The hippocampus, an integral component of the corticolimbic circuitry of the brain, has been recently implicated in the pathophysiology of schizophrenia. This article has employed quantitative morphometric techniques to determine whether abnormalities of posterior hippocampal cross-sectional area, as well as the number, size, and degree of disarray of pyramidal neurons were present in 9 control and 14 schizophrenic subjects. Seven schizophrenic patients showed evidence of superimposed mood disturbance (schizoaffective type), while the remaining seven were a mixture of paranoid, undifferentiated, and catatonic types. All morphometric measurements were conducted under strictly blind conditions; stepwise multiple regression and analyses of covariance were used to evaluate the effects of various confounding variables. There were no differences in the cross-sectional size of the hippocampus or degree of neuronal disarray between the two groups. Similarly, the number of pyramidal neurons was also the same in sectors Cornu Ammonis (CA) 2, CA 3, and CA 4 for the controls and schizophrenic subjects. In CA 1, the schizophrenic subjects without mood disturbances showed a significant reduction (36%) of pyramidal neuron numbers when compared with those of both controls and patients with mood disturbance. Pyramidal neurons were smaller in all sectors of the schizophrenic specimens, CA 1 (p less than or equal to 0.01), CA 2 (p less than or equal to 0.01), CA 3 (p less than or equal to 0.01), and CA 4 (p less than or equal to 0.005), but there were no differences with respect to the presence of mood disturbances. Corrections for the effects of age, postmortem interval, fixation interval, hypoxia, and neuroleptic exposure did not alter the pattern in the data. The significance of a smaller size of hippocampal pyramidal neurons in this group of schizophrenic specimens is unclear, but it is consistent with the suggestions of other laboratories that there may be altered function of this brain region in chronically psychotic individuals.

Recent postmortem investigations have suggested that schizophrenia may involve a defect in associative information processing in the upper layers of limbic cortex. One of these studies reported that vertical processes visualized with antibodies against the neurofilament 200K subunit (NFP-200K) of the axon cytoskeleton were increased in density in layer II and upper portions of layer IIIa of the cingulate region of schizophrenic individuals. Based on this latter finding, it was hypothesized that there may be a superbundance of associative afferents to this region. To explore this possibility further, an immunoperoxidase localization of the amino acid glutamate has been employed to visualize vertical fibers in layers II and IIIa of postmortem anterior cingulate cortex in both normal controls (n = 15) and schizophrenics (n = 17). Vertical fibers were distinguished according to small or large calibers and were differentially counted with a blind computer-assisted technique. The schizophrenic group showed a markedly higher density (77.8%) of small-caliber glutamate-immunoreactive vertical fibers when compared to controls; the density of large-caliber vertical fibers also showed a similar, though smaller (30.2%), increase in the schizophrenic group. There were no differences in the density of either small- or large-caliber processes in prefrontal cortex of the two groups. The effects of age, postmortem interval, fixation, and neuroleptic exposure do not account for the differences between the normal and schizophrenic subjects. Taking together their small caliber, vertical orientation, localization in superficial layers, and marked glutamate immunoreactivity, it seems plausible that the fibers showing an increased density in schizophrenics may be glutamatergic afferents, possibly ones that are associative in nature.(ABSTRACT TRUNCATED AT 250 WORDS)

Several reports have suggested that chronic haloperidol (HAL) treatment induces ultrastructural changes in synapses of substantia nigra, corpus striatum, and medial prefrontal cortex (mPFC) of rat brain. The effects of HAL on specific cortical transmitter systems, however, are not well characterized. Recent studies have indicated that there may be a loss of gamma-aminobutyric acid (GABA)ergic cells in anterior cingulate cortex of schizophrenic subjects and this hypothesis has prompted interest in the question of whether dopamine receptor antagonists, such as HAL, may influence the activity of this transmitter system. This current report describes a quantitative light microscopic analysis of GABA-immunolabeled axosomatic terminals in mPFC of rats treated with HAL decanoate (0.5 mg/kg/day, i.m.) for a period of 4 months. GABA-containing terminals were visualized with an avidin-biotin immunoperoxidase method for localizing anti-GABA antibodies. Computer-assisted image processing was employed to determine the total number of pixels representing GABA-immunoreaction product in axon terminals that were in direct apposition to pyramidal cell bodies. Drug-treated animals showed a significant increase in the number of pixels representing GABA-immunoreaction product in axosomatic terminals of layers II, III, VI, and VI (93%, 63%, 31%, and 43%, respectively). These data are consistent with the idea that chronic HAL administration may be associated with a significant increase in the amount of GABA present in terminals surrounding pyramidal neurons of rat mPFC. The fact that GABA-containing terminals showed the greatest increase in layer II is not consistent with the known distribution of dopamine afferents to this region which is lowest in superficial laminae. Based on the laminar distribution of non-dopaminergic receptor types that have a high affinity for HAL, the effect of this drug on GABAergic transmission could potentially involve changes that are mediated through mechanisms in which 5-HT2 or sigma opiate receptors play a role.