Neonatal hypoxia is a medical condition resulting from the deprivation of oxygen to a newborn infant lasting long enough to cause harm, usually to the brain. It remains a serious condition that causes significant mortality and morbidity, including seizures. Outcomes for babies suffering from neonatal hypoxia and seizures are particularly poor, with a 23% mortality rate (1 million neonatal deaths per year worldwide) and survivors developing long-lasting neurological disorders (e.g., cerebral palsy, epilepsy, learning disabilities). Current diagnostics and treatment have limitations; it is an unmet medical need to better understand the pathophysiology of neonatal hypoxia to find novel diagnostic tools that can support more effective treatments. Magnetic resonance imaging (MRI) is the gold standard for the assessment of brain injury in infants with neonatal hypoxia and has been used to characterise the effect of neonatal hypoxia in mouse models which include an ischemic element. However, to date no work has been completed to characterise these effects in a purely hypoxic model.

We evaluate the spatial and temporal pathophysiology of the brain using a mouse model of neonatal hypoxia. At postnatal day 7 (P7), pups are placed in a hypoxia chamber at 34°C and are subjected to hypoxic (5% oxygen / 95% nitrogen) or normoxic (room air) conditions for 15 minutes. Two weeks post-hypoxia (P21), all mice undergo gelatine training for 2-3 days. One group of hypoxia-exposed mice are fed jellies containing an anti-inflammatory compound (MCC950, 75 mg/kg) twice a day for five days, while all others receive control jellies (n=12 per group). Between P28 and P40, mice perform a battery of behavioural tests, including the light/dark box, open field, and novel object location. 5 weeks after hypoxia, mice receive a 0.4 mmol/kg dose of manganese chloride (MnCl2) and 24 hours later T1- and T2-weighted scans are obtained using a 9.4T MRI scanner. All statistical comparisons between groups use a Kruskal-Wallis non-parametric test.

We evaluate anatomical changes over time after hypoxia and compare them with control and MCC950-treated littermates. T1-weighted scans do not show a change in volume in the cortex or hippocampus, while T2-weighted images show an increase (p=0.0271) in the volume of cytotoxic oedema particularly in the hippocampus, showing that the inflammatory process may be persistent over time. Vehicle-treated hypoxia mice display significant increases in latency to exit the light chamber (p=0.0063) and open field immobility (p=0.0132) relative to controls, as well as a decrease in time spent with the object in the novel location (p=0.0039). This level of impairment strongly correlates with the volume of oedema observed in the hippocampus. MCC950-treated hypoxia mice display no significant behavioural or anatomical differences relative to either group, but their behavioural phenotype is more similar to that of controls.

In conclusion, neonatal hypoxia has profound and lasting consequences in the mouse brain characterised by hippocampal inflammation, which leads to behavioural impairments and long-term neurodevelopmental deficits. Treating hypoxia-exposed mice with MCC950 two weeks post-hypoxia has shown indications of reducing this inflammation and subsequent behavioural impairment, but a larger sample size is still required to provide greater statistical power.