The effect of chronic prenatal hypoxia on the development of mature neurons in the cerebellum

An inadequate intrauterine environment is thought to affect fetal brain development [1, 2]. Very low birth weight, which is affected by intrauterine growth restriction (IUGR), is associated with a regional reduction in brain volume [3] and an increased risk of white matter injury [4].

Abnormal brain development is associated with neurological disorders such as cerebral palsy [5], schizophrenia [6], and cognitive [7] and learning/memory [8] deficits. These neurological problems are affected by conditions that cause chronic placental insufficiency (CPI), such as malnutrition, hypoxia, and acidemia [9]. Several models of CPI have been used to determine the relationship between neurological disorders and abnormal brain development. For example, Duncan et al. reported that late gestational CPI affects the morphology and neurotrophin expression of the postnatal sheep brain [10]. Moreover, IUGR induced by CPI in the guinea pig was shown to affect the maturation of myelin [11].

In a recent guinea pig model of CPI, which was induced by uterine artery ligation, the growth of the cellular and neuropil layers were found to be reduced in the cerebellum, without concomitant changes in the expressions of immunoreactivity to either brain-derived neurotrophic factor (BDNF)-IR or tropomyosin receptor kinase B (TrkB)-IR [12]. Some studies have implicated hypoxia in structural alterations of the cerebellum during development. First, significant reductions have been observed in the width of the molecular and inner granular layers [13]. Second, the size and number of Purkinje cells were also found to be reduced [14]. Unlike changes in other layers of the cerebellum, the changes in the cerebellar external granular layer (EGL) induced by chronic prenatal hypoxia are not well understood. Granule cells comprise half of the neurons in the central nervous system, and their precursors are dispersed within the cerebellar EGL [15].

The effects of hypoxia on the development of mature neurons in the EGL were investigated in the present study using a guinea pig CPI model.

The guinea pig is a useful model of brain development because it has a relatively long gestation period (with term occurring at approximately 67 dg) and most of the development occurs in utero. Lobules I and VIII were selected for examination because they are the first and one of the last lobules respectively, to mature [18]. Unilateral uterine artery ligation was performed at 30 dg, and we studied fetuses at 50 and 60 dg. It has been reported previously that the cerebellum of 30-dg fetuses appeared immature and consisted of undifferentiated neurons [16]. Furthermore, Jansson et al. [19] reported that fetal weight and placental blood flow were decreased in guinea pigs at 45, 55, and 65 dg. Therefore, we expected that neurogenesis in the guinea pig cerebellum would be reduced at 50 and 60 dg. However, the relative proportions of NeuN-IR cells did not differ significantly between control and GR animals at 50 and 60 dg, and the pattern of immunoreactivity was similar between 50 and 60 dg. On the other hand, the density of NeuN-IR cells was increased in GR fetuses at 60 dg.

NeuN is a neuronal-specific nuclear protein [20] and NeuN immunoreactivity is used to obtain independent estimates of total numbers of neuronal and nonneuronal cells [21]. At 50 dg, the proportion of NeuN-IR cells relative to the total number of cells and the density of NeuN-IR cells in the EGL did not differ between normal and hypoxic cerebella. This suggests that chronic hypoxia does not have a comprehensive effect on neuronal cell formation at 50 dg. However, a significant reduction in the number of NeuN-positive neuronal nuclei was demonstrated in the developing chick brain [22] and Rees et al. [23] reported a reduction in the density of cells undergoing mitosis in the EGL layer of hypoxemic fetal sheep compared to controls. The differences in the findings between the present study and others may be at least partly attributable to the use of different species of experimental model and to the degree of hypoxia imposed. The expression of Hif1α is induced by hypoxia-ischemia [24]. Therefore, while the presence of HIF1α immunoreactivity in the fetal tissue at 50 dg demonstrates that at that time the fetus had been in a hypoxic condition, it does not indicate whether that hypoxia was mild or severe. Jansson et al. performed unilateral uterine artery ligation at 30 to 32 dg in the guinea pig [19]. Placenta blood flow was reduced by 25% at 45 dg and by almost 50% at 55 dg in the ligated horn [19]. Further experiments are needed to determine the degree of hypoxia that is established using this ligation model, and the degree to which it varies between animals at the same gestation time points. However, since the major developmental events in the guinea pig brain occur before 48 dg [25], it is possible that ligation of the uterine artery affects brain development at 50 dg.

At 60 dg, the proportion of NeuN-IR cells did not differ between control and GR cerebella. However, the density of NeuN-IR cells was increased in GR fetuses, indicating a larger total cell volume. The density of HIF1α-IR cells did not differ between 60 and 50 dg, which implies that the hypoxic condition had not changed between these two time periods. The TUNEL assay findings demonstrate that cell death was not affected by the hypoxia induced in this model. Pax6 is a neuronal stem cell marker [26] and NeuroD is implicated in the differentiation of neurons [27]; we observed no cell immunoreactivity to either of these markers, which suggests that neuronal maturation is already complete at this stage (that is, 60 dg). We performed immunohistochemistry with Pax6 and NeuroD at 30 dg, and observed a few Pax6-IR and NeuroD-IR cells. However, the number of sections was insufficient for providing representative data. An earlier-gestation guinea pig model is required to study the early stages of neuron formation. Since in normal brain development the granule cells in the cerebellar EGL migrate medially and the EGL gradually disappears [28], it is possible that the migration of mature neurons was delayed at this time.

At one week after birth, NeuN-IR cells were only rarely observed in the EGL under both the GR and control conditions. As mentioned above, this is the usual process of cerebellum development. In our study, the observed prenatal alterations to the EGL did not persist postnatally. More studies with animals of different ages are required to establish how chronic prenatal hypoxia affects the mature cerebellum.

The findings of this study suggest that chronic prenatal hypoxia affects the process of neuronal production late in fetal life, but that this effect does not persist postnatally. The approach used in this study is useful for extending our understanding of neurogenesis in the EGL, and the findings may be useful for elucidating the brain injury caused by prenatal hypoxia.