Effects of antenatal inflammation and postnatal oxygen fluctuation on developing white matter in a rodent model of prematurity

Abstract

Inflammation and oxidative stress are increasingly recognised as important independent mediators of preterm brain injury and have been implicated in the pathogenesis of cerebral palsy and cognitive impairment. Such exposures are common for the premature infant in whom infection and inflammatory morbidities occur in around 60%. Furthermore, many preterm infants require oxygen therapy and respiratory support due to lung immaturity. Epidemiological and experimental studies indicate that in addition to the independent effects of inflammation and extreme hyperoxia on the developing brain, inflammation preconditions the developing brain resulting in variable injury when exposed to subsequent hypoxia-ischaemia. However experimental studies employing exposure to more modest oxygen fluctuations are lacking. This thesis characterises a clinically relevant model of prematurity where the developing brain is exposed to low grade inflammation and oxygen fluctuation around a hyperoxic mean. We hypothesise that antenatal inflammation and postnatal oxygen fluctuation, both alone and in combination, have detrimental effects on developing white matter. Pregnant dams received intraperitoneal lipopolysaccharide (LPS) or saline on G18 and G19. Dams and their pups were then reared in room air or fluctuating hyperoxia (circa 10kPa) for seven days. We measured longitudinal brain and body growth in different experimental groups to 12 weeks. Whole brains were examined for mRNA expression of inflammatory cytokines (TNFα, IL-1β, IL-6 and IL-10) and markers of oxidative injury (iNOS, SOD2). To determine the effect of perinatal insults on developing white matter, we analysed the expression of myelin basic protein (MBP) and glial fibrillary acidic protein (GFAP) in the internal and external capsule. We also examined white matter tracts for differences in microglia (CD68), oligodendrocyte progenitor cells (NG2), oligodendroglial cells (Olig2) and cell death (cleaved caspase3). Behavioural studies (Morris Watermaze Test, Elevated Plus Test and Open Field Test) were undertaken at 12 weeks of age to detect any long-term functional difference between the groups. Antenatal inflammation reduces both brain and body growth at P7. This normalises by P14 unless this inflammatory insult has been followed by postnatal oxygen fluctuation, where brain and body growth restriction persists until P14. We defined our inflammatory response at P1 following antenatal inflammation and did not observe elevation of mRNA at P1. We demonstrated increased SOD2 at this time point, indicating a reparative process. At P7 we observed a significant reduction in the oxidative response following combined exposure to antenatal inflammation and postnatal oxygen fluctuation, indicating a potential limit to, or suppression of, the reparative process. In terms of white matter injury, antenatal inflammation reduces myelination at P7. There is no synergistic effect of inflammation and oxygen fluctuation on MBP immunohistochemistry at P7. However, MBP mRNA expression is increased in pups exposed to both insults compared to those exposed to inflammation alone suggesting that the oxygen fluctuation may stimulate MBP production in response to oxidative injury. MBP mRNA levels and protein expression have all normalised by P14. We observed a reduction in total cell number in the external capsule and corpus callosum in the dual insult group, without an increase in caspase. In keeping with other studies we detected no effect of our perinatal insults on NG2+ve oligodendrocytes. Olig2+ve cell numbers were also consistent between experimental groups. In further characterisation of the cellular response, antenatal inflammation followed by postnatal oxygen fluctuation resulted in a decrease in GFAP mRNA at P7, an effect which was reversed and significantly increased by P14 suggesting delayed activation of the innate immune system. No difference was observed in microglial numbers between experimental groups. There was however, increased microglial cell death (CD68 + caspase) in the group exposed to antenatal inflammation. When this insult was combined with postnatal oxygen fluctuation there was a comparative decrease in microglial cell death, which may reflect an earlier peak of microglial cell death, due to an increased and sustained inflammatory stimulus. Morris Watermaze testing demonstrated that pups exposed to both insults took longer than controls to locate the hidden platform on day 1, which is a measure of spatial learning. The Elevated Plus Test and Open Field Test demonstrated that pups exposed to both insults were less anxious and took more risks than pups exposed to single insults. In conclusion, within a clinically relevant preterm model, antenatal inflammation transiently disrupts both brain and body growth and myelination of the motor tracts of the developing brain. Moreover, when combined with postnatal oxygen fluctuation, detrimental effects on growth are amplified and sustained. Decreased cell numbers are also observed within white matter tracts. In terms of long term functionality, these pups display disinhibition of behaviour as young adults. Collectively, this thesis demonstrates that synergistic actions of common low-grade perinatal insults may alter normal neurodevelopment, and that this may carry a risk of neurodevelopmental sequelae for preterm infants.