It has long been recognized that intracisternal injection of kaolin is an effective method for producing hydrocephalus in rodents and that the result is variable and somewhat unpredictable . In this study, intracisternal injection of kaolin resulted in the development of hydrocephalus in about 90% of the rats. Despite the fact that the same volume of kaolin was injected, the rats developed varying degrees of ventricular dilatation which we have classified into mild, moderate and severe. The brains of the 3-week old rats used in this study were developmentally similar to that of a human infant . Cisternal injection of kaolin causes a chemical arachnoiditis and results in obstructive hydrocephalus from impairment of flow of cerebrospinal fluid in the basal cisterns; this model is therefore similar to post-meningitic hydrocephalus in the human infant. With neonatal meningitis being one of the commonest causes of hydrocephalus in Africa , this is an appropriate experimental model for studying this disease. After an initial weight loss, hydrocephalic rats regained weight at the same velocity as controls, but mean body weight remained lower. This was most likely due to loss of appetite and reduced feeding activity in the early phase of induction of hydrocephalus. Del Bigio  has previously stated that delayed growth is one of the first signs of hydrocephalus onset.
The reduced motor activity of the hydrocephalic rats in the open field was most effectively illustrated by a significant reduction in vertical movements (rearing). Whereas the animals were freely mobile and exhibited preservation of horizontal movements, the reduced tendency for exploratory rearing may also be manifestation of hindlimb motor weakness, which could be due to damage of the periventricular axons. Although rearing activity was correlated with ventricular enlargement, the overall association was not statistically significant. This may be an area requiring further exploration, given the proximity of lower limb fibres to the ventricular system in mammals. The forelimb grip test, which directly measures the muscular strength in the forelimbs, lends weight to the observations in the open field test. The graduated reduction in muscular strength as the ventricular size increased observed here is similar to that reported from previous studies which used other tests of motor function like the rotating cylinder and ability to transverse a narrow beam [7, 17]. Hydrocephalus is associated with damage to axons in the periventricular white matter including long tracts that project to the spinal cord , and manifests as a delayed myelination process in neurologically immature rats . The relationship between ventricular distension and forelimb motor control has not been extensively studied. Our findings suggest that in this model, the forelimb grip test is a valuable measure of the impact of hydrocephalus on motor function in the proximal limbs. Learning and memory, as tested by the Morris water maze was observed to be impaired. The impairment of spatial learning ability , which was shown by the prolonged escape latency observed in the hydrocephalic rats, suggests that hydrocephalus impairs the function of the hippocampus and its connections i.e. fimbria, alveus and fornix.
The progressive flattening and discontinuation of the ependymal lining with increasing ventriculomegaly was distinct. From cuboid cells with cilia and round nuclei, the cells gradually lost their cilia and became flattened, stretched or disrupted with gaps seen in the epithelium in severe hydrocephalus – this was most obvious at the dorso-lateral angle of the lateral ventricle. The subependymal layer showed periventricular reactive astrogliosis with hypertrophy and a non-significant increase in the number astrocytes in the mildly hydrocephalic rats; however it was decreased in the moderate and severely hydrocephalic rats. This could mean that the neuroinflammatory process diminishes as the hydrocephalic process progresses; it could also be due to a generalized apoptosis taking place as hydrocephalus progresses, which affects all the brain cells including astrocytes. Fukushima et al. reported an increase in production of new brain cells which were almost exclusively differentiated into astrocytes in hydrocephalic rats; however the study did not distinguish between the various degrees of ventriculomegaly. The progressive thinning of the corpus callosum with increasing ventriculomegaly implies that there is loss of the periventricular white matter secondary to pressure from the expanding ventricles. Oedema of the periventricular white matter with consequent enlargement of the extracellular spaces occurred in keeping with findings from other studies ; this is most likely due to stasis of extracellular fluid which normally drains into the CSF. These spaces increase as the ventriculomegaly progressed and thus demonstrates that the more severe the accumulation in the ventricles, the worse the oedema.
Oligodendrocytes in the periventricular white matter were reduced in quantity as the ventriculomegaly increased and the corpus callosum thinned out. The adjacent parietal cortex also showed this decline in oligodendrocytes which was most marked in the severely hydrocephalic rats. The loss of oligodendrocytes in hydrocephalus has been shown by Del Bigio and Zhang  to be due to cell death, caused by chronic ischaemia, toxicity from accumulation of glutamate, and inflammatory white matter injury.
Proliferation of astrocytes usually occurs secondary to central nervous system injury  and has been reported in hydrocephalus. In this study, the increased GFAP staining in the periventricular regions suggest that the periventricular white matter is an important site of brain injury in hydrocephalus, possibly from the stretch and compression which accompanies ventriculomegaly. Other factors such as hypoxia, ischaemia and cellular death, all present in hydrocephalus, could also be contributory. Gliosis occurred in both gray and white matter, but was more marked in the periventricular white matter. The role of periventricular astrocytic proliferative change in hydrocephalic brain injury is supported by studies that demonstrated increased brain GFAP ribonucleic acid (RNA) levels (measured by western blot) with progression of hydrocephalus in kaolin models of kittens , and increased CSF levels of GFAP in human hydrocephalic patients .
Reactive astrocytes secrete inhibitory molecules such as proteoglycans which create a hostile environment for remyelination of injured axons . Together with microglia cells, the astrocytes also form a glial scar (that is, astrocytic proliferation and intracellular accumulation of glial filaments) which may play a major role in the chronicity of hydrocephalus in children as their brains are less pliable than normal . There are similar reports of astrocytic activation observed in hydrocephalic Texas (H-Tx) rat models of congenital hydrocephalus . Significantly increased expression of common pro-inflammatory cytokines have been reported in structures distant from the kaolin deposits but adjacent to the expanded ventricles such as the frontal cortex, parietal cortex, caudate- putamen, and hippocampus with normal expression in the medulla immediately adjacent to the kaolin deposits . This lends weight to the belief that the neuroinflammation observed is secondary to the hydrocephalic process and not the direct consequence of kaolin injection. This has implications for putative pharmacological treatment of hydrocephalus.
Microglial cells are the primary immune effector cells of the brain parenchyma. They play a major role in the response to brain injury from a variety of causes including hydrocephalus. The characteristic morphological changes in microglia activation, that is, an altered morphology with withdrawal of the dendritic branches were observed in this study.
The brain parenchyma and periventricular white matter suffer various insults from the hydrocephalic process. The pressure from the fluid stretches the overlying cortex, stretches and sometimes damages the axons through ischaemia, the process of remyelination is limited by astrocytic proteoglycans and glial scar formation and the oligodendrocytes themselves are also reduced in number from necrosis/ apoptosis [16, 21, 28]. Each one of these processes needs to be addressed for the management of the hydrocephalic patient to be successful. Death of the oligodendrocytes is pivotal as they are needed for a successful remyelination process and since they do not seem to be regenerated [20, 21], intervention needs to be timely.
In this study, the hippocampus appeared grossly normal and its cell population seemed unaltered on routine H&E staining (figure not shown), however, it was clearly functionally altered as manifested by the impairment in spatial learning and memory. This is however not surprising since changes may be substructural. It has been shown that biochemical composition and synaptic potentials are altered in hippocampal neurons in hydrocephalic rats . Immunochemical staining of cholinergic neurons in the hippocampus, revealed a strong negative correlation between ventricular size, population of AChE (acetylcholinesterase) and ChAT (choline acetyltransferase) – immunosensitive cells in the hippocampus and spatial memory impairment . Tashiro et al.  also reported progressive functional impairment of glutamic acid decarboxylase (GAD), parvalbumin (PV) and calbindin D28K neurons in the cerebrum and hippocampus observed by immunostaining while Nissl staining of the same tissues revealed almost no morphological changes.
The periventricular white matter seems to be the primary target in hydrocephalus with oedema, thinning and reduced cell populations observed which correspond to the degree of ventriculomegaly. However, the parietal cortex is also thinned and compressed by the increased ventricular size.