MDMA-assisted psychotherapy for PTSD: Growing evidence for memory effects mediating treatment efficacy

The application of MDMA in conjunction with psychotherapy has in recent years seen a resurgence of clinical, scientific, and public interest in the treatment of posttraumatic stress disorder (PTSD). Clinical trials have shown promising safety and efficacy, but the mechanisms underlying this treatment form remain largely unestablished. This article explores recent preclinical and clinical evidence suggesting that the treatment ’ s efficacy may be influenced by the mnemonic effects of MDMA. We review data on the effects of MDMA on fear extinction and fear reconsolidation and the utility of these processes for PTSD treatment. We corroborate our findings by incorporating research from cognitive psychology and psychopharmacology and offer recommendations for future research .


Introduction
In recent years, there has been a revitalization of scientific, clinical, and public interest in the employment of psychedelic substances adjunctive with psychotherapeutic interventions for the treatment of mental disorders.Although not conventionally regarded as a traditional psychedelic substance, 3,4-methylenedioxymethamphetamine (MDMA) in conjunction with psychotherapy has re-emerged as a promising treatment modality for posttraumatic stress disorder (PTSD).PTSD is a trauma-and stressor-related disorder characterized by re-experiencing, avoidance, and negative alterations in cognition, mood, arousal, and reactivity pertaining to the traumatic experience(s) (American Psychiatric Association, 2013).In preliminary clinical trials of MDMA-assisted psychotherapy, the findings have been compelling with approximately half of all participants with chronic, moderate-to-severe, and treatmentrefractory PTSD no longer qualifying for the diagnosis one-to-two months subsequent to termination of treatment (Jerome et al., 2020;Mithoefer et al., 2019) and two-thirds of all participants no longer qualifying 12 months after treatment cessation (Gorman et al., 2020;Jerome et al., 2020).Long-term follow-up studies show that these improvements are sustainable at one-to-six years following treatment termination with an average of four years (Jerome et al., 2020;Mithoefer et al., 2013).
Studies within cognitive psychology suggest that the memory system exerts a significant influence in mediating PTSD development and maintenance (Rubin et al., 2008a).In this article, we will review preclinical and clinical data from studies on the mnemonic effects of MDMA and augment our analyses with studies from cognitive psychology on the role of memory in PTSD.By combining this literature, we propose that some mechanisms of action can be the effects of MDMA on traumatic memories.Before discussing the findings of MDMA and memory, we commence our review with an examination of the role of a traumatic memory in mediating PTSD with a particular focus on opposing perspectives on memories of traumatic experiences and their influence on PTSD pathogenesis, which will augment our later discussions.

The role of memory in PTSD
For several decades, it has been proposed within the psychotraumatic literature that memories of traumatic experiences are generally fragmented, incoherent, and disorganized.The cause primarily revolves around impaired processing and integration of the traumatic experience into memory, which results in disrupted recollection of the event (Berntsen, 2009;Porter and Birt, 2001;Zoellner and Bittenger, 2004).It is posited that a traumatic experience undergoes a bifurcation into two different memory systems resulting in impaired voluntary but enhanced involuntary access to the traumatic memory.These effects are more pronounced for traumatic memories in individuals with PTSD than without (Rubin et al., 2008b).This notion is prominent in several clinical theories of PTSD (e.g.Brewin et al., 1996;Ehlers and Clark, 2000;Foa et al., 1989;Janoff-Bulman, 1989;van der Kolk and Fisler, 1995) (see Dalgleish, 2004, for a review).Impaired traumatic memory is also one of the core symptoms of PTSD enumerated in the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) (American Psychiatric Association, 2013) and the International Classification of Diseases (ICD-11) (World Health Organization, 2019).
Multiple studies (e.g.Amir et al., 1998;Foa et al., 1995;Halligan et al., 2003;Murray et al., 2002;van der Kolk and Fisler, 1995) have supported this assumption of impaired memory of traumatic experiences in individuals with PTSD, e.g. by measuring and comparing coherence and articulation of trauma narratives.Special trauma-exclusive mechanisms such as peritraumatic dissociation and repression are adduced as causes of impaired processing and integration of the traumatic event into memory, which these clinical theories of PTSD associate with PTSD pathogenesis and maintenance (Berntsen and Rubin, 2007;Berntsen et al., 2003;Rubin et al., 2008b).The notion that special traumaexclusive mechanisms result in disrupted encoding and integration of traumatic experiences in memory yielding differential accessibility for recollection is often designated as the "Special Mechanisms View" of traumatic memories (Rubin et al., 2008b).
In recent years, research has begun to sow doubts and challenge this idea about the nature of traumatic memories.While the findings of multiple studies have supported this assumption of impaired memory (e. g.Amir et al., 1998;Foa et al., 1995;Halligan et al., 2003;Murray et al., 2002;van der Kolk and Fisler, 1995), multiple other studies and reviews have failed to support and/or replicate these findings (e.g.Berntsen and Rubin, 2014;Berntsen et al., 2003;Gray and Lombardo, 2001;Lorenzzoni et al., 2014;Porter and Birt, 2001;Reviere and Bakeman, 2001).Many of the studies supporting the "Special Mechanisms View" have been criticized for, among other factors, poor methodology, diverging or inadequate operationalizations and measurements of narrative components such as "fragmentation", lack of controlling for relevant confounds, an absence of testing alternative hypotheses, and an absence of proper control groups (Berntsen, 2009(Berntsen, , 2015;;Berntsen and Rubin, 2014;Berntsen et al., 2003;Crespo and Fernández-Lansac, 2016;O'Kearney and Perrott, 2006;Rosen and Lilienfeld, 2008;Zoellner and Bittenger, 2004).While it is acknowledged that instances of impoverished traumatic memory can occur, the causes of these instances are erroneously adduced as trauma-exclusive mechanisms rather than basic mechanisms that affect all types of memories (Berntsen and Rubin, 2014;McNally et al., 2004;Rosen and Lilienfeld, 2008;Rubin et al., 2008a).
Contrary to the assumption of disrupted traumatic memory, studies support that, in most cases, traumatic memories are particularly wellremembered with pronounced vividness, emotional intensity, and sensory reliving (Berntsen et al., 2003;Lorenzzoni et al., 2014;Porter and Birt, 2001;Reviere and Bakeman, 2001).Traumatic experiences are distinctive, rare, emotionally intense, frequently rehearsed, and involve the release of stress hormones.These are among the mechanisms posited to contribute to enhanced encoding and consolidation of a traumatic experience into memory, which subsequently results in potentiated accessibility of the traumatic memory for both voluntary and involuntary recall (Berntsen, 2009;Rubin et al., 2008b).These factors contribute to the traumatic experience being perceived as a turning point in the victim's life story by profoundly changing their life course and a reference point for generating expectations for future events and interpretations of other experiences.In most cases, the traumatic experience becomes a core, emblematic component of the trauma victim's identity (Berntsen andRubin, 2006a, 2007).
Basic factors such as personality traits, gender, educational level, socioeconomic status, and more, in addition to the aforementioned mechanisms, interact in influencing the memory of a traumatic event (Rubin et al., 2008a).Traumatic memories are influenced by an interplay of factors pertaining to the traumatic event and the individual experiencing the event.These factors influence the memory of any type of experience, not just traumatic experiences.They are not inherently pathological or trauma-exclusive mechanisms that selectively impact traumatic memories, for which there is limited evidence to support.This view that the memory of a traumatic experience is influenced by basic rather than trauma-special mechanisms is often denoted as the "Basic Mechanisms View" of traumatic memories (Berntsen andRubin, 2006a, 2007;Rubin et al., 2008a;Rubin et al., 2008b).
PTSD is distinct from most disorders in the DSM-5 (American Psychiatric Association, 2013) and the ICD-11 (World Health Organization, 2019) given that its aetiology is specified.Clinicians and researchers can causally presume that one or multiple traumatic experiences preceded the development of PTSD (traumatic experience → PTSD).In "The Mnemonic Model" of PTSD, Rubin et al. (2008a) introduce an extra component in this conceptualization of PTSD, which is the memory of the event and reactions to it, which change with time and situation (traumatic experience → traumatic memory → PTSD).The model presupposes that basic rather than special mechanisms account for the memory of a traumatic event (Rubin et al., 2008a).Indeed, memory mediates several key symptoms of PTSD.Arousal and distress from reminders, intrusive recollections, avoidance, and flashbacks of the traumatic event are all predicated on some form of memory representation of the traumatic experience (Berntsen and Rubin, 2006b;Simonelli, 2013).Some researchers have even characterized PTSD as a memoryrelated disorder (Dalgleish, 2004;Rubin et al., 2008a).The human memory is a complex system and is often conceptualized in many different and often dichotomous and independent subsystems, e.g.implicit versus explicit memory (Gaensbauer, 2002).However, it is important to stress that memory is holistic and integrative with complex and reverberating interconnections between various regions of the brain.Traumatic representations can be registered, encoded, and subsequently expressed through multiple channels, e.g.affectively, behaviourally, verbally, and symbolically (Gaensbauer, 2002).
Implicit memory of the traumatic event has been documented to mediate PTSD symptoms as exemplified by findings from paediatric PTSD, i.e. in infants and toddlers without matured cognitive abilities, including the capability for explicit traumatic memory storage and recollection (Coates, 2016;Gaensbauer, 2002;Paley and Alpert, 2003;Simonelli, 2013;Woolgar et al., 2022).Nonetheless, findings from cognitive psychology suggest that explicit memory, particularly the autobiographical memory system, exerts a significant role in mediating PTSD pathogenesis (Rubin et al., 2008a).Autobiographical memories are memories of personally significant experiences from one's past.They are the foundation of one's identity, personality, self, and schemas (Conway and Pleydell-Pearce, 2000;Williams et al., 2008;Williams et al., 2007).The autobiographical memory system is posited to begin emerging from the pre-school years (i.e.three-to-five years of age) onwards (Fivush, 2011).PTSD has been proposed to reflect a disorganization of autobiographical memory and identity induced by the traumatic event (Berntsen and Rubin, 2006b).The autobiographical memory system permits the expression of personal experiences in the form of verbal recall and ensures the formation of retrievable and coherent autobiographical narratives of traumatic events (Simonelli, 2013).
This view that PTSD reflects a disorganization of the autobiographical memory system is in accordance with multiple findings within cognitive psychology.PTSD has been associated with overgeneral retrieval of autobiographical memories, and overgeneral autobiographical memory following traumatic exposure is also predictive of subsequent PTSD development (Bryant, 2015;Lorenzzoni et al., 2014;Moore and Zoellner, 2007;Ono et al., 2016).This view is also in accordance with research into the centrality of a traumatic experience for one's identity and life story.The Centrality of Event Scale (CES) by Berntsen andRubin (2006a, 2007) is a tool to measure the degree to which a traumatic or stressful event is central to one's identity, a turning point in one's life story, and a reference point in one's autobiographical memory for the organization and interpretation of other memories and expectations for the future (Berntsen andRubin, 2006a, 2007).These are common characteristics of highly accessible memories, particularly traumatic memories, and the CES is therefore a measure of the general accessibility of a traumatic memory.The CES should therefore correlate with the ease with which a traumatic memory is involuntarily recalled and voluntarily accessed (Rubin et al., 2008a;Rubin et al., 2008b).The CES and its individual components (i.e.centrality, reference point, selfdefinition) have been positively correlated with PTSD symptoms (e.g.intrusions, avoidance, hyperarousal) and with enhanced recollection of the traumatic event (e.g.sensorial vividness, emotional intensity, visceral reactions), even when controlling for potentially confounding variables such as comorbid disorders (Berntsen andRubin, 2006a, 2007;Berntsen and Rubin, 2008;Berntsen et al., 2003;Boals, 2010;Gehrt et al., 2018;Lorenzzoni et al., 2014).
These findings contradict the assumption commonly emanating from the "Special Mechanisms View" of impoverished integration of traumatic experiences into memory and differential subsequent accessibility for recall (Berntsen andRubin, 2006a, 2007).There is an inadequate and unpersuasive empirical basis for this assumption of a relationship between impaired traumatic memory and PTSD.Indeed, the inability to recall important aspects of a traumatic experience correlates poorly with other PTSD symptoms (Rubin et al., 2008b).There is a positive correlation between the level of PTSD symptoms and the degree to which a traumatic memory is perceived as central to the trauma victim's life story and identity and serves as a cognitive reference point for other experiences.PTSD symptoms decrease when traumatic memory accessibility decreases and increase when accessibility increases (Rubin et al., 2008a;Rubin et al., 2008b).In short, PTSD symptoms correlate positively with the traumatic incident being highly accessible for recollection; central to one's identity, self, and life story; and a reference point in one's autobiographical memory.Based on these findings that the accessibility and centrality of a memory contribute to PTSD symptom severity, it is a corollary that an intervention that successfully diminishes the accessibility of the traumatic memory also targets PTSD symptomatology.Accordingly, Rubin et al. (2008a) found support for this in their review of several studies of brain injury-induced amnesia, childhood amnesia, and pharmacologically induced amnesia of the traumatic memory (Rubin et al., 2008a).These studies support that inducing amnesia, insofar as it targets traumatic memory, can abate PTSD symptom severity.These reviewed autobiographical memory alterations (i.e.overgeneralization, enhanced centrality, enhanced accessibility) all significantly relate to PTSD development and maintenance.In contrast, these alterations are less pronounced in traumatized individuals without PTSD (Lorenzzoni et al., 2014).
Collectively, these findings support the involvement of the autobiographical memory system in mediating PTSD development and maintenance.However, by emphasizing the importance of the autobiographical memory system, we do not disregard the involvement of other memory systems or non-mnemonic mechanisms in mediating PTSD symptoms.For example, incidents of PTSD in infants and toddlers without matured cognitive abilities negate that PTSD development is fully predicated upon the autobiographical memory system.These results suggest that the autobiographical memory system is influential to, but not the sole determiner of, PTSD development and maintenance.It is also important to acknowledge that the biological and psychological mechanisms of memory, including learning, consolidation, and storage of information, still remain largely empirically unfolded (Camina and Güell, 2017;Crystal and Glanzman, 2013;Kandel, 2009;Ortega-de San Luis and Ryan, 2022).PTSD has been associated with alterations in multiple biological, social, psychological, and behavioural domains, including epigenetic, genetic, molecular, cellular, and brain changes (Bremner, 2006;Cao-Lei et al., 2022;Girgenti et al., 2017;Kim et al., 2018).Understanding PTSD development, maintenance, and treatment, including the relationship between trauma memory and PTSD, necessitates an integrative, cross-disciplinary approach that combines knowledge from various scientific fields beyond what an exclusive focus on the autobiographical memory system can offer.
To recapitulate, clinical theories of PTSD often posit that special trauma-specific mechanisms result in inadequate integration of traumatic experiences into memory and impaired voluntary recall of the event, but studies cumulatively do not support this assumption.On the contrary, traumatic memories often exhibit pronounced accessibility for both voluntary and involuntary recollection.The memory of a traumatic experience is more parsimoniously claimed to be governed by basic mechanisms that affect the memory of any other type of experience, including non-traumatic experiences.The autobiographical memory system exerts a significant influence in mediating PTSD development and maintenance following traumatic exposure.The perception of a traumatic event as central to one's identity, self, and life story; the use of the traumatic event as a reference point in one's autobiographical memory; and enhanced recollection of the traumatic memory all correlate positively with PTSD symptoms.It therefore follows that targeting the centrality and accessibility of the traumatic memory could benefit PTSD treatment.

The pharmacology of MDMA
In order to examine the effects of MDMA on traumatic memory, we first commence with an examination of the pharmacological properties of the substance.Synthesizing the empirical literature pertaining to the pharmacological properties of MDMA will augment understanding the mechanisms involved in mediating its effects on memory.
In humans, studies have namely employed a pre-treatment paradigm, e.g. using serotonin transporter (SERT) or noradrenaline transporter (NET) inhibitors, to elucidate the pharmacological properties of MDMA and their associations with the psychological and physiological effects of the substance.While research has often implicated various neurochemicals and neuromodulatory systems in mediating the effects of MDMA, the effects of MDMA on serotonergic (5-HT), noradrenergic (NE), and dopaminergic (DA) processes have been the principal targets of the pharmacological studies.The pharmacological properties of MDMA pertaining to the 5-HT, NE, and DA systems and their relationships with the effects of MDMA in humans will be the main focus of the following section.However, we will also briefly review clinical literature pertaining to the involvement of cortisol and oxytocin in the effects of MDMA as these neurohormones will be relevant later on.
Generally, binding at the NET and SERT appear to account for most of the psychological and physiological effects of MDMA (Farré et al., 2007;Hysek et al., 2012b;Hysek et al., 2011;Hysek et al., 2012d;Liechti et al., 2000a;Liechti and Vollenweider, 2000a;Liechti and Vollenweider, 2000b;Tancer and Johanson, 2007).Stimulation of the α 2 receptors and ventricular release of NE appear inconsequential to the effects of MDMA (Hysek et al., 2012a).β and α 1 receptors have been found to contribute to the cardiostimulant and thermogenic effects of MDMA (Hysek et al., 2013;Hysek et al., 2012c;Hysek et al., 2010).One study found a minor influence of postsynaptic 5-HT 1A receptors on the subjective effects of MDMA (Hasler et al., 2009) while other studies found none (Kuypers et al., 2014;van Wel et al., 2011van Wel et al., , 2012) ) using pindolol as pre-treatment.However, since pindolol does not fully bind at either pre-or postsynaptic 5-HT 1A receptors (Rabiner et al., 2000), it may not have effectively blocked them.Some subjective and to a lesser extent physiological effects have been attributed to 5-HT 2A receptors (Liechti et al., 2000b;van Wel et al., 2012).
Findings suggest that SERT-mediated 5-HT release is more involved in the mood-related and subjective effects (e.g.Liechti et al., 2000a;Tancer and Johanson, 2007) and NET-mediated NE release is more involved in the stimulant-like and cardiovascular effects (e.g.Hysek et al., 2011) of MDMA.Two studies reported the involvement of D 2 receptors in some subjective and mood effects using haloperidol (i.e. a D 2 receptor antagonist) as pre-treatment (Liechti et al., 2001;Liechti and Vollenweider, 2000a).However, haloperidol administration alone increased self-rated state anxiety and decreased scores of well-being in one of these two studies (Liechti and Vollenweider, 2000a), which limits the validity of their conclusion.This highlights an important caveat with the pre-treatment studies: The administration of pharmacological substances as pre-treatment interventions could result in the introduction of novel effects and side effects that confound delineating the roles of individual neurotransmitter receptors, neurohormones, or monoamine transporters in mediating the effects of MDMA.In one study using clinically healthy human participants, intravenous injection of 1.4 mg of haloperidol 10 minutes before oral administration of 1.5 mg/kg MDMA failed to significantly impact the effects of MDMA on pre-pulse inhibition and startle habituation, but it did significantly reduce scores of selfreported positive mood and mania-like experience (Liechti et al., 2001).In another study, oral administration of 60 mg immediate-release methylphenidate, which elevates DA and NE levels by inhibition of their corresponding transporters, an hour before 125 mg MDMA administration produced the same subjective effects as MDMA alone, including self-reported measures of mood rating.It did, however, increase cardiovascular and adverse effects (Hysek et al., 2014b).These results suggest that binding at the dopamine transporter (DAT) and DA release exert a limited influence on the effects of MDMA.However, it is important to emphasize that other indirect processes may affect dopaminergic mechanisms upon administration of MDMA.For example, NE and 5-HT release through NET-and SERT-binding, respectively, and 5-HT 2 receptor activation have also been associated with MDMA-induced increased DA release in rats, independent of DAT-binding (Gudelsky and Nash, 1996;Gudelsky and Yamamoto, 2008;Gudelsky et al., 1994).
Numerous animal studies, especially in rodents, have also explored the pharmacological properties of MDMA.While non-human translational models unarguably offer rich avenues for neurobiological research, it is also important to emphasize the notorious difficulties of inter-special translational research in MDMA (see de la Torre and Farré, 2004;Easton and Marsden, 2006;Green et al., 2012, for reviews).For example, studies have associated MDMA with impaired pre-pulse inhibition in rodents (Marti et al., 2019;Stove et al., 2010), but with enhanced pre-pulse inhibition in humans (Liechti et al., 2001;Vollenweider et al., 1999).Therefore, translational behavioural models should be used with caution as the validity of generalizing preclinical results to humans in vivo may be limited.
The pharmacological profile of MDMA in vivo and in vitro in animals and in vitro in human cells show similar trends as those outlined previously.MDMA promotes the release of the three monoamines more potently than it inhibits their reuptake.There is a clear trend towards greater release and inhibition of reuptake of 5-HT and NE than DA through binding and inverting at their corresponding transporters.In terms of receptors, studies report higher affinity for 5-HT 2B , 5-HT 1A , 5-HT 2A , and α 2 receptors compared with other receptor subtypes, but binding at SERT and NET appear to account for most of the effects of MDMA (Battaglia et al., 1988;Green et al., 2003;Gudelsky and Yamamoto, 2008;Hysek et al., 2012d;Rothman and Baumann, 2002;Simmler et al., 2012;Verrico et al., 2007).
Clinical studies have found MDMA to increase cortisol (e.g.Kuypers et al., 2013Kuypers et al., , 2014) ) and oxytocin (see Dumont & Verkes, 2006, for a review) levels.Increased cortisol levels were associated with an index of subjective rewarding effect in one study (Harris et al., 2002), but others failed to find any association between increased cortisol levels and the subjective, prosocial, and social cognitive effects of MDMA (Bershad et al., 2017;Hysek et al., 2012bHysek et al., , 2014a;;Kuypers et al., 2013).Oxytocin receptor gene variants (Vizeli & Liechti, 2018) and increased oxytocin levels (Dumont et al., 2009) have been implicated in some prosocial effects of MDMA, and a recent study (Atila et al., 2023) found that MDMA-induced increased oxytocin concentrations were associated with the subjective prosocial, empathetic, and anxiolytic effects of MDMA, suggestive of a more paramount role of oxytocin in mediating the subjective effects of MDMA.However, others have failed to find any significant correlations between increased oxytocin levels and the social cognitive, social behavioural, empathogenic, and affective effects of MDMA (Kirkpatrick et al., 2014a(Kirkpatrick et al., , 2014b;;Hysek et al., 2012bHysek et al., , 2014a;;Kuypers et al., 2014;Schmid et al., 2014).
To recapitulate, findings from human studies suggest that binding at SERT and NET account for most of the effects of MDMA.Serotonergic and noradrenergic mechanisms appear more involved than dopaminergic mechanisms in mediating the effects of MDMA.Preclinical findings show similar patterns.There is a paucity of clinical findings supporting the involvement of increased cortisol levels in mediating the psychological effects of MDMA, and findings are mixed in terms of the influence of increased oxytocin levels on the subjective effects induced by MDMA.The pharmacological profile of MDMA remains notoriously enigmatic with limited and tentative extant evidence from both human participants and non-human subjects for any definite conclusion pertaining to its pharmacological profile.

Fear reconsolidation and fear extinction
In the previous section, we discussed some of the findings pertaining to the pharmacological profile of MDMA.In the following, we expound upon empirical literature on the processes of fear extinction and fear reconsolidation before examining recent preclinical and clinical findings of the effects of MDMA on these processes.
Common proposals of the underlying cognitive mechanisms of MDMA-assisted psychotherapy are effects on fear extinction and reconsolidation (Feduccia et al., 2018;Feduccia et al., 2019;Feduccia and Mithoefer, 2018;Hake et al., 2019;Maples-Keller et al., 2022;Vizeli et al., 2022).During a traumatic event, inherently aversive, traumarelated stimuli may emerge concurrent with inherently non-aversive, neutral stimuli and elicit fear responses.In Pavlovian fear conditioning literature, these neutral stimuli (conditioned stimuli, CS) may subsequently inherit the ability of the trauma-related stimuli (unconditioned stimuli, US) to elicit fear responses (conditional response, CR).Learning this association is known as fear acquisition, but the subsequent storage of this association in long-term memory is called fear consolidation (Auchter et al., 2017).
Fear extinction involves exposing a conditioned subject to the fear responses-eliciting, but inherently non-aversive, conditioned stimuli in the absence of the inherently aversive, unconditioned stimuli to diminish the ability of the conditioned stimuli to elicit fear responses (Auchter et al., 2017;Myers and Davis, 2007).The result is not an alteration in the original association between the conditioned stimuli and fear responses (i.e.unlearning), but the formation of a new association (i.e.new learning) between the conditioned stimuli and the absence of fear responses (Vaverková et al., 2020).This new association takes precedence over the original association upon subsequent exposure to the conditioned stimuli (Kida, 2019;Myers and Davis, 2007).
Following the consolidation of a memory, activation of the memory trace is posited to instigate a trinary sequence of processes: retrieval, labilization to maintenance or modification, and restabilization (i.e.reconsolidation) of the memory into a more solidified state.This sequence is known as reconsolidation of memory.It has been proposed that a memory can be altered by modifying it during the labile state or by obstructing its reconsolidation following activation (Auchter et al., 2017;Kida, 2019;Vaverková et al., 2020).Thus, reconsolidation and extinction differ mechanistically as extinction involves the formation of a new inhibitory memory while reconsolidation involves alterations of the original memory.
Patients with PTSD often display extinction impairments (Maren and Holmes, 2016).There is theoretical and clinical support that enhancing fear extinction, e.g. by prolonged exposure (PE) therapy, is beneficial for PTSD treatment (Kida, 2019).However, extinction does not affect the original fear contingencies.Relapse can therefore be common as the original fear contingencies can resurface (Maren and Holmes, 2016), unlike reconsolidation interference which targets and modifies the original memory trace itself.Indeed, in some studies, pharmacologically induced post-retrieval amnesia of fear memories appear to produce more long-lasting reductions in fear memory expression than extinction (Kindt and Elsey, 2023).Thus, reconsolidation disruption may be more effective in some instances than enhancement of extinction.The conditions necessary for disrupting the reconsolidation of memories and the specific aspects of the memory that are targeted by reconsolidation disruption remain elusive (Kindt and Elsey, 2023;Vaverková et al., 2020).
Extinction and reconsolidation appear contingent upon an extensive array of common and differential biological, psychological, and contextual processes (Alberini et al., 2006;Fiorenza et al., 2011;Gonzalez et al., 2019;Suzuki et al., 2004;Vaverková et al., 2020) (see Baldi and Bucherelli, 2015, for review).Fear extinction and reconsolidation both require re-exposure to stimuli triggering the recall of the fear memory.Both processes also necessitate a mismatch between the original memory and the novel experience (i.e. a violation of expectations), which is known as a prediction error (Vaverková et al., 2020).Extinction and reconsolidation are putatively mutually distinct and exclusive processes (Baldi and Bucherelli, 2015;Merlo et al., 2014;Vaverková et al., 2020), but others have proposed that there is some interaction between the two processes (Suzuki et al., 2004).

The effects of MDMA on memory: recent evidence on extinction and reconsolidation
Several studies have been conducted to examine the mnemonic effects of MDMA.Out of these studies, we identified four preclinical studies (Arluk et al., 2022;Hake et al., 2019;Young et al., 2015;Young et al., 2017) and two clinical studies (Maples-Keller et al., 2022;Vizeli et al., 2022) that examined the effects of MDMA on fear extinction and/ or fear reconsolidation.In the following, we expound upon the findings of these studies before discussing their benefits for PTSD treatment within the context of MDMA-assisted psychotherapy.
In Young et al. (2015), the researchers employed a translational behavioural model using mice as subjects to examine the effects of MDMA on fear memory and the pharmacological mechanisms underlying these memory effects.For fear conditioning, mice were exposed to one or multiple pairings of a single tone as the conditioned stimulus (CS) and a foot shock as the aversive, unconditioned stimulus (US).Upon exposure to the conditioned stimulus (i.e.tone), fear responses were measured in the percentage of time spent freezing (i.e. the presence or absence of non-respiratory movements) relative to the time spent in the experimental apparatuses (Young et al., 2015).Mice were exposed to fear conditioning on day 1, extinction training on day 3, and extinction testing on day 4 (and day 10 for assessment of long-term effects) (Young et al., 2015).Young et al. (2015) injected 0, 3.0, 5.6, and 7.8 mg/kg MDMA intraperitoneally into the mice 30 minutes before extinction training.Young et al. (2015) found that 7.8 mg/kg significantly reduced conditioned freezing during extinction training.Mice in both the 5.6 and 7.8 mg/kg groups exhibited significantly reduced freezing at extinction testing 24 hours after extinction training.The extinction enhancements observed 24 hours after extinction training in the 7.8 mg/kg group were persistent at 10 days follow-up.The extinction effects of 7.8 mg/kg MDMA were observed when extinction testing was carried out in the same context as extinction training as well as when it was carried out in a different, novel context, reflecting a generalization of extinction.7.8 mg/kg MDMA administered immediately following extinction training did not affect freezing the following day (Young et al., 2015).Young et al. (2015) also examined whether the observed effects were due to interference with reconsolidation.7.8 mg/kg MDMA administered 30 minutes before re-exposure to a single conditioned stimulus significantly reduced freezing, but the effects did not persist 24 hours later at extinction testing, indicating no effects on reconsolidation (Young et al., 2015).In mice that exhibited enhanced fear extinction following MDMA in combination with extinction training, brain-derived neurotropic factor (BDNF) mRNA expression was increased in the basolateral amygdala but not in the medial prefrontal cortex one-hour post-extinction.BDNF expression was not increased in animals that did not exhibit enhanced extinction (Young et al., 2015).
Infusing 7.8 mg/kg MDMA directly into the infralimbic subregion of the medial prefrontal cortex or basolateral amygdala (i.e.subregions required for extinction learning) in combination with extinction training increased extinction when tested 24 hours later (Young et al., 2015).Young et al. (2015) found that disrupting BDNF signalling in the basolateral amygdala by injecting 0.2 μg BDNF-neutralizing antibody prevented the effects of MDMA on extinction.Accordingly, studies have also shown that BDNF in the basolateral amygdala aids in the consolidation of extinction memory (Young et al., 2015).MDMA administration only increased BDNF expression in the basolateral amygdala when MDMA was combined with extinction training, which could explain the inefficacy of post-extinction training administration of MDMA on extinction.In conclusion, Young et al. (2015) found that MDMA combined with extinction training enhanced the extinction of conditioned fear responses mediated by BDNF expression in the basolateral amygdala.Young et al. (2015) hypothesized that MDMA's ability to rapidly increase NE and 5-HT levels and binding at 5-HT 2A receptors could mediate its effects on BDNF and thereby fear extinction.These pharmacodynamic hypotheses were investigated by Young et al. (2017).They examined the effects of 3.0, 5.6, and 7.8 mg/kg MDMA on fear extinction using auditory fear conditioning to measure effects on freezing behaviour as well as fear-potentiated startle (FPS).Conditioned freezing is a more stereotypic fear behaviour in rodents while FPS is a highly conserved fear behaviour across species, including in humans.For the extinction of conditioned freezing, Young et al. (2017) employed a similar procedure to Young et al. (2015).Mice were exposed to cued fear conditioning on day 1, extinction training on day 3, and extinction testing on day 4. Mice were exposed to multiple tones (CS) followed by foot shocks (US) for fear conditioning.Saline or MDMA was administered 30 minutes before extinction training in these experiments (Young et al., 2017).For the experiments with FPS, mice were placed in startle chambers on days 1-3 for acclimation and exposed to multiple white noise startle stimuli.On day 4, mice were given an acclimation period in the chamber followed by multiple startle stimuli and then multiple trials of an auditory tone (CS) followed by the startle stimuli (Young et al., 2017).On day 5, mice underwent fear conditioning in the form of multiple CS+US pairings.FPS was assessed on day 6 by exposing mice to interspersed startle stimuli alone and CS plus startle tone pairings.On day 7, the effects of MDMA on within-session extinction of FPS were assessed by administering vehicle or MDMA 20 minutes before placing mice in the startle chamber.Extinction testing was conducted on day 8 (Young et al., 2017).
In multiple experiments, Young et al. (2017) administered selective pharmacological inhibitors of NET (reboxetine: 0, 1, 5, and 10 mg/kg), DAT (RTI-336: 0, 3, 5, and 10 mg/kg), SERT (citalopram: 0, 5, 10, and 20 mg/kg), or a control drug vehicle 30 minutes prior to MDMA treatment.0, 3, 5.6, and 7.8 mg/kg fenfluramine (i.e. a 5-HT-releasing agent, agonist of various 5-HT receptors, and positive modular of σ 1 receptors (Samanta, 2022)) was injected 30 minutes before extinction training in the absence of MDMA administration to evaluate the influence of 5-HT release on fear extinction.Young et al. (2017) also examined the effects of chronic citalopram.For experiments on the effects of citalopram on freezing, mice were treated with 10 mg/kg intraperitoneal citalopram for 22 days following fear conditioning.Extinction training was conducted two days after the final chronic citalopram treatment.In separate FPS experiments, mice underwent fear conditioning on day 1 followed by 22 days of citalopram treatment.Fear conditioning testing was conducted on day 25.Within-session extinction training was conducted on day 26 in which mice were treated with MDMA, and betweenextinction (retention) testing occurred on day 27 (Young et al., 2017).
7.8 mg/kg MDMA significantly reduced freezing during extinction training and testing.Only acute pre-treatment with 10 mg/kg citalopram significantly interfered with MDMA-induced reductions of conditioned freezing at extinction testing.Chronic injections of 10 mg/ kg citalopram blocked MDMA's reductions in freezing during training and testing but did not significantly affect conditioned freezing when administered alone without MDMA (Young et al., 2017).Citalopram administered 24 hours before extinction training did not influence the effects of MDMA on freezing at extinction testing.Fenfluramine dosedependently increased freezing at extinction training, but these effects did not persist the following day upon re-exposure to the conditioned stimuli.Mice treated with injections of 1 mg/kg of the selective 5-HT 2A/ 2C receptor agonist 2,5-dimethoxy-4-iodoamphetamine (DOI) 48 hours prior to the last chronic dose of citalopram exhibited significantly less DOI-induced head-twitch behaviour.This behaviour is argued to be dependent on 5-HT 2A receptors (Young et al., 2017).Young et al. (2017) therefore administered the selective 5-HT 2A receptor antagonist MDL-100,907 30 minutes prior to MDMA treatment to evaluate the influence of 5-HT 2A receptors in MDMA-induced extinction effects.MDL-100,907 inhibited the effects of MDMA at extinction training and significantly reduced the effects of MDMA on conditioned freezing 24 hours later.
All three doses of MDMA administered before extinction training significantly reduced FPS at extinction training, but only 7.8 mg/kg MDMA resulted in significantly less FPS the subsequent day at extinction testing.The reductions of FPS by MDMA were obstructed by chronic citalopram treatment (Young et al., 2017).Acute treatment with fenfluramine did not influence the effects of MDMA on fear extinction, suggesting a role outside potent 5-HT release.NET and DAT are also unlikely to be involved as neither reboxetine nor RTI-336 administered in combination with MDMA failed to interact with MDMA-induced enhanced fear extinction.However, it cannot be dismissed that other noradrenergic or dopaminergic mechanisms are involved in the observed effects of MDMA besides direct binding at NET and DAT (Young et al., 2017).Young et al. (2017) hypothesized that chronic citalopram treatment downregulated SERT availability and function and 5-HT 2A receptor expression and/or function.In sum, the findings of Young et al. (2017) show that MDMA decreased fear responses, and in combination with extinction training, MDMA enhanced long-lasting fear extinction as measured by reduced conditioned freezing and FPS.Young et al. (2017) hypothesized based on their pre-treatment experiments that SERT, 5-HT 2A receptors, and potentially other pharmacodynamic mechanisms mediated these extinction effects.
Hake et al. ( 2019) conducted a study using rats as subjects to examine the effects of MDMA on fear extinction and reconsolidation, which would evaluate the results of Young et al. (2015Young et al. ( , 2017)).Rats were exposed to either auditory or contextual fear conditioning followed by fear extinction training 24 hours later.Rats received either saline, or 1, 3, 5, or 10 mg/kg MDMA 30 minutes before auditory fear extinction training and either saline, or 3 or 5 mg/kg MDMA 30 minutes before contextual fear extinction training (Hake et al., 2019).Auditory or contextual fear extinction memory tests were employed 24 hours later, but an auditory fear extinction memory test was also conducted 7 days later.To examine the effects of these doses of MDMA on reconsolidation, cued fear memories were reactivated in the absence of fear extinction by either a single auditory conditioned stimulus or by re-exposure to the conditioned context for three minutes immediately after reactivation or 6 hours later (Hake et al., 2019).Saline or MDMA was administered after either type of re-exposure.Fear memory tests were identical to the fear extinction memory tests and occurred 24 hours and 7 days after extinction training.Rats were exposed to 10 seconds of an auditory stimulus (CS) followed by a foot shock (US) and then 1-minute intertrial-intervals for a total of 4 or 5 blocks of trials in the auditory fear experiments.In the contextual conditioning paradigm, rats were exposed to either 5 (in the contextual fear extinction experiment) or (in the contextual fear reconsolidation experiment) foot shocks (US).Freezing in the contextual fear experiments was measured in minutes (Hake et al., 2019).
In Hake et al. (2019), 3 and 10 mg/kg MDMA administered during auditory fear extinction training interfered with the retrieval of fear extinction memory at testing 24 hours later by increasing freezing.mg/kg MDMA almost reached similar significance.These increases in freezing ceased at extinction testing 7 days later.No effects were observed for contextual fear extinction.No effects on reconsolidation were observed 24 hours later, but 7 days later, 3 and 5 mg/kg MDMA administered immediately after reactivation yielded less freezing than the saline group during the third trial block of the auditory fear memory test (Hake et al., 2019).For contextual reconsolidation, 5 mg/kg MDMA administered immediately after reactivation disrupted reconsolidation 24 hours later at the fifth and final 2-minute block.Rats that received mg/kg immediately after reactivation froze less than the 3 mg/kg group during the first and third 4-minute block and less than all other groups at the second 4-minute block 7 days later.Administration of MDMA hours after reactivation yielded no significant differences compared to the saline group, which suggests that MDMA affects fear responses through reconsolidation processes when administered during the reconsolidation window.3 mg/kg MDMA did not differ at any point from saline.Thus, the effects of MDMA on reconsolidation were more pronounced 7 days than 24 hours after reactivation (Hake et al., 2019).In sum, MDMA interfered with fear extinction and disrupted fear reconsolidation in certain conditions, both findings in contrast with those of Young et al. (2015Young et al. ( , 2017)).Arluk et al. (2022) conducted four experiments to examine the biobehavioural foundations of MDMA in a translational behavioural model of PTSD using rats as subjects.In the first experiment, rats were exposed to predator scent stress (PSS) or sham-PSS.Seven days later (i.e. on day 7), 5 mg/kg MDMA or saline was administered 30 minutes prior to exposure to a trauma cue.Seven days after this (i.e. on day 14), behaviour was assessed in the elevated plus maze and acoustic startle response tests.Based on the rats' stress responses in these tests, rats were grouped into one out of three behavioural response groups, i.e. extreme, partial, and minimal behavioural responders (Arluk et al., 2022).On day 15, rats were exposed to situational reminders, and freezing behaviour was assessed.One day later (i.e. on day 16), rats were sacrificed for morphological staining of their brains to measure the neurochemical and neuroanatomical effects of MDMA (Arluk et al., 2022).In the second experiment, Arluk et al. (2022) sought to investigate the effects of MDMA on reconsolidation processes.Rats were exposed to PSS, and seven days later (i.e. on day 7), MDMA or saline was administered with or without a trauma cue.Another two groups of rats were exposed to PSS, and MDMA or saline was then administered six days later (i.e. on day 6), a day before exposure to the situational reminder on day 7, which would evaluate the effects of MDMA on fear responses when MDMA treatment was unpaired from the traumatic reminder.Similar behavioural assessments were conducted as in experiment 1 (Arluk et al., 2022).Arluk et al. (2022) found that MDMA paired with a trauma cue produced beneficial effects on measures of behavioural stress and decreased the severity of behavioural stress responses.These effects on stress responses are dependent on MDMA being administered in conjunction with trauma reactivation.Thus, this provides preliminary support for the anxiolytic effects of MDMA being mediated by effects on reconsolidation processes.
In the third experiment, Arluk et al. (2022) used Lewis rats who are characterized by hypoactive and hyporeactive HPA responses and exhibit greater susceptibility to experimentally induced PTSDphenotype.The pathophysiology of Lewis rats, particularly the HPA axis response to stressors, is therefore proposed to provide an efficient and practical model of the human PTSD phenotype for experimental procedures (Arluk et al., 2022;Cohen et al., 2006;Goswami et al., 2013).MDMA treatment had no long-term behavioural effects in the third experiment, suggesting a role of glucocorticoids in mediating the beneficial effects of MDMA on stress-related behaviour (Arluk et al., 2022).In the fourth experiment, Arluk et al. (2022) found that the anxiolytic effects of MDMA were reduced by pre-treatment with 7.5 mg of the glucocorticoid receptor antagonist RU486, 5 mg/kg of the 5-HT 2A receptor antagonist ketanserin, and 0.3 mg/kg of the partial 5-HT 1A receptor agonist pindolol, but not by 3 mg/kg of the selective 5-HT 2C receptor antagonist SB-242084.These results support the involvement of glucocorticoid, 5-HT 2A , and 5-HT 1A , but not 5-HT 2C , receptors in the anxiolytic effects of MDMA.Arluk et al. (2022) hypothesized that the observed effects may be due to effects on reconsolidation and not extinction due to the duration of memory reactivation, age of trauma memory, and time span between reactivation and behavioural testing.The beneficial effects of MDMA paired with a trauma cue accompanied prominent increases in the dendritic arborization of granular neurons in the dentate gyrus and a decrease in pyramidal neurons in the basolateral amygdala (Arluk et al., 2022).The absence of glucocorticoid receptor reactivity, either due to genetics using Lewis rats or due to pharmacological blockade using RU486, inhibited the anxiolytic effects of MDMA.Arluk et al. (2022) therefore speculate that the HPA axis mediated the anxiolytic effects of MDMA.Similarly, 5-HT 1A and 5-HT 2A receptors also appeared to mediate these anxiolytic effects.Activation of these receptors can increase corticosterone concentrations, further suggesting a potential role of the HPA axis in these anxiolytic effects.According to Arluk et al. (2022), glucocorticoid receptor antagonists and agonists have both been shown to impair reactivated memories.Arluk et al. (2022) therefore propose that there may exist an inverted U-shaped dose-response relationship between corticosterone and memory processes for memory impairment, and that the increased levels of corticosterone induced by MDMA mediate its disruptive effects on memory reconsolidation (Arluk et al., 2022).One explanation could be that this relationship is due to the enhanced inhibition of the negative cortisol feedback loop in PTSD patients which can result in increased responsiveness of the glucocorticoid receptors (Yehuda and Golier, 2009).
Further supporting the potential involvement of the HPA axis in mediating the clinical effects of MDMA, a recent study (Lewis et al., 2023) investigated epigenetic changes in HPA genes in response to MDMA-assisted psychotherapy for PTSD.Epigenetic alterations have been proposed to underlie neuroendocrine abnormalities in PTSD and have been proposed as risk factors for developing PTSD following trauma exposure (Lewis et al., 2023).According to Lewis et al. (2023), findings from multiple studies suggest that certain epigenetic alterations on HPA axis genes may be markers or potential predictors of the treatment responses in PTSD.Lewis et al. (2023) measured changes in DNA methylation of three HPA axis genes (i.e.NR3C1, FKBP5, and CRHR1) before and after MDMA-assisted psychotherapy in patients with PTSD.Changes in DNA methylation of these three genes have been associated with treatment responses in PTSD (Lewis et al., 2023).Lewis et al. (2023) found that DNA methylation changes in 37 sites of these three genes predicted treatment responses, but only one site in the NR3C1 gene (i.e.cg01391283) and one site in the CRHR1 gene (i.e.cg08276280) predicted treatment responses following false discovery rate correction.Of these two sites, only NR3C1 cg01391283 showed a significantly greater change in DNA methylation in the MDMA group (n = 16) compared with the placebo group (n = 7).The NR3C1 gene encodes for glucocorticoid receptors that are part of the HPA axis and thereby influence the regulation of stress responses.In sum, Lewis et al. (2023) propose that MDMA-assisted psychotherapy may allow for a state of epigenetic malleability with the potential of modifying epigenetics (in this case, DNA methylation of HPA axis genes) underlying PTSD symptom reduction.
Evaluating the results of Lewis et al. (2023) and Arluk et al. (2022), MDMA-assisted psychotherapy may cause epigenetic alterations underlying glucocorticoid receptor functioning.This might increase the dendritic arborization of granular neurons in the dentate gyrus and decrease pyramidal neurons in the basolateral amygdala, which affect the reconsolidation of memories.These changes in the dentate gyrus and basolateral amygdala were associated with the beneficial effects of MDMA treatment paired with a trauma cue in Arluk et al. (2022).As such, an underlying mechanism of MDMA-assisted psychotherapy may be its interference with the reconsolidation of traumatic memories which is mediated by epigenetic and functional alterations of the HPA axis, more specifically of glucocorticoid functioning.
Maples-Keller et al. ( 2022) examined the effects of MDMA on fear extinction training and retention in healthy human participants using an FPS paradigm.Extinction training was conducted 24 hours after fear conditioning, and extinction retention was tested 48 hours after extinction training.Participants received 100 mg MDMA orally two hours prior to extinction training.Because individual differences in fear extinction processes exist, Maples-Keller et al. ( 2022) also conducted subsequent analyses to classify participants in one of two groups based on the degree to which they exhibited return of fear or retention of extinction learning, which allow for comparison of extinction retainers and non-retainers in the MDMA versus the placebo group.All participants acquired and extinguished fear.Pre-extinction training administration of MDMA did not interfere with fear extinction learning, and acute MDMA administration did not affect extinction learning or retention.This effect, however, could be due to a floor effect in startle responding according to Maples-Keller et al. (2022) as the total sample exhibited almost complete extinction of FPS by the end of the extinction blocks.Six out of 17 in the MDMA group exhibited retained fear extinction versus 0/17 in the placebo group.In sum, Maples-Keller et al. ( 2022) did not find any effect of MDMA on fear extinction enhancement in humans, in contrast with the preclinical findings of Young et al. (2015Young et al. ( , 2017)).The prevalence of extinction retainers was significantly greater in the MDMA group than in the placebo group, suggestive of intra-individual differences between study participants mediating extinction retention.Vizeli et al. (2022) also examined the effects of 125 mg of orally administered MDMA on fear extinction in healthy human participants.Participants either received MDMA or a placebo (mannitol) and underwent fear conditioning using skin conductance response (SCR) or FPS paradigms with aversive, auditory stimuli and an air puff to the throat, respectively.First, participants would undergo fear conditioning using either paradigm.In the following hour, participants would receive either MDMA or a placebo.Two hours following MDMA or placebo administration, the participants would undergo corresponding extinction training.Extinction testing was conducted 24 hours after the instruction and acquisition sessions (i.e.19-20 hours after extinction training).Oxytocin release has been shown to enhance fear extinction in humans (Eckstein et al., 2015), and Vizeli et al. (2022) therefore examined whether changes in oxytocin levels contributed to the effects of MDMA on extinction.MDMA significantly increased extinction recall measured using SCR, and the acute effects of MDMA were associated with increased extinction recall in this paradigm.No association was observed between oxytocin levels and extinction recall.No effects were observed in the FPS paradigm, in accordance with the findings of Maples-Keller et al. (2022).
Using a similar cut-off method as Maples-Keller et al. ( 2022) to identify extinction retainers versus non-retainers yielded no significant difference in the prevalence in either group (Vizeli et al., 2022).Vizeli et al. (2022) speculate that this difference could be due to them using a stronger and longer aversive, unconditioned stimulus compared with Maples-Keller et al. ( 2022) and/or because of the timing of the fear acquisition and recall phases.In Maples-Keller et al. ( 2022), the fear acquisition was conducted 24 hours before and recall training 48 hours after extinction learning.In Vizeli et al. (2022), the fear acquisition was conducted 2 hours before and recall training 24 hours after extinction learning.According to Vizeli et al. (2022), SCR is a measure of the electrical conductance of the skin and changes as a result of emotional arousal, but it is less specific for measuring emotional valence.Additionally, PTSD symptoms are more strongly associated with deficits in extinction using FPS than SCR (Glover et al., 2011).Vizeli et al. (2022) also noted that the absence of effects using FPS could be due to a ceiling effect given the relatively high extinction recall rate in the placebo group, and that increased sympathomimetic activation could result in increased sweating, which could interfere with accurate SCR measurements during extinction training.Furthermore, due to the proximity of the fear conditioning and fear extinction training sessions, Vizeli et al. (2022) also speculated that the fear extinction training may have activated reconsolidation processes, which MDMA may have interfered with.Vizeli et al. ( 2022) also hypothesized that MDMA may have interfered with the consolidation of the fear acquisition due to MDMA being administered shortly after the fear conditioning.In sum, the results of Vizeli et al. (2022) show that MDMA enhanced fear extinction retention as measured by SCR, but not FPS.Young et al. (2015) administered MDMA 30 minutes prior to reexposure to a single conditioned stimulus, which significantly reduced freezing, but these effects did not persist 24 hours later at extinction training, which they interpreted as no effects on reconsolidation.Hake et al. (2019) propose that the timing of the experiment by Young et al. (2015) may not have allowed them to adequately capture the effects of MDMA on reconsolidation.In Hake et al. (2019), the effects on reconsolidation were observed when MDMA was administered after memory reactivation, and although some significant effects were observed at the contextual fear memory tests 24 hours later, these effects were even more pronounced seven days later, and only seven days later were significant effects observed at the auditory fear memory tests.Similarly, the effects of MDMA on fear behaviour in Arluk et al. (2022) were observed when tested 7-9 days after MDMA administration 30 minutes prior to trauma re-exposure.As such, and as Hake et al. (2019) hypothesized, Young et al. (2015) may not have allowed adequate time for observing the effects of MDMA on reconsolidation as 24 hours may have been too proximal to the memory reactivation.Hake et al. (2019) also propose that inter-species differences between mice and rats could account for the discrepant responses to MDMA observed between their study and those of Young et al. (2015Young et al. ( , 2017)).Based on a review by Easton and Marsden (2006), Hake et al. (2019) write that the serotonergic system appears more sensitive to MDMA in rats, and the dopaminergic system appears more sensitive to MDMA in mice.Thus, serotonergic mechanisms could explain the memory-disruptive effects observed in Hake et al. (2019).Considering the limited evidence of dopaminergic mechanisms mediating the effects of MDMA in humans and the reviewed evidence suggesting a more serotonergic-and noradrenergic-dominated neuropharmacological profile in humans, one may speculate that the results with rats as subjects are more generalizable to humans than those with mice as subjects.
Wan Aasim et al. (2017) examined the stereoselective protein binding of MDMA enantiomers in human plasma, rat serum, and mouse serum.Protein bindings were similar between human plasma and rat serum, but lower in mouse serum.Stereoselectivity in protein binding was found only in mouse serum with S-MDMA being bound to a higher extent than R-MDMA (Wan Aasim et al., 2017).MDMA enantiomers have been shown to display different pharmacodynamics due to differences in binding affinities to receptor sites.S-MDMA has been shown to induce greater release of dopamine and serotonin, produce more stimulant-like effects, and a higher rate of elimination, whereas R-MDMA is associated more with hallucinogenic properties (Wan Aasim et al., 2017).As such, differences in stereoselectivity and the MDMA enantiomers administered can limit the translational value of interspecies studies of MDMA, especially when translating findings from mice to humans.Earlier, it was already demonstrated that mice metabolize MDMA differently from rats (Green et al., 2012).Differences in pharmacodynamic and pharmacokinetic responses to MDMA could therefore account for the differential effects on memory observed between studies of mice versus rats and humans.
Hake et al. ( 2019) also write that "5-HT signaling through the 5-HT2family receptors can decrease BDNF" in rats, and that "the increase in BDNF produced by MDMA in mice is dependent on DA".Hake et al. (2019) hypothesized that the discrepant findings between their study and those of Young et al. (2015Young et al. ( , 2017) ) could also be due to serotonergic mechanisms decreasing BDNF in rats, which impaired fear extinction in Hake et al. (2019), and dopaminergic mechanisms increasing BDNF in mice, which enhanced fear extinction in Young et al. (2015Young et al. ( , 2017)).This hypothesis will be examined further in the following section.

Neuroplasticity, learning, and memory
BDNF has been implicated in several learning and memory processes, including fear consolidation, extinction, and reconsolidation (Gonzalez et al., 2019;Notaras and van den Buuse, 2020).One study (Holze et al., 2020) examined the effects of MDMA on BDNF plasma concentrations in humans but did not find significant effects.However, it has been suggested that the use of serum or whole blood to determine BDNF levels is more reliable and accurate than plasma (Elfving et al., 2010;Zhou et al., 2017).Holze et al. (2020) may therefore not have achieved valid measures of BDNF levels.
Four studies (Hemmerle et al., 2012;Koprich et al., 2003;Martínez-Turrillas et al., 2006;Soleimani Asl et al., 2017) investigated the effects of MDMA on BDNF in rats, but with vastly different methodologies and regions of interest, i.e. chronic versus acute administration, timing, and the brain regions and subregions investigated.None of these studies examined BDNF levels in the basolateral amygdala, which Young et al. (2015) found to be involved in mediating fear extinction in their experiment.Hemmerle et al. (2012) injected rats with four doses of mg/kg MDMA subcutaneously at 2-hour intervals between each dose and examined their brains 1, 7, and 24 hours after the last of the four doses.Hemmerle et al. (2012) found a time-dependent increase in BDNF levels in several brain regions, including the prefrontal cortex, frontal cortex, striatum, and more.In terms of hippocampal subfields, Hemmerle et al. (2012) found elevated BDNF levels in the CA1 region only hours after administration and decreased levels in the CA3 region and dentate gyrus only 1-and 7-hours post-administration.
Soleimani Asl et al. ( 2017) injected rats intraperitoneally with either single doses of 10 mg/kg MDMA or 1 ml saline in the acute group.In the chronic group, rats received 10 mg/kg MDMA or 1 ml saline for two consecutive days per week for two months.Rats were killed 24 hours later in the acute group and one week later in the chronic group after which their hippocampi were removed for analyses.Soleimani Asl et al. (2017) found that rats administered acute MDMA treatment showed a more prominent reduction of hippocampal BDNF than those who received chronic treatment.Martínez-Turrillas et al. (2006) investigated the effects of intraperitoneal injection of 10 mg/kg MDMA on BDNF expression in the hippocampus and frontal cortex of rats.They found that BDNF expression in the frontal cortex was significantly increased 24 hours post-injection and further increased 48 hours post-injection, but expression returned to control levels after seven days.BDNF expression was significantly decreased in the CA1, CA3, and dentate gyrus of the hippocampus 48 hours post-injection and further decreased 7 days postinjection.Koprich et al. (2003) reported increased BDNF levels in the frontal cortex, striatum, hippocampus, and brainstem of rats after chronic treatment with 20 mg/kg subcutaneous injections of MDMA twice daily for 10 days.Neurochemistry tests of brain samples were conducted the day after the last dose.The findings of these studies are diverging, but there is a trend towards increased BDNF expression in most of the examined brain regions, including the prefrontal cortex, frontal cortex, parietal cortex, entorhinal cortex, and striatum, and decreased expression in the CA1, CA3, and/or dentate gyrus of the hippocampus following MDMA administration.
In Young et al. (2015), MDMA administration only increased BDNF expression in the basolateral amygdala of mice when combined with extinction training.In Hemmerle et al. (2012), Koprich et al. (2003), Martínez-Turrillas et al. (2006), andSoleimani Asl et al. (2017), administration of MDMA in rats produced induction of BDNF levels in most brain regions.These increases were observed in the absence of any training paradigm.Methodological heterogeneity between studies could therefore influence the effects of MDMA on BDNF and thereby potentially account for differences related to the mnemonic effects of MDMA.Hake et al. (2019) claimed that serotonin signalling through 5-HT 2 receptors can decrease BDNF in rats based on two studies (Vaidya et al., 1997;Vaidya et al., 1999).However, Vaidya et al. (1997) reported increased BDNF expression in rats in most of the brain regions examined except in the dentate gyrus after intraperitoneal injection of 8 mg/kg DOI (i.e. a selective 5-HT 2A/2C agonist).Vaidya et al. (1999) found that blockade of 5-HT 2A receptors using 1 mg/kg MDL-100,907 (i.e. a selective 5-HT 2A receptor antagonist) inhibited stress-induced downregulation of BDNF expression in the dentate gyrus and CA3 subfields of the hippocampus in rats.Preclinical studies of 5-HT-releasing agents injected into rats show decrements in BDNF expression being restricted to some hippocampal subfields with either no effect on or increases in BDNF expression in other brain regions (Coppell et al., 2003;Khundakar and Zetterström, 2006;Zetterström et al., 1999).Preclinical studies using 5-HT 2A agonists such as lysergic acid diethylamide (LSD) and ayahuasca report similar findings of induction of BDNF expression in various brain regions (see de Vos et al., 2021;Jaggar and Vaidya, 2018, for reviews).
Importantly, findings from studies examining the effects of 5-HT 2A receptor activation on BDNF expression are not unequivocal.One in vitro study found that the 5-HT 2A agonists LSD, DOI, and DMT did not increase the expression of BDNF transcript, and while they doubled BDNF protein levels, these increases were not statistically significant (Ly et al., 2018).Another study found increased hippocampal BDNF expression following 5 μg/1 μL intracerebroventricular injection of the 5-HT 2A agonist TCB-2 in rats treated with streptozotocin-induced memory deficits with suppressed BDNF expression (Afshar et al., 2018).Jiang et al. (2016) found increased BDNF protein levels and mRNA expression in several brain regions, including the hippocampal CA1 and CA3 subfields, in rats following 3 mg/kg intraperitoneal injection of DOI (i.e. a 5-HT 2A/2C receptor agonist).They also found decreased levels and expression following 5 mg/kg intraperitoneal injection of ketanserin (i.e. a 5-HT 2A receptor antagonist) (Jiang et al., 2016).As such, while Vaidya et al. (1997Vaidya et al. ( , 1999) ) found DOI to upregulate BDNF expression in cortical regions and downregulate expression in the hippocampus, Jiang et al. (2016) found DOI to increase BDNF expression in all the examined brain regions.Jiang et al. (2016) hypothesize these differences can be due to methodological differences related to the used stress paradigms, the brain regions examined, and/or the technique to determine BDNF levels and expression.In sum, the available data is equivocal but suggests that agonistic activation of 5-HT 2A receptors and 5-HT-releasing agents increase cortical BDNF expression and either have no effect on or decrease hippocampal BDNF expression.Hake et al. (2019) postulate that 5-HT signalling through the 5-HT 2family receptors can decrease BDNF based on the results of Vaidya et al. (1997Vaidya et al. ( , 1999)), but neither Vaidya et al. (1997Vaidya et al. ( , 1999) ) nor other studies using 5-HT 2A agonists or 5-HT-releasing agents support their claim.Rather, these studies exhibit a trend in the opposite direction (i.e.increased BDNF expression in most regions except for the hippocampus).Interestingly, Young et al. (2015), whose findings pertaining to fear extinction Hake et al. (2019) sought to evaluate, wrote that MDMA binds directly to 5-HT 2A receptors that have been observed to increase BDNF signalling, based on the results of Vaidya et al. (1997).As such, the hypothesis of Hake et al. (2019) likely stems from a misinterpretation of Vaidya et al. (1997Vaidya et al. ( , 1999) ) and contradicts findings from most studies in that field.Hake et al. (2019) also claim that the increase in BDNF produced by MDMA in mice is dependent on DA, based on the results of Mouri et al. (2017).Mouri et al. (2017) reported increased BDNF expression in multiple brain regions, albeit not in the hippocampus, following subcutaneous injection of 10 mg/kg MDMA.However, they only examined the influence of dopaminergic mechanisms on BDNF expression in the nucleus accumbens.Using subcutaneous injections of 0.1 mg/kg of the D 1 antagonist SCH23390 and 2 mg/kg of the D 2 antagonist raclopride as pre-treatments, Mouri et al. (2017) found that D 1 and D 2 receptors mediate MDMA-induced decrements in BDNF expression in the nucleus accumbens of mice (Mouri et al., 2017).Thus, Mouri et al. (2017) only found that DA is involved in BDNF expression in a single brain region in mice, which contrasts with the wider claim of Hake et al. (2019).
In sum, the differential effects of MDMA on BDNF expression do not provide a compelling explanation for the diverging results of studies with mice versus rats.The impetus remains to connect the reviewed preclinical data on BDNF expression with the results of studies examining the effects of MDMA on memory in humans.The collated studies of the effects of MDMA on BDNF expression in rats show some mixed results, but they point towards decreased hippocampal and increased cortical BDNF expression following administration of various doses of MDMA ranging from a single 10 mg/kg injection to multiple 10-20 mg/ kg injections for several days or weeks.Findings from studies of 5-HT 2A receptor agonists and 5-HT-releasing agents show similarly mixed results, but also point towards decreased hippocampal and increased cortical BDNF expression.It can be hypothesized that the effects of MDMA on BDNF are mediated by serotonergic mechanisms, but more research is needed for more definitive conclusions pertaining to the effects of MDMA on BDNF, and how these effects may relate to memory functioning in humans.
The only obvious difference between the effects of MDMA on BDNF expression in mice and rats from the reviewed studies is the trend towards decreased hippocampal BDNF expression in rats and the absence of changes in hippocampal BDNF expression in mice reported by Mouri et al. (2017).One study found that intrahippocampal infusion of a function-blocking anti-BDNF antibody following extinction learning impaired reconsolidation (Radiske et al., 2015).This could lend credence to a hypothesis that MDMA-induced decrements of BDNF in some hippocampal subfields result in impaired reconsolidation in rats.This hypothesis, however, rests upon a precipice of assumptions derived from inconsistent results from heterogeneous studies.The extant evidence currently does not allow for more than conjectures about changes in BDNF expression mediating the memory effects of MDMA.
Overall, the methodological heterogeneity of studies examining the effects of MDMA on fear extinction and reconsolidation cumbers more definitive conclusions appertaining to the mnemonic effects of MDMA.Some have proposed that MDMA may act on both fear extinction and reconsolidation (e.g.Arluk et al., 2022;Hake et al., 2019;Singleton et al., 2022), but as the two processes are said to be mutually exclusive, the impetus remains to elucidate exactly how each would manifest in conjunction with each other.More definite conclusions pertaining to the involvement of fear extinction versus reconsolidation, and the mechanisms they are predicated upon, in MDMA-assisted psychotherapy for the treatment of PTSD currently do not allow for more than conjectures with the available evidence.More research is needed to uncover the effects of MDMA on memory and the mechanisms mediating these effects.

Summary of the effects of MDMA on memory
To recapitulate, Young et al. (2015Young et al. ( , 2017) ) showed that MDMA in combination with extinction training can enhance the extinction of conditioned fear in mice.The extinction enhancements persisted over multiple days and generalized to a novel context, and these enhancements were hypothesized to be mediated by MDMA affecting serotonergic mechanisms that increase BDNF in the basolateral amygdala.Hake et al. (2019) found that MDMA interfered with fear extinction in rats.Both Hake et al. (2019) and Arluk et al. (2022) found that MDMA reduced fear responses when administered in conjunction with fear memory reactivation, suggesting interference with fear memory reconsolidation as the mediating mechanism.This disruption of reconsolidation is hypothesized to be due to MDMA decreasing BDNF via serotonergic mechanisms according to Hake et al. (2019) or by the influence of MDMA on serotonergic mechanisms and the HPA axis according to Arluk et al. (2022).Maples-Keller et al. (2022) found no significant effect of MDMA on fear extinction measured using FPS in healthy humans, although more participants in the MDMA group exhibited retained fear extinction than in the placebo group.Vizeli et al. (2022) found that MDMA enhanced fear extinction in healthy humans, but only when measured using SCR, not FPS.
As previously mentioned, PTSD symptoms are more strongly associated with deficits in extinction using FPS than SCR, and while conditioned freezing is more stereotypic fear behaviour in rodents, FPS is a highly conserved fear behaviour across species, including in humans.Studies using FPS may therefore provide a more reliable measure of fear behaviour than studies using conditioned freezing and SCR.Both Maples-Keller et al. ( 2022) and Vizeli et al. (2022) involved healthy human participants.As mentioned earlier, PTSD is associated with changes in various biological, psychological, social, and behavioural domains, including epigenetic, genetic, cellular, molecular, and brain changes.Fear conditioning in healthy humans may therefore not approximate the same perturbations as those observed in patients with PTSD.It cannot be ruled out that the mnemonic effects of MDMA may differ between participants with PTSD and those without due to the perturbations associated with PTSD.In contrast, rodent models can circumvent the limitations of using human participants.Rodent models of PTSD have been shown to induce core neurobiological phenotypes of PTSD (Aspesi and Pinna, 2019;Deslauriers et al., 2018;Verbitsky et al., 2020).However, caution is encouraged when generalizing preclinical findings to humans as inter-species differences can yield differential results.This has, as we previously mentioned, been demonstrated before when comparing the effects of MDMA on rodents versus humans (e.g. on prepulse inhibition).
In the preclinical studies (Arluk et al., 2022;Hake et al., 2019;Young et al., 2015;Young et al., 2017), the timing, sequence, dosage, and frequency of MDMA administration relative to extinction training and/ or fear memory reactivation were varied.In the clinical studies, only a single dose was used.The use of single doses contrasts with the clinical trials of MDMA-assisted psychotherapy in which one-to-three full doses or split doses are administered.The preclinical studies may therefore better model the dosing paradigm utilized in clinical trials of MDMAassisted psychotherapy.Findings from two clinical (Maples-Keller et al., 2022;Vizeli et al., 2022) and two preclinical (Young et al., 2015;Young et al., 2017) studies point towards MDMA being able to enhance fear extinction.The results of Young et al. (2015) suggest that this effect on fear extinction may only work in combination with extinction training.MDMA administration alone may therefore not be sufficient for enhancing extinction in patients with PTSD.Instead, it may be necessary to combine it with psychotherapy to activate the fear associations and to provide corrective, adaptive information to associate conditioned stimuli with.Enhancement of extinction would unarguably be beneficial for PTSD treatment, particularly given the extinction deficits often observed in patients with PTSD.
MDMA-assisted psychotherapy could therefore ameliorate extinction deficiencies in participants with PTSD.This effect could provide corrective, adaptive information for the development of novel contingencies that take precedence over the former maladaptive contingencies pertinent to the traumatic event.The results of Young et al. (2015Young et al. ( , 2017) ) further suggest that these effects can persist over multiple days and generalize to other stimuli related to the conditioned stimuli, which would argue against the extinction enhancements being confined to the therapeutic sessions.Nonetheless, while extinction-based approaches undoubtedly offer rich avenues for treatment of PTSD, the obvious limitation remains of the original fear contingencies resurfacing, and the patient thereby relapsing.Extinction does not modify the original contingencies, hence the risk of impermanence of extinction.Reconsolidation, in contrast, does, and interference with the reconsolidation process presents an alternative that compensates for this limitation of extinction.
We previously expounded upon the diverging findings between Young et al. (2015, 20217) and Arluk et al. (2022) and Hake et al. (2019).Multiple possible explanations were provided for these discrepancies.Young et al. (2015) may not have allowed adequate time for detecting effects on reconsolidation.The extinction enhancements in Young et al. (2015Young et al. ( , 2017) ) could be due to neurobiological differences between mice and rats, including differential effects of MDMA on dopaminergic and serotonergic mechanisms and different pharmacodynamic and pharmacokinetic responses between rats and mice.Humans and rats did not display stereoselectivity in protein binding, in contrast with mice, and serotonergic and noradrenergic mechanisms appear to account for most of the effects of MDMA in humans with limited evidence of the involvement of dopaminergic mechanisms.As such, while these differences do not provide definitive conclusions regarding the involvement of fear extinction and fear reconsolidation as mediating mechanisms of MDMA-assisted psychotherapy, the results of studies using rats as subjects (Arluk et al., 2022;Hake et al., 2019) may offer more valid translational value to interpretation of effects in humans with PTSD.
It is therefore possible that an underlying mechanism of MDMAassisted psychotherapy is the ability of MDMA to interfere with the reconsolidation of traumatic memories to reduce their pathological impact on the patient.Indeed, the assertion that MDMA attenuates or modifies memories of traumatic experiences is in accordance with other preclinical and clinical studies reporting impairments in learning and memory following MDMA administration.For example, multiple preclinical studies have associated various dosages of acute intraperitoneal or subcutaneous MDMA administration with deficits in spatial learning and memory, reinforcement-based learning, episodic memory performance, reference memory, and working memory (Barrionuevo et al., 2000;Byrne et al., 2000;Harper et al., 2006;Harper et al., 2005;Kay et al., 2010;Moyano et al., 2004).In clinical studies, findings also show transient impairments in verbal, spatial, and working memory following oral administration of 75 mg (Kuypers et al., 2013;Kuypers andRamaekers, 2005, 2007;Kuypers et al., 2011;van Wel et al., 2011), 75 mg plus a subsequent 50 mg (Kuypers et al., 2008), and 100 mg (Stough et al., 2012) of MDMA.
Of the studies that focused on the effects of MDMA on emotional or fear memories, but not in the context of fear extinction or reconsolidation, Pantoni et al. (2022) examined the effects of MDMA on learning and memory in mice.Mice were assigned to groups based on the amount of MDMA they received, ranging from 0 to 10 mg/kg intraperitoneally (Pantoni et al., 2022).MDMA or saline was given 30 minutes before a 10-minute training session in which a single auditory stimulus was paired with a foot shock, and training was followed by a post-shock test for measuring effects on short-term memory.Seven days after training, mice were returned to the training context for assessment of their contextual memory.One day after context testing, mice were placed in an alternative context for tone testing.Fear responses were measured in time spent freezing (Pantoni et al., 2022).Mice administered 10 mg/kg MDMA exhibited reduced freezing at training, reflective of short-term memory impairment.Pantoni et al. (2022) found that this impairment was not confounded by the effects of MDMA on either nociception or locomotor activity.Mice administered 3 and 10 mg/kg MDMA also exhibited reduced freezing at context and tone testing (Pantoni et al., 2022).Doss et al. (2018) exposed human participants to emotional images to examine the effects of 1 mg/kg MDMA on learning and memory.They found that MDMA attenuated the encoding and recollection of salient details of emotional memories, but not the overall ability to recollect the occurrence of these emotional events (Doss et al., 2018).Carhart-Harris et al. (2014) examined the effects of 100 mg orally administered MDMA on the recollection of favourite and worst autobiographical memories in healthy human participants.MDMA significantly attenuated negative affect during the recollection of the participants' worst autobiographical memories.From the qualitative reports of some participants, this effect made it easier for the participants to approach their traumatic experiences and reinterpret them.Favourite memories were also experienced as more vivid, emotionally intense, and positive following MDMA administration (Carhart-Harris et al., 2014).
These findings suggest that MDMA may carry the potential to inhibit recall of certain aspects of a memory and/or modify the emotional valence of the memory trace itself.If indeed MDMA affects the reconsolidation of traumatic memories, the nature of these effects remains unestablished.Some clinical and preclinical studies of pharmacological reconsolidation disruption show patterns of alterations in the emotional characteristics of traumatic memories and fear memories, i.e. diminished negative emotional valence and reduced fear responses (see Beckers and Kindt, 2017;Dunbar and Taylor, 2017;Kindt and Elsey, 2023, for reviews).Some studies on reconsolidation disruption have also found declarative memory of fear experiences to be intact (Kindt and Elsey, 2023).As such, extant findings do not support that reconsolidation disruption can affect the memory of a traumatic memory as a whole.Similarly, the available evidence also does not indicate that MDMA induces a general amnestic effect on the traumatic memory, i.e. by making all aspects of the memory more difficult to recall.
Carhart- Harris et al. (2014) found that MDMA altered the recall of emotional aspects of memories.These effects, according to Carhart- Harris et al. (2014), are consistent with increased serotonergic functioning and with MDMA inducing a positivity bias.Accordingly, multiple studies have found that MDMA causes an affective bias as manifested by increased recognition of positive information (e.g.happy countenances) and/or decreased recognition of negative information (e.g.angry or fearful countenances) (e.g.Bedi et al., 2010;Bedi et al., 2009;Dolder et al., 2018;Hysek et al., 2012b;Wardle and de Wit, 2014).MDMA has also been found to increase comfort when reporting emotional autobiographical memories (Baggott et al., 2016) and to facilitate more direct and open communication between individuals in an emotional relationship (Greer and Tolbert, 1986).Additionally, MDMA has also been found to promote positive changes in mood, emotional states, attitudes, goals, and beliefs (Greer and Tolbert, 1986).Barone et al. (2019) examined the qualitative effects of MDMA-assisted psychotherapy on patients with PTSD.Some participants reported experiencing greater resilience and safety when accessing their traumatic experiences, whereas they were previously too overwhelmed and exerted avoidant behaviours.It does not appear from these findings that MDMA affected participants' ability to recall their traumatic experiences, but rather their thoughts and emotions pertaining to these experiences (Barone et al., 2019).
In sum, findings do not support that MDMA induces a general amnestic effect.Instead, two hypotheses remain more probable.First, it can be hypothesized that MDMA via reconsolidation processes may promote a window of malleability in which adaptive information can be incorporated into the patient's traumatic memory.Although the science behind and findings related to reconsolidation updating are still tentative and at times diverging, some reconsolidation studies have shown that fear memories can be updated with corrective, adaptive information during the reconsolidation window (Bellfy and Kwapis, 2020;Schiller et al., 2010;Silva and Gräff, 2023).Carhart-Harris et al. (2014) found that MDMA affected the experienced emotions of recalled autobiographical memories.The positivity bias might contribute to the experience of more positive emotional valence and/or decreased negative emotional valence upon MDMA consumption.In the context of PTSD, a positivity bias is useful given that a core symptom of PTSD is the experience of persistent negative emotional states and persistent inability to experience positive emotions (American Psychiatric Association, 2013).In the context of MDMA-assisted psychotherapy, it can therefore be hypothesized that MDMA allows for updating the memory of the traumatic experience during the reconsolidation process.There is limited evidence to suggest that reconsolidation updating affects declarative aspects of negative memories or that MDMA affects declarative aspects of patients' traumatic memories.As such, we hypothesize that MDMA may promote a transient window in which corrective, emotional information can be incorporated into the original traumatic memory trace.MDMA could promote the development of more positive emotions, goals, and attitudes in patients with PTSD given its serotonininduced subjective effects, which could affect the reconstruction of emotional aspects of their traumatic experience more adaptively.Indeed, autobiographical memories are constructive, and memory recall is influenced by the current attitudes, goals, and concerns of the individual (Rubin et al., 2008a).
One mechanism through which MDMA may modify the emotional characteristics of memories through reconsolidation is by targeting fear contingencies.A positivity bias could manifest in diminished negative emotional valence (e.g.fear and anxiety) and/or increased positive emotional valence (e.g.safety, empathy, perceived self-resilience).Recalling a traumatic memory without being overwhelmed by fear and anxiety responses could provide a sufficient mismatch between expected and actual expectations (i.e. a prediction error).Some (Kindt and Elsey, 2023;Vaverková et al., 2020) have proposed prediction error to be a necessary (but not sufficient) condition for successful reconsolidation interference.Thus, MDMA may affect the emotional characteristics of traumatic memories by modifying the fear contingencies associated with the traumatic memories to incorporate more adaptive information.This would be in accordance with studies of reconsolidation disruption that have found decreased expression of fear responses to conditioned stimuli, reflective of decreased negative emotional valence attributed to the conditioned (and related) stimuli (Kindt and Elsey, 2023).
Another hypothesis is that MDMA interferes with the reconsolidation of traumatic memory by temporarily or permanently impairing memory of emotionally salient aspects of the traumatic experience.As such, instead of updating the memory with corrective information, MDMA may diminish the ability to recall emotional aspects of the traumatic event, i.e. target the accessibility of the memory.Reconsolidation updating would occur during the stage of labilization, and this amnestic effect on the emotional aspects of memory would occur by obstructing the reconsolidation stage of the memory.The findings of Arluk et al. (2022), Hake et al. (2019), Doss et al. (2018), Pantoni et al. (2022), andCarhart-Harris et al. (2014) suggest that MDMA can diminish recall of emotional information and/or reduce fear responses related to aversive memories.These effects were observed in the absence of any intervention aimed at modifying recall or subjective appraisal of the event (e.g.extinction training).MDMA may therefore reduce fear responses by targeting the reconsolidation of traumatic memory, but instead of modifying emotional aspects of the traumatic memory, it obstructs the recollection of emotional information pertaining to the memory.The mechanism(s) through which it diminishes recollection of emotional information of traumatic memories remain(s) undetermined.The emotional information that could be targeted also remains to be uncovered.Obstructing recall of emotional information is also in accordance with studies of pharmacological reconsolidation disruption in which recollection of emotional aspects of memories is impaired (Dunbar and Taylor, 2017).The result would be a decrease in fear responses upon subsequent exposure to trauma-related stimuli and/or recall of the traumatic memory.
In addition to the emotional positivity bias, MDMA has also been found to increase prosocial attitudes, speech, and behaviours, including mutual trust with the therapist, cooperation, openness, closeness to others, emotional empathy, and direct and open communication (e.g.Baggott et al., 2016;Baggott et al., 2015;Bedi et al., 2014;Bedi et al., 2010;Frye et al., 2014;Gabay et al., 2018;Gabay et al., 2019;Greer and Tolbert, 1986).Patients have also reported reduced fear responses and greater experience of safety when engaging with their traumatic experience and less avoidant behaviours during their sessions with MDMA (Barone et al., 2019;Greer and Tolbert, 1986;Greer and Tolbert, 1998).Given that PTSD is characterized by interpersonal difficulties, hypervigilance, hyperarousal, and avoidant behaviours (American Psychiatric Association, 2013), these aforementioned effects could bolster the patient's ability to engage with their traumatic experience without being emotionally overwhelmed and without applying avoidant coping mechanisms.These effects would unarguably be conducive to efforts of fear extinction and/or reconsolidation interference by bolstering the patient's ability to engage with their traumatic experience.
Taken together, studies have reported that MDMA enhances fear extinction and/or disrupts fear reconsolidation.Hypotheses about the processes that mediate the effects of MDMA on either of these two mechanisms are currently limited or conjectural.Whether one or both of these processes are involved in mediating the efficacy of MDMA-assisted psychotherapy for the treatment of PTSD remains undetermined.MDMA may allow for enhanced extinction of fear conditioning pertaining to the traumatic event.It could also be hypothesized that MDMA may allow for a state in which a participant's traumatic memory can be modified with more adaptive information or that MDMA may disrupt recollection of emotional aspects of a traumatic memory.
Emotional aspects of traumatic memories are often remembered more frequently and are the most easily recalled (Berntsen, 2009), contribute to the severity of PTSD symptoms (Rubin et al., 2008a;Rubin et al., 2008b), and contribute to the consolidation and accessibility of traumatic memories.These effects could potentiate the centrality and importance of the traumatic memory for the trauma victim.Extinguishing fear contingencies, adaptive modification of emotional aspects of the traumatic memory, and/or obstructing recall of emotional aspects of the traumatic memory could diminish the centrality of the traumatic event for the patient's identity, self, and life story and diminish the accessibility of aspects of the traumatic memory for recollection.These effects would all be viabilities for PTSD treatment.

Conclusion and recommendations for future research
An accumulating number of studies within cognitive psychology have substantiated the important role of the traumatic memory in mediating PTSD development and maintenance.In particular, enhanced traumatic memory accessibility for recall and centrality to identity and life story correlate positively with PTSD symptomatology.Recent years have seen a resurgence of the clinical augmentation of MDMA to psychotherapeutic interventions for PTSD.The clinical efficacy of this treatment method has been underpinned in multiple trials, but the mechanisms mediating its efficacy remain largely unknown.Some hypothesized mechanisms include MDMA enhancing fear extinction and/ or interfering with the reconsolidation of traumatic memories.We explored these hypotheses in this article.It could be hypothesized that MDMA in conjunction with psychotherapy enhances the extinction of conditioned fear.Another hypothesis is that MDMA in conjunction with psychotherapy interferes with the reconsolidation of traumatic memory.This would allow for adaptive modification and/or diminished recollection of emotional aspects of the traumatic memory.Overall, the extant evidence does not allow fully discrediting one of these two proposed mechanisms over the other, but some preclinical findings favour interference with reconsolidation as a mediating mechanism.
Future research is needed to uncover the effects of MDMA on memory, and how these effects could mediate the treatment efficacy of MDMA-assisted psychotherapy.Further, the biological underpinnings and the pharmacological profile of MDMA mediating these potential mnemonic effects remain to be elucidated.Some have proposed several mechanisms that could mediate the influence of MDMA on memory, including the 5-HT system, BDNF, and the HPA axis, but the involvement of these mechanisms needs to be fully evaluated.The pharmacological profile of MDMA is still largely enigmatic with only tentative evidence, and more research is encouraged to uncover its pharmacology and the biological and psychological processes that MDMA instigates.Furthermore, research on the reconsolidation of memories is still in its incipiency.The conditions necessary to induce reconsolidation disruption, and the aspects of a memory that are targeted by such disruption, are still not fully understood nor elucidated.
PTSD has been associated with an extensive array of changes, including epigenetic, genetic, cellular, molecular, and brain changes, that have not been fully accounted for in this article.Instead, this article focuses primarily on the cognitive (i.e.autobiographical memoryrelated) aspects of PTSD.Accordingly, this article primarily examined the neurohormonal and monoamine effects of MDMA in relation to its mnemonic effects.It cannot be ruled out that other processes such as molecular or neuroanatomical changes can also be implicated in its mnemonic effects.We also do not exclude the possibility that the treatment efficacy of MDMA-assisted psychotherapy is mediated by other mechanisms that are not related to the mnemonic effects of MDMA or that other memory systems, besides the autobiographical memory system, could be implicated.
If one mechanism of action behind MDMA-assisted psychotherapy is disrupted reconsolidation of traumatic memories, future research is needed to explicate the influence of this effect on traumatic memories.The CES can be used to correlate changes in PTSD symptomatology to changes in the perceived centrality of the traumatic event in the patient's life story and identity.Autobiographical memory measures could be used to assess pre-and post-treatment changes in the patient's traumatic memories.One such measure is the Autobiographical Memory Questionnaire (AMQ) (Rubin et al., 2010;Rubin et al., 2003), which measures various components of autobiographical memories, including emotional, narrative, and sensory qualities, availability for recall, and specificity.Another measure is the Memory of Experiences Questionnaire (MEQ) (Sutin and Robins, 2007), which measures ten phenomenological properties of autobiographical memories, including vividness, accessibility, emotional intensity, and sensory details.These tests could aid in understanding how MDMA-assisted psychotherapy affects core components of the patient's traumatic memories.
It can also be necessary to measure pre-and post-treatment changes in non-traumatic, autobiographical memories to assess how MDMA affects memory functioning in general.MDMA has also been associated with impairments in various forms of attention (de la Torre et al., 2000;Hasler et al., 2009;Stough et al., 2012) and verbal, spatial, and working memory (Kuypers et al., 2008;Kuypers et al., 2013;Kuypers andRamaekers, 2005, 2007;Kuypers et al., 2011;Stough et al., 2012;van Wel et al., 2011) (see Pantoni and Anagnostaras, 2019, for a review of preclinical findings of the cognitive effects of MDMA).Prospective autobiographical memory research on MDMA-assisted psychotherapy should involve controlling the potentially confounding impact of these deficits.

Declaration of Competing Interest
Kim P. C. Kuypers is a principal investigator on research projects, the present article not included, that are sponsored by Mindmed and MAPS, which are companies that are developing psychedelic medicines, and she is a paid member of the scientific advisory board of Clerkenwell Health.