Antidepressant effects of psilocybin
in animal models of depression
Literature thesis 5th of December 2018 F. M. Nelissen, 11787899
University of Amsterdam, the Netherlands Supervisor
Assoc. prof. B. Elfving Århus University, Denmark
Co-assessor Dhr. prof. dr. Lucassen
Swammerdam Institute for Life Sciences, the Netherlands Program
MSc in Brain and Cognitive Sciences University of Amsterdam, the Netherlands
Table of contents
TABLE OF CONTENTS ... 2 ABSTRACT ... 3 LIST OF ABBREVIATIONS ... 4 INTRODUCTION ... 5 CURRENT ANTIDEPRESSANT TREATMENTS ARE INADEQUATE ... 5 PSILOCYBIN IS A PROMISING NEW FAST-ACTING ANTIDEPRESSANT ... 5 SEVERAL PROBLEMS IN PATIENT STUDIES PREVENT CONCLUSIVE RESULTS REGARDING THE ANTIDEPRESSANT EFFECTS OF PSILOCYBIN ... 6 NEW GOAL: INVESTIGATING THE MECHANISM OF ACTION OF PSILOCYBIN IN ANIMAL MODELS OF DEPRESSION ... 7 1. THE NEUROBIOLOGY OF DEPRESSION ... 9 INTRODUCING THE NEUROTROPHIC HYPOTHESIS OF DEPRESSION ... 9 NEUROTROPHIC FACTOR BDNF ORCHESTRATES NEUROPLASTICITY ... 10 BDNF REGULATES HIPPOCAMPAL NEUROGENESIS ... 11 CHRONIC STRESS AFFECT NEUROPLASTICITY, RESULTING IN DEPRESSION ... 12 THE NEUROTROPHIC HYPOTHESIS EXPLAINS THERAPEUTIC DELAY OF SSRIS ... 12 BDNF, ORCHESTRATOR OF NEUROPLASTICITY, AS POSSIBLE TARGET OF PSILOCYBIN ... 13 2. PSYCHEDELICS AFFECT NEUROPLASTICITY ... 14 SEROTONERGIC SIGNALLING INDUCES BDNF EXPRESSION ... 14 CLASSICAL PSYCHEDELICS AFFECT NEUROPLASTICITY THROUGH BDNF ... 15 BEHAVIOURAL EFFECTS OF CLASSICAL PSYCHEDELICS IN ANIMAL MODELS OF DEPRESSION ... 16 ROLE OF BDNF EXPRESSION IN ANTIDEPRESSANT EFFECTS OF PSILOCYBIN IS UNKNOWN ... 17 3. ANIMAL MODELS OF DEPRESSION ... 18 CRITERIA TO ASSESS THE VALIDITY OF ANIMAL MODELS ... 18 ANIMAL MODELS OF DEPRESSION ... 20 Behavioural read-outs of depression ... 20 Species, strain and genetics ... 20 Environmental factors leading to depression ... 23 ALL DESCRIBED MODELS ARE SUITABLE TO STUDY BDNF EXPRESSION ... 25 4. RESEARCH PROPOSAL ... 26 DISCUSSION ON USABILITY OF THE DESCRIBED MODELS OF DEPRESSION ... 26 1.0 MG/KG PSILOCYBIN IS PREDICTED TO INDUCE ANTIDEPRESSANT EFFECTS ... 27 EXPERIMENT 1: BDNF EXPRESSION POTENTIALLY INCREASED BY PSILOCYBIN ... 28EXPERIMENT 2: PSILOCYBIN POTENTIALLY RESCUES CMS-INDUCED BDNF EXPRESSION DECREASE ... 29
DISCUSSION ... 32
Abstract
Depression is a major contributor to the overall global burden of disease, but available therapies are inadequate. Psilocybin, the classic psychedelic found in magic mushrooms, is a potential new antidepressant drug that appears to be fast-acting and enduring after a single dose in randomized-controlled trials. Yet, the molecular mechanisms contributing to these effects are unknown. This thesis aimed to investigate how animal models of depression can contribute to understanding the putative antidepressant effects of psilocybin. The current available literature was consulted to answer this question. As depression is a multifactorial disorder, there are multiple hypotheses regarding its etiology. One of the theories is the neurotrophic hypothesis of depression, where brain-derived neurotrophic factor (BDNF) is a key player. Several classical psychedelics induce changes in BDNF expression, possibly explaining their suggested antidepressant effects. Today, it is generally accepted that genetic susceptibility and stress contribute to the development of depression. Based on that, a number of rodent models are used in depression research. Among them, the genetic models, the Flinders Sensitive Line (FSL) rats and the Wistar Kyoto (WKY) rats, as well as the stress-induced models such as chronic mild stress (CMS) and the Learned Helplessness (LH) paradigms are the most widely used. Until now only one study has explored the antidepressant effects of psilocybin in an animal model of depression (Nichols & Hibicke, 2018). They have reported an enduring antidepressant effect 32 days after one dose of psilocybin in WKY rats. To address this lacuna, we suggested two experiments investigating the changes in BDNF expression and depressive-like behaviour after one dose of psilocybin in the genetic models and the CMS model. A better understanding of the antidepressant effects of psilocybin will hopefully contribute to the development of other fast-acting, enduring, non-hallucinogenic antidepressants.
List of abbreviations
5-MeO-DMT 5-methoxy-N,N-dimethyltryptamine
AC Adenyl cyclase
BDNF Brain-derived neurotrophic factor DFP Diisopropylfluorophosphate CMS Chronic mild stress
CREB cAMP response element-binding protein
DAG Diacylglycerol
DMT Dimethyltryptamine
DOI 2,5-Dimethoxy-4-iodoamphetamine
DR Dorsal raphe nucleus
ELISA Enzyme-linked immunosorbent assay
EPM Elevated plus maze
FRL Flinders Resistant Line FSL Flinders Sensitive Line
FST Forced swim test
HPA axis Hypothalamic–pituitary–adrenal axis IP3 Inositol 1,4,5-trisphosphate
LH Learned helplessness
LSD D-lysergic acid diethylamide MAPK Mitogen-activated protein kinase mTOR Mammalian target of rapamycin
NAc Nucleus accumbens
OFT Open field test
PFC Prefrontal cortex
PI3K Phosphatidyl inositol-3 kinase
PKA Protein kinase A
PKB Protein kinase B
PKC Protein kinase C
PLCγ Phospholipase Cγ
qPCR Quantitative polymerase chain reaction
REM Rapid eye movement
SERT Sodium-dependent serotonin transporter SNRI Serotonin-norepinephrine reuptake inhibitor SPT Sucrose preference test
SSRI Selective serotonin reuptake inhibitor TRD Treatment-resistant depression
UTR Untranslated region
WHO World Health Organisation
Introduction
Current antidepressant treatments are inadequate
Depression is a common mental disorder currently with more than 300 million afflicted people (WHO, 2018). Twice as many women as men are depressed (Hankin, 2009) and in 2017, World Health Organisation (WHO) proclaimed that depression was the leading cause of disability worldwide, and a major contributor to the overall global burden of disease (WHO, 2018). Current treatment options appear far from ideal, as the overall cumulative remission rate of the first-line treatments, including drugs and cognitive behavioural therapy, is 67% (Rush et al., 2006). First-line recommendations for antidepressant drugs are mostly selective serotonin reuptake inhibitors (SSRIs) and serotonin-norepinephrine reuptake inhibitors (SNRIs) (Kennedy et al., 2016), despite of two major disadvantages.
First of all, the antidepressant response is delayed. The effects in the first two weeks of treatment are assumed to be placebo effects (Bauer et al., 2013). The persistent drug response occurs mainly after three to four weeks, but maximum symptom reduction may take up to eight to ten weeks (Bauer et al., 2013). In light of the high risk of suicide that is associated with depression (Bhimji & Dulebohn, 2018), this slow response to treatment is undesirable. Instead, a fast-acting therapy with a rapid time of onset is necessary.
Secondly, at least 30% of depressed patients do not respond to first-line treatment with an antidepressant (Bauer et al., 2013). In the United Kingdom, 55% of patients that have taken antidepressants for at least six weeks continue to experience depressive symptoms and are diagnosed with treatment-resistant depression (TRD) (Thomas et al., 2013). Following an inadequate response to treatment, several strategies are possible. These include increasing the dose, switching to another antidepressant or combining the antidepressant with another antidepressant, other agents that enhance the antidepressant efficacy, psychotherapeutic intervention or non-pharmacological biological therapies (Bauer et al., 2013). Yet, the evidence substantiating these strategies is minimal (Wiles et al., 2018). So, to enable treatment of all depressed patients, new therapy strategies with higher efficiency rates must be developed.
Psilocybin is a promising new fast-acting antidepressant
Recently, there has been renewed interest in the possibly antidepressant properties of psychedelic drugs. The psychedelics that are most investigated are the classical psychedelics: psilocybin, dimethyltryptamine (DMT) and D-lysergic acid diethylamide (LSD) (review by Rucker, Iliff, & Nutt, 2017). These compounds all have substantial serotonin 2A receptor agonist properties
(Carhart-Harris & Goodwin, 2017). A systematic review based on 151 clinical trials investigating antidepressive, anxiolytic, and anti-addictive effects of these classical psychedelics performed between 1990 and 2015 concluded that these compounds appear useful pharmacological tools in treating drug addiction, anxiety and mood disorders (Dos Santos et al., 2016). Regarding antidepressant properties, LSD does not seem very successful. Some antidepressant effects have been found after treatment with ayahuasca – the brew made of plants containing DMT (Osorio Fde et al., 2015; Sanches et al., 2016). But of the classical psychedelics, psilocybin has shown the most promising antidepressant effects and has therefore been researched the most in patients with depressive symptoms (Dos Santos et al., 2016; Griffiths et al., 2016; Ross et al., 2016).
Psilocybin, the classic hallucinogen found in magic mushrooms, is a potential new antidepressant drug that appears to be fast acting. Randomized-controlled trials in depressed patients showed that a single dose combined with psychotherapy decreases depression and anxiety and increases quality of life within 24 hours (Griffiths et al., 2016; Ross et al., 2016). These effects persist even six months after treatment. Considering that these effects are rapid, robust and enduring, psilocybin appears a promising antidepressant drug.
Several problems in patient studies prevent conclusive results
regarding the antidepressant effects of psilocybin
Unfortunately, not much is known about psilocybin for a variety of reasons. The first studies investigating psilocybin were performed between 1950 and 1970 and are now considered unreliable, because they were methodologically sub-optimal (review by Rucker et al., 2017). Then, the registration of psilocybin as a Schedule 1 drug in 1967 defined psilocybin as having no medical use and an extremely high abuse potential (review by Rucker et al., 2017). This hampered human experiments almost entirely, because of the lack of a clinical focus (review by Tylš, Páleníček, & Horáček, 2014). Despite this prohibition, the interest in human experiments using psychedelics has been revived in the last 20 years and psilocybin is currently one of the most used psychedelics in human experiments (review by Tylš et al., 2014).
Three double-blind randomized-controlled trials have investigated the effects of psilocybin on end-of-life anxiety in patients with advanced-stage cancer (Griffiths et al., 2016; Grob et al., 2011; Ross et al, 2016) and an open label trial has studied the effects on TRD (Carhart-Harris et al., 2016). Even though all studies show rapid and sustained antidepressant effects of psilocybin, the results should be interpreted with caution. There are three major points of criticism regarding these studies that have to be taken into
recruiting patients with interest for psychedelics, who will therefore be predisposed to positive experiences. Indeed, the percentages of patients with previous experience with psychedelics were high (45% in Griffiths et al., 2016; 25 % in Grob et al, 2011; 55% in Ross et al, 2016). Secondly, the obvious changes in subjective experiences during the altered state of consciousness induced by psychedelics cause unblinding of the patients. This results in a demand characteristic: the participants are aware of their expected behaviour or the expected findings (Carhart-Harris & Goodwin, 2017). Furthermore, the outcome measures are entirely subjective, similar to almost all studies investigating antidepressants in humans. Together, these factors leads to a major possible bias towards finding reliable beneficial effects of psychedelics in depressed patients.
To conclude, even though the findings in all these patient studies are promising, the problem is that it is difficult to disentangle the relative contribution of the context, the internal state of the patient, the offered psychotherapy and psilocybin to the putative therapeutic effect.
New goal: Investigating the mechanism of action of psilocybin in
animal models of depression
To be able to move forward in the field of antidepressant effects of psychedelics, a different approach is needed. The exact molecular mechanism of psilocybin in depressed patients is still unknown. However, we know from the literature that animal models of depression can be used to gain insight into the underlying molecular mechanisms of antidepressants. Currently, no
in vivo research regarding the putative antidepressant effects of psilocybin in
animal models of depression has been published in scientific journals. One preliminary study found an antidepressant effect of psilocybin in Wistar Kyoto (WKY) rats, a genetic depression model (Nichols & Hibicke, 2018). However, because this study is described in an abstract of the 19th Meeting of International Society for Serotonin Research, no specific details are available. Exploring the antidepressant-like effect of psilocybin in animal models of depression will lead to new insights regarding the molecular mechanisms of psilocybin. Subsequently, this will potentially catalyse further research regarding new, fast-acting, enduring antidepressants that are possibly non-hallucinogenic.
This thesis focuses on: Which role can animal models of depression play in understanding the putative antidepressant-like effect of psilocybin? The current available literature was consulted to find a relevant theoretical framework of depression. Then, we explored what is currently known about the molecular mechanism of psilocybin. With this knowledge, we designed appropriate experiments to answer the question. The findings are described in four chapters. The first chapter describes one of the currently most prominent neurobiological hypotheses of depression: the neurotrophic
hypothesis. Chapter two presents current research regarding the effects of psychedelics on depressive-like neurobiology and behaviour. Chapter three portrays the most widely used animal models of depression relevant to the neurotrophic hypothesis and their strengths and limitations. In chapter four, these findings are integrated and combined into suggested future experiments. With these suggestions, we hope to contribute to the development of the research regarding psilocybin as a possible antidepressant.
1. The neurobiology of depression
Depression is a multifactorial disease of which the etiology remains largely unknown. Therefore, several possible hypotheses are being investigated currently. This chapter aims to define depression within a specific theoretical framework of one of these hypotheses. This framework is necessary to provide direction for the further investigation of the large body of available literature.
Introducing the neurotrophic hypothesis of depression
Recently, a paradigm shift has taken place in the field of depression research. For a long time, decreased availability of monoamine neurotransmitters such as serotonin and noradrenaline in the nervous system was thought to underlie the development of depression (review by Liu, Liu, Wang, Zhang, & Li, 2017). Antidepressants would increase the synaptic serotonin or noradrenaline concentration, resulting in reduction of depressive symptoms. Although this monoamine hypothesis has been very influential for psychiatry in the past, it has been critiqued mainly because of the therapeutic delay. Namely, even though SSRI treatment immediately increases monoamine transmission, symptom reduction takes several weeks (review by Krishnan & Nestler, 2008). This led to the exploration of other possible hypotheses that might be able to explain the pathogenesis of depression and late onset of SSRIs and eventually lead to new, more effective treatments.
One of these new hypotheses is the neurotrophic hypothesis, which has been intensively investigated and has been described in a number of reviews (Duman & Li, 2012; Haase & Brown, 2015; Liu et al., 2017; Pittenger & Duman, 2008). Substantiated by both preclinical and clinical evidence, it sheds new light on how changes in neurogenesis and neuroplasticity might contribute to the development of depression. Neurogenesis is the formation of new neurons in the brain, mainly in the hippocampus (Wang et al., 2011). Neuroplasticity, defined as the brains capability to change and adapt, implies that the brain undergoes dynamic physiological changes following interactions of the organism with the environment (Gulyaeva, 2017). There are multiple scales on which neuroplasticity can be observed, of which learning and memory would be the highest level. On the cellular level, neuroplasticity consists roughly out of three processes: neuritogenesis, spinogenesis and synaptogenesis.
This chapter first introduces the key regulators of these processes. Then, it describes how changes in neuroplasticity and hippocampal neurogenesis explain both the contribution of chronic stress to the development of
depression. Finally, the therapeutic delay of current antidepressant treatments is explained in light of the neurotrophic hypothesis.
Neurotrophic factor BDNF orchestrates neuroplasticity
Changes in neuroplasticity are regulated by neurotrophic factors: growth factors that are expressed during neurodevelopment and regulate plasticity in the adult brain (review by Krishnan & Nestler, 2008). A large protein interaction network underlies these changes in neuroplasticity (Waller et al., 2017). However, there is one neurotrophic factor that stands out the most: brain-derived neurotrophic factor (BDNF), a key regulator of synaptic plasticity (review by Liu et al., 2017). Its effects are both structural and functional and range from short-term to long lasting, affecting excitatory and inhibitory synapses in many brain regions. Its actions are mediated by the TrkB receptor and influence the activation of several signalling pathways in the glutamatergic and serotonergic neurotransmitter systems (Gulyaeva, 2017). The TrkB receptor homodimerizes upon binding its ligand BDNF. Because the TrkB receptor is a tyrosin kinase, this leads to multiple phosphorylations that trigger three major pathways: the Ras-mitogen-activated protein kinase (MAPK) pathway, the phospholipase Cγ (PLCγ) pathway and the phosphatidyl inositol-3 kinase (PI3K) pathway (Figure 1) (Yoshii & Constantine-Paton, 2010; Eliwa, Belzung & Surget, 2017). These three pathways are described here. However, this is highly simplified, since all protein networks related to neuroplasticity are highly complex (Waller et al., 2017).
The MAPK-pathway is activated through activation of Ras, which triggers downstream kinases including MAPK. MAPK influences neuroplasticity by increasing the activity of proteins involved in protein synthesis and by activating cAMP response element-binding protein (CREB) (Yoshii & Constantine-Paton, 2010). CREB is a transcription factor, influencing the expression of a variety of genes involved in synaptic plasticity, including
BDNF. Further increase of BDNF expression stabilizes synaptic changes and
establishes a feed-forward loop, because BDNF induces translocation of MAPK to the nucleus. The activation of nuclear substracts induced by MAPK further increases CREB activation, leading to even more BDNF expression.
Activation of the PLCγ pathway stimulates hydrolysis of phosphatidylinositol 4,5-bisphosphate into diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP3). These molecules induce protein kinase C (PKC) activation and Ca2+ releases from intracellular stores respectively. The elevated Ca2+ levels increase adenyl cyclase (AC) activity, catalysing the transformation of ATP into second messenger cAMP. Elevation of cAMP levels leads to activation of CREB-dependent transcription. This pathway also directly contributes to synaptic plasticity by translocating GLuR1 subunits to
synapses (review by Pittenger & Duman, 2008; Yoshii & Constantine-Paton, 2010).
Activation of PI3K leads to activation of protein kinase B (PKB), a protein involved in several processes, including cell survival and protein translation. The latter is induces by activation of mammalian target of rapamycin (mTOR), a key regulator in protein synthesis. Since this pathway also regulates the transport of synaptic proteins, it is essential in long-term maintenance of synaptic plasticity (Yoshii & Constantine-Paton, 2010).
So, activation of BDNF leads to the activity of these three pathways, contributing to synaptic plasticity through gene transcription, protein translation and transport.
Figure 1. Simplified cartoon of post-synaptic pathways activated by brain-derived neurotrophic factor (BDNF): the Ras-mitogen-activated protein kinase (MAPK) pathway, the phospholipase Cγ (PLCγ) pathway and the phosphatidyl inositol-3 kinase (PI3K) (adjusted from Yoshii & Constantine-Paton, 2010).
BDNF regulates hippocampal neurogenesis
Besides being a key regulator in several processes underlying neuroplasticity, BDNF is suggested to be involved in neurogenesis. Neurogenesis is a process consisting of four stages: (1) proliferation of progenitor cells, (2) differentiation of these cells to neurons or glia cells, (3)
maturation, which includes neuroplasticity processes such as neuritogenesis,
spinogenesis, functional development of synapses and circuit integration and (4) survival (Eliwa et al., 2017). In humans, the hippocampus is the primary brain region that performs neurogenesis (Wang et al., 2011).
Neuroplasticity in the hippocampus is of particular interest in the etiology of depression, because this area is involved in many cognitive functions. The dorsal part of the hippocampus in rodents (posterior in humans) is associated
with spatial navigation and episodic memory and the ventral part (anterior in humans) is preferentially involved with emotional reactivity, anxiety-related behaviours and the stress response (Eliwa et al., 2017). Hippocampal neurogenesis plays an important role in the development of depression in animal models, because neurogenesis-deficient mice show depressive-like behaviour, such as increased behavioural despair and decreased sucrose preference (Snyder, Soumier, Brewer, Pickel, & Cameron, 2011). Several studies provide evidence for the positive contribution of BDNF to hippocampal neurogenesis (review by Numakawa, Odaka, & Adachi, 2017). For example, mice with truncated 3’ untranslated region (UTR) BDNF mRNA present several deficits in neurogenesis, including impaired differentiation and maturation (Waterhouse et al., 2012). Therefore, BDNF is suggested to be involved in hippocampal neurogenesis, even though the direct mechanism has not yet been identified.
Chronic stress affect neuroplasticity, resulting in depression
The neurotrophic hypothesis of depression argues that negative effects of chronic stress on neuroplasticity and neurogenesis lead to the development of depression. Chronic stress is a risk factor for the development of depression (review by Krishnan & Nestler, 2008). It is known to induce both hippocampus and prefrontal cortex (PFC) atrophy and glia loss (review by Duman & Li, 2012). In depressed patients, volume reduction of the hippocampus and PFC, decreased neuron size and loss of glia cells are observed (review by Duman & Li, 2012). More specifically, stress has negative effects on neuroplasticity: increased circulating cortisol levels affect hippocampal neurons through the glucocorticoid receptors, leading to decreased spinogenesis, synaptogenesis and increased apoptosis (reviews by Pittenger & Duman, 2008; Liu et al., 2017). Stress also negatively influences hippocampal neurogenesis (Eliwa et al., 2017; Lucassen et al., 2015). As decreased BDNF levels are observed in in the hippocampus and PFC following physical or social stress in rodent models (reviews by Krishnan & Nestler, 2008; Castrén & Rantamäki, 2010), BDNF signalling is suggested to be involved in these stress-induced changes in neuroplasticity.
The neurotrophic hypothesis explains therapeutic delay of SSRIs
Besides explaining the contribution of stress to the etiology of depression, the neurotrophic hypothesis accounts for the therapeutic delay of SSRI treatment. SSRI treatment influences the protein interaction networks related to synaptic plasticity, signal transduction and neurodevelopment (Waller et al., 2017). BDNF is suggested to be both necessary and sufficient for the effects of antidepressants on hippocampal neurogenesis (review by Pittenger & Duman, 2008; Eliwa et al., 2017).
Chronic SSRI treatment influences neurogenesis over the entire dorso-ventral axis of the hippocampus (Eliwa et al., 2017). Serotonin positively regulates the proliferation, maturation and survival of adult-born hippocampal neurons (Eliwa et al., 2017). SSRIs stimulate serotonergic neurotransmission by blocking the reuptake of serotonin by the presynaptic serotonin receptor (SERT), resulting in increased availability of synaptic serotonin (Eliwa et al., 2017). Progenitor cells and immature neurons express several serotonin receptors: the 5-HT1A, 5-HT2, and 5-HT4 receptor (review by Haase & Brown, 2015). Therefore, their development is positively affected by elevated serotonin levels. Namely, the elevation of synaptic serotonin induces gene expression and release of neurotrophic factors, such as BDNF (review by Haase & Brown, 2015). TrkB-related signalling pathways and BDNF expression are increased substantially in the PFC and hippocampus after various antidepressant treatments (Molteni et al., 2006; Eliwa et al., 2017). Increased BDNF expression results in a feed forward loop, because chronic administration of BDNF increases the serotonergic transmission (review by Deltheil et al., 2008). This way, BDNF expression slowly increases synaptic plasticity and hippocampal neurogenesis. Indeed, the therapeutic delay is similar to the time it takes for new hippocampal neurons to become fully functional (review by Mahar, Bambico, Mechawar, & Nobrega, 2014). This would explain the symptom reduction observed during week six or seven of antidepressant treatment. One major limitation of this hypothesis is that it remains unclear whether blocking hippocampal neurogenesis actually induces depressive like symptoms (review by Mahar et al., 2014).
BDNF, orchestrator of neuroplasticity, as possible target of
psilocybin
The neurotrophic hypothesis of depression states that altered balances of neurotrophic factors, due to chronic stress, contribute to the etiology of depression. Chronic stress induces depressive-like behaviour and decreases neuroplasticity and BDNF levels in the hippocampus and PFC. Antidepressant SSRI treatments in their turn increase BDNF expression and are suggested to induce antidepressant effects through increasing neuroplasticity and neurogenesis. BDNF is a key regulator in neuroplasticity and is suggested to play an important role in hippocampal neurogenesis. This led us to investigate BDNF expression as potential key player in antidepressant effects.
2. Psychedelics affect neuroplasticity
As BDNF expression is defined as a relevant process to investigate depression, it would be interesting to know how psilocybin influences BDNF expression. Unfortunately, this is currently unknown. However, the relation between dissociative anesthetic ketamine and BDNF expression is well studied. Ketamine is not a classical psychedelic, but an NMDA receptor antagonist. Classical psychedelics are (partial) agonists of the 5-HT2A receptor and have a high affinity for the 5-HT1A receptor (review by Nichols, 2016). These receptors are mostly expressed in pyramidal neurons in the cortex and not in the mesolimbic pathway (Olson, 2018). Therefore, classical psychedelics are considered non-addictive, making them an attractive substitute for ketamine.
Recently, many studies have shown that ketamine induces fast-acting, enduring antidepressant effects (review by Sattar et al., 2018). A large body of evidence suggests that ketamine causes these effects by inducing neuroplasticity through TrkB-BDNF signalling. For example, a single infusion of ketamine increases BDNF expression in the hippocampus (Caffino et al., 2016). Besides, miR-206, a miRNA regulating BDNF expression, is down regulated in ketamine-treated rats (Yang et al., 2014). A recent study investigating the anxiety-like behaviour of rats upon micro dosing with ketamine or psilocin – the pharmacologically active form of psilocybin – found similar effects of these drugs in several paradigms, suggesting that the drugs work though a similar downstream mechanism (Horsley, Páleníček, Kolin, & Valeš, 2018).
Therefore, this chapter describes the link between serotonergic signalling and BDNF expression. Then, it explores currently known effects of classical psychedelics on neuroplasticity, which are also suggested to be regulated by BDNF. Subsequently, the currently known antidepressant behavioural effects of classical psychedelics in animal models of depression are described. Together, these findings form an indication of the possible effects of psilocybin on BDNF expression.
Serotonergic signalling induces BDNF expression
Psilocybin binds to the 5-HT2A receptor (Gallaher, Chen, & Shih, 1993), which induces serotonergic signalling. Serotonergic signalling and BDNF expression are closely connected in the brain through the cAMP-protein kinase A (PKA)-pathway. Serotonin is produced mainly by neurons in the dorsal raphe nucleus (DR) (review by Mahar et al., 2014). These serotonergic DR neurons practically innervate all serotonin receptor expressing corticolimbic structures involved in stress responses and mood regulation,
including the PFC, amygdala, hippocampus and nucleus accumbens (NAc). Activation of these receptors elicits the activation of the PKA through elevating cAMP levels. This results in phosphorylation of CREB, which then promotes BDNF expression (Eliwa et al., 2017). This way, serotonin influences BDNF expression, contributing to the regulation of neuronal plasticity, stress-responses and antidepressants efficacy (Rafa-Zabłocka, Kreiner, Bagińska, & Nalepa, 2018).
Classical psychedelics affect neuroplasticity through BDNF
The influence of classical psychedelics on neuroplasticity has not been studied extensively. Ly et al., (2018) performed the only study to date directly testing the hypothesis that classical psychedelics promote structural and functional neural plasticity. They assessed the effects of DMT, LSD, psilocin, 2,5-Dimethoxy-4-iodoamphetamine (DOI) and several other non-classical psychedelics on neuritogenesis, spinogenesis and synaptogenesis in vitro and
in vivo. Their findings are of major importance to our research question, so
their most relevant results are described here.
Classical psychedelics induced neuritogenesis, spinogenesis and synaptogenesis in vitro and in vivo. DMT, LSD and psilocin increased neuritogenesis, measured as dendritic arbor complexity in rat cortical cultures (Ly et al., 2018). LSD also increased the number of dendritic branches in vivo in Drosophila class 1 sensory neurons. DMT and LSD also induced spinogenesis. In rat cortical cultures, the number of dendritic spines per unit length was increased and the distribution of spine morphology was shifted from mature to immature spines. In vivo, a DMT injection also increased spine density in rat PFC cortical pyramidal neurons and caused functional enduring effects in pyramidal neurons (Ly et al., 2018). Cells from human cerebral organoids that were treated with 5-methoxy-N,N-dimethyltryptamine (5-MeO-DMT), a molecule similar to DMT, differentially expressed proteins involved in the formation and maturation of dendritic spines (Dakic et al., 2017), further substantiating the neuroplasticity inducing capacities of classical psychedelics. Finally, DMT, LSD and DOI induced synaptogenesis in
vitro by increasing the synapse density and not the synapse size. The changes
induced by DMT were again enduring (Ly et al., 2018). These findings suggest that classical psychedelics promote neuroplasticity.
Then, the signalling molecules required for these changes in neuroplasticity were identified in rat cortical cultures (Ly et al., 2018). First, the involvement of BDNF was demonstrated, as treatment with BDNF increased neuritogenesis. Moreover, 24 hours after DMT and LSD treatment, the protein levels doubled, although not significantly. The TrkB and mTOR dependency was determined as well. Selective TrkB antagonist ANA-12 completely abrogated the psychedelic-induced stimulation of neuritogenesis and spinogenesis. The ability of psychedelics to stimulate neuritogenesis was
also blocked by rapamycin, an mTOR inhibitor. Finally, the involvement of the 5-HT2A receptor was demonstrated. Selective 5-HT2A antagonist ketanserin blocked the neuroplasticity promoting abilities of DMT and LSD and the ability of psilocin to increase the number of dendritic branches. These findings suggest that activation of TrkB, mTOR and 5-HT2A are required for the neuroplasticity promoting effects of classical psychedelics (Ly et al., 2018).
Behavioural effects of classical psychedelics in animal models of
depression
Following up on these findings, several studies have investigated the antidepressant behavioural effects in some animal models of depression.
One study investigated the effects of LSD in an animal model of depression (Buchborn, Schröder, Höllt, & Grecksch, 2014). The bulbectomized rat model was used, which resembles the human disease well and is known to respond to chronic, but not acute antidepressant treatments (review by Czéh, Fuchs, Wiborg, & Simon, 2016). Chronic administration of LSD normalized the deficits in active avoidance learning induced by the bulbectomy, similar to chronic antidepressant treatments (Buchborn et al., 2014).
More researchers investigated the behavioural effects of DMT on depressive-like behaviour and anxiety. In the forced swim test (FST) (see chapter 3), rats that received DMT injections six hours and one hour prior to testing spend less time immobile and more time swimming compared to controls. Since DMT was shown to reduce locomotion in another paradigm, the increased swimming can be seen as true antidepressant behaviour (Cameron, Benson, Dunlap, & Olson, 2018). The evidence regarding possible anxiogenic effects of DMT is contradictory. For example, rats showed anxiogenic effects in novelty induced locomotion and elevated plus maze (EPM) paradigms one hour after treatment (Cameron et al., 2018). However, no changes were found in the EPM upon chronic ayahuasca administration (Favaro, Yonamine, Soares, & Oliveira, 2015). Furthermore, DMT induced anxiogenic effects during contextual and cue-induced fear conditioning (Cameron et al., 2018). Acute DMT treatment before training increased the percentage of time animals spent freezing immediately after receiving the shock. These results indicate contradicting antidepressant and anxiogenic effects of DMT. Upon chronic ayahuasca administration, an increase of the contextual fear conditioning response was maintained over time, suggesting that ayahuasca interferes with fear associations (Favaro et al., 2015).
Other researchers investigated the effects of psilocybin on anxiety-like and depressive-like behaviour. Microdosing of psilocin induced acute anxiogenic effects in the EPM (Horsley et al., 2018). However, preliminary results show that one dose of psilocybin induces enduring antidepressant and anxiolytic behaviour in the EPM in Sprague Dawley rats 42 days after dosing (Nichols &
measured weekly in the FST for 32 days in genetically depressed WKY rats (see chapter 3). So even though psilocybin seems to have acute anxiogenic effects, these preliminary data suggest that psilocybin indeed produces long-lasting antidepressive behaviour in vivo.
Together, these behavioural findings suggest that psilocybin is a promising classical psychedelic in inducing enduring, antidepressant effects. Even though the acute effects seem to be anxiogenic, the enduring antidepressant effects appear to confirm the positive findings in human studies. Unfortunately, these studies do not investigate the involvement of BDNF expression in these behavioural changes.
Role of BDNF expression in antidepressant effects of psilocybin is
unknown
Previous chapters have described that BDNF is an important regulator of neuroplasticity and is therefore suggested to be important in the etiology of depression. Classical psychedelics have been shown to alter neuroplasticity through BDNF expression and to induce antidepressant behaviour in animal models of depression. However, only one preliminary study investigated enduring antidepressant behaviour after psilocybin treatment in a rat depression model (Nichols & Hibicke, 2018). They reported enduring antidepressant behaviour, but the involvement of BDNF expression has not been investigated.
3. Animal models of depression
Previous chapters have summarized promising results regarding antidepressant effects of psilocybin. These findings include behavioural and neurochemical results. To cohere these findings and to further develop this line of research regarding the molecular mechanism of the antidepressant effects of psilocybin, several essential links and relations have to be elucidated. For example, future studies could investigate the effects of psilocybin on depressive-like behaviour and BDNF expression in the same animals. This will be a first step towards unravelling the working mechanism of the putative antidepressant effects of psilocybin.
To be able to perform these experiments, an appropriate experiment design is essential. When conducting in vivo research with animal models of depression, three large decisions have to be made. Firstly, the species, strain and genetic background of the animal have to be chosen, determining the predisposition to developing the disorder. Secondly, environmental factors possibly leading to development of a depressive-like phenotype have to be selected. The final step is the selection of the appropriate experiments and read-outs of depressive-like behaviour and/or biomarkers relevant to the research question and the chosen model. This chapter first describes the criteria used to indicate the translational power of animal models. Then, the most widely used animal models of depression and read-outs relevant to our research question are presented, with their advantages and disadvantages according to those criteria. This discussion will serve as a starting point for the proposed future experiment in the next chapter.
Criteria to assess the validity of animal models
To be able to decide which animal model to use to answer a specific question, Willner (1984) proposed three criteria: predictive validity, face validity and construct validity. Originally, predictive validity referred to a human-animal correlation of therapeutic outcomes (review by Willner, 1984), but later it has been broadened, as it can also refer to the ability of the model to predict specific disease outcomes (Belzung & Lemoine, 2011). Face validity was initially described as the resemblance of the animal model to human disease, both symptomatic and therapeutically (review by Willner, 1984). This means that the features the model has in common with the human condition are specific to the disease and that no clinically irrelevant features are present in the model. To meet the construct validity criterion, the theoretical rationale of how dysfunctional processes arise in animal models and in the human disease has to be similar at the behavioural or cognitive level and in etiology (review by Willner, 1984). These well-known criteria have been used for many
years to assess the translational power of a variety of animal models in psychiatric research.
Belzung and Lemoine (2011) integrated these criteria into a broader framework, leading to five major criteria that more specifically refer to translational features of animal models. First, they describe the development of a disease as follows. The initial organism is defined by their genetic code. At this point, the organism can be genetically vulnerable to the disease already. Early environmental factors can also transform the initial organism into a vulnerable organism. Subsequently, triggering factors in adulthood can turn the vulnerable organism into a pathological organism, observable by behavioural symptoms and biomarkers. The differences between the initial and the pathological organism define the cognitive and behavioural mechanisms underlying the disease. Therapeutic drugs can turn the pathological organism back into a vulnerable organism by diminishing the mechanism of action underlying the behavioural symptoms and biomarkers, resulting in their reduction.
Within this framework, several criteria for validity are represented, that are summarized here (for a more abundant overview, see Belzung & Lemoine, 2011). Homological validity refers to the similarity between humans and the chosen species and strain of the species. Pathogenic validity refers to the similarity of the mechanisms leading to the disease. Pathogenic validity can be divided into two categories, dependent on the type of factor that is interacting with the organism; Ontopathogenic validity refers to the interaction of early environmental factors with the initial organism, leading to the vulnerable organism. Triggering validity assesses the similarity of the mechanisms that produce the pathological organism after interaction of either the initial or vulnerable animal with triggering factors. The mechanistic validity refers to the similarity of the disease mechanism itself. This mechanism is what is producing the behavioural symptoms and the biomarkers, but also the target mechanism of effective therapeutic drugs. The similarity of the behavioural symptoms is assessed in the ethological validity criterion and the similarity of the biomarkers is assessed in the biomarker validity. Together, these two criteria are the face validity criterion: assessing the similarity of what is observed in the pathological organism and in human patients. And finally,
pathological validity describes the relation between the action of triggering
factors (induction validity) or therapeutic agents (remission validity) and the disease. It refers to the interaction between these factors or agents with the observed effects of the disease (behavioural symptoms and biomarkers) and not with the actual mechanism leading to the disease. Together, these criteria can assess the translational power animal models in a very specific way. Therefore, they form the theoretical framework used to assess which animal model of depression is the optimal choice to detect antidepressant effects of psilocybin.
Animal models of depression
The number of possible animal models of depression is substantial. For an extensive review of animal models of depression currently available, including the link between the symptoms observed in depressed patients and the representative phenotype of the animals and a list of the appropriate tests to measure the behavioural or physiological changes of the animals, see Czéh et al. (2016). In this thesis, only the four most widely used rat models relevant to our research question are discussed, as describing all possible animal models of depression is beyond the scope of this thesis.
Behavioural read-outs of depression
To be able to assess depressive-like behaviour and the antidepressant effects of specific agents on this behaviour, numerous read-outs have been developed. Here, the FST and sucrose preference test (SPT) are depicted, because these two tests are relevant for the animal models of depression described later in this chapter.
The FST is a short paradigm inducing behavioural despair. The rodents are placed in a cylinder filled with water, so they are forced to swim to survive this inescapable situation. If the rodent does not swim, but floats and only moves to keep its head above the water, this behaviour is referred to as immobility. Immobility is a measurement of behavioural despair and therefore represents depressive-like behaviour. Antidepressants would decrease the time animals spent immobile. The FST is mostly used as a screening tool for putative antidepressants, because it is said to have high remission validity (review by Cryan, Valentino, & Lucki, 2005). However, this is disputed, because many agents with antidepressant effects in this model are not effective in depressed patients (review by Yin, Guven, & Dietis, 2016).
The SPT is an example of a reward related test used to measure anhedonia. Anhedonia is defined as “the inability to experience pleasure from rewarding or enjoyable activities” (Liu et al., 2018, p. 1) and is a key symptom of depression. In this test, food and water deprived rodents can drink freely from a bottle containing a sucrose solution for one hour (Palmfeldt, Henningsen, Eriksen, Müller, & Wiborg, 2016). Then, the sucrose preference is calculated based on measurements of the weight of the bottle containing the sucrose solution before and after the SPT. A decrease in sucrose preference in the experimental group relative to control indicates anhedonia.
Species, strain and genetics
In the etiology of depression, genetics play an important role. The heritability of liability to depression is approximately 40% for women and 30% for men (Kendler, Gatz, Gardner, & Pedersen, 2006; Viktorin et al., 2016). Every rat strain has its own unique genetic background as well. Some of these genetic backgrounds lead to high vulnerability to developing depression
(review by Belzung, 2014). Therefore, deciding the appropriate strain of rats is important to reach the homological validity criterion. The genetic background can be established by selective breeding. This means the animals selected upon a specific behaviour or read-out are bred for several generations (review by Czéh et al., 2016). Continuous breeding leads to animals with identical genotypes, which is useful for research. The Flinders Sensitive Line (FSL) rats and the WKY rats are the most popular inbred strains, so they are described here.
The FSL rat has a high validity as depression model
The FSL rat was selectively bred from Sprague Dawley rats for its increased response to an anticholinesterase agent diisopropylfluorophosphate (DFP) (review by Overstreet & Wegener, 2013). Its control strain, the Flinders Resistant Line (FRL), is more resistant to DFP than the FSL rats, but less than Sprague Dawley rats (Neumann et al., 2011). Many of the core symptoms of depression can be reproduced in the FSL rats (review by Overstreet & Wegener, 2013). Neumann et al. (2011) listed the most important variables of depressed patients and the according characteristic of the FSL rat (table 1). Based on the similarities, this strain can be considered as having an adequate level of face validity.
Behaviourally, when comparing the FSL rats to the FRL rats, the FSL rats show reduced appetite and psychomotor functioning (review by Czéh et al., 2016), and higher immobility in the FST – that is reversed upon antidepressant treatment (Chen, Madsen, Wegener, & Nyengaard, 2010). However, no changes in reward related tests - the SPT or intracranial self-stimulation - are observed in non-stressed FSL rats, thus anhedonia is not present in this strain (review by Overstreet & Wegener, 2013). Signs of anhedonia are present in FSL rats upon chronic mild stress (CMS) (described below) (Pucilowski, Overstreet, Rezvani, & Janowsky, 1993). When scrutinizing the neurochemical changes in the FSL rats, deficits are found in the serotonergic, cholinergic, dopaminergic and neuropeptide Y systems (review by Czéh et al., 2016).
To be able to select the most appropriate animal model for identifying possible changes of BDNF expression upon psilocybin administration, the BDNF biomarker validity is relevant. Furthermore, the interaction of antidepressants with the biomarker is relevant, so the remission validity is important too. In FSL rats, BDNF levels are increased in serum and whole blood and decreased in the hippocampus (Elfving, Plougmann, Müller, et al., 2010). Correspondingly, the volume and the number of hippocampal neurons and synapses are smaller (Chen et al., 2010). These numbers are increased upon chronic antidepressant treatment, suggesting that the antidepressant reverses the suppression of hippocampal neurogenesis and synaptogenesis (Chen et al., 2010). As, the BDNF response in FSL rats is representative for the
human disease; the FSL rats reach the biomarker and remission validity criterion.
Face validity Patients FSL rats
Activity Psychomotor retardation Reduced bar pressing for rewards
Passive stress coping Yes Yes (increased immobility FST)
Anxiety Often increased No anxiety
Anhedonia Yes Yes (following stress)
Appetite Reduced Reduced
Weight Weight loss Lower bodyweight
Cognitive performance Reduced Yes/no (dependent on test)
Social behaviour Often abnormal Enhanced affective aggression
REM sleep Elevated Elevated
HPA axis dysregulation Yes Yes
Cardiovascular morbidity Increased Increased
Monoamine dysregulation Suggested Altered
Amygdalar responsiveness Increased Not assessed
Pain perception Altered Probably altered
Predictive validity
Tricyclic antidepressants Effective Effective Selective serotonin-reuptake
inhibitors Effective Effective
Monoamine uptake inhibitors Effective Effective Atypical antidepressants Effective Effective
Benzodiazepines Effective Effective
Etiological validity
Learned fearful associations Increased Increased
Genetic predisposition Suggested Cholinergic sensitivity
Table 1. Summary table depicting the validity of the FSL rat. The validity terminology of Willner
(1984) is used here. Etiological validity refers to causative conditions (adjusted from Neumann et al. (2011)).
The validity of WKY rats is moderate for depression, but high for TRD
The WKY rat strain was initially developed as the normotensive control strain for the spontaneously hypertensive rat, but later turned out to have similarities to depressed patients (review by Czéh et al., 2016). WKY rats do not show robust antidepressant responses in the FST upon several antidepressant treatments (review by Willner & Belzung, 2015). Therefore, the WKY rat was proposed to be an appropriate model to study TRD (reviews by Overstreet & Wegener, 2013; Willner & Belzung, 2015). A recent study validated this by reporting that WKY rats that are subjected to CMS meet the four criteria for animal models of TRD: “a phenotypic resemblance to a risk factor for depression; enhanced response to stress; nonresponse to
al., 2018). As psilocybin induced antidepressant effects in TRD patients (Carhart-Harris et al., 2016), the WKY rat appears a relevant model to study the antidepressant-like effects of psilocybin. Accordingly, fast-acting, enduring antidepressant behaviour in the FST is measured in WKY rats upon treatment with ketamine (Tizabi, Bhatti, Manaye, Das, & Akinfiresoye, 2012) and psilocybin (Nichols & Hibicke, 2018). Therefore, WKY rats might have moderate remission and homological validity for depression, but appear to be appropriate to measure antidepressant-like effects of psilocybin. Besides, antidepressant treatment induced hippocampal BDNF expression (Hurley et al., 2013), thus WKY rats reach the remission validity criterion too.
Environmental factors leading to depression
The neurotrophic hypothesis emphasizes the effects of stress on the development of depression. Therefore, to reach the pathological validity criterion, this chapter only describes the available depression models that use a certain type of stressor as environmental factor leading to depression-like features. A recent review by Yin et al. (2016) lists all available stress-induced models of depression, including the used stressors, main advantages and disadvantages based on literature. They found that the three most widely used animal models of depression are the FST, the learned helplessness (LH) and the CMS model. However, it is argued that the FST is not an appropriate stress model of depression, but only a fast screening for antidepressant compounds (Nestler & Hyman, 2010). Therefore, we have decided to classify the FST as a read-out of depressive-like behaviour, as described above. Thus, only the LH and CMS models are described here.
The LH model has a high validity
The LH model is developed to mimic the feeling of helplessness in depressed patients. It is based on the cognitive theory of depression (review by Clark & Beck, 2010). This theory states that depressed patients have maladaptive expectations and ideas regarding the self, the world and the future. These expectations and ideas are constantly activated, confirmed and strengthened by negative experiences in life that match the expectations and ideas. This results in a feeling of helplessness and contributes to the etiology of depression. In LH models, the helplessness is manifested as giving up to escape from a stressful situation, because of multiple uncontrollable stressful experiences (review by Vollmayr & Gass, 2013). The depressive-like behaviour is induced by several random inescapable foot shock or tail shocks (review by Yin et al., 2016), which creates negative expectations about the future. 24 hours later, the animals receive escapable shocks in a box with a door they can use to exit. Native animals will actively avoid the shock and leave the box. However, learned helplessness-induced animals will fail to escape. This contributes to a high triggering validity.
The face validity criterion is also met, because many symptoms observed in depressed patients are present in this model, such as anhedonia and increased corticosterone levels (review by Vollmayr & Gass, 2013). The LH model is one of the few models that show the therapeutic delay; Antidepressants improve the escaping behaviour of the rats, but only after chronic administration (review by Vollmayr & Gass, 2013). Besides, non-antidepressant drugs do not alter the escaping behaviour. This contributes to excellent remission validity.
In the LH paradigm, BDNF expression and protein levels are lower in mice that respond to the LH paradigm, compared to mice that appeared to be resistant to the paradigm (Su, Su, Hsiao, & Gean, 2016). Antidepressive treatment reduced the depressive-like behaviour and reversed the decrease of BDNF expression. Therefore, the LH model reaches the biomarker and remission validity criterion.
The CMS model has a high validity, but reproduction is challenging
The CMS model has been developed to optimally meet the pathogenic validity criterion (review by Willner, 2005). It provokes anhedonia, by semi-randomly exposing the animals to a variety of stressors over a period of several weeks. During intervals between the stressors, the SPT is performed. Because of the long-term development of depressive-like symptoms after chronic exposure to a variety of stressors, the triggering validity of this model is high. Besides, almost all antidepressants reduce depressive-like behaviour triggered by CMS (review by Willner, 2016). Therefore, the remission validity of CMS can be considered as high. A limitation of this model is that multiple aspects of the protocols used by different researchers can diverge extremely, which inevitably contributes to the inconsistent findings generated with the CMS model (review by Yin et al, 2016). First of all, the possible types of stressors are abundant, including restraint stress, disturbed day-night cycle and water deprivation. Besides, the time and amount of stress applied to the animal can vary, as can the time and amount of the intervals. Finally, the SPT is only an example of a possible read-out. Any reward related test could be used to measure anhedonia.
BDNF expression is decreased in the CMS model compared to controls, and this decrease is reversed by antidepressant treatment (First et al., 2013; Molteni, Rossetti, Savino, Racagni, & Calabrese, 2016). A study investigating the influence of one type of stressor (chronic restraint stress) on depression reported that the chronic stress did not induce significant changes in BDNF and TrkB expression levels compared to controls (Chiba et al., 2012). This highlights the importance of using several different stressors in modelling depression, which is an important characteristic of the CMS model. Thus the CMS model reaches the biomarker and remission validity criterion.
All described models are suitable to study BDNF expression
This chapter has given summary of four widely used rat models of depression and their translational power according to the validity criteria proposed by Belzung and Lemoine (2011). There is not one model of depression that models all aspects perfectly. Therefore, this chapter aimed to select and describe the most appropriate rat models for identifying possible changes of BDNF expression upon psilocybin administration. These models are the genetic models of depression, the FSL and WKY rats, and the chronic-stress models, the LH and CMS models. This selection is based on literature on depression models, the biomarker validity regarding their BDNF expression and remission validity with regards to the reversal of BDNF expression decreases after currently available antidepressant treatments.
4. Research proposal
We aimed to answer the question of how animal models can contribute to investigate the potential antidepressant effects of psilocybin. Based on previously presented available literature, this chapter discusses the appropriate animal model of depression for this question. Then, it describes the appropriate dose of psilocybin. Finally, two experiments are suggested to test the hypothesis that a single dose of psilocybin reverses a decrease in hippocampal BDNF expression over time.
Discussion on usability of the described models of depression
Each time a research question is asked, the most appropriate model has to be selected based on available literature, the question, the practical possibilities in the lab and the ethical approval of the experiments. In chapter three, four animal models of depression were identified that met the biomarker validity criterion of reduced hippocampal BDNF expression levels and the remission validity criterion, because the BDNF expression reduction could be reversed by antidepressant treatments in all four models. However, several other factors of these models have to be taken into account when selecting the most appropriate model for this research question. This section argues which of them is the most appropriate genetic and stress-induced depression model.
When comparing the genetic rat models of depression regarding homological validity, they appear similar because they are both selectively bred rat models. The face, biomarker and remission validity of the FSL rat are high. Most importantly, the neurotrophic hypothesis related variables are representative for depressed patients (Elfving, Plougmann, Müller, et al., 2010). Even though some antidepressants induced BDNF expression in WKY rats, the remission validity is lower, because of the variable responsiveness to antidepressant treatments. Therefore, WKY is suggested to model TRD. However, this does not cause a problem since positive results with psilocybin have been shown in TRD patients (Carhart-Harris et al., 2016) and WKY rats (Nichols & Hibicke, 2018). This preliminary finding of enduring antidepressant effects of psilocybin in WKY rats also advocates in favour of the WKY rat. The FSL and WKY rats both show substantial advantages over each other. Therefore, based on these arguments, no decision can be made between them. Consequently, both genetic models are included in the proposed future experiment.
The LH and CMS models are both remarkable chronic-stress models. Both show excellent levels of ethological, biomarker, face and remission validity. The largest difference between the two can be found in the triggering validity
criterion. Both use uncontrollable stressful events as a triggering factor, but in the CMS model, the stressor changes. This adds an extra dimension of stress to the model, because it is unpredictable. This is an advantage over the LH model, because it resembles the etiology of depression in people. A disadvantage is that using several different stressors could result in heterogeneity of involved brain circuitries, complicating the experiment. Besides, using several different stressors contributes to the replication difficulty of the experiments. However, based on the validity criteria described in chapter 2, the CMS model is the most appropriate chronic stress model to answer this research question.
1.0 mg/kg psilocybin is predicted to induce antidepressant effects
Finding the right dose of psilocybin is essential for the success of the experiments. When the dose is too low, the induction of neuroplasticity might not be enough to result in detectable differences in BDNF expression or changes in depressant-like behaviour. On the other hand, when the dose is too high, animals become ataxic (Geyer, Light, & Rose 1979; Halberstadt, Koedood, Powell, & Geyer, 2011).
In clinical studies, the effective dose of psilocybin was determined to be between 0.045 and 0.429 mg/kg body weight orally or 1 to 2 mg per adult intravenously (review by Tylš et al., 2014). The three successful clinical trials described in the introduction used 0.2 mg/kg (Grob et al., 2011), a moderate dose of 0.3 mg/kg (Ross et al., 2016) and high dosage of 0.31 or 0.43 mg/kg (Griffiths et al., 2016). To mimic the observed effects of drugs in patients in animal models, the appropriate dose has to be calculated. Several aspects have to be taken into account, such as differences in body weight, body surface area and species-specific pharmacokinetics. Allometric scaling is a widely used empirical method to calculate the appropriate dose of drugs in rats based on the normalization of the dose to the body surface area (review by Nair & Jacob, 2016). However, psilocin has a higher affinity for the human 5-HT2A receptor compared to the rat 5-HT2A receptor (Gallaher et al., 1993). This obstructs the allometric scaling calculation of the dosage.
Therefore, instead of calculating the dose based on patient studies, the dose can be based on other experimental animal studies. In vitro, changes in neuroplasticity were found using a concentration of 10 µM (0.1% DMSO) (Ly et al., 2018). During behavioural experiments with rodents, the used dosage of psilocybin and psilocin is generally between 0.25 and 10 mg/kg (review by Tylš et al., 2014). Even micro dosing with 0.05 or 0.075 mg/kg produced some observable effects in Wistar rats in the EPM paradigm (Horsley et al., 2018). A recent experiment showed that an intraperitoneal injection of 1.0 mg/kg results in antidepressant effects in Sprague Dawley rats and WKY rats 42 days later (Nichols & Hibicke, 2018). This experiment is the most relevant to
our research question, so we propose a single intraperitoneal injection of 1.0 mg/kg psilocybin.
Experiment 1: BDNF expression potentially increased by psilocybin
We propose to begin with an experiment with the genetic rat models, because these models are less complicated and labour-intensive. The animals are easy to breed and no additional actions are necessary to induce the depressive-like state. Two rat strains are investigated: FSL rats and WKY rats. The corresponding controls are a FRL group and a Wistar group respectively. The independent variable is a single intraperitoneal injection of 1.0 mg/kg of psilocybin. One dependent variable is the hippocampal BDNF expression. To determine the BDNF expression levels, both endogenous protein concentration and mRNA levels are analysed. The BDNF protein concentration is measured by two complementing techniques to ensure robust results. These techniques are an enzyme-linked immunosorbent assay (ELISA) as described by Elfving, Plougmann, and Wegener (2010) and Western blotting according to Rogóż, Kamińska, Pańczyszyn-Trzewik, and Sowa-Kućma (2017). The mRNA levels are determined using real-time quantitative polymerase chain reaction (qPCR) as described by Elfving, Plougmann, Müller, et al. (2010). The other dependent variable is depressive-like behaviour. Since FSL rats do not present anhedonic behaviour, the selected depressive-like behaviour in this model is the level of behavioural despair measured as time spent immobile in the FST.
FSL rats, FRL rats, WKY rats and Wistar rats that are five weeks old are habituated to their cages for three weeks. Then, the basal level of depressive-like behaviour is determined by performing the FST (Figure 2). The rats are also subjected to the open field test (OFT) right before the FST. In this test, the time and velocity of their movements are measured, to make sure the immobility in the FST is not caused by dysfunctional locomotion. Then, of each group, rats are sacrificed to be able to measure the basal hippocampal BDNF expression. After this experiment, the rats of each strain are divided into an experimental and a control group. Three days later, the experimental group receives a single intraperitoneal injection of 1.0 mg/kg of psilocybin and the control group receives saline. The peak of behavioural changes is observed at 30 to 90 minutes (review by Tylš et al., 2014). Therefore, one day after the injection the time spent immobile in the FST is tested, preceded by an OFT. At this time point, the behaviour observed is most likely not due to effects of remaining psilocybin in the brain. To be able to measure the enduring effects of psilocybin, these tests are repeated four weeks after the psilocybin injection. After this final test, the rats are sacrificed to measure the hippocampal BDNF protein and mRNA levels.
