Altered spatial learning, cortical plasticity and hippocampal anatomy in a neurodevelopmental model of schizophrenia-related endophenotypes

Abstract

Adult rats exposed to the DNA-methylating agent methylazoxymethanol on embryonic day 17 show a pattern of neurobiological deficits that model some of the neuropathological and behavioral changes observed in schizophrenia. Although it is generally assumed that these changes reflect targeted disruption of embryonic neurogenesis, it is unknown whether these effects generalise to other antimitotic agents administered at different stages of development. In the present study, neurochemical, behavioral and electrophysiological techniques were used to determine whether exposure to the antimitotic agent Ara-C later in development recapitulates some of the changes observed in methylazoxymethanol (MAM)-treated animals and in patients with schizophrenia. Male rats exposed to Ara-C (30 mg/kg/day) at embryonic days 19.5 and 20.5 show reduced cell numbers and heterotopias in hippocampal CA1 and CA2/3 regions, respectively, as well as cell loss in the superficial layers of the pre- and infralimbic cortex. Birth date labeling with bromodeoxyuridine reveals that the cytoarchitectural changes in CA2/3 are a consequence rather that a direct result of disrupted cortical neurogenesis. Ara-C-treated rats possess elevated levels of cortical dopamine and DOPAC (3,4-didyhydroxypheylacetic acid) but no change in norepinephrine or serotonin. Ara-C-treated rats are impaired in their ability to learn the Morris water maze task and showed diminished synaptic plasticity in the hippocampocortical pathway. These data indicate that disruption of neurogenesis at embryonic days 19.5 and 20.5 constitutes a useful model for the comparative study of deficits observed in other gestational models and their relationship to cognitive changes observed in schizophrenia.

Fig. 1

Cytoarchitectural changes in the hippocampus of Ara-C-treated rats. (A,B) Cresyl violet images of the CA2/3 regions of the dorsal hippocampus from control (A) and Ara-C-treated (B) rats to illustrate the disruption in the laminar organization of CA2/3 pyramidal neurons in Ara-C-treated rats (indicated by an arrowhead in B). Scale bar: 200 μm. (C–F) Pyramidal neuron cell counts and regional volume measurements in the hippocampus of adult rats treated with saline or Ara-C at E19.5 and 20.5. Cresyl violet images of the dorsal hippocampus from control (C) and Ara-C-treated (D) rats. Arrowheads and asterisks denote specific regions of CA1 and CA2/3, respectively, in the dorsal hippocampus used to compile cell counts. Scale bar: 1 mm. (E,F) Bar graphs illustrating the change in pyramidal cell number (C) and corresponding change in regional volume (D) of CA2/3 and CA1 region of the dorsal and ventral hippocampus. *P < 0.05, Student's t-test.

Fig. 2

BrdU immunolabeling of the dorsal hippocampus and frontal cortex. (A,B) Representative coronal sections from the dorsal hippocampus. Arrowheads in A indicate BrdU-labeled nuclear profiles in the hilus and CA1 regions in saline-treated control rats (B). Note that the corresponding regions in Ara-C-treated rats are essentially devoid of BrdU-labeled profiles. The sparse BrdU labeling in the CA2/3 region of saline-treated rats (arrow) indicates that few neurons in this region were generated at the time of Ara-C injection (A). Note, however, that those few BrdU-labeled profiles in the CA2/3 region are missing in the Ara-C-treated rat. (C,D) BrdU labeling in the PrL-IL cortex. Although prominent BrdU immunolabeling of nuclear profiles was observed in saline-treated rats in the superficial cell layers of the cortex (C, arrows), only faint labeling of occasional nuclear profiles could be detected in Ara-C-treated rats (D, arrows). The results suggest that many superficial layer neurons generated between E19.5 and E20.5 are missing in the Ara-C-treated rats. Scale bar: 500 μm (for A-D). (E,F) High magnification of BrdU immunoreactivity in the PrL-IL cortex of saline- (E) and Ara-C-treated rats (F). Both darkly labeled (black arrowhead) and lightly filled (white arrowhead) nuclear profiles are observed in saline-treated rats, while only lightly labeled BrdU profiles (white arrowheads) are observed in the Ara-C-treated rats. Scale bar: 10 μm. (G,H) Estimates of the number of BrdU-labeled profiles in the ventral hippocampus (G) and PrL-IL cortex (H) of saline- and AraC-treated rats. These regions correspond to the respective sites of electrical stimulation and recording in the LTP experiments. *P < 0.05.

Fig. 3

Effects of prenatal Ara-C exposure on DA, DOPAC and homovanillic acid levels in the cortex. Tissue levels, expressed in pg/mg of tissue were determined in extracts obtained from adult rats. **P< 0.005, *P <0.05, Student's t-test.

Fig. 4

Acquisition phase of the Morris water maze task in saline- and Ara-C-treated rats. Both saline- (open circles) and Ara-C- (filled circles) treated rats improved performance across the 4-day acquisition of the task. However, saline-treated rats improved at a faster rate, with significantly faster times than Ara-C-treated rats on days 2, 3 and 4 (*P< 0.05, Fisher's LSD). No significant differences were observed between treatment groups on the first day of training.

Fig. 5

Effect of Ara-C treatment on exploration of the target platform's area in the probe trial of the Morris water maze test. Saline- (white) and Ara-C- (black) treated animals did not differ on performance in the probe test as measured by the number of crosses over the area formerly occupied by the platform (A) or the mean distance to the area formerly occupied by the target platform (C). However, when compared by start location, Ara-C rats did show significantly decreased platform crosses (P<0.05; B) and a greater mean distance to the target platform (N and E; * P < 0.05; D).

Fig. 6

Ara-C-treated rats exhibit diminished expression of in vivo hippocampal–prefrontal cortex long-term potentiation. (A) Schematic representation of recording (left panel) and stimulating (right panel) electrodes from control (open circles) and Ara-C-treated (filled circles) rats. (B) Representative cortical field potentials evoked by test stimuli applied before (solid line) and after (dashed line) high-frequency stimulation of the CA1 region of the hippocampus. The peak amplitude (arrow) and latency to onset (horizontal line) of negative going component of the evoked potential was expressed as a percentage of an average value obtained prior to tetanic stimulation. (C) Comparison of the percentage change in the amplitude (mean ± SEM) of the negative component of the cortical evoked potential in saline- (open circles) and Ara-C-treated rats. Control rats showed a 50% increase in field potential amplitude following two tetanic train stimulations (onset indicated by two vertical arrows). By contrast, Ara-C-treated rats showed an approximately 20% increase in evoked potential amplitude. Group differences were significant at all but three time points (*P < 0.05, Holm-Sidak).