Neurophysiological mechanisms of cortical plasticity impairments in schizophrenia and modulation by the NMDA receptor agonist D-serine

Abstract

Schizophrenia is associated with deficits in cortical plasticity that affect sensory brain regions and lead to impaired cognitive performance. Here we examined underlying neural mechanisms of auditory plasticity deficits using combined behavioural and neurophysiological assessment, along with neuropharmacological manipulation targeted at the N-methyl-D-aspartate type glutamate receptor (NMDAR). Cortical plasticity was assessed in a cohort of 40 schizophrenia/schizoaffective patients relative to 42 healthy control subjects using a fixed reference tone auditory plasticity task. In a second cohort (n = 21 schizophrenia/schizoaffective patients, n = 13 healthy controls), event-related potential and event-related time-frequency measures of auditory dysfunction were assessed during administration of the NMDAR agonist d-serine. Mismatch negativity was used as a functional read-out of auditory-level function. Clinical trials registration numbers were NCT01474395/NCT02156908 Schizophrenia/schizoaffective patients showed significantly reduced auditory plasticity versus healthy controls (P = 0.001) that correlated with measures of cognitive, occupational and social dysfunction. In event-related potential/time-frequency analyses, patients showed highly significant reductions in sensory N1 that reflected underlying impairments in θ responses (P < 0.001), along with reduced θ and β-power modulation during retention and motor-preparation intervals. Repeated administration of d-serine led to intercorrelated improvements in (i) auditory plasticity (P < 0.001); (ii) θ-frequency response (P < 0.05); and (iii) mismatch negativity generation to trained versus untrained tones (P = 0.02). Schizophrenia/schizoaffective patients show highly significant deficits in auditory plasticity that contribute to cognitive, occupational and social dysfunction. d-serine studies suggest first that NMDAR dysfunction may contribute to underlying cortical plasticity deficits and, second, that repeated NMDAR agonist administration may enhance cortical plasticity in schizophrenia.

Keywords: clinical trials; cortical plasticity; imaging; schizophrenia; social cognition.

© The Author (2016). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: journals.permissions@oup.com.

Figures

Figure 1

Model of paradigms. (A) Model of paradigm used in the behavioural assessment experiment. (B) Model of paradigm used in the neurophysiological and the NMDAR mechanisms study. Each session began with a pre-training MMN. Schizophrenia patients (n = 21) then received d-serine (60 mg/kg) or placebo/no intervention ∼30 min prior to the auditory plasticity session. Neurophysiology was recorded during auditory plasticity sessions, followed by post-treatment MMN. Each patients with schizophrenia subject completed three sessions, ∼1 week apart, while healthy controls (HC, n = 13) completed one session each without d-serine. Usable behavioural (auditory plasticity) data were available from all 21 patients with schizophrenia, yielding 63 schizophrenia sessions (three per subject) and 13 healthy controls (one session each, 13 total sessions). A portion of time-frequency and EEG data were excluded for poor data quality. Time-frequency data were excluded from 2 of 13 healthy control sessions (yielding 11 remaining) and 11 patients with schizophrenia sessions (yielding 52 remaining). Useable schizophrenia sessions included 16 receiving placebo, 21 during first exposure d-serine, and 15 during their second or third consecutive weekly exposure. MMN data were excluded from one healthy control, yielding 12 remaining, and four schizophrenia sessions (three placebo and one d-serine session), yielding 19 placebo, 25 first d-serine exposures and 15 consecutive exposures.

Figure 2

Behavioural assessment of auditory plasticity. Line graph of plasticity for schizophrenia and controls for the random (A) and fixed (B) conditions. (C) Bar graph for final JND threshold (%Δf) from trials 70–80. (D) Bar graph for plasticity (%Δf, mean trial 20–30 to 70–80). Scatter plot for JND thresholds versus working memory (E), auditory emotion recognition (F) or C-TOPP composite (reading) (G). Text in scatter plot shows correlations after control for group (partial r). Error bars represent standard error of the mean. ***P < 0.001.

Figure 3

Neurophysiology of auditory plasticity. (A) Grand average ERP waveforms for patients and controls during plasticity session. (B) Time-frequency plots for controls (left) and schizophrenia patients (right) for ITC (top) and baseline corrected single-trial power (bottom). The timing of S1 and S2 presentation, as well as sensory response, retention, and motor preparation intervals are as illustrated. (C–E) Bar graph for controls and schizophrenia patients for N1 (C) and θ-ITC (D) and single-trial power (E) for the indicated time window and frequency band. Insets for D and E are topographical plots (headmaps) for θ-ITC (top) and β-power (bottom) during the preparation interval, showing the expected localization over frontocentral scalp for both sets of measures. *P < 0.05; ***P < 0.001.

Figure 4

Effects of d-serine on auditory plasticity. (A) Line graph of % change in NMDA across sessions for the indicated treatment orders. (B) Bar graph of plasticity for controls and for schizophrenia patients receiving indicated treatments. (C) Grand average ERP waveforms for d-serine or placebo for ERP to plasticity session. (E) Time-frequency plots for placebo (left) and two consecutive sessions of d-serine (right) for ITC (top) and single-trial power (bottom). (F) Bar graph for controls and schizophrenia patients for θ-ITC during the preparation interval (indicated by asterisk in C). (G) Scatter plot for final threshold versus motor-preparation interval θ-ITC. Text in scatter plot shows correlations after control for group (partial r). Error bars indicate standard error of the mean. *P < 0.05; ***P < 0.001. N.S. = not significant.

Figure 5

Effects of d-serine on MMN. (A) Voltage topography maps for MMN pre-training (top) and post-training (bottom) shown at peak latencies. Analysed electrode noted by red circles. (B) Bar graph for controls and schizophrenia patients for pre-training amplitude (left) and change (right) in amplitude. Error bars indicate standard error of the mean. (C) Scatter plot for JND% change versus change in MMN amplitude to the trained tone. *P < 0.05. N.S. = not significant.

Figure 6

Schematic of distributed hierarchical model of auditory plasticity in schizophrenia, modified from Daikhin and Ahissar (2015) and Javitt and Sweet (2015). We show an auditory cortex pyramidal cell receiving bottom-up input from the thalamic medial geniculate nucleus (MGN), parvalbumin (PV), and somatostatin (SST) interneurons (Womelsdorf et al., 2014), which in-turn receive top-down input from intraparietal or frontoparietal neurons (inset). NMDAR, noted by an asterisk, appears to be involved at multiple levels.