Adolescent‐onset heavy cannabis use associated with significantly reduced glial but not neuronal markers and glutamate levels in the hippocampus, Grace Blest‐Hopley et al., 2019

Adolescent‐onset heavy cannabis use associated with significantly reduced glial but not neuronal markers and glutamate levels in the hippocampus

Grace Blest‐Hopley | Aisling O’Neill | Robin Wilson | Vincent Giampietro |
David Lythgoe | Alice Egerton | Sagnik Bhattacharyya

Addiction Biology, 2019, e12827, 1-11.

DOI: 10.1111/adb.12827



Cannabis use has been associated with adverse mental health outcomes, the neurochemical underpinnings of which are poorly understood. Although preclinical evidence suggests glutamatergic dysfunction following cannabis exposure in several brain regions including the hippocampus, evidence from human studies have been inconsistent. We investigated the effect of persistent cannabis use on the brain levels of N‐acetyl aspartate (NAA) and myo-inositol, the metabolite markers of neurons and glia, the site of the main central cannabinoid CB1 receptor, and the levels of glutamate, the neurotransmitter directly affected by CB1 modulation. We investigated cannabis users (CUs) who started using during adolescence, the period of greatest vulnerability to cannabis effects and focused on the hippocampus, where type 1 cannabinoid receptors (CBR1) are expressed in high density and have been linked to altered glutamatergic neurotransmission. Twenty‐two adolescent‐onset CUs and 21 nonusing controls (NU), completed proton magnetic resonance spectroscopy, to measure hippocampal metabolite concentrations. Glutamate, NAA, and myoinositol levels were compared between CU and NU using separate analyses of covariance. CU had significantly lower myoinositol but not glutamate or NAA levels in the hippocampus compared with NU. Myoinositol levels in CU positively correlated with glutamate levels, whereas this association was absent in NU. Altered myoinositol levels may be a marker of glia dysfunction and is consistent with experimental preclinical evidence that cannabinoid‐induced glial dysfunction may underlie cannabinoid‐induced memory impairments. Future studies using appropriate imaging techniques such as positron emission tomography should investigate whether glial dysfunction associated with cannabis use underlies hippocampal dysfunction and memory impairment in CUs.

KEYWORDS : Cannabis, Glia, H1MRS, myoinositol, spectroscopy, tetrahydrocannabinol



The long‐term use of cannabis, one of the most widely used illicit substances worldwide,1 has been associated with alterations in a number of cognitive domains.2-5 Cannabis use, particularly adolescent onset regular cannabis use, has also been associated with increased risk of onset6,7 and relapse8-10 of psychosis, as well as greater risk of other adverse mental and behavioural outcomes11-13 and longer hospitalizations. 14 However, the neurochemical underpinnings of these behavioural, cognitive, and mental health effects associated with persistent cannabis use remain unclear.15-17

Delta‐9‐tetrahydro-cannabinol (THC), the main psychoactive component in cannabis, is a partial agonist at the type 1 cannabinoid receptors (CBR1),18 which are mostly expressed in GABAergic and glutamatergic neurons.19 Although animal studies have consistently reported evidence of altered dopamine neuro-transmission or dopamine levels in several regions,20-23 this has not been consistently borne out by human studies.15,24 Investigation of the effects of persistent cannabis use on brain glutamate levels or glutamatergic neurotransmission is also of particular interest. Animal studies have reported dysregulation in glutamatergic synapses of the hippocampus, showing increased25 as well as decreased glutamate receptor expression26 and decreased glutamatergic neurotransmission following THC exposure in the hippocampus26-28 and striatum.29 In addition, to their predominant neuronal location, recent work has also demonstrated that CB1 receptors are also present on glial cells, particularly astrocytes, which have a role in endocannabinoid‐mediated communication between neurons and astrocytes.30-32 Furthermore, preclinical evidence has also emerged that activation of glial CB1 receptors in the hippocampus by exogenous cannabinoids induces glutamate release and activation of NMDA receptors31-33 and there is some evidence that astroglial but not neuronal CB1 receptors mediate the memory impairments, which are associated with cannabis use.34 Given the neuronal and glial localisation of CB1 receptors, the main target of THC and other cannabinoids present in cannabis, one would expect regular cannabis use to affect brain metabolite markers in both neurons (ie, N‐acetyl aspartate [NAA]35) and glia (ie, myoinositol36) as well as the levels of glutamate, the neurotransmitter directly affected by modulation of the neuronal and glial CB1 receptors. In line with these and other preclinical evidence (summarized in Colizzi et al., 2016), a number of human studies have employed proton magnetic resonance spectroscopy (1H‐MRS) to investigate a range of neurochemicals in vivo in cannabis users (CUs). Previous studies of long‐term CU have investigated glutamate levels in the basal ganglia,37,38 frontal and parietal white matter,37,39 and anterior cingulate cortex.40,41 These studies
have reported decreased glutamate levels in CU compared with nonusing controls(NU) in most37,39-41 but not all studies.38 A further recent study investigating a modest sized sample reported no significant
difference in hippocampal glutamate levels between CU and NU.42

1H‐MRS studies have also reported lower levels of myoinositol, an astroglial marker involved in glial function and metabolism,36 in the thalamus of CU compared with controls, although this was only observed in the left hemisphere.4 Significantly decreased levels of myoinositol in CU compared with NU have also been observed in the anterior cingulate cortex41 and in a matrix of voxels including basal ganglia, thalamus, temporal, parietal, and occipital lobe.43 Other studies have reported nonsignificant changes in myoinositol in the frontal white matter39 in CU or in the temporal region in cannabis and ecstasy users.44 Studies investigating NAA, a marker of neuronal integrity,35,45 have reported decreased levels in CU compared with NU in the prefrontal cortex46 and hippocampus,47 and inverse relationship between NAA levels and cannabis use in the inferior frontal gyrus.44 However, not all previous studies in CU have reported the concentration of all of these metabolites in the brain.

Another pertinent consideration is the potential confounding effects of other drugs, which are often also used by CU. Alcohol and other drugs of abuse may affect glutamate and other metabolite levels in the brain.48 Among the other common drugs of abuse in regular CUs, cocaine use has been associated with lower levels of glutamate, 49 nicotine dependence has also been associated with changes to the glutamatergic system,50,51 and alcohol use may affect NAA and myoinositol levels.52,53

Therefore, the main objective of the present study was to investigate the effect of persistent cannabis use on the brain levels of NAA and myoinositol, the metabolite markers of neurons and glia, the site of the main central cannabinoid CB1 receptor, and the levels of glutamate, the neurotransmitter directly affected by CB1 modulation. We used 1H‐MRS focusing on a brain voxel that included parts of the left hippocampus to compare CU and NU by taking into account the effects of comorbid use of other drugs of abuse. We focused on the hippocampus, as CBR1 are expressed in high density in this region,54-56 and as animal studies have shown reduced neuronal density in the hippocampus following prolonged exposure toTHC57,58 and modulation of both glial and neuronal CB1 receptors in the hippocampus have been linked to altered glutamatergic neurotransmission.26-28,32,33 Furthermore, glutamate
in the hippocampus is critical for memory encoding,59,60 which is a key cognitive domains affected by cannabis use.61,62 We purposely focused on those who had started using cannabis during adolescence because of previous evidence that adolescence is a period when the brain is particularly vulnerable to the consequences of cannabis use.63,64

Chronic experimental exposure to high doses of THC has been shown to be associated with significant decrease in astroglial, presynaptic, and postsynaptic but not specific neuronal markers and NMDA
receptor levels in the hippocampus of adolescent rats65 consistent with independent evidence of a specific role of astroglial but not neuronal CB1 receptors in working memory impairment associated with cannabis use.34 Therefore, we predicted that adolescent‐onset regular cannabis use would be associated with selective reduction in astroglial (myoinositol) markers and glutamate levels but not neuronal (NAA) markers in the hippocampus as measured using 1H‐MRS in humans.