Familial abnormalities of endocannabinoid signaling in schizophrenia, Dagmar Koethe et al., 2018

Familial abnormalities of endocannabinoid signaling in schizophrenia

Dagmar Koethe, Franziska Pahlisch, Martin Hellmich , Cathrin Rohleder , Juliane K. Mueller, Andreas Meyer-Lindenberg , E. Fuller Torrey, Daniele Piomelli and F. Markus Leweke,




Objectives : Epidemiological and experimental evidence suggests that the endocannabinoid system plays a pathophysiological role in schizophrenia. This is reflected by elevated cerebrospinal levels of the endocannabinoid anandamide in schizophrenia and its initial prodromal states.

Methods : We analyzed plasma concentrations of anandamide, 2-arachidonoyl-sn-glycerol, palmitoylethanolamide and oleoylethanolamide from 25 twin pairs discordant for schizophrenia, six discordant for bipolar disorder and eight healthy twin pairs to determine hereditary traits.

Results : Twin pairs discordant for schizophrenia or bipolar disorder had significantly higher levels
of anandamide and palmitoylethanolamide compared to healthy twins (both P<0.002). Nonaffected twins discordant for schizophrenia, who developed a psychotic disorder within 5 years follow-up showed lower anandamide (P¼0.042) and 2-arachidonoyl-sn-glycerol levels (P¼0.049) than twins who remained healthy.

Conclusions : We suggest that the protective upregulation of endocannabinoid signalling reflects
either a hereditary trait or mirrors a modulating response to genetically influenced cerebral function
involving, e.g., other neurotransmitters or energy metabolism.

KEYWORDS : Psychosis; monozygotic twins; genetic risk; anandamide (AEA); endocannabinoids

1. Introduction

The involvement of the endocannabinoid system (ECS) in the pathogenesis of schizophrenia was initially suggested by the discovery on an association between cannabis use and an increased risk of developing
psychotic symptoms (Zammit et al. 2002). This link was strengthened by two sets of observations. First,
additional epidemiological data indicated that cannabis use before the age of 15 might increase the risk of psychotic symptoms (Arseneault et al. 2002). Second, experimental studies showed that the major psychoactive component of cannabis, delta-9-tetrahydrocannabinol, induced psychotic symptoms in healthy subjects and worsened those in schizophrenia patients (Leweke et al. 2000; D’Souza et al. 2005; Koethe et al. 2006). Further studies on the underlying neurobiological mechanisms have shown that cerebrospinal fluid (CSF) levels of the endocannabinoid anandamide (AEA), which are elevated in acute schizophrenic subjects, are inversely correlated to psychopathology (Leweke et al. 1999; Giuffrida et al. 2004). This was unlikely due to exposure to cannabis since AEA in CSF was negatively related to the use of this drug in schizophrenia (Leweke et al. 2007). Importantly, changes in CSF AEA levels were already noticeable in individuals showing At Risk Mental States (ARMS), a potential early stage of psychosis or schizophrenia. ARMS individuals with lower levels of AEA showed a higher risk for an early transit into psychosis. This upregulation of AEA in ARMS suggests an adaptive if not protective role of AEA in early psychosis (Koethe, Giuffrida, et al. 2009). Up to now it is not clear, if the alteration of AEA is hereditary and may be regarded as a trait and/or risk for the widely genetic group of schizophrenic disorders. Compared to schizophrenia, investigations of the ECS in bipolar affective disorders are limited. No alterations of endocannabinoid levels in CSF and serum were observed in 22 patients suffering affective disorders (Giuffrida et al. 2004). Furthermore, no differences in cannabinoid type-1 receptor (CB1R) density were found in the anterior cingulate cortex of bipolar patients using immunohistochemistry (Koethe et al. 2007). However, molecular genetic- and populationbased studies indicate that schizophrenia and bipolar affective disorder overlap in genetic variants and symptoms (Lichtenstein et al. 2009). Due to these similarities of both disorders a modulation of the ECS might be conceivable. The twin model is a powerful approach for determining the relative contributions of heredity apart
from common and unique environmental influences on variation of disorder-associated abnormalities.
Despite evidence of heterogeneity across studies, meta-analytic results from twin studies of schizophrenia are consistent with the view of schizophrenia as a complex trait that results from genetic and environmental influences. Sullivan et al. (2003) found evidence for substantial additive genetic effects by using a multi-group twin model. The point estimate of heredity in liability to schizophrenia was 81% (95% confidence interval, 73–90%). Notably, there was consistent evidence across these studies for common or shared environmental influences on liability to schizophrenia (point estimate 11% (95% confidence interval, 3–19%). A discordant twin study offers an optimal experimental design to distinguish trait-related risk factors from those related to disorder state. In such a study, twin pairs are recruited on the basis that one twin is affected by a psychiatric disorder while the co-twin is not clinically affected. By contrasting the unaffected, healthy co-twins and completely healthy controls, heritable,
as either genetic or common environmental effects in the co-twins, can be identified. If a given measure is largely related to heritable traits, but not to disorder state, it should be exhibited in both discordant twins. Many intermediate phenotypes of schizophrenia indeed show this pattern, e.g., with respect to brain structure. Brans et al. (2008) showed significant decreases over time in whole-brain, frontal and temporal lobe volumes in patients suffering schizophrenia and their unaffected co-twins compared
to control twins. Koolschijn et al. (2008) reported a decrease in hypothalamus volume in patients and discordant twin pairs compared to healthy controls. Higher within-twin pair similarities in monozygotic compared to dizygotic twin pairs, suggest that the hypothalamic volume might be partly genetically controlled. Hirvonen and Hietala (2011) found an increased caudate D2-receptor density in unaffected monozygotic co-twins compared to unaffected dizygotic co-twins and healthy control twins. This was associated with poor performance on cognitive tasks related to schizophrenia vulnerability in the entire sample. Conversely, a parameter related to disorder state should only differ between discordant twins, as well as to healthy controls. For example, Hirvonen and Hietala (2011) found no evidence for an association between increased genetic risk for schizophrenia and altered D1/D2-receptor balance in the striatum. Here, we investigated whether monozygotic twins discordant for schizophrenia or bipolar disorder display differences in the plasma levels of the endocannabinoids AEA and 2-arachidonoyl-sn-glycerol (2-AG), as well as the endocannabinoid-related fatty acid ethanolamides (FAEs) oleoyl-ethanolamide (OEA) and palmitoylethanolamide (PEA). We hypothesized that plasma levels of AEA and related FAEs are altered in affected schizophrenia twins compared to healthy controls and presumed that   such an alteration is absent in nonaffected monozygotic co-twins.