THC-induced abnormalities may be interrelated, with the Δ9-THC-induced glial loss, indexed as myo-inositol reduction, correlating with the perturbation of the normal diurnal decline in cortisol levels. Future studies in larger samples are needed to systematically examine such relationships as well as to replicate the present results. Further, whether differential sensitivity to the effects of Δ9-THC or cannabis use on glial function and neuroendocrine markers linked to the stress response also underlie differential sensitivity to the association between psychotic disorder and regular cannabis use remains to be tested. Future multimodal neuroimaging studies, potentially incorporating a more direct measure of glial function as indexed using Positron Emission Tomography imaging, therefore need to integrate longitudinal information to track the trajectory of change in the bidirectional neuron-glia communication as well as other potential biomarkers associated with cannabis use and examine the relationship of such change with adverse mental health outcomes such as psychosis associated with cannabis use. Elucidating the astrocyte–glucocorticoid interaction and its role in the glutamate signaling may be important in comprehending how the defects in the bidirectional neuron-glia communication could contribute to cannabis-associated psychosis as well as neuropsychiatric conditions developed in response to other environmental or genetic insults.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0033291720003827
Acknowledgements
The research leading to these results received funding from the People Programme (Marie Curie Actions) of the European Union's Seventh Framework Programme (FP7/2007-2013) under REA grant agreement no. 608 765 granted to Dr Weltens, Prof. Van Oudenhove, Prof. Bhattacharyya, and Prof. Steve Williams. Sagnik Bhattacharyya has been supported by a NIHR Clinician Scientist Award (NIHR CS-11-001) and the MRC (MR/J012149/1). Lukas Van Oudenhove has been supported by a KU Leuven Special Research Fund (BOF, Bijzonder Onderzoeksfonds). Nathalie Weltens has been supported by a KU Leuven BOF Post-Doctoral Mandate. All other authors report no financial support.
The funders had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
The authors acknowledge infrastructure support from the NIHR Mental Health Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.
Conflict of interest
The authors reported no biomedical financial interests or potential conflicts of interest.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.
Clinical trial registration
The present study is not a clinical trial of an IMP; it is a non-therapeutic mechanistic study performed among healthy volunteers to better understand human pathophysiology with reference to the effect of cannabis psychoactive ingredient delta-9-tetrahydrocannabinol (Δ9-THC).
