Abstract:
Maternal-fetal infection and fetal neuroinflammation are highly associated with white matter injury (WMI) and neurodevelopment impairment in preterm infants. At least half of the infants that survive preterm birth experience adverse neurodevelopmental outcomes, and a further 15% develop severe problems such as cerebral palsy. The developing oligodendrocyte (OL) is intrinsically vulnerable to injury and the periventricular white matter is marred by its loss. Increasing evidence suggests inflammatory mediators, produced within the fetal brain in response to infection, contribute to the demise of OLs, the myelinating cells of the central nervous system; however the precise mechanisms involved are poorly understood. The broad objectives of this thesis are to: i) develop a primary culture model of inflammatory WMI derived from a developmentally similar population of fetal ovine glia; ii) confirm that infection and inflammation contribute to glial pathology; and iii) utilise our in vitro model to further our knowledge of the mechanisms downstream of inflammation through novel studies. Cultures of isolated glial cells have been employed in research previously, however mixed glial cultures have not been utilised to the same extent, despite the observation that all glia contribute to the pathology of neuroinflammation. Hence, the first aim was to characterise an in vitro primary culture model of infection/inflammation-mediated preterm WMI. To achieve this we systematically generated a method for isolating glial cells from the forebrains of the day 90-95 (0.65 gestation, term is 145) fetal sheep; primary cultures were exposed to tumour necrosis factor alpha (TNF-α, 50 and 100 ng/mL) and lipopolysaccharide (LPS, 1 μg/mL). In control cultures astrocytes, microglia and a developmental range of OLs were present. Following induced inflammation we identified a lack of astrogliosis and a marked increase in OL apoptosis with all treatments (p < 0.001). Moreover, TNF-α production in LPS treated cultures was markedly increased (p < 0.001 and p < 0.01 at 24 and 48 hours). We utilised primary mixed glial cultures to investigate whether matrix metalloproteinases (MMPs) were influenced by TNF-α and LPS or contributed to inflammation-induced glial injury. We demonstrated that TNF-α and LPS regulate MMP expression in glia, however their inhibition provided no protection to OLs in cultures exposed to TNF-α and only transiently improved OL survival during LPS exposure (p < 0.01 after 72 hours). We also identified a potential advantageous role for MMPs in regulating activated microglia survival after chronic TNF-α or LPS exposure. In an in vitro model of inflammatory OL injury, the effect of TNF-α a nd LPS o n T NF-α/ cyclooxygenase 2 (COX2) dependent glutamate release and glutamate receptor-mediated OL toxicity was investigated. This study confirmed the expression of alpha-amino-3-hydroxy-5- methyl-4-isoxazole-propionate (AMPA) and N-methyl-D-aspartate (NMDA) receptors on developing OLs, and verified that AMPA receptor subunit conformation was altered by TNF- α and LPS resulting in a cellular phenotype more permeable to Ca2+. However, glutamate receptor inhibition following TNF-α or LPS exposure was ineffective in preventing OL loss. Alternatively, developing OLs were protected, and glutamate concentrations effectively reduced, by the inhibition of TNF-α and COX2, suggesting that glutamate may contribute to inflammatory OL injury via non-receptor mediated mechanisms. Translating practical results at the cellular level to therapeutic interventions has many obstacles. Nevertheless, data from our studies provide an important mechanistic insight into the underlying pathophysiology of white matter inflammation and subsequent OL demise. These results and the in vitro model provide a useful tool for future advances in our understanding of the mechanisms of infection/inflammation-mediated preterm brain injury.