Effects of Augmenting N-Methyl-D-Aspartate Receptor Signaling on Working Memory and Experience-Dependent Plasticity in Schizophrenia: An Exploratory Study Using Acute d-cycloserine

Abstract

Cognitive deficits in schizophrenia have been hypothesized to reflect N-methyl-D-aspartate receptor (NMDAR) dysfunction. However, the mechanisms through which the NMDAR contributes to individual cognitive functions differ. To explore how NMDAR signaling relates to specific cognitive deficits in schizophrenia, we tested the effects of enhancing NMDAR signaling on working memory and experience-dependent plasticity using d-cycloserine (DCS). Plasticity was assessed using an EEG paradigm that utilizes high-frequency visual stimulation (HFvS) to induce neural potentiation, and 2 learning tasks, the information integration (IIT) and weather prediction (WPT) tasks. Working memory was assessed using an N-back task. Forty-five schizophrenia patients were randomized to receive a single 100 mg DCS dose (SZ-DCS; n = 24) or placebo (SZ-PLC; n = 21) in a double-blind, between-groups design. Testing occurred on a single day after placebo or DCS administration; baseline values were not obtained. DCS did not affect plasticity, as indicated by similar neural potentiation, and similar IIT and WPT learning between groups. However, among patients who successfully engaged in the working memory task (ie, performed above chance), SZ-DCS (n = 17) showed superior 2-back performance compared to SZ-PLC (n = 16). Interestingly, SZ-DCS also showed larger pre-HFvS neural responses during the LTP task. Notably, this pattern of DCS effects is the opposite of those found in our prior study of healthy adults. Results are consistent with target engagement of the NMDAR by DCS, but suggest that NMDAR signaling was not translated into synaptic plasticity changes in schizophrenia. Results highlight the importance of considering how distinct NMDAR-associated processes contribute to individual cognitive deficits in schizophrenia.

Keywords: NMDA receptor; cognition; learning; long-term potentiation; memory; psychosis.

Fig. 1.

(A) Standard circle black and white checkerboard stimulus presented at 0.83 Hz during visual evoked potential (VEP) assessment blocks and at ~8.87 Hz during high-frequency visual stimulation (HFvS). (B) Time course of the long-term potentiation (LTP) paradigm.

Fig. 2.

(A) Two example trials for the 1-back condition of the N-back task. (B) Two example information integration task (IIT) trials. (C) Two example weather prediction task (WPT) trials. The N-back and IIT were identical in stimuli and trial structure such that the only difference participants experienced was whether they were asked to recall whether stimuli were in the same location on the screen as recently shown stimuli (ie, for the N-back) or learn about the stimuli (ie, for the IIT).

Fig. 3.

(A) Average visual evoked potentials (VEPs) elicited by the checkerboard stimulus in Oz and (B) topography of VEPs for schizophrenia patients who received placebo (SZ-PLC) or DCS (SZ-DCS) across assessment blocks. The VEP included a negative component, C1, and a positive component, P2. (C) Mean (±SE) C1 amplitude was larger across blocks in SZ-DCS compared to SZ-PLC*. (D) Mean (±SE) P2 amplitude was similar in SZ-PLC and SZ-DCS across blocks.

Fig. 4.

(A) Mean ± SE percent correct responses per 80-trial load on the N-back for schizophrenia patients who received placebo (SZ-PLC) or DCS (SZ-DCS) and performed above chance. *SZ-DCS performed significantly better than SZ-PLC during the 2-back. Inset shows percent correct responses when all SZ-PLC (n = 20) and SZ-DCS (n = 23) were included. Mean ± SE percent correct responses per 80-trial blocks for the (B) weather prediction task (WPT) and (C) information integration task (IIT) for SZ-PLC and SZ-DCS.