Mismatch negativity predicts psychotic experiences induced by nmda receptor antagonist in healthy volunteers

Abstract

Background: Previous studies indicate that mismatch negativity (MMN)—a preattentive auditory event-related potential (ERP)–depends on NMDA receptor (NMDAR) functioning. To explore if the strength of MMN generation reflects the functional condition of the NMDAR system in healthy volunteers, we analyzed correlations between MMN recorded before drug administration and subsequent responses to the NMDAR antagonist ketamine or the 5-HT2a agonist psilocybin.

Methods: In two separate studies, MMN was recorded to both frequency and duration deviants prior to administration of ketamine or psilocybin. Behavioral and subjective effects of ketamine and psilocybin were assessed with the Brief Psychiatric Rating Scale and the OAV Scale—a rating scale developed to measure altered states of consciousness. Correlations between ERP amplitudes (MMN, N1, and P2) and drug-induced effects were calculated in each study group and compared between them.

Results: Smaller MMN to both pitch and duration deviants was significantly correlated to stronger effects during ketamine, but not psilocybin administration. No significant correlations were observed for N1 and P2.

Conclusions: Smaller MMN indicates a NMDAR system that is more vulnerable to disruption by the NMDAR antagonist ketamine. MMN generation appears to index the functional state of NMDAR-mediated neurotransmission even in subjects who do not demonstrate any psychopathology.

Introduction

Mismatch negativity (MMN) is an auditory event-related potential (ERP) indexing preattentive detection of stimulus deviance (Näätänen 1995). It is automatically generated with a latency of 100 to 200 ms after the presentation of a stimulus that deviates in one of its physical dimensions (pitch, intensity, duration, location) from preceding frequently repeated standard stimuli Näätänen 1995, Novak et al 1990, Ritter et al 1995. Mismatch negativity is thus the manifestation of a preattentive process that compares the deviant stimulus to the sensory memory trace of the standard stimulus Näätänen 1990, Novak et al 1990, Ritter et al 1995.

Mismatch negativity-like activities can also be recorded in animals and thus are amenable to investigations of its underlying neurobiological substrate Csepe et al 1987, Javitt et al 1994, Javitt et al 1996. Studies in monkeys and cats show that MMN specifically involves activation of the supragranular layer of the primary and secondary auditory cortices (Karmos et al 1998; Javitt et al 1994, Javitt et al 1996). It has been hypothesized that this activation involves excitation of pyramidal neurons as a net result of two processes: 1) reduction of tonic inhibition (disinhibition) due to stimulation by the standard tone, and 2) excitation of NMDA receptor (NMDAR) in response to the deviant stimulus (Schroeder et al 1997; Karmos et al personal communication; Javitt et al 1996). NMDAR functions are uniquely sensitive to membrane potential (Cotman et al 1995); thus, the magnitude of stimulation-induced current flow through NMDAR depends upon the degree of neuronal disinhibition. This property makes the NMDAR uniquely suited for mediating conditional responses such as MMN. Studies in monkeys have indeed demonstrated that intracortical and systemic application of NMDAR antagonists selectively abolishes MMN-like activities without affecting such sensory ERPs as N1 Javitt et al 1996, Javitt et al 1994. In addition, we recently demonstrated that ketamine, a noncompetitive NMDAR antagonist, selectively impairs MMN generation in humans without reducing sensory ERPs (Umbricht et al 2000). Furthermore, ketamine significantly reduced the magnetic counterpart of MMN, MMNm, in healthy volunteers (Sauer et al 2000). The NMDAR antagonist nitrous oxide (N2O) (Jevtovic-Todorovic et al 1998) has also been reported to induce a significant reduction of MMN (Pang et al 1999). In contrast, the psychotomimetic psilocybin does not significantly affect MMN in healthy volunteers, although it reduces N1 (Umbricht et al 2001).

Psilocybin (O-phosphoryl-4-hydroxy-N,N-dimethyltryptamine)—a naturally occurring indoleamine hallucinogen—is readily metabolized to psilocin after ingestion (Hasler et al 1997). In rats, psilocin binds with high affinity to the 5-HT2A (Ki = 6.0 nM) and with moderate affinity to the 5-HT1A receptor (Ki = 190 nM) (McKenna et al 1990). Although indoleamine hallucinogens bind to various subtypes of serotonin receptors, the hallucinogenic and cognitive effects are primarily believed to be mediated through agonist action at the 5-HT2A receptor Aghajanian and Marek 1999a, Vollenweider et al 1998. Moreover, the affinity of indoleamine hallucinogens at the 5-HT2A receptor is highly correlated with their hallucinogenic potency in humans (Aghajanian et al 1999a).

In summary, the available evidence strongly indicates a central role of the NMDAR in MMN generation. Thus, deficits in MMN generation as observed in such neuropsychiatric disorders as schizophrenia Catts et al 1995, Javitt et al 1995, Umbricht et al 1998 may be a manifestation of deficient NMDAR-dependent neurotransmission Olney and Farber 1995, Javitt and Zukin 1991.

Furthermore, these data suggest that even in non-patient subjects the strength of MMN generation may reflect the functional condition of NMDAR systems. To explore this hypothesis, we examined the correlations between MMN recorded prior to drug administration and psychosis-like experiences and behavioral effects induced by the subsequent administration of ketamine or the psychotomimetic psilocybin. Ketamine, and to a lesser extent psilocybin, can induce signs and symptoms in healthy volunteers that mimic some psychotic symptoms of schizophrenia Krystal et al 1994, Vollenweider et al 1998. Thus, we hypothesized that if the strength of MMN indeed reflects the functional condition of the NMDAR system (i.e., its efficiency or redundancy), then the effects of ketamine should be more disruptive. (i.e., stronger in a subject with a smaller MMN at baseline). In other words, an inverse correlation between baseline MMN and ketamine-induced psychosis-like experiences should exist. On the other hand, such a correlation should not be observed between MMN and the effects of psilocybin.

To test this hypothesis, data from two separate studies on the effects of ketamine and psilocybin in healthy volunteers were analyzed. The main goal of both studies was to assess the effects of ketamine or psilocybin, respectively, on MMN generation. The results of these studies have been (Umbricht et al 2000) or will be reported elsewhere (Umbricht unpublished). Here, we only report on the session with active drug administration (ketamine infusion or psilocybin administration, respectively).