Juvenile antioxidant treatment prevents adult deficits in a developmental model of schizophrenia

Abstract

Abnormal development can lead to deficits in adult brain function, a trajectory likely underlying adolescent-onset psychiatric conditions such as schizophrenia. Developmental manipulations yielding adult deficits in rodents provide an opportunity to explore mechanisms involved in a delayed emergence of anomalies driven by developmental alterations. Here we assessed whether oxidative stress during presymptomatic stages causes adult anomalies in rats with a neonatal ventral hippocampal lesion, a developmental rodent model useful for schizophrenia research. Juvenile and adolescent treatment with the antioxidant N-acetyl cysteine prevented the reduction of prefrontal parvalbumin interneuron activity observed in this model, as well as electrophysiological and behavioral deficits relevant to schizophrenia. Adolescent treatment with the glutathione peroxidase mimic ebselen also reversed behavioral deficits in this animal model. These findings suggest that presymptomatic oxidative stress yields abnormal adult brain function in a developmentally compromised brain, and highlight redox modulation as a potential target for early intervention.

Copyright © 2014 Elsevier Inc. All rights reserved.

Figures

Figure 1. A NVHL blocks the adolescent increase in PV interneuron labeling in the PFC

(A) Representative micrographs of prefrontal PV staining in juvenile (P21) and adult (P61) SHAM, NVHL, and NAC-treated NVHL rats. Scale bar is 80 μm. See Suppl. Fig 1 for lesion extent. (B) Bar graphs illustrating PV cell counts using unbiased stereology at P21 (left) and P61 (right) in all four treatment groups. PV cell count increased between P21 and P61 in SHAM, but not in NVHL rats. Juvenile NAC treatment rescued the progression in PV cell numbers in NVHL rats. ANOVA F(5,36)=4.7, p=0.002. Age: F(1,36)=10.2, p=0.003, Lesion: F(1,36)=1.36, p=0.11, Lesion x Age: F(1,36)=6.7, p=0.014, Treatment: F(1,36)=3.9, p=0.057, Treatment x Age: n.s. (C) Diagram illustrating the time course of NAC treatment and juvenile (P21) and adult (P61) assessments. In this and all other figures, data are expressed as mean ± SEM, *p<0.05.

Figure 2. Oxidative stress shown with 8-oxo-dG in the PFC of NVHL rats

(A) Representative micrographs showing double labeling for PV (red) and 8-oxo-dG (green) in the PFC in the four groups at P21. Scale bar is 10 μm. (B) Summary of the datashowing that an NVHL causes a massive increase in 8-oxo-dG labeling in the PFC at P21 that is prevented with juvenile NAC treatment. Top graph illustrates 8-oxo-dG fluorescence intensity and the bottom graph quantifies the number of labeled voxels in each group. ANOVA for 8-oxo-dG intensity: F(3,14)=13.7, p=0.00002, Lesion F(1,14)=18.4, p=0.008, Treatment F(1,14)=9.6, p=0.008, Lesion x Treatment F(1,14)=13.0, p=0.003. (C) Representative micrographs showing double labeling for PV (red) and 8-oxo-dG (green) in the PFC at P61. Scale bar is 10 μm. (D) Summary of the data showing that the NVHL increases 8-oxo-dG in the PFC at P61, which is prevented with juvenile NAC treatment. ANOVA for 8-oxo-dG intensity: F(3,18)=7.8, p=0.001, Lesion F(1,18)=12.5, p=0.002, Treatment F(1,18)=0.13, p=n.s., Lesion x Treatment F(1,18)=10.8, p=0.004.

Figure 3. Oxidative stress in the PFC of adult NVHL rats shown with 3-NT

(A) Representative micrographs showing triple labeling for 3-NT (green), WFA (blue) and PV (red) in the three groups. Scale bar is 100 μm. b, (B) Summary of the data showing that an NVHL causes a significant increase in 3-NT labeling in the PFC at P61 that is prevented with juvenile NAC treatment. Top graph illustrates 3-NT fluorescence intensity in each group. One-way ANOVA for 3-NT intensity revealed a very significant effect of treatment (F(2,53)=85.2, p<0.0001). Comparisons between each pairs using Tukey-Kramer also showed significant differences (P<0.0001) for SHAM versus NVHL and NVHL versus NAC.

Figure 4. The NVHL causes increased oxidative stress in PV, but not CR and CB interneurons, which is prevented by developmental NAC treatment

(A) Micrographs showing 8-oxo-dG labeling (green) of parvalbumin (PV)-, calretinin (CR)-and calbindin (CB)-positive interneurons (red) in the PFC of SHAM, NVHL and NAC-treated NVHL rats. Scale bar is 10 μm. (B) Summary of the data. In PV interneurons, 8-oxo-dG labeling increased following an NVHL lesion, which was prevented with NAC treatment (Treatment: F(2,65)=212.97, p<0.0001). ***p<0.001.

Figure 5. Perineuronal nets (PNN) are reduced in the PFC of adult NVHL rats, but rescued by juvenile NAC treatment

(A) Representative micrographs showing double labeling of PV (red) and Wisteria floribunda agglutinin (WFA; green), which labels PNN. Scale bar is 10 μm. (B) Plots illustrating PV interneuron (PVI) counts (top) and the number of cells co-labeled with PV and WFA (bottom). PVI count is reduced following an NVHL lesion, and this reduction is prevented with juvenile NAC treatment. (Overall effect: F(8,16)=3.8, p=0.01, PVI count: F(2,11)=15.3, p<0.0007). The number of WFA PVI decreases in NVHL rats compared to controls, and this reduction is prevented with juvenile NAC treatment (PNN count: F(2,11)=28.5, p<0.0001). **p<0.01, ***p<0.001.

Figure 6. Electrophysiological deficits are rescued by NAC treatment in NVHL rats

(A) Representative traces of excitatory post-synaptic potentials (EPSP) evoked by superficial layer electrical stimulation in adult PFC before (black trace) and after (green trace) bath application of the D2-agonist quinpirole (5 μM). (B) Neurobiotin-filled layer V pyramidal cell in the PFC; the relative position of the bipolar stimulating electrode and the recording electrode are shown schematically. (C) Bar graphs illustrating the magnitude of EPSP attenuation by quinpirole in slices from SHAM, NVHL, and NAC-treated NVHL rats. In sham rats, quinpirole reduces the size of the synaptic response, whereas in NVHL rats this attenuation is absent. NAC treatment during development reverses this deficit in NVHL animals (ANOVA: F(2,39)=3.328, p=0.046). (D) Traces from in vivo intracellular recordings in PFC pyramidal neurons showing responses to electrical stimulation of the ventral tegmental area (VTA) with trains of 5 pulses at 20 Hz in anesthetized SHAM (top), NVHL (middle), and NAC-treated NVHL (bottom) rats. Each panel is an overlay of 5 traces that illustrate the representative type of response observed in each group, with NVHL showing enhanced firing following VTA stimulation, while firing is sparse in SHAM and NAC-treated NVHL rats. (E) Bar graph illustrating group data for action potential firing in the 500 ms epoch following VTA stimulation in all three groups. ANOVA: F(2,37)=4.5, p<0.05; NVHL firing was higher than in shams (post-hoc Tukey's q=3.9, p<0.05) and higher than in NAC-treated NVHL rats (post-hoc Tukey's q=3.6, p<0.05). In all electrophysiology experiments data from SHAM and NAC-treated SHAM rats were combined as they did not show differences.

Figure 7. Mismatch negativity (MMN) deficits are rescued by NAC treatment

(A) Representative traces of auditory evoked potentials from standard (blue) and deviant (red) stimuli in a sham (n=6; top), NVHL (n=3; middle), and NAC-treated NVHL rat (n=3; bottom). The green box highlights the epoch in which the negativity was measured (35-100 ms following the stimulus). All traces are averages of at least 80 repetitions. (B) Group data comparing MMN measured as the area under the curve in the highlighted region reveal a significant difference among groups (ANOVA: F(2,11)=9.742; p=0.006). The data illustrated are averages from 3 different sessions in each rat. A post-hoc comparison between NVHL and NVHL+NAC revealed a significant difference (Bonferroni test; p=0.005).

Figure 8. Prepulse inhibition deficits were rescued with antioxidant treatment

(A) Prepulse inhibition deficits were observed in NVHL rats when challenged with apomorphine (0.1 mg/kg, i.p.). This deficit was completely reversed with juvenile NAC treatment (Lesion: F(1,42)=3.529 p=0.067, Treatment: F(1,42)=1.644, p=0.207, Lesion x Treatment: F(1,42)=5.730, p=0.021). n=12-16, * p<0.05 compared to NVHL. (B) In another group of rats, NAC was administered starting at P35, stopped at P50, and the rats tested for PPI at P61. The bar graph illustrates PPI at three different prepulse intensities in this group with adolescent NAC treatment. ANOVA: group effect F(2,28)=3.364, p<0.045; post-hoc tests revealed only a trend for a difference in PPI in NVHL compared to SHAM (LSD, p=0.069), and a significant difference between NVHL and NAC-treated NVHL (LSD, p=0.016). (C) Some animals received ebselen from P35 and were tested for PPI at P61. There was a significant lesion effect (F(1,28)=7.11; p=0.013) and a significant lesion status by treatment interaction (F(1,28)=7.09; p=0.013). (D) Another set of animals received apocynin and were tested for PPI. We observed a significant lesion by treatment interaction (F(1,25)=4.8; p=0.038).