Neurophysiological treatment effects of mesdopetam, pimavanserin and clozapine in a rodent model of Parkinson's disease psychosis

Abstract

Psychosis in Parkinson's disease is a common phenomenon associated with poor outcomes. To clarify the pathophysiology of this condition and the mechanisms of antipsychotic treatments, we have here characterized the neurophysiological brain states induced by clozapine, pimavanserin, and the novel prospective antipsychotic mesdopetam in a rodent model of Parkinson's disease psychosis, based on chronic dopaminergic denervation by 6-OHDA lesions, levodopa priming, and the acute administration of an NMDA antagonist. Parallel recordings of local field potentials from eleven cortical and sub-cortical regions revealed shared neurophysiological treatment effects for the three compounds, despite their different pharmacological profiles, involving reversal of features associated with the psychotomimetic state, such as a reduction of aberrant high-frequency oscillations in prefrontal structures together with a decrease of abnormal synchronization between different brain regions. Other drug-induced neurophysiological features were more specific to each treatment, affecting network oscillation frequencies and entropy, pointing to discrete differences in mechanisms of action. These findings indicate that neurophysiological characterization of brain states is particularly informative when evaluating therapeutic mechanisms in conditions involving symptoms that are difficult to assess in rodents such as psychosis, and that mesdopetam should be further explored as a potential novel antipsychotic treatment option for Parkinson psychosis.

Keywords: Antipsychotics; Behavior; High-frequency oscillations; In vivo; Local field-potentials.

Fig. 1

Overview of experiment design and brain structures recorded. A) Experimental timeline (period from lesion to sacrifice corresponds to approximately 3 months). B) Summary of doses, and administration routes/times for the compounds investigated. C) Overview of tip positions for all recording electrodes, as reconstructed from postmortem CT scans (overlapping sites are indicated by higher color saturation [21]. D) Summary of anatomical locations analyzed; colors mark structures with predominantly cognitive/limbic (purple) versus sensorimotor (yellow) related functions.

Fig. 2

Quantification of spontaneous behavior in the open-field. A) Average locomotion displayed per second. B) Frequency of locomotion bout initiation. Black dots correspond to individual recording sessions, and asterisks denote significant differences with respect to baseline (p ​< ​0.05, ANOVA). Error bars indicate standard error of the mean.

Fig. 3

Example of LFP data collection during an experiment and spectral features of the MK-801 PD-P model. A) Example mean spectrogram from a single recording constructed by averaging spectrograms obtained from all differential LFP signals recorded from electrodes in OFC in the lesioned hemisphere. The spectrogram is normalized to the arrhythmic background (colors denote relative power [dB_fractal]). The vertical line marks the administration time for MK-801+vehicle. B) Spectra obtained by averaging all differential LFP signals from electrode pairs in the corresponding structures, across all animals and recordings. The spectra are normalized to the arrhythmic background. Gray: baseline. Red: MK-801+vehicle. Left column: Intact hemisphere. Right: lesioned hemisphere.

Fig. 4

Summary of treatment effects on high-frequency oscillation features. A) Detection rate B) Band power (dBfractal) C) Peak frequency (Hz). D) Map of significant changes in HFO detection rate in the lesioned hemisphere (%). E) Map of significant changes in HFO band power in the lesioned hemisphere (dB_fractal). F) Map of significant changes in HFO peak frequency in the lesioned hemisphere (Hz). Black fields in panel C denote structures where a distinct peak was absent. Blue/red squares in panels D–F mark significant reductions/increases compared to MK-801+vehicle.

Fig. 5

Changes in functional connectivity and brain state entropy. A) Histogram of HFO peak frequency for all episodes where there is simultaneous detection of HFOs in more than one structure. The marked absence of off-diagonal points demonstrates that when simultaneous HFOs are detected, they share the same frequency. B) Distributions of phase relations between 12 example electrodes (three in each structure, color coded in red, blue and orange; located within vStr, and in mPFC, OFC and amyg, respectively) and a reference electrode located in vStr, under three different treatment conditions. C) Changes in phase-synchronization (kappa-values) for electrodes located in different pairs of structures (as indicated by the x- and y-labels) for four treatment conditions compared to MK-801+vehicle (cold colors denote reduced phase-coupling). D) Scatterplot of phase synchronization (kappa values; colorbar denotes counts) for all pairs of electrodes during MK-801+vehicle (x-axis) vs. each of the four treatment conditions. Note a similar reduction in phase-coupling across the different treatments affecting electrode pairs with relatively stronger and weaker synchronization to a comparable degree. E) Distributions of permutation entropy values of the LFP signal for all electrodes in OFC and vStr in the lesioned hemisphere, pooled across animals and recordings, for different conditions. Top: OFC. Bottom: vStr. Dark gray ∗ (asterisk) denotes a significant difference from baseline, and light gray ⋆ (star) denotes a significant difference from MK-801+vehicle. Triangles point to the direction of the significant difference between means.

Fig. 6

Comparison of global brain states induced by mesdopetam, pimavanserin, clozapine, and the selective dopamine D3 antagonist SB-227011-A. A) Concatenated LFP power spectra of six structures (1–300Hz), for vehicle and different treatments (mesdopetam, clozapine, pimavanserin, and SB-227011-A in combination with MK-801). Spectra are normalized to the arrhythmic background. B) Correlation analyses of spectra against baseline and vehicle for different treatments. The circles and whiskers show median values with 25th and 75th percentiles of the corresponding correlation coefficients against vehicle (vertical axis) and baseline (horizontal axis). Statistical comparisons are given for significant differences only (∗p ​< ​0.05; ∗∗p ​< ​0.01, Wilcoxon rank sum test). C) Pairwise spectral correlation coefficients between mesdopetam and the other treatments (mesdopetam 3 ​mg/kg and 10 ​mg/kg were merged for this analysis; Kruskal-Wallis test, ∗∗∗p ​< ​0.001; n.s. p ​> ​0.05).